Baseline MRI predictors of conversion from MCI to probable AD in the ADNI cohort

IU Center for Neuroimaging, Division of Imaging Sciences, Department of Radiology, Indiana University School of Medicine, 950 W Walnut St, R2 E124, Indianapolis, IN 46202, USA.
Current Alzheimer research (Impact Factor: 3.89). 08/2009; 6(4):347-61. DOI: 10.2174/156720509788929273
Source: PubMed


The Alzheimer's Disease Neuroimaging Initiative (ADNI) is a multi-center study assessing neuroimaging in diagnosis and longitudinal monitoring. Amnestic Mild Cognitive Impairment (MCI) often represents a prodromal form of dementia, conferring a 10-15% annual risk of converting to probable AD. We analyzed baseline 1.5T MRI scans in 693 participants from the ADNI cohort divided into four groups by baseline diagnosis and one year MCI to probable AD conversion status to identify neuroimaging phenotypes associated with MCI and AD and potential predictive markers of imminent conversion. MP-RAGE scans were analyzed using publicly available voxel-based morphometry (VBM) and automated parcellation methods. Measures included global and hippocampal grey matter (GM) density, hippocampal and amygdalar volumes, and cortical thickness values from entorhinal cortex and other temporal and parietal lobe regions. The overall pattern of structural MRI changes in MCI (n=339) and AD (n=148) compared to healthy controls (HC, n=206) was similar to prior findings in smaller samples. MCI-Converters (n=62) demonstrated a very similar pattern of atrophic changes to the AD group up to a year before meeting clinical criteria for AD. Finally, a comparison of effect sizes for contrasts between the MCI-Converters and MCI-Stable (n=277) groups on MRI metrics indicated that degree of neurodegeneration of medial temporal structures was the best antecedent MRI marker of imminent conversion, with decreased hippocampal volume (left > right) being the most robust. Validation of imaging biomarkers is important as they can help enrich clinical trials of disease modifying agents by identifying individuals at highest risk for progression to AD.

