Martin Dyrba

Multimodal analysis of functional and structural disconnection in Alzheimer's disease using multiple kernel SVM

Alzheimers disease (AD) patients exhibit alterations in the functional connectivity between spatially segregated brain regions which may be related to both local gray matter (GM) atrophy as well as a decline in the fiber integrity of the underlying white matter tracts. Machine learning algorithms are able to automatically detect the patterns of the disease in image data, and therefore, constitute a suitable basis for automated image diagnostic systems. The question of which magnetic resonance imaging (MRI) modalities are most useful in a clinical context is as yet unresolved. We examined multimodal MRI data acquired from 28 subjects with clinically probable AD and 25 healthy controls. Specifically, we used fiber tract integrity as measured by diffusion tensor imaging (DTI), GM volume derived from structural MRI, and the graph‐theoretical measures ‘local clustering coefficient’ and ‘shortest path length’ derived from resting‐state functional MRI (rs‐fMRI) to evaluate the utility of the three imaging methods in automated multimodal image diagnostics, to assess their individual performance, and the level of concordance between them. We ran the support vector machine (SVM) algorithm and validated the results using leave‐one‐out cross‐validation. For the single imaging modalities, we obtained an area under the curve (AUC) of 80% for rs‐fMRI, 87% for DTI, and 86% for GM volume. When it came to the multimodal SVM, we obtained an AUC of 82% using all three modalities, and 89% using only DTI measures and GM volume. Combined multimodal imaging data did not significantly improve classification accuracy compared to the best single measures alone. Hum Brain Mapp 36:2118–2131, 2015.

Robust automated detection of microstructural white matter degeneration in Alzheimer's disease using machine learning classification of multicenter DTI data.

Diffusion tensor imaging (DTI) based assessment of white matter fiber tract integrity can support the diagnosis of Alzheimer’s disease (AD). The use of DTI as a biomarker, however, depends on its applicability in a multicenter setting accounting for effects of different MRI scanners. We applied multivariate machine learning (ML) to a large multicenter sample from the recently created framework of the European DTI study on Dementia (EDSD). We hypothesized that ML approaches may amend effects of multicenter acquisition. We included a sample of 137 patients with clinically probable AD (MMSE 20.6±5.3) and 143 healthy elderly controls, scanned in nine different scanners. For diagnostic classification we used the DTI indices fractional anisotropy (FA) and mean diffusivity (MD) and, for comparison, gray matter and white matter density maps from anatomical MRI. Data were classified using a Support Vector Machine (SVM) and a Naïve Bayes (NB) classifier. We used two cross-validation approaches, (i) test and training samples randomly drawn from the entire data set (pooled cross-validation) and (ii) data from each scanner as test set, and the data from the remaining scanners as training set (scanner-specific cross-validation). In the pooled cross-validation, SVM achieved an accuracy of 80% for FA and 83% for MD. Accuracies for NB were significantly lower, ranging between 68% and 75%. Removing variance components arising from scanners using principal component analysis did not significantly change the classification results for both classifiers. For the scanner-specific cross-validation, the classification accuracy was reduced for both SVM and NB. After mean correction, classification accuracy reached a level comparable to the results obtained from the pooled cross-validation. Our findings support the notion that machine learning classification allows robust classification of DTI data sets arising from multiple scanners, even if a new data set comes from a scanner that was not part of the training sample.

Cortical thinning and its relation to cognition in amyotrophic lateral sclerosis

Clinical, genetic, and pathological findings suggest a close relationship between amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). We studied the patterns of cortical atrophy across the spectrum between ALS and ALS-FTD. A surface-based morphometry analysis based on an age- and sex-matched sample of 81 ALS patients and 62 healthy control subjects (HC) was conducted. In addition, we used an age-matched subsample of 57 ALS patients and 31 HC to compare cortical thickness between 3 groups of neuropsychologically characterized ALS patients: (1) cognitively unimpaired; (2) cognitively impaired; and (3) ALS-FTD patients. Compared with HC, the entire sample of patients demonstrated cortical thinning in the bilateral precentral gyrus, right precuneus, and right frontal and temporal lobes. ALS-FTD patients showed cortical thinning in regions including the frontal and temporal gyri and the posterior cingulate cortex. Cognitively impaired ALS patients showed cortical thinning in regions largely overlapping with those found in ALS-FTD, but changes were less widespread. In conclusion, the cognitive status of ALS subjects is associated with different patterns of cortical atrophy.

