No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Characterization of major depressive disorder using a multiparametric classification approach based on high resolution structural images
Abstract
Major depressive disorder (MDD) is one of the most disabling mental illnesses. Previous neuroanatomical studies of MDD have revealed regional alterations in grey matter volume and density. However, owing to the heterogeneous symptomatology and complex etiology, MDD is likely to be associated with multiple morphometric alterations in brain structure. We sought to distinguish first-episode, medication-naive, adult patients with MDD from healthy controls and characterize neuroanatomical differences between the groups using a multiparameter classification approach.
We recruited medication-naive patients with first-episode depression and healthy controls matched for age, sex, handedness and years of education. High-resolution T1-weighted images were used to extract 7 morphometric parameters, including both volumetric and geometric features, based on the surface data of the entire cerebral cortex. These parameters were used to compare patients and controls using multivariate support vector machine, and the regions that informed the discrimination between the 2 groups were identified based on maximal classification weights.
Thirty-two patients and 32 controls participated in the study. Both volumetric and geometric parameters could discriminate patients with MDD from healthy controls, with cortical thickness in the right hemisphere providing the greatest accuracy (78%, p ≤ 0.001). This discrimination was informed by a bilateral network comprising mainly frontal, temporal and parietal regions.
The sample size was relatively small and our results were based on first-episode, medication-naive patients.
Our investigation demonstrates that multiple cortical features are affected in medication-naive patients with first-episode MDD. These findings extend the current understanding of the neuropathological underpinnings of MDD and provide preliminary support for the use of neuroanatomical scans in the early detection of MDD.
... ML studies using sMRI to diagnose MDD in adults with MDD have also been reported. Qiu et al. found that alteration of the cortical thickness in the right hemisphere could differentiate first-onset MDD patients from healthy controls, providing an accuracy of 78% [7]. They suggested that morphological alterations in the right hemisphere were more evident than those in the left hemisphere in diagnosing MDD. ...
... ML studies for the diagnosis of depression are rapidly increasing, but most studies have a limitation of a small sample size [5][6][7]64]. In a comparative study of bipolar disorder and MDD, it was difficult to identify the characteristics of bipolar I and II because the bipolar disorder subtypes were not classified in the bipolar disorder patient group. ...
... For example, it is common to use medication-naïve samples in MDD neuroimaging studies. As the effects of medication on neuroimaging findings need to be controlled, many researchers have attempted to compare medication-naïve patients and control groups [7]. MDD is a chronic disease in clinical practice, and many patients suffer from chronic impairments caused by MDD itself. ...
Depressive disorders are highly heterogeneous in nature. Previous studies have not been useful for the clinical diagnosis and prediction of outcomes of major depressive disorder (MDD) at the individual level, although they provide many meaningful insights. To make inferences beyond group-level analyses, machine learning (ML) techniques can be used for the diagnosis of subtypes of MDD and the prediction of treatment responses. We searched PubMed for relevant studies published until December 2021 that included depressive disorders and applied ML algorithms in neuroimaging fields for depressive disorders. We divided these studies into two sections, namely diagnosis and treatment outcomes, for the application of prediction using ML. Structural and functional magnetic resonance imaging studies using ML algorithms were included. Thirty studies were summarized for the prediction of an MDD diagnosis. In addition, 19 studies on the prediction of treatment outcomes for MDD were reviewed. We summarized and discussed the results of previous studies. For future research results to be useful in clinical practice, ML enabling individual inferences is important. At the same time, there are important challenges to be addressed in the future.
... Existing studies have used measures derived from structural MRI and DTI, with depression classification accuracies ranging from 55% and up to and above 90% (reviews in Gao, Calhoun, & Sui, 2018;Kambeitz et al., 2017;Patel, Khalaf, & Aizenstein, 2016). Several studies with regional cortical thickness, surface area and volume measures reported cross-validation accuracies between 75 and 80% (Kipli & Kouzani, 2015;Qiu et al., 2014). Although white matter integrity measures have seen limited application, several studies have also reported accuracies close to 75% (Matsuoka et al., 2017;Schnyer, Clasen, Gonzalez, & Beevers, 2017). ...
... For each of the 10 data sets we performed either leave-one-out (LOOCV), 10-fold or fivefold cross-validation, depending on the size of the data set. Cross-validation was attempted separately with three classifier models, with different feature subdomains (e.g., all brain morphometric measures or only cortical thickness, surface area, volume or subcortical measures), with or without classifier hyperparameter optimisation, and with or without feature selection (e.g., Patel et al., 2015;Qiu et al., 2014;Schnyer et al., 2017;Yang et al., 2018). Where feasible, cross-validation was repeated multiple times with different fold partitions. ...
... Accuracies for all other classification attempts with brain morphometric measures were below 59% and can be found alongside scores for each approach in Table S14. (Kipli & Kouzani, 2015;Matsuoka et al., 2017;Qiu et al., 2014;Sacchet et al., 2015;Schnyer et al., 2017;Yang et al., 2018). MD features in the study appeared more discriminative of F I G U R E 1 Surface area regions consistently selected as decision tree cut features across cross-validation folds in cMDD-STR sample. ...
Major depressive disorder (MDD) has been the subject of many neuroimaging case–control classification studies. Although some studies report accuracies ≥80%, most have investigated relatively small samples of clinically‐ascertained, currently symptomatic cases, and did not attempt replication in larger samples. We here first aimed to replicate previously reported classification accuracies in a small, well‐phenotyped community‐based group of current MDD cases with clinical interview‐based diagnoses (from STratifying Resilience and Depression Longitudinally cohort, ‘STRADL’). We performed a set of exploratory predictive classification analyses with measures related to brain morphometry and white matter integrity. We applied three classifier types—SVM, penalised logistic regression or decision tree—either with or without optimisation, and with or without feature selection. We then determined whether similar accuracies could be replicated in a larger independent population‐based sample with self‐reported current depression (UK Biobank cohort). Additional analyses extended to lifetime MDD diagnoses—remitted MDD in STRADL, and lifetime‐experienced MDD in UK Biobank. The highest cross‐validation accuracy (75%) was achieved in the initial current MDD sample with a decision tree classifier and cortical surface area features. The most frequently selected decision tree split variables included surface areas of bilateral caudal anterior cingulate, left lingual gyrus, left superior frontal, right precentral and paracentral regions. High accuracy was not achieved in the larger samples with self‐reported current depression (53.73%), with remitted MDD (57.48%), or with lifetime‐experienced MDD (52.68–60.29%). Our results indicate that high predictive classification accuracies may not immediately translate to larger samples with broader criteria for depression, and may not be robust across different classification approaches.
... Furthermore, clinical evaluation of MDD relies on the counting of a minimum number of symptoms. This symptom-based approach may result in diagnostic inconsistencies, and the accurate detection of subtle clinical abnormalities in the early stage of the disorder requires medical staff skilled in highly specialized mental health services [13]. A more objective and reliable method, such as structural MRI, could be helpful for diagnosing MDD. ...
... Machine learning consists of a group of methods used to develop prediction models from empirical real-world data that make accurate predictions about unseen new data, and it is a powerful tool in a wide range of biomedical applications, owing to its ability to learn to categorize complex, highdimensional training data and apply the learned classification rules to unseen data [14]. When we consider the heterogeneous and complex clinical features of MDD, it is likely that its neuroanatomical differences would exhibit multiple volumetric and geometric features [13], which may be decipherable with machine learning approaches. An important challenge for adolescent neuroimaging is to identify brain measurements that are useful for predicting the onset of MDD. ...
... Based on these previous structural MRI findings, we can consider the possibility of using machine learning to predict MDD. Machine learning has been used to test the potential of structural MRI as a relevant biomarker of various forms of depression in several studies, including studies of adult patients with MDD [47]; medication-naive adult patients with first-episode depression [13]; pediatric patients with MDD [48]; and adolescent patients with first-episode depression [49], late-life depression [50], and treatment-refractory depression [51]. Wu et al. acquired structural T1-weighted scans from 25 pediatric unipolar depression patients and 26 demographically matched healthy controls. ...
It is hard to differentiate adolescent Major Depressive Disorder (MDD) patients from healthy adolescent controls based on structural MRI research findings, as the clinical characteristics of the patient group are heterogeneous, and the neuroimaging study results are ambiguous. We aimed to determine whether it is possible to reliably train a highly accurate predictive classification algorithm, even with the first onset of drug-naive adolescent MDD, solely using structural magnetic resonance imaging and without using any other clinical data from the patients. We also estimated the probability of the subject belonging to the predicted class to quantify the confidence of the prediction. Medication-naive adolescent patients in their first episode of MDD and healthy volunteers, matched for age, sex, and years of education, were prospectively recruited. Twenty-seven patients and 27 controls participated in the study. The two most significant variables were the standard deviations of intensity of the right ventral diencephalon and thickness of the superior segment of the circular sulcus of the insula. A participant is diagnosed as having MDD when the variation of either intensity in the right ventral diencephalon region or thickness of the superior segment of the circular sulcus of the insula increases. Structural brain changes can be used to build an accurate classification model for machine learning, even when the duration of illness is relatively short and the influence of MDD on the brain structure is minimal.
... Right middle frontal gyrus Guo et al., 2012 Left inferior frontal gyrus, orbital part Jin et al., 2011;Guo et al., 2012;Lord et al., 2012 Rolandic operculum Zhu et al., 2016 Right supplementary motor area Liu et al., 2012 Left superior frontal gyrus, medial Jin et al., 2011 Left median cingulate and paracingulate gyri Guo et al., 2012;Zhu et al., 2016 Right parahippocampal gyrus Qiu et al., 2014 Left lingual gyrus Lord et al., 2012;Qiu et al., 2014 Right paracentral lobule Qiu et al., 2014 Left putamen Lord et al., 2012;Gong et al., 2014 Left inferior temporal gyrus Gong et al., 2014 In the multi-level λ setting, it is important that how to get the optimizing combination of λ. If Enumeration method was adopted, the computation consumption was too huge. ...
... Right middle frontal gyrus Guo et al., 2012 Left inferior frontal gyrus, orbital part Jin et al., 2011;Guo et al., 2012;Lord et al., 2012 Rolandic operculum Zhu et al., 2016 Right supplementary motor area Liu et al., 2012 Left superior frontal gyrus, medial Jin et al., 2011 Left median cingulate and paracingulate gyri Guo et al., 2012;Zhu et al., 2016 Right parahippocampal gyrus Qiu et al., 2014 Left lingual gyrus Lord et al., 2012;Qiu et al., 2014 Right paracentral lobule Qiu et al., 2014 Left putamen Lord et al., 2012;Gong et al., 2014 Left inferior temporal gyrus Gong et al., 2014 In the multi-level λ setting, it is important that how to get the optimizing combination of λ. If Enumeration method was adopted, the computation consumption was too huge. ...
... Right middle frontal gyrus Guo et al., 2012 Left inferior frontal gyrus, orbital part Jin et al., 2011;Guo et al., 2012;Lord et al., 2012 Rolandic operculum Zhu et al., 2016 Right supplementary motor area Liu et al., 2012 Left superior frontal gyrus, medial Jin et al., 2011 Left median cingulate and paracingulate gyri Guo et al., 2012;Zhu et al., 2016 Right parahippocampal gyrus Qiu et al., 2014 Left lingual gyrus Lord et al., 2012;Qiu et al., 2014 Right paracentral lobule Qiu et al., 2014 Left putamen Lord et al., 2012;Gong et al., 2014 Left inferior temporal gyrus Gong et al., 2014 In the multi-level λ setting, it is important that how to get the optimizing combination of λ. If Enumeration method was adopted, the computation consumption was too huge. ...
Brain network analysis has been widely applied in neuroimaging studies. A hyper-network construction method was previously proposed to characterize the high-order relationships among multiple brain regions, where every edge is connected to more than two brain regions and can be represented by a hyper-graph. A brain functional hyper-network is constructed by a sparse linear regression model using resting-state functional magnetic resonance imaging (fMRI) time series, which in previous studies has been solved by the lasso method. Despite its successful application in many studies, the lasso method has some limitations, including an inability to explain the grouping effect. That is, using the lasso method may cause relevant brain regions be missed in selecting related regions. Ideally, a hyper-edge construction method should be able to select interacting brain regions as accurately as possible. To solve this problem, we took into account the grouping effect among brain regions and proposed two methods to improve the construction of the hyper-network: the elastic net and the group lasso. The three methods were applied to the construction of functional hyper-networks in depressed patients and normal controls. The results showed structural differences among the hyper-networks constructed by the three methods. The hyper-network structure obtained by the lasso was similar to that obtained by the elastic net method but very different from that obtained by the group lasso. The classification results indicated that the elastic net method achieved better classification results than the lasso method with the two proposed methods of hyper-network construction. The elastic net method can effectively solve the grouping effect and achieve better classification performance.
