Babak A Ardekani’s research while affiliated with Nathan Kline Institute and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (149)


A comparative study of rigid-body registration algorithms for the alignment of longitudinal structural MRI of the brain
  • Preprint

November 2024

Yazdan Rezaee Jouryabi

·

Reza Lashgari

·

Babak A Ardekani

Longitudinal structural MRI (sMRI) may be used to characterize brain morphological changes over time. A key requirement for this approach is accurate rigid-body alignment of longitudinal sMRI. We have recently developed the automatic temporal registration algorithm (ATRA) for this purpose. ATRA is a landmark-based approach capable of registering dozens of serial sMRI simultaneously in an unbiased manner. The aim of the research presented in this paper was to evaluate the accuracy and inverse-consistency of ATRA in comparison to three commonly used sMRI alignment methods: FSL, FreeSurfer, and ANTS. In the absence of a ground truth, it is only possible to quantitatively determine the degree of discrepancy between two algorithms. We propose that if the discrepancy exceeds a certain threshold, the relative accuracy of the two algorithms could be determined visually. We computed the discrepancy between ATRA and each of the three other methods for the alignment of 150 pairs of sMRI taken roughly one year apart. We visually rated the accuracy of alignments in cases where the discrepancy was greater than .5 mm while the rater was agnostic to the registration method. In those instances, ATRA was considered more accurate than FSL in 46 out of 48 cases, more accurate than FreeSurfer in 6 out of 7 cases, and more accurate than ANTS in all 6 cases. ATRA was also the most inverse-consistent method. In addition to being capable of performing unbiased group-wise registration, ATRA is the most accurate algorithm in comparison to several commonly used rigid-body alignment methods.



Impact of Diffusion Tensor Field Smoothing with Log-Cholesky Metric on Noise Reduction

July 2024

·

8 Reads

·

1 Citation

Frontiers in Biomedical Technologies

Diffusion Tensor Imaging (DTI) is a noise-sensitive method, where a low Signal-to-Noise Ratio (SNR) results in significant errors in the estimated tensor field. Post-reconstruction tensor field smoothing is a simple and effective solution for alleviating this problem. Diffusion tensors can be represented by Symmetric Positive-Definite (SPD) matrices which can be viewed as a Riemannian manifold after defining a suitable metric on the space of SPD matrices. The Log-Cholesky metric is a recently developed concept with several advantages over previously defined metrics, e.g., Frobenius, Log-Euclidean, and affine-invariant metrics. In this work, we implemented a smoothing method based on the Log-Cholesky metric and show its effectiveness as a simple solution to filtering noisy diffusion tensor fields.


Fig. 1. Distribution of the individual values of Corrected Brain Age against Chronological Age (left panels) and Corrected Brain Age Gap (right panels) for the Control group (top panels) and the AUD group (bottom panels). The individual values (marked in dots) above the diagonal line in the left panels represent increased brain age while the values below the diagonal line represent decreased brain age.
Fig. 2. Bar Graphs showing the mean values for Chronological Age, Uncorrected Brain Age, and Corrected Brain Age (top panel) as well as for Uncorrected Brain Age Gap and Corrected Brain Age Gap (bottom panel) for the AUD dataset.
Fig. 3. Distribution of the individual values of Corrected Brain Age against Chronological Age (left panels) and Corrected Brain Age Gap (right panels) for the LoFHD group (top panels) and the HiFHD group (bottom panels). The individual values (marked in dots) above the diagonal line in the left panels represent increased brain age while the values below the diagonal line represent decreased brain age.
Fig. 4. Bar Graphs showing the mean values for Chronological Age, Corrected Brain Age, and Uncorrected Brain Age (top panel) as well as for Brain Age Gap (bottom panel) for the FHD dataset.
Pearson correlations among the brain age measures in the AUD dataset. The correlation values in the lower triangle represent the control group and those in the upper triangle represent the AUD group.