Fractional Anisotropy Changes in Alzheimer's Disease Depend on the Underlying Fiber Tract Architecture: A Multiparametric DTI Study using Joint Independent Component Analysis

Diffusion tensor imaging (DTI) allows the simultaneous measurement of several diffusion indices that provide complementary information on the substrate of white matter alterations in neurodegenerative diseases. These indices include fractional anisotropy (FA) as measure of fiber tract integrity, and the mode of anisotropy (Mode) reflecting differences in the shape of the diffusion tensor. We used a multivariate approach based on joint independent component analysis of FA and Mode in a large sample of 138 subjects with Alzheimers disease (AD) dementia, 37 subjects with cerebrospinal fluid biomarker positive mild cognitive impairment (MCI-AD), and 153 healthy elderly controls from the European DTI Study on Dementia to comprehensively study alterations of microstructural white matter integrity in AD dementia and predementia AD. We found a parallel decrease of FA and Mode in intracortically projecting fiber tracts, and a parallel increase of FA and Mode in the corticospinal tract in AD patients compared to controls. Subjects with MCI-AD showed a similar, but spatially more restricted pattern of diffusion changes. Our findings suggest an early axonal degeneration in intracortical projecting fiber tracts in dementia and predementia stages of AD. An increase of Mode, parallel to an increase of FA, in the corticospinal tract suggests a more linear shape of diffusion due to loss of crossing fibers along relatively preserved cortico-petal and cortico-fugal fiber tracts in AD. Supporting this interpretation, we found three populations of fiber tracts, namely cortico-petal and cortico-fugal, commissural, and intrahemispherically projecting fiber tracts, in the peak area of parallel FA and Mode increase.

Predicting Prodromal Alzheimer's Disease in Subjects with Mild Cognitive Impairment Using Machine Learning Classification of Multimodal Multicenter Diffusion‐Tensor and Magnetic Resonance Imaging Data

Alzheimers disease (AD) patients show early changes in white matter (WM) structural integrity. We studied the use of diffusion tensor imaging (DTI) in assessing WM alterations in the predementia stage of mild cognitive impairment (MCI).

Measuring cortical connectivity in Alzheimer's disease as a brain neural network pathology: Toward clinical applications

OBJECTIVES The objective was to review the literature on diffusion tensor imaging as well as resting-state functional magnetic resonance imaging and electroencephalography (EEG) to unveil neuroanatomical and neurophysiological substrates of Alzheimers disease (AD) as a brain neural network pathology affecting structural and functional cortical connectivity underlying human cognition. METHODS We reviewed papers registered in PubMed and other scientific repositories on the use of these techniques in amnesic mild cognitive impairment (MCI) and clinically mild AD dementia patients compared to cognitively intact elderly individuals (Controls). RESULTS Hundreds of peer-reviewed (cross-sectional and longitudinal) papers have shown in patients with MCI and mild AD compared to Controls (1) impairment of callosal (splenium), thalamic, and anterior-posterior white matter bundles; (2) reduced correlation of resting state blood oxygen level-dependent activity across several intrinsic brain circuits including default mode and attention-related networks; and (3) abnormal power and functional coupling of resting state cortical EEG rhythms. Clinical applications of these measures are still limited. CONCLUSIONS Structural and functional (in vivo) cortical connectivity measures represent a reliable marker of cerebral reserve capacity and should be used to predict and monitor the evolution of AD and its relative impact on cognitive domains in pre-clinical, prodromal, and dementia stages of AD.

Robust Detection of Impaired Resting State Functional Connectivity Networks in Alzheimer's Disease Using Elastic Net Regularized Regression