... Investigating wholebrain structural neuroanatomy as a diagnostic biomarker, Costafreda et al. [19] obtained a modest DA of 67.6% in the discrimination between MDD patients and HC using a SVM-based classifier. Qiu et al. [20] studied a group of drugnäive patients presenting a first-episode of MDD ( = 32) versus HC ( = 32) with a SVM classifier and different combinations of morphometric features. The authors reported overall modest classification accuracies ranging from 50% to 78% depending on the combination of features employed [20]. ...
... Qiu et al. [20] studied a group of drugnäive patients presenting a first-episode of MDD ( = 32) versus HC ( = 32) with a SVM classifier and different combinations of morphometric features. The authors reported overall modest classification accuracies ranging from 50% to 78% depending on the combination of features employed [20]. Only one morphometric MRI study has applied pattern classification techniques in BD, comparing two independent samples of patients with BD type I (BD-I) versus HC [21]. ...
... Congruently with our results, the few studies with structural MRI and neuroanatomical pattern classifiers in mood disorders published to date have achieved lower DA than fMRI studies [17][18][19][20][21]. Also, the literature on structural MRI investigations of BD has consistently shown a great variability of findings, including many negative studies and low reproducibility even across the different meta-analyses published so far [41][42][43]. ...
The presence of psychotic features in the course of a depressive disorder is known to increase the risk for bipolarity, but the early identification of such cases remains challenging in clinical practice. In the present study, we evaluated the diagnostic performance of a neuroanatomical pattern classification method in the discrimination between psychotic major depressive disorder (MDD), bipolar I disorder (BD-I), and healthy controls (HC) using a homogenous sample of patients at an early course of their illness. Twenty-three cases of first-episode psychotic mania (BD-I) and 19 individuals with a first episode of psychotic MDD whose diagnosis remained stable during 1 year of followup underwent 1.5 T MRI at baseline. A previously validated multivariate classifier based on support vector machine (SVM) was employed and measures of diagnostic performance were obtained for the discrimination between each diagnostic group and subsamples of age- and gender-matched controls recruited in the same neighborhood of the patients. Based on T1-weighted images only, the SVM-classifier afforded poor discrimination in all 3 pairwise comparisons: BD-I versus HC; MDD versus HC; and BD-I versus MDD. Thus, at the population level and using structural MRI only, we failed to achieve good discrimination between BD-I, psychotic MDD, and HC in this proof of concept study.
... The multiparameter morphometric neuroimaging approach using both volumetric features (cortical thickness) and geometric features (fractal dimensionality, gyrification, and sulcal depth) provides a holistic neurological assessment useful in elucidating the neurobiological correlates of sexual choking. The volumetric and geometric features reflect different aspects of biological underpinning; hence, they do not necessarily correlate with one another (Qiu et al., 2014). The cortical thickness measure informs changes in gray and white matter volume, whereas the fractal dimensionality reflects how the white matter surface fits space constraints and is used to investigate brain white matter surface complexity (L. ...
... Group comparison of cortical thickness, fractal dimensionality, gyrification, and sulcal depth was performed using CAT12 and analyzed via a non-parametric permutation technique. The threshold-free cluster enhancement (TFCE) was used in the permutation test, which gives cluster-based thresholding for familywise error correction (Qiu et al., 2014). As a result, the TFCE p-value images obtained were fully corrected for multiple comparisons across space. ...
Introduction
Being choked/strangled during partnered sex is an emerging sexual behavior, particularly prevalent among young adult women. Using a multiparameter morphometric imaging approach, we aimed to characterize neuroanatomical differences between young adult women (18–30 years old) who were exposed to frequent sexual choking and their choking naïve controls.
Methods
This cross‐sectional study consisted of two groups (choking [≥4 times in the past 30 days] vs. choking‐naïve group). Participants who reported being choked four or more times during sex in the past 30 days were enrolled in the choking group, whereas those without were assigned to the choking naïve group. High‐resolution anatomical magnetic resonance imaging (MRI) data were analyzed using both volumetric features (cortical thickness) and geometric features (fractal dimensionality, gyrification, sulcal depth).
Results
Forty‐one participants (choking n = 20; choking‐naïve n = 21) contributed to the final analysis. The choking group showed significantly increased cortical thickness across multiple regions (e.g., fusiform, lateral occipital, lingual gyri) compared to the choking‐naïve group. Widespread reductions of the gyrification were observed in the choking group as opposed to the choking‐naïve group. However, there was no group difference in sulcal depth. The fractal dimensionality showed bi‐directional results, where the choking group exhibited increased dimensionality in areas including the postcentral gyrus, insula, and fusiform, whereas decreased dimensionality was observed in the bilateral superior frontal gyrus and pericalcarine cortex.
Conclusion
These data in cortical morphology suggest that sexual choking events may be associated with neuroanatomical alteration. A longitudinal study with multimodal assessment is needed to better understand the temporal ordering of sexual choking and neurological outcomes.
... However, this symptom-based approach may result in diagnostic discrepancy among clinicians since MDD contains heterogeneous, multifaceted factors that have various biological and psychosocial etiologies. Furthermore, an early diagnosis with clinical signs requires professionals who have expertise in highly specialized mental health services such as psychiatrists and psychologists [4]. Therefore, using more objective and reliable diagnostic tools, such as neuroimaging, genetic, immunologic, and neuropsychological assessments, can improve identifying biomarkers. ...
... Support vector machine technique is a type of multivariate pattern analysis technique which learns by computers alone on how to classify complex, high-dimensional data sets and generalizes and categorizes data sets that have not yet been labeled. In general, supportive vector machine includes two stages; the first stage is for the computer system to learn and train well-classified data labeled by a human, while the second stage is to adopt new data sets with this learned system and then reclassify the data sets [4,8]. Apart from the support vector machine technique, other algorithms such as Gaussian process classifier, linear discriminant analysis, and decision tree areas are variously used as well. ...
Mood disorders include all types of depression and bipolar disorder, and mood disorders are sometimes called affective disorders. We will discuss newly developing two issues in affective disorders in children and adolescents. Those are the new diagnostic challenges using neuroimaging techniques in affective disorders and the introduction of disruptive mood dysregulation disorder (DMDD). During the 1980s, mental health professionals began to recognize symptoms of mood disorders in children and adolescents, as well as adults. However, children and adolescents do not necessarily have or exhibit the same symptoms as adults. It is more difficult to diagnose mood disorders in children, especially because children are not always able to express how they feel. Child mental health professionals believe that mood disorders in children and adolescents remain one of the most underdiagnosed mental health problems. We are currently trying to introduce the new diagnostic technique—machine learning in children and adolescents with MDD. We will discuss the current progress in the clinical application of machine learning for MDD. After that, we would also discuss a new challenging diagnosis—DMDD. We are still suffering from a lack of evidence when trying to treat the patients with DMDD. In addition, there are some debates about the diagnostic validity of DMDD. We will explain the current situation of DMDD studies and the future directions in the study of DMDD.
... We obtained CT by calculating the shortest distance from the gray/white boundary to the gr ay/cer ebr ospinal fluid boundary at each vertex. The SA of eac h hemispher e is calculated by adding up the ar ea of all tessellations on the gray matter surface (Deng et al., 2019 ;Qiu et al., 2014 ;Xiao et al., 2023 ). ...
Background
Major depressive disorder (MDD) has different clinical presentations in males and females. However, the neuroanatomical mechanisms underlying these sex differences are not fully understood.
Methods
High-resolution T1-weighted images were acquired from 61 patients with MDD and 61 healthy controls (36 females and 25 males, both). The sex differences in cortical thickness (CT) and surface area (SA) were obtained using the FreeSurfer software and compared between every two groups by post-hoc test. Spearman correlation analysis was also performed to explore the relationships between these regions and clinical characteristics.
Results
In male patients with MDD, the CT of the right precentral was thinner compared to female patients, although not survive Bonferroni correction. The SA of several regions, including right superior frontal, medial orbitofrontal gyrus, inferior frontal gyrus triangle, superior temporal, middle temporal, lateral occipital gyrus, and inferior parietal lobule in female patients with MDD was smaller than that in male patients (p < 0.01 after Bonferroni correction). In female patients, the SA of the right superior temporal (r = 0.438, p = 0.008), middle temporal (r = 0.340, p = 0.043), and lateral occipital gyrus (r = 0.372, p = 0.025) were positively correlated with illness duration.
Conclusion
The current study provides evidence of sex differences in CT and SA in patients with MDD, which may improve our understanding of the sex-specific neuroanatomical changes in the development of MDD.
... This is not surprising, given that the weakly connected state has been closely linked to depression [56,57,72], and the intergroup differences between SA and NS in the present study also pointed to the weakly connected state. Our model, constructed with the weakly connected state for distinguishing MDD patients from HCs, is comparable with previous classification models constructed with cerebral functional features [73,74] and is superior to models constructed with structural features [75,76]. Moreover, our models also had powerful efficiency in stratifying patients with different suicidal risk levels, which supplements previous findings on using structural features to stratify MDD patients with diverse suicide risk [33,77]. ...
Major depressive disorder (MDD) is a severe brain disease associated with a significant risk of suicide. Identification of suicidality is sometimes life-saving for MDD patients. We aimed to explore the use of dynamic functional network connectivity (dFNC) for suicidality detection in MDD patients. A total of 173 MDD patients, including 48 without suicide risk (NS), 74 with suicide ideation (SI), and 51 having attempted suicide (SA), participated in the present study. Thirty-eight healthy controls were also recruited for comparison. A sliding window approach was used to derive the dFNC, and the K-means clustering method was used to cluster the windowed dFNC. A linear support vector machine was used for classification, and leave-one-out cross-validation was performed for validation. Other machine learning methods were also used for comparison. MDD patients had widespread hypoconnectivity in both the strongly connected states (states 2 and 5) and the weakly connected state (state 4), while the dysfunctional connectivity within the weakly connected state (state 4) was mainly driven by suicidal attempts. Furthermore, dFNC matrices, especially the weakly connected state, could be used to distinguish MDD from healthy controls (area under curve [AUC] = 82), and even to identify suicidality in MDD patients (AUC = 78 for NS vs. SI, AUC = 88 for NS vs. SA, and AUC = 74 for SA vs. SI), with vision-related and default-related inter-network connectivity serving as important features. Thus, the dFNC abnormalities observed in this study might further improve our understanding of the neural substrates of suicidality in MDD patients.
... Morphometric and CTA analyses in MDD have demonstrated atrophy in the prefrontal and limbic regions, more specifically in the bilateral superior frontal gyri, inferior frontal gyri, middle temporal gyri, parahippocampal gyri, and hippocampus (Schmaal et al., 2016;Zhao et al., 2014). Interestingly, structural markers can discriminate between healthy individuals and those with MDD (Qiu et al., 2014a), between a first episode of disease and chronic symptoms (Qiu et al., 2014b), and between ongoing MDD and remitted patients (Salvadore et al., 2011). White matter is affected is well, with lower FA values measured in many tracts, including the anterior limb of the left internal capsule (Jia et al., 2014). ...
Psychiatry remains the only medical specialty where diagnoses are still based on clinical syndromes rather than measurable biological abnormalities. As imaging technology and analytical methods evolve, it is becoming clear that subtle but measurable radiological characteristics exist and can be used to experimentally classify psychiatric disorders, predict response to treatment and, hopefully, develop new, more effective therapies. This review highlights advances in neuroimaging modalities that are now allowing assessment of brain structure, connectivity and neural network function, describes technical aspects of the most promising methods, and summarizes observations made in some frequent psychiatric disorders.
... Although MDD-related brain differences were found in several literatures, these studies usually reported small to very small effect sizes. Previous machine learning (ML) studies with structural brain features also show potential for unbiased diagnostic classification (Lebedeva et al., 2017;Patel et al., 2015;Qiu et al., 2014). Features derived from structural magnetic resonance imaging (MRI) have shown promise for MDD case-control classification, with linear or non-linear supported vector machine classifiers achieving accuracies of >70% (Gao et al., 2018). ...
There is increasing interest in using data-driven unsupervised methods to identify structural underpinnings of common mental illnesses, including Major Depressive Disorder (MDD) and associated traits such as cognition. However, studies are often limited to severe clinical cases with small sample sizes and most do not include replication.
Here, we examine two relatively large samples with structural magnetic resonance imaging (MRI), measures of lifetime MDD and cognitive variables: Generation Scotland (GS subsample, N = 980), and UK Biobank (UKB, N = 8900); for discovery and replication, using an exploratory approach. Regional measures of FreeSurfer derived cortical thickness (CT), cortical surface area (CSA), cortical volume (CV) and subcortical volume (subCV) were input into a clustering process, controlling for common covariates. The main analysis steps involved constructing participant K-nearest neighbour graphs, and graph partitioning with Markov Stability to determine optimal clustering of participants. Resultant clusters were i) checked whether they were replicated in an independent cohort, and ii) tested for associations with depression status, and cognitive measures.