+1

Prediction of brain age in individuals with and at risk for alcohol use disorder using brain morphological features
  • Preprint
  • File available

March 2024

·

67 Reads

Brain age measures predicted from structural and functional brain features are increasingly being used to understand brain integrity, disorders, and health. While there is a vast literature showing aberrations in both structural and functional brain measures in individuals with and at risk for alcohol use disorder (AUD), few studies have investigated brain age in these groups. The current study examines brain age measures predicted using brain morphological features, such as cortical thickness and brain volume, in individuals with a lifetime diagnosis of AUD as well as in those at higher risk to develop AUD from families with multiple members affected with AUD (i.e., higher family history density (FHD) scores). The AUD dataset included a group of 30 adult males (mean age = 41.25 years) with a lifetime diagnosis of AUD and currently abstinent and a group of 30 male controls (mean age = 27.24 years) without any history of AUD. A second dataset of young adults who were categorized based on their FHD scores comprised a group of 40 individuals (20 males) with high FHD of AUD (mean age = 25.33 years) and a group of 31 individuals (18 males) with low FHD (mean age = 25.47 years). Brain age was predicted using 187 brain morphological features of cortical thickness and brain volume in an XGBoost regression model; a bias-correction procedure was applied to the predicted brain age. Results showed that both AUD and high FHD individuals showed an increase of 1.70 and 0.09 years (1.08 months), respectively, in their brain age relative to their chronological age, suggesting accelerated brain aging in AUD and risk for AUD. Increased brain age was associated with poor performance on neurocognitive tests of executive functioning in both AUD and high FHD individuals, indicating that brain age can also serve as a proxy for cognitive functioning and brain health. These findings on brain aging in these groups may have important implications for the prevention and treatment of AUD and ensuing cognitive decline.

Download

Fig. 1. Area under the curve analysis: Baseline left DG volume prediction of engagement status.
Hippocampal Subfield Volumes Predict Disengagement from Maintenance Treatment in First Episode Schizophrenia

November 2022

·

102 Reads

·

8 Citations

Schizophrenia Bulletin

Objectives Disengagement from treatment is common in first episode schizophrenia (FES) and is associated with poor outcomes. Our aim was to determine whether hippocampal subfield volumes predict disengagement during maintenance treatment of FES. Methods FES patients were recruited from sites in Boston, New York, Shanghai, and Changsha. After stabilization on antipsychotic medication, participants were randomized to add-on citalopram or placebo and followed for 12 months. Demographic, clinical and cognitive factors at baseline were compared between completers and disengagers in addition to volumes of hippocampal subfields. Results Baseline data were available for 95 randomized participants. Disengagers (n = 38, 40%) differed from completers (n = 57, 60%) by race (more likely Black; less likely Asian) and in more alcohol use, parkinsonism, negative symptoms and more impairment in visual learning and working memory. Bilateral dentate gyrus (DG), CA1, CA2/3 and whole hippocampal volumes were significantly smaller in disengagers compared to completers. When all the eight volumes were entered into the model simultaneously, only left DG volume significantly predicted disengagement status and remained significant after adjusting for age, sex, race, intracranial volume, antipsychotic dose, duration of untreated psychosis, citalopram status, alcohol status, and smoking status (P < .01). Left DG volume predicted disengagement with 57% sensitivity and 83% specificity. Conclusions Smaller left DG was significantly associated with disengagement status over 12 months of maintenance treatment in patients with FES participating in a randomized clinical trial. If replicated, these findings may provide a biomarker to identify patients at risk for disengagement and a potential target for interventions.


