Inter-modality relationship constrained multi-modality multi-task feature selection for Alzheimer's Disease and mild cognitive impairment identification

Abstract

Previous studies have demonstrated that the use of integrated information from multi-modalities could significantly improve diagnosis of Alzheimer's Disease (AD). However, feature selection, which is one of the most important steps in classification, is typically performed separately for each modality, which ignores the potentially strong inter-modality relationship within each subject. Recent emergence of multi-task learning approach makes the joint feature selection from different modalities possible. However, joint feature selection may unfortunately overlook different yet complementary information conveyed by different modalities. We propose a novel multi-task feature selection method to preserve the complementary inter-modality information. Specifically, we treat feature selection from each modality as a separate task and further impose a constraint for preserving the inter-modality relationship, besides separately enforcing the sparseness of the selected features from each modality. After feature selection, a multi-kernel support vector machine (SVM) is further used to integrate the selected features from each modality for classification. Our method is evaluated using the baseline PET and MRI images of subjects obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. Our method achieves a good performance, with an accuracy of 94.37% and an area under the ROC curve (AUC) of 0.9724 for AD identification, and also an accuracy of 78.80% and an AUC of 0.8284 for mild cognitive impairment (MCI) identification. Moreover, the proposed method achieves an accuracy of 67.83% and an AUC of 0.6957 for separating between MCI converters and MCI non-converters (to AD). These performances demonstrate the superiority of the proposed method over the state-of-the-art classification methods.

Keywords: Alzheimer's Disease; Inter-modality relationship; Mild cognitive impairment; Multi-kernel support vector machine; Multi-task feature selection.

Fig. 1

Schematic diagram illustrating the proposed classification framework.

Fig. 2

Comparison of different feature selection models. Fig. 2A shows the model constructed by the L₁ norm, where feature selection is performed independently on each individual modality. Fig. 2B shows the model constructed by the L_2,1 norm, where a common set of features is selected from all modalities. Fig. 2C shows the proposed feature selection model, where, in additon to the common features selected across all modalities, the complementary information conveyed by the modality-specific features is also effectively preserved.

Fig. 3

ROC curves of different methods for classification of AD (left) and MCI (right).

Fig. 4

ROC curves of different methods for classification of MCI subgroups.

Fig. 5

Top ten most discriminative PET regions in the MCI subgroup classification. Of note, different colors in the figure just indicate different brain regions.

Fig. 6

Top ten most discriminative MRI regions in the MCI subgroup classification. Of note, different colors in the figure just indicate different brain regions.

Fig. 7

Performances of four different methods on a longitudinal dataset. Our proposed method achieves the best results, and the performances of all four methods are improved for the later scans since the difference between MCI converters and MCI non-converters becomes larger and larger with aging.