Participants separated into two clusters based on structural brain measurements in GS subsample, with large Cohen’s d effect sizes between clusters in higher order cortical regions, commonly associated with executive function and decision making. Clustering was replicated in the UKB sample, with high correlations of cluster effect sizes for CT, CSA, CV and subCV between cohorts across regions. The identified clusters were not significantly different with respect to MDD case-control status in either cohort (GS subsample: pFDR = 0.2239 − 0.6585; UKB: pFDR = 0.2003 − 0.7690). Significant differences in general cognitive ability were, however, found between the clusters for both datasets; for CSA, CV, subCV (GS subsample: d = 0.2529 − 0.3490, pFDR < 0.005; UKB: d = 0.0868 − 0.1070, pFDR < 0.005).
Our results suggest that there are replicable natural groupings of participants based on cortical and subcortical brain measures, which may be related to differences in cognitive performance, but not to the MDD case-control status.
... Unlike previous findings on adults, only cortical thickness could provide the best accuracy in three adolescent classification models in our study. This is consistent with the findings by Qiu et al., who used an SVM based on various brain morphometric features to distinguish between 32 adult patients with first-episode MDD and 32 HC [61]. They reported that multiple cortical features could discriminate them with cortical thickness providing the highest accuracy. ...
Background
Early diagnosis of adolescent psychiatric disorder is crucial for early intervention. However, there is extensive comorbidity between affective and psychotic disorders, which increases the difficulty of precise diagnoses among adolescents.
Methods
We obtained structural magnetic resonance imaging scans from 150 adolescents, including 67 and 47 patients with major depressive disorder (MDD) and schizophrenia (SCZ), as well as 34 healthy controls (HC) to explore whether psychiatric disorders could be identified using a machine learning technique. Specifically, we used the support vector machine and the leave-one-out cross-validation method to distinguish among adolescents with MDD and SCZ and healthy controls.
Results
We found that cortical thickness was a classification feature of a) MDD and HC with 79.21% accuracy where the temporal pole had the highest weight; b) SCZ and HC with 69.88% accuracy where the left superior temporal sulcus had the highest weight. Notably, adolescents with MDD and SCZ could be classified with 62.93% accuracy where the right pars triangularis had the highest weight.
Conclusions
Our findings suggest that cortical thickness may be a critical biological feature in the diagnosis of adolescent psychiatric disorders. These findings might be helpful to establish an early prediction model for adolescents to better diagnose psychiatric disorders.
... The SVM analysis achieved almost as good performance as the RVM analysis. Similarly, Qiu et al. [33] reported that firstepisode, medication-naïve MDD patients and healthy ...
Major depressive disorder (MDD) causes great decrements in health and quality of life with increments in healthcare costs, but the causes and pathogenesis of depression remain largely unknown, which greatly prevent its early detection and effective treatment. With the advancement of neuroimaging approaches, numerous functional and structural alterations in the brain have been detected in MDD and more recently attempts have been made to apply these findings to clinical practice. In this review, we provide an updated summary of the progress in translational application of psychoradiological findings in MDD with a specified focus on potential clinical usage. The foreseeable clinical applications for different MRI modalities were introduced according to their role in disorder classification, subtyping, and prediction. While evidence of cerebral structural and functional changes associated with MDD classification and subtyping was heterogeneous and/or sparse, the ACC and hippocampus have been consistently suggested to be important biomarkers in predicting treatment selection and treatment response. These findings underlined the potential utility of brain biomarkers for clinical practice.
... We identified a downgraded and contracted connectome hierarchy in MDD, indicated by a lessexplained variance in the functional connectome and a narrower distribution range of gradient scores. Disturbances in the architecture of the macroscale functional brain network have recently been considered critical in the pathology of depression 21,38,39 . Notably, although the changes in the topology of the functional connectomes in MDD remain inconsistent in direction across the literature, likely as a result of differences in patient inclusion criteria and methodological choices 40 , many studies have demonstrated a replicable pattern towards a randomized configuration of the functional brain networks in patients with MDD 23,40,41 . ...
Background
Patients with major depressive disorder (MDD) exhibit concurrent deficits in sensory processing and high-order cognitive functions such as self-awareness and rumination. Connectome mapping studies have suggested a principal primary-to-transmodal gradient in functional brain networks, supporting the spectrum from sensation to cognition. However, whether this principal connectome gradient is disrupted in patients with MDD and how this disruption is associated with gene expression profiles remain unclear.
Methods
Using a large cohort of resting-state functional magnetic resonance imaging data from 2,234 participants (1,150 patients with MDD and 1,084 healthy controls) recruited at 10 sites, we investigated MDD-related alterations in the principal connectome gradient. We further used Neurosynth and postmortem gene expression data to assess the cognitive functions and transcriptional profiles related to the gradient alterations in MDD, respectively.
Results
Relative to controls, patients with MDD exhibited abnormal global topography of the principal primary-to-transmodal gradient, as indicated by reduced explanation ratio, gradient range, and gradient variation (Cohen’s d = −0.16∼-0.21). Focal alterations of gradient scores were mostly in the primary systems involved in sensory processing and in the transmodal systems implicated in high-order cognition. The transcriptional profiles explained 53.9% of the spatial variance in the altered gradient patterns, with the most correlated genes enriched in transsynaptic signaling and calcium ion binding.
Conclusions
These results highlight the dysfunction of the core connectome hierarchy in MDD and its linkage with gene expression profiles, providing insights into the neurobiological and molecular genetic underpinnings of sensory-cognitive deficits in this disorder.
... Qiu et al. recruited medication-naive adult patients with first-episode depression and healthy controls and tried to characterize MDD using a multiparametric classification approach based on high-resolution structural images. They reported that regions including the lateral occipital gyrus contributed to the identification of patients with MDD in the cortical thickness discrimination map [22]. In another study using different research methods such as regional cerebral blood flow single photon emission computed tomography on drug-naïve children with a diagnosis of MDD, occipital brain perfusion deficits were observed [23]. ...
Major depressive disorder is a major mental disorder affecting adolescents. Cortical thickness provides a sensitive measure of age-associated changes. Previous studies using cortical thickness analysis reported inconsistent results on brain structural changes in adolescent major depressive disorder. The neuroanatomical substrates of major depressive disorder in adolescents are not fully understood. We aimed to compare the anatomical structures of the brain in first-onset drug-naïve adolescents with major depressive disorder to normal controls. Twenty-seven first-episode drug-naïve adolescents with major depressive disorder and an equal number of age-matched control subjects were scanned on a 3T MRI scanner. Comparisons between those two groups were performed using surface-based morphometry analysis for cortical thickness and volumetric analysis of subcortical gray matter. The correlations between morphometric indexes and clinical measures (Hamilton depression rating scale score or children's depression inventory score) were also calculated. We found that the cortical area is thinner in major depressive disorder patients than in controls, specifically in the left occipital area (precuneus and cuneus, cluster-level family-wise corrected P < 0.05). The hippocampus volume was also smaller in major depressive disorder patients than in the control group. No significant correlations were found between morphometric indexes (average cortical thickness extracted from the left precuneus cluster and hippocampal volume) and clinical measures. The left occipital cortical regions may have a role in the pathophysiology of adolescent major depressive disorder, and the involvement of the hippocampus is important for pathogenic changes even in the early stages of major depressive disorder.
... It has revealed patterns of brain abnormalities that differentiate patient groups, but to date it has limited clinical translation particularly for single patients. 9 This method had been applied to both structural or functional imaging in a number of psychiatric disorders including schizophrenia, 10 depression, 11 and obsessive compulsive disorder (OCD). 12,13 In recent years, more advanced algorithms such as deep learning (DL) have been increasingly used to investigate the neuroimaging features of psychiatric and neurological disorders. ...
Psychoradiology is an emerging field that applies radiological imaging technologies to psychiatric conditions. In the past three decades, brain imaging techniques have rapidly advanced understanding of illness and treatment effects in psychiatry. Based on these advances, radiologists have become increasingly interested in applying these advances for differential diagnosis and individualized patient care selection for common psychiatric illnesses. This shift from research to clinical practice represents the beginning evolution of psychoradiology. In this review, we provide a summary of recent progress relevant to this field based on their clinical functions, namely the (1) classification and subtyping; (2) prediction and monitoring of treatment outcomes; and (3) treatment selection. In addition, we provide guidelines for the practice of psychoradiology in clinical settings and suggestions for future research to validate broader clinical applications. Given the high prevalence of psychiatric disorders and the importance of increased participation of radiologists in this field, a guide regarding advances in this field and a description of relevant clinical work flow patterns help radiologists contribute to this fast-evolving field.
... On the other hand, the positive correlation between supramarginal cortical thickness and HRSD scores suggests that one should be cautious when making the assumption that reductions in thickness will equate to reductions in depressive symptoms. Whereas a recent study reported that left supramarginal cortical thickness was increased in drug-naïve, first-episode MDD patients as compared to healthy controls 41 , there was no difference between MDD patients and healthy controls in our study. This raises the question on whether the positive correlation between left supramarginal cortical thickness and HRSD scores had clinical impact on MDD patients. ...
Recent studies have reported that methylation of the brain-derived neurotrophic factor (BDNF) gene promoter is associated with major depressive disorder (MDD). This study aimed to investigate the association between cortical thickness and methylation of BDNF promoters as well as serum BDNF levels in MDD. The participants consisted of 65 patients with recurrent MDD and 65 age- and gender-matched healthy controls. Methylation of BDNF promoters and cortical thickness were compared between the groups. The right medial orbitofrontal, right lingual, right lateral occipital, left lateral orbitofrontal, left pars triangularis, and left lingual cortices were thinner in patients with MDD than in healthy controls. Among the MDD group, right pericalcarine, right medical orbitofrontal, right rostral middle frontal, right postcentral, right inferior temporal, right cuneus, right precuneus, left frontal pole, left superior frontal, left superior temporal, left rostral middle frontal and left lingual cortices had inverse correlations with methylation of BDNF promoters. Higher levels of BDNF promoter methylation may be closely associated with the reduced cortical thickness among patients with MDD. Serum BDNF levels were significantly lower in MDD, and showed an inverse relationship with BDNF methylation only in healthy controls. Particularly the prefrontal and occipital cortices seem to indicate key regions in which BDNF methylation has a significant effect on structure.
... Cortical structures may be the primary components of this circuit by regulating emotional and cognitive processes (e.g., prefrontal cortex and cingulate cortex [32,33]). Although some reports have summarized structural changes in MDD (e.g., frontal cortex [2], parietal cortex [34,35], and cingulate cortex [16][17][18]) the distinct alterations of cortical structures need to be thoroughly studied in MDD. In the present study, we hypothesized that MDD might have atypical cortical patterns involving cortical thickness, surface area or cortical folding, potentially located in the prefrontal, parietal and cingulate regions. ...
Major depressive disorder (MDD) is accompanied by atypical brain structure. This study first presents the alterations in the cortical surface of patients with MDD using multidimensional structural patterns that reflect different neurodevelopment. Sixteen first-episode, untreated patients with MDD and 16 matched healthy controls underwent a magnetic resonance imaging (MRI) scan. The cortical maps of thickness, surface area, and gyrification were examined using the surface-based morphometry (SBM) approach. Increase of cortical thickness was observed in the right posterior cingulate region and the parietal cortex involving the bilateral inferior, left superior parietal and right paracentral regions, while decreased thickness was noted in the parietal cortex including bilateral pars opercularis and left precentral region, as well as the left rostral-middle frontal regions in patients with MDD. Likewise, increased or decreased surface area was found in five sub-regions of the cingulate gyrus, parietal and frontal cortices (e.g., bilateral inferior parietal and superior frontal regions). In addition, MDD patients exhibited a significant hypergyrification in the right precentral and supramarginal region. This integrated structural assessment of cortical surface suggests that MDD patients have cortical alterations of the frontal, parietal and cingulate regions, indicating a vulnerability to MDD during earlier neurodevelopmental process.
... Cortical structures may be the primary components of this circuit by regulating emotional and cognitive processes (e.g., prefrontal cortex and cingulate cortex [32,33]). Although some reports have summarized structural changes in MDD (e.g., frontal cortex [2], parietal cortex [34,35], and cingulate cortex [16][17][18]) the distinct alterations of cortical structures need to be thoroughly studied in MDD. In the present study, we hypothesized that MDD might have atypical cortical patterns involving cortical thickness, surface area or cortical folding, potentially located in the prefrontal, parietal and cingulate regions. ...