Figure 4. The distribution of minimal depth among the trees of the forest for the significant variables is shown in different colors for each level of minimal depth. The mean minimal depth in the distribution for each variable is marked by a vertical black bar overlapped by a value label inside a box. Based on the mean minimal depth values, the importance list comprised 2 BIS scores, 13 FC, and 1 neuropsychological score, which contributed to the RF classification of AUD and CTL individuals. The lower mean minimal depth of a feature represents a higher number of observations (participants) categorized in a specific group based on the feature. The number of trees for a feature represents the total number of decision trees in which a split occurs on the feature (see Table 2 for details about the ROI numbers (1-34) represented in the FC variables).
Demographic and clinical characteristics of the sample.
Brain regions of interest (ROI) analyzed for reward network functional connectivity.
Cont.
Comparison of neuropsychological variables between AUD or CTL group using one- way ANOVA.
Differentiating Individuals with and without Alcohol Use Disorder Using Resting-State fMRI Functional Connectivity of Reward Network, Neuropsychological Performance, and Impulsivity Measures

April 2022

·

149 Reads

·

12 Citations

Behavioral Sciences

Individuals with alcohol use disorder (AUD) may manifest an array of neural and behavioral abnormalities, including altered brain networks, impaired neurocognitive functioning, and heightened impulsivity. Using multidomain measures, the current study aimed to identify specific features that can differentiate individuals with AUD from healthy controls (CTL), utilizing a random forests (RF) classification model. Features included fMRI-based resting-state functional connectivity (rsFC) across the reward network, neuropsychological task performance, and behavioral impulsivity scores, collected from thirty abstinent adult males with prior history of AUD and thirty CTL individuals without a history of AUD. It was found that the RF model achieved a classification accuracy of 86.67% (AUC = 93%) and identified key features of FC and impulsivity that significantly contributed to classifying AUD from CTL individuals. Impulsivity scores were the topmost predictors, followed by twelve rsFC features involving seventeen key reward regions in the brain, such as the ventral tegmental area, nucleus accumbens, anterior insula, anterior cingulate cortex, and other cortical and subcortical structures. Individuals with AUD manifested significant differences in impulsivity and alterations in functional connectivity relative to controls. Specifically, AUD showed heightened impulsivity and hypoconnectivity in nine connections across 13 regions and hyperconnectivity in three connections involving six regions. Relative to controls, visuo-spatial short-term working memory was also found to be impaired in AUD. In conclusion, specific multidomain features of brain connectivity, impulsivity, and neuropsychological performance can be used in a machine learning framework to effectively classify AUD individuals from healthy controls.


Figure 3. Boxplots of the 1st principal component scores in the AUD and control groups. The scores were significantly different (p < 0.02) between groups, as determined by a nonparametric Mann-Whitney U test.
Significant clusters identified during successful NoGo BOLD responses and their MNI coordinates, sizes in mm 3 , locations at the peak/trough, and activation directions.
Cont.
Statistical Nonparametric fMRI Maps in the Analysis of Response Inhibition in Abstinent Individuals with History of Alcohol Use Disorder

April 2022

·

45 Reads

·

3 Citations

Behavioral Sciences

Inhibitory impairments may persist after abstinence in individuals with alcohol use disorder (AUD). Using traditional statistical parametric mapping (SPM) fMRI analysis, which requires data to satisfy parametric assumptions often difficult to satisfy in biophysical system as brain, studies have reported equivocal findings on brain areas responsible for response inhibition, and activation abnormalities during inhibition found in AUD persist after abstinence. Research is warranted using newer analysis approaches. fMRI scans were acquired during a Go/NoGo task from 30 abstinent male AUD and 30 healthy control participants with the objectives being (1) to characterize neuronal substrates associated with response inhibition using a rigorous nonparametric permutation-based fMRI analysis and (2) to determine whether these regions were differentially activated between abstinent AUD and control participants. A blood oxygen level dependent contrast analysis showed significant activation in several right cortical regions and deactivation in some left cortical regions during successful inhibition. The largest source of variance in activation level was due to group differences. The findings provide evidence of cortical substrates employed during response inhibition. The largest variance was explained by lower activation in inhibition as well as ventral attentional cortical networks in abstinent individuals with AUD, which were not found to be associated with length of abstinence, age, or impulsiveness.


