Discriminative subnetwork mining for multiple thresholded connectivity-networks-based classification of mild cognitive impairment

• Published in: 2014 International Workshop on Pattern Recognition in Neuroimaging pp. 1 - 4
• Main Authors: Fei Fei, Biao Jie, Lipeng Wang, Daoqiang Zhang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2014
• Subjects: Accuracy; Alzheimer's disease; Data mining; Kernel; Testing; Training
• DOI: 10.1109/PRNI.2014.6858518

Recent studies on brain connectivity networks have suggested that many brain diseases, such as, Alzheimer's disease (AD) and mild cognitive impairment (MCI), are related with large-scale connectivity networks, rather than individual brain regions. However, it is challenging to find those networks from the whole connectivity network due to the complexity of brain networks. In this paper, we propose a novel method to mine the discriminative subnetworks for classifying MCI patients from healthy controls (HC). Specifically, we first apply multiple thresholds to generate multiple thresholded connectivity networks, and extract a set of frequent subnetworks from each of the two groups (i.e., MCI and HC), respectively. Then, we measure the discriminative ability of those frequent subnetworks using graph-kernel-based classification method and select the most discriminative subnetworks for subsequent classification. The results on the functional connectivity networks of 12 MCI and 25 HC show that our method can obtain a competitive results compared with state-of-the-art methods on MCI classification.