Major depressive disorder (MDD) is accompanied by atypical brain structure. This study first presents the alterations in the cortical surface of patients with MDD using multidimensional structural patterns that reflect different neurodevelopment. Sixteen first-episode, untreated patients with MDD and 16 matched healthy controls underwent a magnetic resonance imaging (MRI) scan. The cortical maps of thickness, surface area, and gyrification were examined using the surface-based morphometry (SBM) approach. Increase of cortical thickness was observed in the right posterior cingulate region and the parietal cortex involving the bilateral inferior, left superior parietal and right paracentral regions, while decreased thickness was noted in the parietal cortex including bilateral pars opercularis and left precentral region, as well as the left rostral-middle frontal regions in patients with MDD. Likewise, increased or decreased surface area was found in five sub-regions of the cingulate gyrus, parietal and frontal cortices (e.g., bilateral inferior parietal and superior frontal regions). In addition, MDD patients exhibited a significant hypergyrification in the right precentral and supramarginal region. This integrated structural assessment of cortical surface suggests that MDD patients have cortical alterations of the frontal, parietal and cingulate regions, indicating a vulnerability to MDD during earlier neurodevelopmental process.
... This analysis revealed significant region group interaction (Wald 2 15.25, p 0.035), indicating that the differences in classification indeed differed across prefrontal cortical regions (Fig. 3C,D). To gain additional insight into the topographical layout of the regions that successfully classified subjects who were more or less reward dependent, we generated discrimination maps (Mourão-Miranda et al., 2012; Qiu et al., 2014) for left (Fig. 4A) and right (Fig. 4B) superior frontal gyri. These maps display the spatial distribution of the weight vectors used in the classification of reward dependence, or in other words the weight of each voxel in discriminating between subjects who were more or less reward dependent . ...
Interindividual differences in the effects of reward on performance are prevalent and poorly understood, with some individuals being more dependent than others on the rewarding outcomes of their actions. The origin of this variability in reward dependence is unknown. Here, we tested the relationship between reward dependence and brain structure in healthy humans. Subjects trained on a visuomotor skill-acquisition task and received performance feedback in the presence or absence of reward. Reward dependence was defined as the statistical trial-by-trial relation between reward and subsequent performance. We report a significant relationship between reward dependence and the lateral prefrontal cortex, where regional gray-matter volume predicted reward dependence but not feedback alone. Multivoxel pattern analysis confirmed the anatomical specificity of this relationship. These results identified a likely anatomical marker for the prospective influence of reward on performance, which may be of relevance in neurorehabilitative settings.
... Evidence from neuroimaging studies also suggests that chronic pain and stress-related psychiatric disorders are associated with structural and functional reorganisation of cortical structures (Burgmer et al., 2009;Flor et al., 2001;Karl et al., 2001;Pleger et al., 2006;Hayes et al., 2012;Asami et al., 2008;Qiu et al., 2013). There is evidence for the involvement of S1 in the modulation of the sensory aspects of pain perception (Bushnell et al., 1999). ...
... Similarly, Costafreda et al. (2009a) accurately identified 71% of MDD patients, before treatment, that responded fully to cognitive behavioral therapy (CBT) from whole-brain patterns of brain activity induced once more by a sad facial processing task. Brain structure, including gray and white matter measures, has also been found to be highly predictive of MDD (Costafreda et al., 2009b;Gong et al., 2011;Mwangi et al., 2012;Qiu et al., 2013). ...
Pattern recognition applied to whole-brain neuroimaging data, such as functional Magnetic Resonance Imaging (fMRI), has been successful at discriminating psychiatric patients from healthy subjects. However, predictive patterns obtained from whole-brain voxel-based features are difficult to interpret in terms of the underlying neurobiology. As is generally accepted, many psychiatric disorders, such as depression and schizophrenia, are brain connectivity disorders. Therefore, pattern recognition based on network models should provide more scientific insight and potentially more powerful predictions than voxel-based approaches. Here, we build a sparse network-based discriminative modelling framework, based on Gaussian graphical models and L1-norm regularised linear Support Vector Machines (SVM). The proposed framework provides easier pattern interpretation in terms of underlying network changes between groups, and we illustrate our technique by classifying patients with depression and controls, using fMRI data from a sad facial processing task.
Machine learning (ML) techniques have gained popularity in the neuroimaging field due to their potential for classifying neuropsychiatric disorders. However, the diagnostic predictive power of the existing algorithms has been limited by small sample sizes, lack of representativeness, data leakage, and/or overfitting. Here, we overcome these limitations with the largest multi-site sample size to date (N = 5365) to provide a generalizable ML classification benchmark of major depressive disorder (MDD) using shallow linear and non-linear models. Leveraging brain measures from standardized ENIGMA analysis pipelines in FreeSurfer, we were able to classify MDD versus healthy controls (HC) with a balanced accuracy of around 62%. But after harmonizing the data, e.g., using ComBat, the balanced accuracy dropped to approximately 52%. Accuracy results close to random chance levels were also observed in stratified groups according to age of onset, antidepressant use, number of episodes and sex. Future studies incorporating higher dimensional brain imaging/phenotype features, and/or using more advanced machine and deep learning methods may yield more encouraging prospects.
Background
Research from the past has shown that the human immunodeficiency virus (HIV) can quickly enter the central nervous system after seroconversion, and that roughly 50% of HIV patients may experience neurological problems. Application of combined antiretroviral therapy (cART) can systemically inhibit viral replication, partially restoring immune functions, but it is unable to entirely eradicate viral proteins in the brain. The influence of HIV on brain functioning and behavioral symptoms is still completely unknown, despite extensive research into the functional and anatomical abnormalities in the brainof HIV patients.
Methods
We gathered resting-state functional MRI data from 77 individuals (42 HIV patients (with behavioral data) and 35 healthy controls) from Beijing YouAn Hospital, Capital Medical University. We identified a constrained primary-to-transmodal gradient and an extended sensorimotor-to-visual gradient using functional connectome gradient analysis.
Results
According to group comparison analysis, the HIV patients had higher sensorimotor-to-visual and sensorimotor-to-visual spatial variation in the posterior cingulate cortex and a lower gradient score of primary-to-transmodal in the middle frontal gyrus. These two abnormal functional gradients of HIV patients were related to individual decreased abstract/executive processing abilities (planning, reasoning, set switching, flexible thinking, and updating, etc.) and clinical symptoms (CD4), as well as topological efficiency of brain functional network.
Conclusion
When taken as a whole, our findings describe the failure of the brain's functional hierarchical architecture in HIV patients, offering a novel perspective on the neurological mechanisms driving the virus.
Background
The development of machine learning models for aiding in the diagnosis of mental disorder is recognized as a significant breakthrough in the field of psychiatry. However, clinical practice of such models remains a challenge, with poor generalizability being a major limitation.
Methods
Here, we conducted a pre-registered meta-research assessment on neuroimaging-based models in the psychiatric literature, quantitatively examining global and regional sampling issues over recent decades, from a view that has been relatively underexplored. A total of 476 studies ( n = 118,137) were included in the current assessment. Based on these findings, we built a comprehensive 5-star rating system to quantitatively evaluate the quality of existing machine learning models for psychiatric diagnoses.
Results
A global sampling inequality in these models was revealed quantitatively (sampling Gini coefficient ( G ) = 0.81, p < .01), varying across different countries (regions) (e.g., China, G = 0.47; the USA, G = 0.58; Germany, G = 0.78; the UK, G = 0.87). Furthermore, the severity of this sampling inequality was significantly predicted by national economic levels ( β = − 2.75, p < .001, R 2 adj = 0.40; r = − .84, 95% CI: − .41 to − .97), and was plausibly predictable for model performance, with higher sampling inequality for reporting higher classification accuracy. Further analyses showed that lack of independent testing (84.24% of models, 95% CI: 81.0–87.5%), improper cross-validation (51.68% of models, 95% CI: 47.2–56.2%), and poor technical transparency (87.8% of models, 95% CI: 84.9–90.8%)/availability (80.88% of models, 95% CI: 77.3–84.4%) are prevailing in current diagnostic classifiers despite improvements over time. Relating to these observations, model performances were found decreased in studies with independent cross-country sampling validations (all p < .001, BF 10 > 15). In light of this, we proposed a purpose-built quantitative assessment checklist, which demonstrated that the overall ratings of these models increased by publication year but were negatively associated with model performance.
Conclusions
Together, improving sampling economic equality and hence the quality of machine learning models may be a crucial facet to plausibly translating neuroimaging-based diagnostic classifiers into clinical practice.
Importance:
Neuroimaging-based artificial intelligence (AI) diagnostic models have proliferated in psychiatry. However, their clinical applicability and reporting quality (ie, feasibility) for clinical practice have not been systematically evaluated.
Objective:
To systematically assess the risk of bias (ROB) and reporting quality of neuroimaging-based AI models for psychiatric diagnosis.
Evidence review:
PubMed was searched for peer-reviewed, full-length articles published between January 1, 1990, and March 16, 2022. Studies aimed at developing or validating neuroimaging-based AI models for clinical diagnosis of psychiatric disorders were included. Reference lists were further searched for suitable original studies. Data extraction followed the CHARMS (Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A closed-loop cross-sequential design was used for quality control. The PROBAST (Prediction Model Risk of Bias Assessment Tool) and modified CLEAR (Checklist for Evaluation of Image-Based Artificial Intelligence Reports) benchmarks were used to systematically evaluate ROB and reporting quality.
Findings:
A total of 517 studies presenting 555 AI models were included and evaluated. Of these models, 461 (83.1%; 95% CI, 80.0%-86.2%) were rated as having a high overall ROB based on the PROBAST. The ROB was particular high in the analysis domain, including inadequate sample size (398 of 555 models [71.7%; 95% CI, 68.0%-75.6%]), poor model performance examination (with 100% of models lacking calibration examination), and lack of handling data complexity (550 of 555 models [99.1%; 95% CI, 98.3%-99.9%]). None of the AI models was perceived to be applicable to clinical practices. Overall reporting completeness (ie, number of reported items/number of total items) for the AI models was 61.2% (95% CI, 60.6%-61.8%), and the completeness was poorest for the technical assessment domain with 39.9% (95% CI, 38.8%-41.1%).
Conclusions and relevance:
This systematic review found that the clinical applicability and feasibility of neuroimaging-based AI models for psychiatric diagnosis were challenged by a high ROB and poor reporting quality. Particularly in the analysis domain, ROB in AI diagnostic models should be addressed before clinical application.
Background:
Previous studies have found qualitative structural and functional brain changes in major depressive disorder (MDD) patients. However, most studies ignored the complementarity of multisequence MRI neuroimaging features and cannot determine accurate biomarkers.
Purpose:
To evaluate machine-learning models combined with multisequence MRI neuroimaging features to diagnose patients with MDD.
Study type:
Prospective.
Subjects:
A training cohort including 111 patients and 90 healthy controls (HCs) and a test cohort including 28 patients and 22 HCs.
Field strength/sequence:
A 3.0 T/T1-weighted imaging, resting-state functional MRI with echo-planar sequence, and single-shot echo-planar diffusion tensor imaging.
Assessment:
Recruitment and integration were used to reflect the dynamic changes of functional networks, while gray matter volume and fractional anisotropy were used to reflect the changes in the morphological and anatomical network. We then fused features with significant differences in functional, morphological, and anatomical networks to evaluate a random forest (RF) classifier to diagnose patients with MDD. Furthermore, a support vector machine (SVM) classifier was used to verify the stability of neuroimaging features. Linear regression analyses were conducted to investigate the relationships among multisequence neuroimaging features and the suicide risk of patients.
Statistical tests:
The comparison of functional network attributes between patients and controls by two-sample t-test. Network-based statistical analysis was used to identify structural and anatomical connectivity changes between MDD and HCs. The performance of the model was evaluated by receiver operating characteristic (ROC) curves.
Results:
The performance of the RF model integrating multisequence neuroimaging features in the diagnosis of depression was significantly improved, with an AUC of 93.6%. In addition, we found that multisequence neuroimaging features could accurately predict suicide risk in patients with MDD (r = 0.691).
Data conclusion:
The RF model fusing functional, morphological, and anatomical network features performed well in diagnosing patients with MDD and provided important insights into the pathological mechanisms of MDD.
Evidence level:
1.
Technical efficacy:
Stage 2.
Background:
In this modern era, depression is one of the most prevalent mental disorder from which millions of individuals are affected today. The symptoms of depression are heterogeneous and often coincide with other disorders such as bipolar disorder, Parkinson's, schizophrenia, etc. It is a serious mental illness that may lead to other health problems if left untreated. Currently, identifying individuals with depression is totally based on the expertise of the clinician's experience. In order to assist clinicians in identifying the characteristics and classifying depressed people, different types of data modalities and machine learning techniques have been incorporated by researchers in this field. This study aims to find the answers of some important questions related to the trend of publications, data modality, machine learning models, dataset usage, pre-processing techniques and feature extraction and selection techniques that are prevalent and guide the direction of future research on depression diagnosis.