A New Approach to Symmetric Registration of Longitudinal Structural MRI of the Human Brain

March 2022

·

5 Reads

·

10 Citations

Journal of Neuroscience Methods

Background This paper presents the Automatic Temporal Registration Algorithm (ATRA) for symmetric rigid-body registration of longitudinal T1-weighted three-dimensional MRI scans of the human brain. This is a fundamental processing step in computational neuroimaging. New Method The notion of leave-one-out consistent (LOOC) landmarks with respect to a supervised landmark detection algorithm is introduced. An automatic algorithm is presented for identification of LOOC landmarks on MRI scans. Multiple sets of LOOC landmarks are identified on each volume and a Generalized Orthogonal Procrustes Analysis of the landmarks is used to find a rigid-body transformation of each volume into a common space where the volumes are aligned precisely. Results Qualitative and quantitative evaluations of ATRA registration accuracy were performed using 2012 volumes from 503 subjects (4 longitudinal volumes/subject), and on a further 120 volumes acquired from 3 normal subjects (40 longitudinal volumes/subject). Since the ground truth registrations are unknown, we devised a novel method for showing that ATRA’s registration accuracy is at least better than 0.5▒mm translation or 0.5∘ rotation. Comparison with Existing Method(s) In comparison with existing methods, ATRA does not require any image preprocessing (e.g., skull-stripping or intensity normalization) and can handle conditions where rigid-body motion assumptions are not true (e.g., movement in eyes, jaw, neck) and brain tissue loss over time in neurodegenerative diseases. In a systematic comparison with the FSL FLIRT algorithm, ATRA provided faster and more accurate registrations. Conclusions The algorithm is symmetric, in the sense that any permutation of the input volumes does not change the transformation matrices, and unbiased, in that all volumes undergo exactly one interpolation operation, which precisely aligns them in a common space. There is no interpolation bias and no reference volume. All volumes are treated exactly the same. The algorithm is fast and highly accurate.


Fig. 1. Cohort selection flowchart.
Fig. 3. (Continued)
Fig. 4. Variable importance of dementia risk prediction in MCI: (a) Xgboost PH model, (b) Cox PH model.
Survival Analysis in Cognitively Normal Subjects and in Patients with Mild Cognitive Impairment Using a Proportional Hazards Model with Extreme Gradient Boosting Regression

December 2021

·

480 Reads

·

11 Citations

Background: Evaluating the risk of Alzheimer's disease (AD) in cognitively normal (CN) and patients with mild cognitive impairment (MCI) is extremely important. While MCI-to-AD progression risk has been studied extensively, few studies estimate CN-to-MCI conversion risk. The Cox proportional hazards (PH), a widely used survival analysis model, assumes a linear predictor-risk relationship. Generalizing the PH model to more complex predictor-risk relationships may increase risk estimation accuracy. Objective: The aim of this study was to develop a PH model using an Xgboost regressor, based on demographic, genetic, neuropsychiatric, and neuroimaging predictors to estimate risk of AD in patients with MCI, and the risk of MCI in CN subjects. Methods: We replaced the Cox PH linear model with an Xgboost regressor to capture complex interactions between predictors, and non-linear predictor-risk associations. We endeavored to limit model inputs to noninvasive and more widely available predictors in order to facilitate future applicability in a wider setting. Results: In MCI-to-AD (n = 882), the Xgboost model achieved a concordance index (C-index) of 84.5%. When the model was used for MCI risk prediction in CN (n = 100) individuals, the C-index was 73.3%. In both applications, the C-index was statistically significantly higher in the Xgboost in comparison to the Cox PH model. Conclusion: Using non-linear regressors such as Xgboost improves AD dementia risk assessment in CN and MCI. It is possible to achieve reasonable risk stratification using predictors that are relatively low-cost in terms of time, invasiveness, and availability. Future strategies for improving AD dementia risk estimation are discussed.


