Multimodal manifold-regularized transfer learning for MCI conversion prediction

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.

Keywords: Manifold regularization; Mild cognitive impairment conversion; Multimodal classification; Sample selection; Semi-supervised learning; Transfer learning.

Fig. 1

The system diagram of our framework for MCI conversion prediction using the proposed multimodal manifold-regularized transfer learning (M2TL) method

Fig. 2

Comparison of the ROC curves of the proposed M2TL method and the competing methods (DTSVM, LapSVM, and SVM) for MCI-C/MCI-NC classification using multi-modality and single-modality, respectively

Fig. 3

The changes of accuracies of M2TL+SS, M2TL, DTSVM+SS and DTSVM with respect to the used number of samples from the auxiliary domain

Fig. 4

The changes of accuracies of M2TL+SS, M2TL, LapSVM+SS, and LapSVM with respect to the used number of unlabeled samples

Fig. 5

Classification accuracies of M2TL+SS and M2TL methods using a feature selection based on t-test statistics (namely M2TL+SS(+t-test) and M2TL(+t-test)), with respect to the different number of selected features for the multimodal case. Here, `MRI+PET' denotes the selected MRI and PET features. For comparison, the classification accuracies of M2TL+SS and M2TL methods without feature selection are also provided by the two dash lines