Methods:
This systematic review was conducted using a broad range of articles from two major databases: IEEE Xplore and PubMed. Studies ranging from the years 2011 to April 2021 were retrieved from the databases resulting in a total of 590 articles (53 articles from IEEE Xplore database and 537 articles from PubMed database). Out of those, the articles which satisfied the defined inclusion criteria were investigated for further analysis.
Results:
A total of 135 articles were identified and analysed for this review. A high growth in the number of publications has been observed in recent years. Furthermore, a significant diversity in the use of data modalities and machine learning classifiers has also been noted in this study. fMRI data with SVM classifier was found to be the most popular choice among researchers. In most of the studies, data scarcity and small sample size, particularly for neuroimaging data are major concerns. The use of identical data pre-processing tools for similar data modality can be seen. This study also provides statistical analysis of the current framework with respect to the modality, machine learning classifier, sample size and accuracy by applying one-way ANOVA and the Tukey - Kramer test.
Conclusion:
The results indicate that an effective fusion of machine learning techniques with a potential data modality has a promising future for assisting clinicians in automatic depression diagnosis.
Background:
Postpartum depression (PPD) is a serious postpartum mental health problem worldwide. However, the cortical structural alterations in patients with PPD remain unclear. This study investigated the cortical structural alterations of PPD patients through multidimensional structural patterns and their potential correlations with clinical severity.
Methods:
High-resolution 3D T1 structural images were acquired from 21 drug-naive patients with PPD and 18 healthy postpartum women matched for age, educational level, and body mass index. The severity of PPD was assessed by using the Hamilton Depression Scale (HAMD) and Edinburgh Postnatal Depression Scale (EPDS) scores. Cortical morphological parameters including cortical thickness, surface area, and mean curvature were calculated using the surface-based morphometric (SBM) method. General linear model (GLM) analyses were performed to evaluate the relationship of cortical morphological parameters with clinical scales.
Results:
In the present study, PPD patients showed a thinner cortical thickness in the right inferior parietal lobule compared with the healthy controls. Increased surface area was observed in the left superior frontal gyrus, caudal middle frontal gyrus, middle temporal gyrus, insula, and right supramarginal cortex in PPD patients. Likewise, PPD patients exhibited a higher mean curvature in the left superior and right inferior parietal lobule. Furthermore, increased cortical surface area in the left insula had a positive correlation with EPDS scores, and higher mean curvature in the left superior parietal lobule was negatively correlated with EPDS scores.
Limitations:
First, SBM cannot reflect the changes of subcortical structures that are considered to play a role in the development of PPD. Second, the sample size of this study is small. These positive results should be interpreted with caution. Third, this cross-sectional study does not involve a comparison of structural MRI before and after pregnancy.
Conclusions:
The complex cortical structural alterations of patients with PPD mainly involved the prefrontal and parietal regions. The morphometric alterations in these specific regions may provide promising markers for assessing the severity of PPD.
This study aimed to identify biomarkers of major depressive disorder (MDD), by relating neuroimage‐derived measures to binary (MDD/control), ordinal (severe MDD/mild MDD/control), or continuous (depression severity) outcomes. To address MDD heterogeneity, factors (severity of psychic depression, motivation, anxiety, psychosis, and sleep disturbance) were also used as outcomes. A multisite, multimodal imaging (diffusion MRI [dMRI] and structural MRI [sMRI]) cohort (52 controls and 147 MDD patients) and several modeling techniques—penalized logistic regression, random forest, and support vector machine (SVM)—were used. An additional cohort (25 controls and 83 MDD patients) was used for validation. The optimally performing classifier (SVM) had a 26.0% misclassification rate (binary), 52.2 ± 1.69% accuracy (ordinal) and r = .36 correlation coefficient (p < .001, continuous). Using SVM, R² values for prediction of any MDD factors were <10%. Binary classification in the external data set resulted in 87.95% sensitivity and 32.00% specificity. Though observed classification rates are too low for clinical utility, four image‐based features contributed to accuracy across all models and analyses—two dMRI‐based measures (average fractional anisotropy in the right cuneus and left insula) and two sMRI‐based measures (asymmetry in the volume of the pars triangularis and the cerebellum) and may serve as a priori regions for future analyses. The poor accuracy of classification and predictive results found here reflects current equivocal findings and sheds light on challenges of using these modalities for MDD biomarker identification. Further, this study suggests a paradigm (e.g., multiple classifier evaluation with external validation) for future studies to avoid nongeneralizable results.
Volumetric and morphometric studies have demonstrated structural abnormalities related to chronic epilepsies on a cohort- and population-based level. On a single-patient level, specific patterns of atrophy or cortical reorganization may be widespread and heterogeneous but represent potential targets for further personalized image analysis and surgical therapy. The goal of this study was to compare morphometric data analysis in 37 patients with temporal lobe epilepsies with expert-based image analysis, pre-informed by seizure semiology and ictal scalp EEG. Automated image analysis identified abnormalities exceeding expert-determined structural epileptogenic lesions in 86% of datasets. If EEG lateralization and expert MRI readings were congruent, automated analysis detected abnormalities consistent on a lobar and hemispheric level in 82% of datasets. However, in 25% of patients EEG lateralization and expert readings were inconsistent. Automated analysis localized to the site of resection in 60% of datasets in patients who underwent successful epilepsy surgery. Morphometric abnormalities beyond the mesiotemporal structures contributed to subtype characterisation. We conclude that subject-specific morphometric information is in agreement with expert image analysis and scalp EEG in the majority of cases. However, automated image analysis may provide non-invasive additional information in cases with equivocal radiological and neurophysiological findings.
Unlike neurologic conditions, such as brain tumors, dementia, and stroke, the neural mechanisms for all psychiatric disorders remain unclear. A large body of research obtained with structural and functional magnetic resonance imaging, positron emission tomography/single photon emission computed tomography, and optical imaging has demonstrated regional and illness-specific brain changes at the onset of psychiatric disorders and in individuals at risk for such disorders. Many studies have shown that psychiatric medications induce specific measurable changes in brain anatomy and function that are related to clinical outcomes. As a result, a new field of radiology, termed psychoradiology, seems primed to play a major clinical role in guiding diagnostic and treatment planning decisions in patients with psychiatric disorders. This article will present the state of the art in this area, as well as perspectives regarding preparations in the field of radiology for its evolution. Furthermore, this article will (a) give an overview of the imaging and analysis methods for psychoradiology; (b) review the most robust and important radiologic findings and their potential clinical value from studies of major psychiatric disorders, such as depression and schizophrenia; and (c) describe the main challenges and future directions in this field. An ongoing and iterative process of developing biologically based nomenclatures with which to delineate psychiatric disorders and translational research to predict and track response to different therapeutic drugs is laying the foundation for a shift in diagnostic practice in psychiatry from a psychologic symptom-based approach to an imaging-based approach over the next generation. This shift will require considerable innovations for the acquisition, analysis, and interpretation of brain images, all of which will undoubtedly require the active involvement of radiologists. (©) RSNA, 2016 Online supplemental material is available for this article.
Background
At present, we do not have any biological tests which can contribute towards a diagnosis of depression. Neuroimaging measures have shown some potential as biomarkers for diagnosis. However, participants have generally been from the same ethnic background while the applicability of a biomarker would require replication in individuals of diverse ethnicities.
Aims
We sought to examine the diagnostic potential of the structural neuroanatomy of depression in a sample of a wide ethnic diversity.
Method
Structural magnetic resonance imaging (MRI) scans were obtained from 23 patients with major depressive disorder in an acute depressive episode (mean age: 39.8 years) and 20 matched healthy volunteers (mean age: 38.8 years). Participants were of Asian, African and Caucasian ethnicity recruited from the general community.
Results
Structural neuroanatomy combining white and grey matter distinguished patients from controls at the highest accuracy of 81% with the most stable pattern being at around 70%. A widespread network encompassing frontal, parietal, occipital and cerebellar regions contributed towards diagnostic classification.
Conclusions
These findings provide an important step in the development of potential neuroimaging-based tools for diagnosis as they demonstrate that the identification of depression is feasible within a multi-ethnic group from the community.
Declaration of interests
C.H.Y.F. has held recent research grants from Eli Lilly and Company and GlaxoSmithKline. L.M. is a former employee and stockholder of Eli Lilly and Company.
Copyright and usage
© The Royal College of Psychiatrists 2016. This is an open access article distributed under the terms of the Creative Commons Non-Commercial, No Derivatives (CC BY-NC-ND) licence.
Results: were significantly reliable. White matter regions showing major contributions favoring SAD over HC were located in the genu and splenium of the corpus callosum, the left uncinate fasciculus, the left inferior longitudinal fasciculus, the left inferior fronto-occipital fasciculus, bilateral frontal gyri and the left occipital lobe. Whereas, white matter in bilateral anterior cingula, the left middle cerebellar peduncle and the left inferior parietal lobule showed more contributions to diagnose HC than to diagnose SAD.
Background
At present there are no objective, biological markers that can be used to reliably identify individuals with post-traumatic stress disorder (PTSD). This study assessed the diagnostic potential of structural magnetic resonance imaging (sMRI) for identifying trauma-exposed individuals with and without PTSD.
Method
sMRI scans were acquired from 50 survivors of the Sichuan earthquake of 2008 who had developed PTSD, 50 survivors who had not developed PTSD and 40 healthy controls who had not been exposed to the earthquake. Support vector machine (SVM), a multivariate pattern recognition technique, was used to develop an algorithm that distinguished between the three groups at an individual level. The accuracy of the algorithm and its statistical significance were estimated using leave-one-out cross-validation and permutation testing.
Results
When survivors with PTSD were compared against healthy controls, both grey and white matter allowed discrimination with an accuracy of 91% (p < 0.001). When survivors without PTSD were compared against healthy controls, the two groups could be discriminated with accuracies of 76% (p < 0.001) and 85% (p < 0.001) based on grey and white matter, respectively. Finally, when survivors with and without PTSD were compared directly, grey matter allowed discrimination with an accuracy of 67% (p < 0.001); in contrast the two groups could not be distinguished based on white matter.
Conclusions
These results reveal patterns of neuroanatomical alterations that could be used to inform the identification of trauma survivors with and without PTSD at the individual level, and provide preliminary support to the development of SVM as a clinically useful diagnostic aid.
Background Group-level results suggest that relative to healthy controls (HCs), ultra-high-risk (UHR) and first-episode psychosis (FEP) subjects show alterations in neuroanatomy, neurofunction and cognition that may be mediated genetically. It is unclear, however, whether these groups can be differentiated at single-subject level, for instance using the machine learning analysis support vector machine (SVM). Here, we used a multimodal approach to examine the ability of structural magnetic resonance imaging (sMRI), functional MRI (fMRI), diffusion tensor neuroimaging (DTI), genetic and cognitive data to differentiate between UHR, FEP and HC subjects at the single-subject level using SVM.
Method Three age- and gender-matched SVM paired comparison groups were created comprising 19, 19 and 15 subject pairs for FEP versus HC, UHR versus HC and FEP versus UHR, respectively. Genetic, sMRI, DTI, fMRI and cognitive data were obtained for each participant and the ability of each to discriminate subjects at the individual level in conjunction with SVM was tested.
Results Successful classification accuracies (p < 0.05) comprised FEP versus HC (genotype, 67.86%; DTI, 65.79%; fMRI, 65.79% and 68.42%; cognitive data, 73.69%), UHR versus HC (sMRI, 68.42%; DTI, 65.79%), and FEP versus UHR (sMRI, 76.67%; fMRI, 73.33%; cognitive data, 66.67%).
Conclusions The results suggest that FEP subjects are identifiable at the individual level using a range of biological and cognitive measures. Comparatively, only sMRI and DTI allowed discrimination of UHR from HC subjects. For the first time FEP and UHR subjects have been shown to be directly differentiable at the single-subject level using cognitive, sMRI and fMRI data. Preliminarily, the results support clinical development of SVM to help inform identification of FEP and UHR subjects, though future work is needed to provide enhanced levels of accuracy.
The tutorial starts with an overview of the concepts of VC dimension and structural risk minimization. We then describe linear Support Vector Machines (SVMs) for separable and non-separable data, working through a non-trivial example in detail. We describe a mechanical analogy, and discuss when SVM solutions are unique and when they are global. We describe how support vector training can be practically implemented, and discuss in detail the kernel mapping technique which is used to construct SVM solutions which are nonlinear in the data. We show how Support Vector machines can have very large (even infinite) VC dimension by computing the VC dimension for homogeneous polynomial and Gaussian radial basis function kernels. While very high VC dimension would normally bode ill for generalization performance, and while at present there exists no theory which shows that good generalization performance is guaranteed for SVMs, there are several arguments which support the observed high accuracy of SVMs, which we review. Results of some experiments which were inspired by these arguments are also presented. We give numerous examples and proofs of most of the key theorems. There is new material, and I hope that the reader will find that even old material is cast in a fresh light.
