TY - GEN
T1 - Combination of grey matter and white matter features for early prediction of posttraumatic stress disorder
AU - Wang, Si
AU - Hu, Hao
AU - Su, Shanshan
AU - Liu, Luyan
AU - Wang, Zhen
AU - Wang, Qian
AU - Shen, Dinggang
N1 - Publisher Copyright:
© Springer International Publishing AG 2016.
PY - 2016
Y1 - 2016
N2 - Posttraumatic stress disorder (PTSD) is a prevalent psychiatric disorder. In previous researches, there are few studies about structural and functional alterations of the whole brain simultaneously about PTSD prediction. Early alterations could provide evidence of early diagnosis and treatment. Early diagnosis of PTSD plays an important role during the treatment. In this work, we extract discriminant features from multi-modal images and implement classification-based prediction for PTSD onset. Specifically, discriminant features are a collection of measures derived from grey matter (GM) and white matter (WM). We choose cortical thickness of GM and three descriptions of WM connection which are fiber count, fractional anisotropy (FA), and mean diffusivity (MD). After applying automated anatomical labeling (AAL) to parcellate the whole brain into 90 regions-of-interest (ROIs), the descriptions can be quantified. Then, a weighted clustering coefficient of every ROI connected with the remaining ROIs is extracted as feature. GM features and WM features are combined and selected automatically, which are later utilized by support vector machine (SVM) for early identification of the patients. The classification accuracy is around 79.86 % as the area of receiver operating characteristic (ROC) curve is 0.816 evaluated via dual leave-one-out cross-validation.
AB - Posttraumatic stress disorder (PTSD) is a prevalent psychiatric disorder. In previous researches, there are few studies about structural and functional alterations of the whole brain simultaneously about PTSD prediction. Early alterations could provide evidence of early diagnosis and treatment. Early diagnosis of PTSD plays an important role during the treatment. In this work, we extract discriminant features from multi-modal images and implement classification-based prediction for PTSD onset. Specifically, discriminant features are a collection of measures derived from grey matter (GM) and white matter (WM). We choose cortical thickness of GM and three descriptions of WM connection which are fiber count, fractional anisotropy (FA), and mean diffusivity (MD). After applying automated anatomical labeling (AAL) to parcellate the whole brain into 90 regions-of-interest (ROIs), the descriptions can be quantified. Then, a weighted clustering coefficient of every ROI connected with the remaining ROIs is extracted as feature. GM features and WM features are combined and selected automatically, which are later utilized by support vector machine (SVM) for early identification of the patients. The classification accuracy is around 79.86 % as the area of receiver operating characteristic (ROC) curve is 0.816 evaluated via dual leave-one-out cross-validation.
UR - http://www.scopus.com/inward/record.url?scp=84989902725&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-46257-8_60
DO - 10.1007/978-3-319-46257-8_60
M3 - Conference contribution
AN - SCOPUS:84989902725
SN - 9783319462561
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 560
EP - 567
BT - Intelligent Data Engineering and Automated Learning - 17th International Conference, IDEAL 2016, Proceedings
A2 - Zhang, Daoqiang
A2 - Gao, Yang
A2 - Yin, Hujun
A2 - Li, Bin
A2 - Li, Yun
A2 - Yang, Ming
A2 - Klawonn, Frank
A2 - Tallón-Ballesteros, Antonio J.
PB - Springer Verlag
T2 - 17th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2016
Y2 - 12 October 2016 through 14 October 2016
ER -