Download full-text


Available from: John D West,
50 Reads
    • "The brain regions and correlations listed in Table 6 have been reported to have abnormal alterations in MCI and AD patients, as well as in AD conversion process, indicating our findings are consistent with previous publications, such as studies in following regions: parahippocampal gyri, lingual gyri and cingulum cortex [60] [61] [62] [63] [64], insula [65] [66], inferior temporal gyri and superior temporal gyri [63] [67] [68], cuneus, parietal lobule, precuneus cortex, postcentral gyri and cingulate cortex [33, 68–71], supramarginal gyri, angular gyri, temporal lobe, and occipital cortex [66] [68] [72]. In addition, the correlations mentioned above are located either in the same hemisphere, or widely spread over the whole brain, suggesting the abnormalities caused by MCI and dementia have affected the entire brain rather than certain areas. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Brain network occupies an important position in representing abnormalities in Alzheimer's disease (AD) and mild cognitive impairment (MCI). Currently, most studies only focused on morphological features of regions of interest without exploring the interregional alterations. In order to investigate the potential discriminative power of a morphological network in AD diagnosis and to provide supportive evidence on the feasibility of an individual structural network study, we propose a novel approach of extracting the correlative features from magnetic resonance imaging, which consists of a two-step approach for constructing an individual thickness network with low computational complexity. Firstly, multi-distance combination is utilized for accurate evaluation of between-region dissimilarity; and then the dissimilarity is transformed to connectivity via calculation of correlation function. An evaluation of the proposed approach has been conducted with 189 normal controls, 198 MCI subjects, and 163 AD patients using machine learning techniques. Results show that the observed correlative feature suggests significant promotion in classification performance compared with cortical thickness, with accuracy of 89.88% and area of 0.9588 under receiver operating characteristic curve. We further improved the performance by integrating both thickness and apolipoprotein E ɛ4 allele information with correlative features. New achieved accuracies are 92.11% and 79.37% in separating AD from normal controls and AD converters from non-converters, respectively. Differences between using diverse distance measurements and various correlation transformation functions are also discussed to explore an optimal way for network establishment.
    Journal of Alzheimer's disease: JAD 10/2015; DOI:10.3233/JAD-150311 · 4.15 Impact Factor
  • Source
    • "pdf . et al. 2008 ; Li et al. 2012 ; Misra et al. 2009; Risacher et al. 2009 ; Wang et al. 2011b ) , and used off the - shelf machine learning tools to discriminate MCI - C from MCI - NC . However , those methods were short of high performance for clinical use . "
    [Show abstract] [Hide abstract]
    ABSTRACT: As the early stage of Alzheimer's disease (AD), mild cognitive impairment (MCI) has high chance to convert to AD. Effective prediction of such conversion from MCI to AD is of great importance for early diagnosis of AD and also for evaluating AD risk pre-symptomatically. Unlike most previous methods that used only the samples from a target domain to train a classifier, in this paper, we propose a novel multimodal manifold-regularized transfer learning (M2TL) method that jointly utilizes samples from another domain (e.g., AD vs. normal controls (NC)) as well as unlabeled samples to boost the performance of the MCI conversion prediction. Specifically, the proposed M2TL method includes two key components. The first one is a kernel-based maximum mean discrepancy criterion, which helps eliminate the potential negative effect induced by the distributional difference between the auxiliary domain (i.e., AD and NC) and the target domain (i.e., MCI converters (MCI-C) and MCI non-converters (MCI-NC)). The second one is a semi-supervised multimodal manifold-regularized least squares classification method, where the target-domain samples, the auxiliary-domain samples, and the unlabeled samples can be jointly used for training our classifier. Furthermore, with the integration of a group sparsity constraint into our objective function, the proposed M2TL has a capability of selecting the informative samples to build a robust classifier. Experimental results on the Alzheimer's Disease Neuroimaging Initiative (ADNI) database validate the effectiveness of the proposed method by significantly improving the classification accuracy of 80.1 % for MCI conversion prediction, and also outperforming the state-of-the-art methods.
    Brain Imaging and Behavior 02/2015; DOI:10.1007/s11682-015-9356-x · 4.60 Impact Factor
  • Source
    • "For hypothesis testing, GWR was analyzed with a vertex-byvertex general linear model (GLM) between conversion groups adjusting for apolipoprotein ε4 (APOE4) carrier status (i.e., positive defined as e2/e4, e3/e4, e4/e4 versus negative defined as e2/e2, e2/e3, e3/e3 (Yip et al. 2005)), scanner type (Han et al. 2006), cortical thickness (i.e., a marker of brain aging (Risacher et al. 2009)), and hippocampal volume (i.e., a marker of neurodegeneration (Braak et al. 2006)). To assess the stability of our observations, a Monte Carlo simulation technique was applied using 5000 iterations of repeated random sampling. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The clinical relevance of gray/white matter contrast ratio (GWR) in mild cognitive impairment (MCI) remains unknown. This study examined baseline GWR and 3-year follow-up diagnostic status in MCI. Alzheimer's Disease Neuroimaging Initiative MCI participants with baseline 1.5 T MRI and 3-year follow-up clinical data were included. Participants were categorized into two groups based on 3-year follow-up diagnoses: 1) non-converters (n = 69, 75 ± 7, 26 % female), and 2) converters (i.e., dementia at follow-up; n = 69, 75 ± 7, 30 % female) who were matched on baseline age and Mini-Mental State Examination scores. Groups were compared on FreeSurfer generated baseline GWR from structural images in which higher values represent greater tissue contrast. A general linear model, adjusting for APOE-status, scanner type, hippocampal volume, and cortical thickness, revealed that converters evidenced lower GWR values than non-converters (i.e., more degradation in tissue contrast; p = 0.03). Individuals with MCI who convert to dementia have lower baseline GWR values than individuals who remain diagnostically stable over a 3-year period, statistically independent of cortical thickness or hippocampal volume.
    Brain Imaging and Behavior 06/2014; 9(2). DOI:10.1007/s11682-014-9291-2 · 4.60 Impact Factor
Show more