Neuroimaging studies examining the effects of aging and neuropsychiatric disorders on the cerebral cortex have largely been based on measures of cortical volume. Given that cortical volume is a product of thickness and surface area, it is plausible that measures of volume capture at least 2 distinct sets of genetic influences. The present study aims to examine the genetic relationships between measures of cortical surface area and thickness. Participants were men in the Vietnam Era Twin Study of Aging (110 monozygotic pairs and 92 dizygotic pairs). Mean age was 55.8 years (range: 51-59). Bivariate twin analyses were utilized in order to estimate the heritability of cortical surface area and thickness, as well as their degree of genetic overlap. Total cortical surface area and average cortical thickness were both highly heritable (0.89 and 0.81, respectively) but were essentially unrelated genetically (genetic correlation = 0.08). This pattern was similar at the lobar and regional levels of analysis. These results demonstrate that cortical volume measures combine at least 2 distinct sources of genetic influences. We conclude that using volume in a genetically informative study, or as an endophenotype for a disorder, may confound the underlying genetic architecture of brain structure.
OBJECTIVE: Elevated levels of glucocorticoids in depression have been hypothesized to be associated with damage to the hippocampus, a brain area involved in learning and memory. The purpose of this study was to measure hippocampal volume in patients with depression. METHOD: Magnetic resonance imaging was used to measure the volume of the hippocampus in 16 patients with major depression in remission and 16 case-matched nondepressed comparison subjects. RESULTS: Patients with depression had a statistically significant 19% smaller left hippocampal volume than comparison subjects, without smaller volumes of comparison regions (amygdala, caudate, frontal lobe, and temporal lobe) or whole brain volume. The findings were significant after brain size, alcohol exposure, age, and education were controlled for. CONCLUSIONS: These findings are consistent with smaller left hippocampal volume in depression.
Previous neuromorphometric investigations of major depressive disorder (MDD) have reported abnormalities in gray matter in several regions, although the results have been inconsistent across studies. Some discrepancies in the results across studies may reflect design limitations such as small sample sizes, whereas others may reflect biological variability that potentially manifests as differences in clinical course. For example, it remains unclear whether the abnormalities found in persistently depressed MDD subjects extend to or persist in patients who experience prolonged remission. The aim of the present study was to investigate gray matter (GM) differences in unmedicated, currently-depressed participants (dMDD) and unmedicated, currently-remitted (rMDD) participants with MDD compared to healthy controls (HC).
Connectivity Map of the Brain
The growing appreciation that clinically abnormal behaviors in children and adolescents may be influenced or perhaps even initiated by developmental miscues has stoked an interest in mapping normal human brain maturation. Several groups have documented changes in gray and white matter using structural and functional magnetic resonance imaging (fMRI) in cross-sectional and longitudinal studies. Dosenbach et al. (p. 1358 ) developed an index of resting-state functional connectivity (that is, how tightly neuronal activities in distinct brain regions are correlated while the subject is at rest or even asleep) from analyses of three independent data sets (each based on fMRI scans of 150 to 200 individuals from ages 6 to 35 years old). Long-range connections increased with age and short-range connections decreased, indicating that networks become sparser and sharper with brain maturation.
Autism spectrum disorder (ASD) is a neurodevelopmental condition with multiple causes, comorbid conditions, and a wide range in the type and severity of symptoms expressed by different individuals. This makes the neuroanatomy of autism inherently difficult to describe. Here, we demonstrate how a multiparameter classification approach can be used to characterize the complex and subtle structural pattern of gray matter anatomy implicated in adults with ASD, and to reveal spatially distributed patterns of discriminating regions for a variety of parameters describing brain anatomy. A set of five morphological parameters including volumetric and geometric features at each spatial location on the cortical surface was used to discriminate between people with ASD and controls using a support vector machine (SVM) analytic approach, and to find a spatially distributed pattern of regions with maximal classification weights. On the basis of these patterns, SVM was able to identify individuals with ASD at a sensitivity and specificity of up to 90% and 80%, respectively. However, the ability of individual cortical features to discriminate between groups was highly variable, and the discriminating patterns of regions varied across parameters. The classification was specific to ASD rather than neurodevelopmental conditions in general (e.g., attention deficit hyperactivity disorder). Our results confirm the hypothesis that the neuroanatomy of autism is truly multidimensional, and affects multiple and most likely independent cortical features. The spatial patterns detected using SVM may help further exploration of the specific genetic and neuropathological underpinnings of ASD, and provide new insights into the most likely multifactorial etiology of the condition.
Major depressive disorder is associated with both structural and functional alterations in the emotion regulation network of the central nervous system. The relation between structural and functional changes is largely unknown. Therefore, we sought to determine the relation between structural differences and functional alterations during the recognition of emotional facial expressions.
We examined 13 medication-free patients with major depression and 15 healthy controls by use of structural T1-weighted high-resolution magnetic resonance imaging (MRI) and functional MRI during 1 session. We set the statistical threshold for the analysis of imaging data to p < 0.001 (uncorrected).
As shown by voxel-based morphometry, depressed patients had reductions in orbitofrontal cortex volume and increases in cerebellar volume. Additionally, depressed patients showed increased activity during emotion recognition in the middle frontal cortex, caudate nucleus, precuneus and lingual gyrus. Within this cerebral network, the orbitofrontal volumes were negatively correlated in depressed patients but not in healthy controls with changes in blood oxygen level-dependent signal in the middle frontal gyrus, caudate nucleus, precuneus and supplementary motor area.
Our results are limited by the relatively small sample size.
This combined functional and structural MRI study provides evidence that the orbitofrontal cortex is a key area in major depression and that structural changes result in functional alterations within the emotional circuit. Whether these alterations in the orbitofrontal cortex are also related to persistent emotional dysfunction in remitted mental states and, therefore, are related to the risk of depression needs further exploration.
Here, we examine morphological changes in cortical thickness of patients with Alzheimer's disease (AD) using image analysis algorithms for brain structure segmentation and study automatic classification of AD patients using cortical and volumetric data. Cortical thickness of AD patients (n = 14) was measured using MRI cortical surface-based analysis and compared with healthy subjects (n = 20). Data was analyzed using an automated algorithm for tissue segmentation and classification. A Support Vector Machine (SVM) was applied over the volumetric measurements of subcortical and cortical structures to separate AD patients from controls. The group analysis showed cortical thickness reduction in the superior temporal lobe, parahippocampal gyrus, and enthorhinal cortex in both hemispheres. We also found cortical thinning in the isthmus of cingulate gyrus and middle temporal gyrus at the right hemisphere, as well as a reduction of the cortical mantle in areas previously shown to be associated with AD. We also confirmed that automatic classification algorithms (SVM) could be helpful to distinguish AD patients from healthy controls. Moreover, the same areas implicated in the pathogenesis of AD were the main parameters driving the classification algorithm. While the patient sample used in this study was relatively small, we expect that using a database of regional volumes derived from MRI scans of a large number of subjects will increase the SVM power of AD patient identification.
Depression is experienced as a persistent low mood or anhedonia accompanied by behavioural and cognitive disturbances which impair day to day functioning. However, the diagnosis is largely based on self-reported symptoms, and there are no neurobiological markers to guide the choice of treatment. In the present study, we examined the prognostic and diagnostic potential of the structural neural correlates of depression.
Subjects were 37 patients with major depressive disorder (mean age 43.2 years), medication-free, in an acute depressive episode, and 37 healthy individuals. Following the MRI scan, 30 patients underwent treatment with the antidepressant medication fluoxetine or cognitive behavioural therapy (CBT). Of the patients who subsequently achieved clinical remission with antidepressant medication, the whole brain structural neuroanatomy predicted 88.9% of the clinical response, prior to the initiation of treatment (88.9% patients in clinical remission (sensitivity) and 88.9% patients with residual symptoms (specificity), p = 0.01). Accuracy of the structural neuroanatomy as a diagnostic marker though was 67.6% (64.9% patients (sensitivity) and 70.3% healthy individuals (specificity), p = 0.027).
The structural neuroanatomy of depression shows high predictive potential for clinical response to antidepressant medication, while its diagnostic potential is more limited. The present findings provide initial steps towards the development of neurobiological prognostic markers for depression.
The brain disturbances that place a person at risk for developing depression are unknown. We imaged the brains of 131 individuals, ages 6 to 54 years, who were biological descendants (children or grandchildren) of individuals identified as having either moderate to severe, recurrent, and functionally debilitating depression or as having no lifetime history of depression. We compared cortical thickness across high- and low-risk groups, detecting large expanses of cortical thinning across the lateral surface of the right cerebral hemisphere in persons at high risk. Thinning correlated with measures of current symptom severity, inattention, and visual memory for social and emotional stimuli. Mediator analyses indicated that cortical thickness mediated the associations of familial risk with inattention, visual memory, and clinical symptoms. These findings suggest that cortical thinning in the right hemisphere produces disturbances in arousal, attention, and memory for social stimuli, which in turn may increase the risk of developing depressive illness.
Interpreting brain image experiments requires analysis of complex, multivariate data. In recent years, one analysis approach that has grown in popularity is the use of machine learning algorithms to train classifiers to decode stimuli, mental states, behaviours and other variables of interest from fMRI data and thereby show the data contain information about them. In this tutorial overview we review some of the key choices faced in using this approach as well as how to derive statistically significant results, illustrating each point from a case study. Furthermore, we show how, in addition to answering the question of 'is there information about a variable of interest' (pattern discrimination), classifiers can be used to tackle other classes of question, namely 'where is the information' (pattern localization) and 'how is that information encoded' (pattern characterization).
A novel approach to correcting for intensity nonuniformity in magnetic resonance (MR) data is described that achieves high performance without requiring a model of the tissue classes present. The method has the advantage that it can be applied at an early stage in an automated data analysis, before a tissue model is available. Described as nonparametric nonuniform intensity normalization (N3), the method is independent of pulse sequence and insensitive to pathological data that might otherwise violate model assumptions. To eliminate the dependence of the field estimate on anatomy, an iterative approach is employed to estimate both the multiplicative bias field and the distribution of the true tissue intensities. The performance of this method is evaluated using both real and simulated MR data.
Accurate and automated methods for measuring the thickness of human cerebral cortex could provide powerful tools for diagnosing and studying a variety of neurodegenerative and psychiatric disorders. Manual methods for estimating cortical thickness from neuroimaging data are labor intensive, requiring several days of effort by a trained anatomist. Furthermore, the highly folded nature of the cortex is problematic for manual techniques, frequently resulting in measurement errors in regions in which the cortical surface is not perpendicular to any of the cardinal axes. As a consequence, it has been impractical to obtain accurate thickness estimates for the entire cortex in individual subjects, or group statistics for patient or control populations. Here, we present an automated method for accurately measuring the thickness of the cerebral cortex across the entire brain and for generating cross-subject statistics in a coordinate system based on cortical anatomy. The intersubject standard deviation of the thickness measures is shown to be less than 0.5 mm, implying the ability to detect focal atrophy in small populations or even individual subjects. The reliability and accuracy of this new method are assessed by within-subject test-retest studies, as well as by comparison of cross-subject regional thickness measures with published values.
Highly accurate surface models of the cerebral cortex are becoming increasingly important as tools in the investigation of the functional organization of the human brain. The construction of such models is difficult using current neuroimaging technology due to the high degree of cortical folding. Even single voxel misclassifications can result in erroneous connections being created between adjacent banks of a sulcus, resulting in a topologically inaccurate model. These topological defects cause the cortical model to no longer be homeomorphic to a sheet, preventing the accurate inflation, flattening, or spherical morphing of the reconstructed cortex. Surface deformation techniques can guarantee the topological correctness of a model, but are time-consuming and may result in geometrically inaccurate models. In order to address this need we have developed a technique for taking a model of the cortex, detecting and fixing the topological defects while leaving that majority of the model intact, resulting in a surface that is both geometrically accurate and topologically correct.
Current treatments for depression are inadequate for many individuals, and progress in understanding the neurobiology of depression is slow. Several promising hypotheses of depression and antidepressant action have been formulated recently. These hypotheses are based largely on dysregulation of the hypothalamic-pituitary-adrenal axis and hippocampus and implicate corticotropin-releasing factor, glucocorticoids, brain-derived neurotrophic factor, and CREB. Recent work has looked beyond hippocampus to other brain areas that are also likely involved. For example, nucleus accumbens, amygdala, and certain hypothalamic nuclei are critical in regulating motivation, eating, sleeping, energy level, circadian rhythm, and responses to rewarding and aversive stimuli, which are all abnormal in depressed patients. A neurobiologic understanding of depression also requires identification of the genes that make individuals vulnerable or resistant to the syndrome. These advances will fundamentally improve the treatment and prevention of depression.