FIGURE 1 | Inclusion diagram.
The Neutrophil to Lymphocyte Ratio Is Associated With the Risk of Subsequent Dementia in the Framingham Heart Study

November 2021

·

41 Reads

·

30 Citations

Frontiers in Aging Neuroscience

Objective: Active neutrophils are important contributors to Alzheimer’s disease (AD) pathology through the formation of capillary stalls that compromise cerebral blood flow (CBF) and through aberrant neutrophil signaling that advances disease progression. The neutrophil to lymphocyte ratio (NLR) is a proxy of neutrophil-mediated inflammation, and higher NLR is found in persons diagnosed with clinical AD. The objective of this study was to investigate whether increased NLR in older adults is independently associated with the risk of subsequent dementia. Methods: We examined associations of baseline NLR with incident dementia risk in the community-based Framingham Heart Study (FHS) longitudinal cohorts. The association between NLR and risk of dementia was evaluated using the cumulative incidence function (CIF) and inverse probability-weighted Cox proportional cause-specific hazards regression models, with adjustment for age, sex, body mass index (BMI), systolic and diastolic blood pressure, diabetes, current smoking status, low-density lipoprotein (LDH), high-density lipoprotein (LDL), total cholesterol, triglycerides, and history of cardiovascular disease (CVD). Random forest survival models were used to evaluate the relative predictive value of the model covariates on dementia risk. Results: The final study sample included 1,648 participants with FHS (average age, 69 years; 56% women). During follow-up (median, 5.9 years), we observed 51 cases of incident dementia, of which 41 were AD cases. Results from weighted models suggested that the NLR was independently associated with incident dementia, and it was preceded in predictive value only by age, history of CVD, and blood pressure at baseline. Conclusion: Our study shows that individuals with higher NLR are at a greater risk of subsequent dementia during a 5.9-year follow-up period. Further evaluating the role of neutrophil-mediated inflammation in AD progression may be warranted.


Citations (79)


... 37,38 DIR integrates structural data from DTI with functional connectivity from resting-state fMRI, revealing how brain structure influences function, while DTI provides detailed maps of white matter tracts, aiding in understanding the physical pathways underlying functional connections. 39 DIR identifies functional networks and integrates them with structural connectivity data, while DTI visualizes white matter tracts and quantifies microstructural properties, providing the structural basis of functional networks. Both techniques offer a comprehensive approach to brain mapping, with DIR simultaneously assessing structural and functional connectivity and DTI widely used in clinical research to investigate neurological conditions and disease progression. ...

Reference:

Future Health Comparative analysis of double inversion recovery (DIR) and diffusion tensor imaging (DTI) sequences for the detection of brain white matter diseases
Impact of Diffusion Tensor Field Smoothing with Log-Cholesky Metric on Noise Reduction
  • Citing Article
  • July 2024

Frontiers in Biomedical Technologies

... prodromal to chronic stages of the illness and that the CA1 may be a target subregion affected by the illness in the prodromal and early stages (29,30). Several previous studies have reported the associations of structural changes in hippocampal subregions with antipsychotic treatment in patients with first-episode psychosis (31)(32)(33). However, the association with clozapine treatment is still uncertain, mainly due to a very limited number of previous studies on this issue. ...

Hippocampal Subfield Volumes Predict Disengagement from Maintenance Treatment in First Episode Schizophrenia

Schizophrenia Bulletin

... As shown in Figure 3b, the average age of the participants in the datasets ranges mostly between 40 and 50 years. Figure 3c shows a predominant male representation, with 10 studies (~26%) exclusively using male datasets (Bordier et al. 2022;Deng et al. 2022;Duan et al. 2022;Kamarajan et al. 2022;Kamarajan et al. 2020;Kim et al. 2017;Song et al. 2021;Wang, Fan, et al. 2018;Wang, Zhao, et al. 2018;Yang et al. 2021). Table 3 provides an overview of the studies' most commonly used analytical methods. ...

Differentiating Individuals with and without Alcohol Use Disorder Using Resting-State fMRI Functional Connectivity of Reward Network, Neuropsychological Performance, and Impulsivity Measures

Behavioral Sciences

... Mounting evidence highlights a close relationship between activated platelets and the risk of AD. A cohort longitudinal study illustrated that subjects free of antiplatelet treatment followed-up for 20 years are at higher risk for the development of AD [61]. Many studies indicate that activated platelets promote the conversion of soluble Aβ to the neurotoxic amyloid plaques [62,63]. ...

