Link prediction boosted psychiatry disorder classification for functional connectivity network

Weiwei Li; Xue Mei; Hao Wang; Yu Zhou; Jiashuang Huang

doi:10.1117/12.2267698

8 February 2017 Link prediction boosted psychiatry disorder classification for functional connectivity network

Weiwei Li, Xue Mei, Hao Wang, Yu Zhou, Jiashuang Huang

Proceedings Volume 10225, Eighth International Conference on Graphic and Image Processing (ICGIP 2016); 102252F (2017) https://doi.org/10.1117/12.2267698
Event: Eighth International Conference on Graphic and Image Processing, 2016, Tokyo, Japan

Abstract

Functional connectivity network (FCN) is an effective tool in psychiatry disorders classification, and represents cross-correlation of the regional blood oxygenation level dependent signal. However, FCN is often incomplete for suffering from missing and spurious edges. To accurate classify psychiatry disorders and health control with the incomplete FCN, we first ‘repair’ the FCN with link prediction, and then exact the clustering coefficients as features to build a weak classifier for every FCN. Finally, we apply a boosting algorithm to combine these weak classifiers for improving classification accuracy. Our method tested by three datasets of psychiatry disorder, including Alzheimer’s Disease, Schizophrenia and Attention Deficit Hyperactivity Disorder. The experimental results show our method not only significantly improves the classification accuracy, but also efficiently reconstructs the incomplete FCN.

Citation Download Citation

Weiwei Li, Xue Mei, Hao Wang, Yu Zhou, and Jiashuang Huang "Link prediction boosted psychiatry disorder classification for functional connectivity network", Proc. SPIE 10225, Eighth International Conference on Graphic and Image Processing (ICGIP 2016), 102252F (8 February 2017); https://doi.org/10.1117/12.2267698

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available