To examine structural abnormalities in subregions of the prefrontal cortex in elderly patients with depression, the authors explored differences in gray matter, white matter, and CSF volumes by applying a parcellation method based on magnetic resonance imaging (MRI).
Twenty-four elderly patients with major depression and 19 group-matched comparison subjects were studied with high-resolution MRI. Cortical surface extraction, tissue segmentation, and cortical parcellation methods were applied to obtain volume measures of gray matter, white matter, and CSF in seven prefrontal subregions: the anterior cingulate, gyrus rectus, orbitofrontal cortex, precentral gyrus, superior frontal cortex, middle frontal cortex, and inferior frontal cortex.
Highly significant bilateral volume reductions in gray matter were observed in the anterior cingulate, the gyrus rectus, and the orbitofrontal cortex. Depressed patients also exhibited significant bilateral white matter volume reductions and significant CSF volume increases in the anterior cingulate and the gyrus rectus. Finally, the depressed group showed significant CSF volume reductions in the orbitofrontal cortex relative to the comparison subjects. None of the other regions examined revealed significant structural abnormalities.
The prominent bilateral gray matter deficits in the anterior cingulate and the gyrus rectus as well as the orbitofrontal cortex may reflect disease-specific modifications of elderly depression. The differential pattern of abnormalities detected in the white matter and CSF compartments imply that distinct etiopathological mechanisms might underlie the structural cortical changes in these regions.
A novel approach to correcting for intensity nonuniformity in magnetic resonance (MR) data is described that achieves high performance without requiring a model of the tissue classes present. The method has the advantage that it can be applied at an early stage in an automated data analysis, before a tissue model is available. Described as nonparametric nonuniform intensity normalization (N3), the method is independent of pulse sequence and insensitive to pathological data that might otherwise violate model assumptions. To eliminate the dependence of the field estimate on anatomy, an iterative approach is employed to estimate both the multiplicative bias field and the distribution of the true tissue intensities. The performance of this method is evaluated using both real and simulated MR data.
The modular organization of nervous systems is a widely documented principle of design for both vertebrate and invertebrate brains of which the columnar organization of the neocortex is an example. The classical cytoarchitectural areas of the neocortex are composed of smaller units, local neural circuits repeated iteratively within each area. Modules may vary in cell type and number, in internal and external connectivity, and in mode of neuronal processing between different large entities; within any single large entity they have a basic similarity of internal design and operation. Modules are most commonly grouped into entities by sets of dominating external connections. This unifying factor is most obvious for the heterotypical sensory and motor areas of the neocortex. Columnar defining factors in homotypical areas are generated, in part, within the cortex itself. The set of all modules composing such an entity may be fractionated into different modular subsets by different extrinsic connections. Linkages between them and subsets in other large entities form distributed systems. The neighborhood relations between connected subsets of modules in different entities result in nested distributed systems that serve distributed functions. A cortical area defined in classical cytoarchitectural terms may belong to more than one and sometimes to several distributed systems. Columns in cytoarchitectural areas located at some distance from one another, but with some common properties, may be linked by long-range, intracortical connections.
Many structural features of the mammalian central nervous system can be explained by a morphogenetic mechanism that involves mechanical tension along axons, dendrites and glial processes. In the cerebral cortex, for example, tension along axons in the white matter can explain how and why the cortex folds in a characteristic species-specific pattern. In the cerebellum, tension along parallel fibres can explain why the cortex is highly elongated but folded like an accordion. By keeping the aggregate length of axonal and dendritic wiring low, tension should contribute to the compactness of neural circuitry throughout the adult brain.
In this chapter we consider bounds on the rate of uniform convergence. We consider upper bounds (there exist lower bounds as well (Vapnik and Chervonenkis, 1974); however, they are not as important for controlling the learning processes as the upper bounds).
We present a technique for automatically assigning a neuroanatomical label to each location on a cortical surface model based on probabilistic information estimated from a manually labeled training set. This procedure incorporates both geometric information derived from the cortical model, and neuroanatomical convention, as found in the training set. The result is a complete labeling of cortical sulci and gyri. Examples are given from two different training sets generated using different neuroanatomical conventions, illustrating the flexibility of the algorithm. The technique is shown to be comparable in accuracy to manual labeling.
We present a technique for automatically assigning a neuroanatomical label to each voxel in an MRI volume based on probabilistic information automatically estimated from a manually labeled training set. In contrast to existing segmentation procedures that only label a small number of tissue classes, the current method assigns one of 37 labels to each voxel, including left and right caudate, putamen, pallidum, thalamus, lateral ventricles, hippocampus, and amygdala. The classification technique employs a registration procedure that is robust to anatomical variability, including the ventricular enlargement typically associated with neurological diseases and aging. The technique is shown to be comparable in accuracy to manual labeling, and of sufficient sensitivity to robustly detect changes in the volume of noncortical structures that presage the onset of probable Alzheimer's disease.
The neurons of the human cerebral cortex are arranged in a highly folded sheet, with the majority of the cortical surface area buried in folds. Cortical maps are typically arranged with a topography oriented parallel to the cortical surface. Despite this unambiguous sheetlike geometry, the most commonly used coordinate systems for localizing cortical features are based on 3-D stereotaxic coordinates rather than on position relative to the 2-D cortical sheet. In order to address the need for a more natural surface-based coordinate system for the cortex, we have developed a means for generating an average folding pattern across a large number of individual subjects as a function on the unit sphere and of nonrigidly aligning each individual with the average. This establishes a spherical surface-based coordinate system that is adapted to the folding pattern of each individual subject, allowing for much higher localization accuracy of structural and functional features of the human brain.
Introduction:
Cortical thickness mapping is a widely used method for the analysis of neuroanatomical differences between subject groups. We applied power analysis methods over a range of image processing parameters to derive a model that allows researchers to calculate the number of subjects required to ensure a well-powered cross-sectional cortical thickness study.
Methods:
0.9-mm isotropic T1 -weighted 3D MPRAGE MRI scans from 98 controls (53 females, age 29.1 ± 9.7 years) were processed using Freesurfer 5.0. Power analyses were carried out using vertex-wise variance estimates from the coregistered cortical thickness maps, systematically varying processing parameters. A genetic programming approach was used to derive a model describing the relationship between sample size and processing parameters. The model was validated on four Alzheimer's Disease Neuroimaging Initiative control datasets (mean 126.5 subjects/site, age 76.6 ± 5.0 years).
Results:
Approximately 50 subjects per group are required to detect a 0.25-mm thickness difference; less than 10 subjects per group are required for differences of 1 mm (two-sided test, 10 mm smoothing, α = 0.05). Sample size estimates were heterogeneous over the cortical surface. The model yielded sample size predictions within 2-6% of that determined experimentally using independent data from four other datasets. Fitting parameters of the model to data from each site reduced the estimation error to less than 2%.
Conclusions:
The derived model provides a simple tool for researchers to calculate how many subjects should be included in a well-powered cortical thickness analysis.
Standard univariate analysis of neuroimaging data has revealed a host of neuroanatomical and functional differences between healthy individuals and patients suffering a wide range of neurological and psychiatric disorders. Significant only at group level however these findings have had limited clinical translation, and recent attention has turned toward alternative forms of analysis, including Support-Vector-Machine (SVM). A type of machine learning, SVM allows categorisation of an individual's previously unseen data into a predefined group using a classification algorithm, developed on a training data set. In recent years, SVM has been successfully applied in the context of disease diagnosis, transition prediction and treatment prognosis, using both structural and functional neuroimaging data. Here we provide a brief overview of the method and review those studies that applied it to the investigation of Alzheimer's disease, schizophrenia, major depression, bipolar disorder, presymptomatic Huntington's disease, Parkinson's disease and autistic spectrum disorder. We conclude by discussing the main theoretical and practical challenges associated with the implementation of this method into the clinic and possible future directions.
Previous meta-analyses of structural MRI studies have shown diffuse cortical and sub-cortical abnormalities in unipolar depression. However, the presence of duplicate publications, recruitment of particular age groups and the selection of specific regions of interest means that there is uncertainty about the balance of current research. Moreover, the lack of systematic exploration of highly significant heterogeneity has prevented the generalisability of finding. A systematic review and random-effects meta-analysis was carried out to estimate effect sizes. Possible publication bias, and the impact of various study design characteristics on the magnitude of the observed effect size were systematically explored. The aim of this study was 1) to include structural MRI studies systematically comparing unipolar depression with bipolar disorder and healthy volunteers; 2) to consider all available structures of interest without specific age limits, avoiding data duplication, and 3) to explore the influence of factors contributing to the measured effect sizes systematically with meta-regression analyses. Unipolar depression was characterised by reduced brain volume in areas involved in emotional processing, including the frontal cortex, orbitofrontal cortex, cingulate cortex, hippocampus and striatum. There was also evidence of pituitary enlargement and an excess of white matter hyperintensity volume in unipolar depression. Factors which influenced the magnitude of the observed effect sizes were differences in methods, clinical variables, pharmacological interventions and sample age.
In this paper, we focus on the retrospective topology correction of surfaces. We propose a technique to accurately correct the spherical topology of cortical surfaces. Specifically, we construct a mapping from the original surface onto the sphere to detect topological defects as minimal nonhomeomorphic regions. The topology of each defect is then corrected by opening and sealing the surface along a set of nonseparating loops that are selected in a Bayesian framework. The proposed method is a wholly self-contained topology correction algorithm, which determines geometrically accurate, topologically correct solutions based on the magnetic resonance imaging (MRI) intensity profile and the expected local curvature. Applied to real data, our method provides topological corrections similar to those made by a trained operator.
The authors used magnetic resonance imaging and an image analysis technique known as cortical pattern matching to map cortical gray matter deficits in elderly depressed patients with an illness onset after age 60 (late-onset depression).
Seventeen patients with late-onset depression (11 women and six men; mean age=75.24, SD=8.52) and 17 group-matched comparison subjects (11 women and six men; mean age=73.88, SD=7.61) were included. Detailed spatial analyses of gray matter were conducted across the entire cortex by measuring local proportions of gray matter at thousands of homologous cortical surface locations in each subject, and these patterns were matched across subjects by using elastic transformations to align sulcal topography. To visualize regional changes, statistical differences were mapped at each cortical surface location in three dimensions.
The late-onset depression group exhibited significant gray matter deficits in the right lateral temporal cortex and the right parietal cortex, where decreases were most pronounced in sensorimotor regions. The statistical maps also showed gray matter deficits in the same regions of the left hemisphere that approached significance after permutation testing. No significant group differences were detected in frontal cortices or any other anatomical region.
Regionally specific decreases of gray matter occur in late-onset depression, supporting the hypothesis that this subset of elderly patients with major depression presents with certain unique neuroanatomical abnormalities that may differ from patients with an earlier onset of illness.
Choosing the appropriate neuroimaging phenotype is critical to successfully identify genes that influence brain structure or function. While neuroimaging methods provide numerous potential phenotypes, their role for imaging genetics studies is unclear. Here we examine the relationship between brain volume, grey matter volume, cortical thickness and surface area, from a genetic standpoint. Four hundred and eighty-six individuals from randomly ascertained extended pedigrees with high-quality T1-weighted neuroanatomic MRI images participated in the study. Surface-based and voxel-based representations of brain structure were derived, using automated methods, and these measurements were analysed using a variance-components method to identify the heritability of these traits and their genetic correlations. All neuroanatomic traits were significantly influenced by genetic factors. Cortical thickness and surface area measurements were found to be genetically and phenotypically independent. While both thickness and area influenced volume measurements of cortical grey matter, volume was more closely related to surface area than cortical thickness. This trend was observed for both the volume-based and surface-based techniques. The results suggest that surface area and cortical thickness measurements should be considered separately and preferred over gray matter volumes for imaging genetic studies.
Functional neuroimaging studies have implicated the insular cortex in emotional processing, including the evaluation of one's own emotion, as well as in the neurobiology of major depressive disorder (MDD). Nevertheless, it remains largely unknown whether MDD patients exhibit morphologic changes of the insular cortex, and whether such changes reflect state or trait markers of the disorder.
We delineated the anterior and posterior insular cortices using magnetic resonance imaging in 29 currently depressed patients (mean age=32.5 years, 7 males), 27 remitted depressed patients (mean age=35.1 years, 9 males), and 33 age- and gender-matched healthy control subjects (mean age=34.0 years, 12 males).
Both current and remitted MDD patients showed significant volume reduction of the left anterior insular cortex as compared with healthy controls, but there was no group difference in the posterior insular cortex volume. Insular volumes did not correlate with the severity of depressive symptoms. Furthermore, the presence of melancholia and co-morbidity with anxiety disorders did not affect insular cortex volumes.