Platelet Function Is Associated With Dementia Risk in the Framingham Heart Study

Journal of the American Heart Association

... Amongst these are the FSL (Jenkinson & Smith, 2001;Jenkinson et al., 2002), FreeSurfer (Reuter et al., 2010), and ANTS (Avants et al., 2008). Another recently develop method is the automatic temporal registration algorithm (ATRA) by (Ardekani, 2022), which is landmark-based with some important advantages over the other methods, e.g., the ability to register dozens of images simultaneously in an inverse-consistent and unbiased manner. ...

A New Approach to Symmetric Registration of Longitudinal Structural MRI of the Human Brain
  • Citing Article
  • March 2022

Journal of Neuroscience Methods

... To estimate the probability of AD in individuals with MCI, Khajehpiri et al. created a proportional hazards model with an Xgboost regressor based on demographic, cognitive, genetic, and neuroimaging features. They achieved a concordance index (C-index) of 0.845 [18]. Sarica et al. used XST, Conditional Survival Forests (CSF), and RSF to forecast AD progression, based on clinical, neuroimaging, and demographic data. ...

Survival Analysis in Cognitively Normal Subjects and in Patients with Mild Cognitive Impairment Using a Proportional Hazards Model with Extreme Gradient Boosting Regression

... A growing number of studies from large American, Chinese, and UK cohorts provide stark evidence for elevated neutrophil-and monocyte-to-lymphocyte ratios in the persons at greater risk for dementia. These studies broadly implicate the NLR and GLR in the development of dementia over 6-8 years evidenced by odds ratios ranging from 1.12 to 1.50 (Chou et al., 2023;Ramos-Cejudo et al., 2021;Zhang et al., 2022). These ratios might provide some insight into the oxidative, phagocytic, and inflammatory byproducts and processes involved in neutrophil expansion ; Sayed et al., 2020;R. ...

The Neutrophil to Lymphocyte Ratio Is Associated With the Risk of Subsequent Dementia in the Framingham Heart Study

Frontiers in Aging Neuroscience

... Instead, in CHR, only the hippocampal tail was smaller compared to CTRs, especially in those not converting to psychosis. This might mean that volume reduction in tail of CHR is not specific to psychosis or it might be related to the effect of antidepressants 55,56 However, there was no association between the use of antidepressants and tail volume in the patient groups. Also, it is possible that some CHR-NC transition to psychosis only after the one-year follow-up period. ...

Effect of citalopram on hippocampal volume in first-episode schizophrenia: Structural MRI results from the DECIFER trial
  • Citing Article
  • April 2021

Psychiatry Research Neuroimaging

... The findings suggest that low doses of these SGAs are effective for rapid symptom reduction in early psychosis, underscoring the need for further research on long-term outcomes and side effects. Wang et al. (2021) explored the effects of aripiprazole in 95 patients with FES on hippocampal volume and inflammation in early schizophrenia through a two-part study: a 12-month trial comparing aripiprazole with other SGAs in first-episode schizophrenia (FES) patients, and an analysis of inflammation and oxidative stress markers in medication-naive schizophrenia patients beginning antipsychotic treatment. Aripiprazole was associated with a slight increase in hippocampal volume, unlike the decrease seen with other antipsychotics, and was linked to reduced levels of certain inflammatory markers, suggesting its potential protective role against hippocampal atrophy and associated inflammation. ...

Association of Aripiprazole With Reduced Hippocampal Atrophy During Maintenance Treatment of First-Episode Schizophrenia
  • Citing Article
  • March 2021

Journal of Clinical Psychopharmacology

... We have summarized the critical acquisition parameters for the datasets in Table 1 . For each image, we set the origin (i.e., (0,0,0) coordinate) to correspond to the anterior commissure (AC) using acpcdetect v2 ( Ardekani, 2018 ;Ardekani et al., 1997 ;Ardekani and Bachman, 2009 ) (available at: https://www.nitrc.org/projects/art/ ). ...

A New Approach to Symmetric Registration of Longitudinal Structural MRI of the Human Brain