The large number of multicollinear regional features that are provided by resting state (rs) fMRI data requires robust feature selection to uncover consistent networks of functional disconnection in Alzheimers disease (AD). Here, we compared elastic net regularized and classical stepwise logistic regression in respect to consistency of feature selection and diagnostic accuracy using rs-fMRI data from four centers of the “German resting-state initiative for diagnostic biomarkers” (psymri.org), comprising 53 AD patients and 118 age and sex matched healthy controls. Using all possible pairs of correlations between the time series of rs-fMRI signal from 84 functionally defined brain regions as the initial set of predictor variables, we calculated accuracy of group discrimination and consistency of feature selection with bootstrap cross-validation. Mean areas under the receiver operating characteristic curves as measure of diagnostic accuracy were 0.70 in unregularized and 0.80 in regularized regression. Elastic net regression was insensitive to scanner effects and recovered a consistent network of functional connectivity decline in AD that encompassed parts of the dorsal default mode as well as brain regions involved in attention, executive control, and language processing. Stepwise logistic regression found no consistent network of AD related functional connectivity decline. Regularized regression has high potential to increase diagnostic accuracy and consistency of feature selection from multicollinear functional neuroimaging data in AD. Our findings suggest an extended network of functional alterations in AD, but the diagnostic accuracy of rs-fMRI in this multicenter setting did not reach the benchmark defined for a useful biomarker of AD.

Basal Forebrain and Hippocampus as Predictors of Conversion to Alzheimer's Disease in Patients with Mild Cognitive Impairment - A Multicenter DTI and Volumetry Study

BACKGROUND Hippocampal grey matter (GM) atrophy predicts conversion from mild cognitive impairment (MCI) to Alzheimers disease (AD). Pilot data suggests that mean diffusivity (MD) in the hippocampus, as measured with diffusion tensor imaging (DTI), may be a more accurate predictor of conversion than hippocampus volume. In addition, previous studies suggest that volume of the cholinergic basal forebrain may reach a diagnostic accuracy superior to hippocampal volume in MCI. OBJECTIVE The present study investigated whether increased MD and decreased volume of the hippocampus, the basal forebrain and other AD-typical regions predicted time to conversion from MCI to AD dementia. METHODS 79 MCI patients with DTI and T1-weighted magnetic resonance imaging (MRI) were retrospectively included from the European DTI Study in Dementia (EDSD) dataset. Of these participants, 35 converted to AD dementia after 6-46 months (mean: 21 months). We used Cox regression to estimate the relative conversion risk predicted by MD values and GM volumes, controlling for age, gender, education and center. RESULTS Decreased GM volume in all investigated regions predicted an increased risk for conversion. Additionally, increased MD in the right basal forebrain predicted increased conversion risk. Reduced volume of the right hippocampus was the only significant predictor in a stepwise model combining all predictor variables. CONCLUSION Volume reduction of the hippocampus, the basal forebrain and other AD-related regions was predictive of increased risk for conversion from MCI to AD. In this study, volume was superior to MD in predicting conversion.

The ε4 genotype of apolipoprotein E and white matter integrity in Alzheimer's disease.

In this multicenter study, we investigated a possible association between the APOE ε4 allele and white matter (WM) integrity in Alzheimers disease (AD) using diffusion tensor imaging (DTI).

Combining DTI and MRI for the automated detection of alzheimer’s disease using a large european multicenter dataset

Diffusion tensor imaging (DTI) allows assessing neuronal fiber tract integrity in vivo to support the diagnosis of Alzheimers disease (AD). It is an open research question to which extent combinations of different neuroimaging techniques increase the detection of AD. In this study we examined different methods to combine DTI data and structural T1-weighted magnetic resonance imaging (MRI) data. Further, we applied machine learning techniques for automated detection of AD. We used a sample of 137 patients with clinically probable AD (MMSE 20.6 ±5.3) and 143 healthy elderly controls, scanned in nine different scanners, obtained from the recently created framework of the European DTI study on Dementia (EDSD). For diagnostic classification we used the DTI derived indices fractional anisotropy (FA) and mean diffusivity (MD) as well as grey matter density (GMD) and white matter density (WMD) maps from anatomical MRI. We performed voxel-based classification using a Support Vector Machine (SVM) classifier with tenfold cross validation. We compared the results from each single modality with those from different approaches to combine the modalities. For our sample, combining modalities did not increase the detection rates of AD. An accuracy of approximately 89% was reached for GMD data alone and for multimodal classification when GMD was included. This high accuracy remained stable across each of the approaches. As our sample consisted of mildly to moderately affected patients, cortical atrophy may be far progressed so that the decline in structural network connectivity derived from DTI may not add additional information relevant for the SVM classification. This may be different for predementia stages of AD. Further research will focus on multimodal detection of AD in predementia stages of AD, e.g. in amnestic mild cognitive impairment (aMCI), and on evaluating the classification performance when adding other modalities, e.g. functional MRI or FDG-PET.