Although there was no difference in the insular cortex volume between medicated and unmedicated patients, a comprehensive investigation of medication effects was not possible, as complete data (e.g., dose, duration) were not available.
These findings suggest that the morphologic abnormality of the anterior insular cortex, which plays a major role in introspection and emotional control, may be a trait-related marker of vulnerability to major depression, supporting the notion that MDD involves pathological alterations of limbic and related cortical structures.
Patients with major depressive disorder are found to show selective attention biases towards mood-congruent information. Although previous studies have identified various structural changes in the brains of these patients, it remains unclear whether the structural abnormalities are associated with these attention biases. In this study, we used voxel-based morphometry (VBM) to explore the structural correlates of attention biases towards depression-related stimuli.
Seventeen female patients with major depressive disorder and 17 female healthy controls, matched on age and intelligence, underwent magnetic resonance imaging (MRI). They also performed positive-priming (PP) and negative-priming (NP) tasks involving neutral and negative words that assessed selective attention biases. The reaction time (RT) to a target word that had been attended to or ignored in a preceding trial was measured on the PP and NP tasks respectively. The structural differences between the two groups were correlated with the indexes of attention biases towards the negative words.
The enhanced facilitation of attention to stimuli in the PP task by the negative valence was only found in the depressed patients, not in the healthy controls. Such attention biases towards negative stimuli were found to be associated with reduced gray-matter concentration (GMC) in the right superior frontal gyrus, the right anterior cingulate gyrus and the right fusiform gyrus. No differential effect in inhibition of attention towards negative stimuli in the NP task was found between the depressed patients and the healthy controls.
Specific structural abnormalities in depression are associated with their attention biases towards mood-congruent information.
We aimed to investigate structural abnormalities in first-episode remitted geriatric depression (RGD) using optimized voxel-based morphometry (VBM) in closely matched patients and healthy controls, and examining the relationship of performances on neuropsychological tests with regional white matter volumes.
Forty subjects with first-episode RGD and 36 well-matched healthy controls were recruited for this study and neuropsychological tests and magnetic resonance imaging (MRI) were conducted on the subjects. The differences in regional white matter volume were determined between these two groups by optimized VBM.
The white matter volumes of left inferior parietal lobule and right inferior frontal gyrus were significantly larger in patients with RGD relative to healthy controls. RGD patients performed significantly worse in the delayed recall of RAVLT, Trail Making Test A and B (seconds), and Symbol Digit Modalities Test when compared with the control group (all P<0.01). And there was a significant positive correlation between white matter volume of right inferior frontal gyrus and Trail Making Test A (r=0.319, P=0.045) in patients with RGD.
This study is cross-sectional, therefore it cannot determine whether increased white matter volume is a state marker or trait marker of RGD.
These results reveal that RGD is associated with larger white matter volumes of left inferior parietal lobule and right inferior frontal gyrus, and the right inferior frontal gyrus may thus be involved in the pathophysiology of executive function in RGD.
Unravelling the pathophysiology of depression is a unique challenge. Not only are depressive syndromes heterogeneous and their aetiologies diverse, but symptoms such as guilt and suicidality are impossible to reproduce in animal models. Nevertheless, other symptoms have been accurately modelled, and these, together with clinical data, are providing insight into the neurobiology of depression. Recent studies combining behavioural, molecular and electrophysiological techniques reveal that certain aspects of depression result from maladaptive stress-induced neuroplastic changes in specific neural circuits. They also show that understanding the mechanisms of resilience to stress offers a crucial new dimension for the development of fundamentally novel antidepressant treatments.
To compare the morphology of the cerebral cortex and its characteristic pattern of gyri and sulci in individuals with and without schizophrenia, T1-weighted magnetic resonance scans were collected, along with clinical and cognitive information, from 33 individuals with schizophrenia and 30 healthy individuals group-matched for age, gender, race and parental socioeconomic status. Sulcal depth was measured across the entire cerebral cortex by reconstructing surfaces of cortical mid-thickness (layer 4) in each hemisphere and registering them to the human PALS cortical atlas. Group differences in sulcal depth were tested using methods for cluster size analysis and interhemispheric symmetry analysis. A significant group difference was found bilaterally in the parietal operculum, where the average sulcal depth was shallower in individuals with schizophrenia. In addition, group differences in sulcal depth showed significant bilateral symmetry across much of the occipital, parietal, and temporal cortices. In individuals with schizophrenia, sulcal depth in the left hemisphere was correlated with the severity of impaired performance on tests of working memory and executive function.
We investigated the changes of sulcal shape (average mean curvature in folded regions and sulcal depth) in mild cognitive impairment (MCI) and Alzheimer's disease (AD) using quantitative surface-based methods in a large sample of magnetic resonance imaging data. Moreover, we observed their relationships with cortical thickness and gyral white matter (WM) volume, while considering age effect. This study involved 85 normal controls (n [men/women]: 36/49, age [mean+/-SD]: 71.1+/-4.9 years), and 100 MCI (44/56, 71.8+/-6.5) and 145 AD subjects (53/92, 72.7+/-7.3). We found significantly lower average mean curvature (greater sulcal widening) and shallower sulcal depth with disease progression from controls to MCI and MCI to AD. The most remarkable change in MCI and AD was sulcal widening observed in the temporal lobe (average mean curvature, control [mean]: 0.564, MCI: 0.534 (5.3% decrease from control), AD: 0.486 (13.8% and 9.0% decrease from control and MCI respectively)). Of the four measurements, the sulcal widening measurement showed the highest sensitivity in revealing group differences between control and MCI, which might be useful for detecting early dementia. Significant reductions in cortical thickness and gyral WM volume also occurred in MCI and AD. Multiple regression analysis demonstrated that a wider and shallower sulcal shape was primarily associated with decreased cortical thickness and gyral WM volume in each group. Age-related trends for the sulcal shape were not strongly found when cortical thickness and gyral WM volume were considered.
Examined the influence of diagnostic subtype of depression on perceptual asymmetry for dichotic listening and visual tachistoscopic tasks. A total of 65 unmedicated patients with major depressive disorders and 30 normal controls were tested on a verbal and nonverbal task in each modality. Patients diagnosed according to the DSM-III with melancholia had abnormal perceptual asymmetry for dichotic nonsense syllable and complex tone tasks. In contrast, patients having a nonmelancholic "atypical depression" (reactivity of mood with preserved pleasure capacity and associated features) did not differ from normal controls on these tasks, but had an increased incidence of left handedness. Bipolar depression (history of hypomania) differed from unipolar depression in showing abnormal perceptual asymmetry for a tachistoscopic dot enumeration task. Alterations of perceptual asymmetry in melancholia and bipolar depression were consistent with hypothesized right hemisphere dysfunction.
During development the human cortex changes from a smooth lissencephalic structure to one that is highly convoluted. Increases in the degree of cortical folding are associated with brain size only for the first part of brain growth; during the second half, differences in cortical folding match those of brain size, resulting in no change in the degree of folding. When the degree of cortical folding is studied as a function of age, a brief postnatal overshoot, an effect of brain size, is observed. The analysis suggests that the mechanical hypothesis of cortical buckling can best explain the degree of cortical folding, but that other hypotheses, like gyrogenesis, are required to explain the placement and orientation of sulci. The adult asymptote in degree of cortical folding is associated with the onset and disappearance of single subplate lamina, suggesting that subplate:cortical plate associations should be examined as causal for gyrification. Areas whose sulci differ in length between the two hemispheres have similar degrees of convolutedness, supporting interpretations that the sizes of gyri are asymmetric in the two hemispheres. The ontogenetic data support the thesis that human cortical proportions evolved when the brain enlarged in size and that the process was not one of neoteny.
The Global Burden of Disease Study, a comprehensive regional and global assessment of mortality and disability from 107 diseases and injuries and 10 risk factors, is an example of an evidence-based input to public health policy debate. The study, which includes projections of the burden through the year 2020, uses the disability-adjusted life year as a composite measure of years of life lost due to premature mortality and years lived with disability. Future patterns of death and disability are likely to change dramatically because of aging of the world's population, the epidemic of tobacco-related disease, the human immunodeficiency virus epidemic, and the likely reduction in death rates from communicable diseases in children.
The modular organization of nervous systems is a widely documented principle of design for both vertebrate and invertebrate brains of which the columnar organization of the neocortex is an example. The classical cytoarchitectural areas of the neocortex are composed of smaller units, local neural circuits repeated iteratively within each area. Modules may vary in cell type and number, in internal and external connectivity, and in mode of neuronal processing between different large entities; within any single large entity they have a basic similarity of internal design and operation. Modules are most commonly grouped into entities by sets of dominating external connections. This unifying factor is most obvious for the heterotypical sensory and motor areas of the neocortex. Columnar defining factors in homotypical areas are generated, in part, within the cortex itself. The set of all modules composing such an entity may be fractionated into different modular subsets by different extrinsic connections. Linkages between them and subsets in other large entities form distributed systems. The neighborhood relations between connected subsets of modules in different entities result in nested distributed systems that serve distributed functions. A cortical area defined in classical cytoarchitectural terms may belong to more than one and sometimes to several distributed systems. Columns in cytoarchitectural areas located at some distance from one another, but with some common properties, may be linked by long-range, intracortical connections.
The surface of the human cerebral cortex is a highly folded sheet with the majority of its surface area buried within folds. As such, it is a difficult domain for computational as well as visualization purposes. We have therefore designed a set of procedures for modifying the representation of the cortical surface to (i) inflate it so that activity buried inside sulci may be visualized, (ii) cut and flatten an entire hemisphere, and (iii) transform a hemisphere into a simple parameterizable surface such as a sphere for the purpose of establishing a surface-based coordinate system.
Several properties of the cerebral cortex, including its columnar and laminar organization, as well as the topographic organization of cortical areas, can only be properly understood in the context of the intrinsic two-dimensional structure of the cortical surface. In order to study such cortical properties in humans, it is necessary to obtain an accurate and explicit representation of the cortical surface in individual subjects. Here we describe a set of automated procedures for obtaining accurate reconstructions of the cortical surface, which have been applied to data from more than 100 subjects, requiring little or no manual intervention. Automated routines for unfolding and flattening the cortical surface are described in a companion paper. These procedures allow for the routine use of cortical surface-based analysis and visualization methods in functional brain imaging.
We present a technique for automatically assigning a neuroanatomical label to each voxel in an MRI volume based on probabilistic information automatically estimated from a manually labeled training set. In contrast to existing segmentation procedures that only label a small number of tissue classes, the current method assigns one of 37 labels to each voxel, including left and right caudate, putamen, pallidum, thalamus, lateral ventricles, hippocampus, and amygdala. The classification technique employs a registration procedure that is robust to anatomical variability, including the ventricular enlargement typically associated with neurological diseases and aging. The technique is shown to be comparable in accuracy to manual labeling, and of sufficient sensitivity to robustly detect changes in the volume of noncortical structures that presage the onset of probable Alzheimer's disease.
Depression is characterized by functional insufficiency of the right hemisphere combined with its physiological overactivation. This paradox can be solved in the frame of the general concept of brain laterality. According to the present assumption, the left hemisphere organizes any information in an unambiguous monosemantic context, and this process requires an additional activation of the brain cortex in order to restrict natural relationships between objects and events. On the contrary, the right hemisphere organizes any information in the polysemantic context based on the simultaneous capture of the numerous natural relationships between elements of information. In healthy creative subjects this process does not require additional physiological activation of the cortex. In depression the physiological overactivation of the right hemisphere reflects the unsuccessful effort to overcome its functional insufficiency.
The orbitofrontal cortex (OFC) plays a major role in neuropsychologic functioning including exteroceptive and interoceptive information coding, reward-guided behavior, impulse control, and mood regulation. This study examined the OFC and its subdivisions in patients with MDD and matched healthy control subjects.
Magnetic resonance imaging (MRI) was performed on 31 unmedicated MDD and 34 control subjects matched for age, gender, and race. Gray matter volumes of the OFC and its lateral and medial subdivisions were measured blindly.
The MDD patients had smaller gray matter volumes in right medial [two-way analysis of covariance F(1,60) = 4.285; p =.043] and left lateral OFC [F(1,60) = 4.252; p =.044]. Left lateral OFC volume correlated negatively with age in patients but not in control subjects. Male, but not female patients exhibited smaller left and right medial OFC volumes compared with healthy control subjects of the same gender.
These findings suggest that patients with MDD have reduced OFC gray matter volumes. Although this reduction might be important in understanding the pathophysiology of MDD, its functional and psychopathologic consequences are as yet unclear. Future studies examining the relationship between specific symptomatic dimensions of MDD and OFC volumes could be especially informative.