基于磁共振的腦網(wǎng)絡(luò)研究進(jìn)展報(bào)告

胡德文++雷皓++沈輝++康彥

摘 要：這篇報(bào)告主要介紹有關(guān)腦網(wǎng)絡(luò)研究的最新進(jìn)展，該研究主要研究范圍包括：腦網(wǎng)絡(luò)研究的理論與方法、小動(dòng)物的實(shí)驗(yàn)研究和各種臨床應(yīng)用。在腦網(wǎng)絡(luò)的理論與方法研究方面，主要研究動(dòng)態(tài)因果模型（DCM），分析了一種基于隨機(jī)濾波理論的新的DCM參數(shù)估計(jì)方法。也將基于機(jī)器學(xué)習(xí)的多體素模式識(shí)別方法應(yīng)用于靜息態(tài)腦功能網(wǎng)路的模式分類。該方法有望為精神類疾病的臨床診斷提供潛在的生物學(xué)標(biāo)記。其次，在小動(dòng)物研究方面，基于DTI成像，深入分析了樹(shù)鼩的解剖學(xué)網(wǎng)絡(luò)，以期為精神類疾病的研究提供有效的動(dòng)物模型。最后，運(yùn)用腦網(wǎng)絡(luò)的理論與方法，開(kāi)展重度抑郁癥、行為成癮和吸煙成癮等各類精神類疾病的病理學(xué)研究。獲得的結(jié)果不僅驗(yàn)證了方法的可靠性，而且為這些疾病的病理成因和神經(jīng)機(jī)制提供了新的認(rèn)識(shí)。

關(guān)鍵詞：腦網(wǎng)絡(luò) 磁共振成像 精神病學(xué) 動(dòng)物模型

Report of Advances in Studies of Brain Networks Based on Magnetic Resonance Imaging （MRI）

Hu Dewen1 Lei Hao2 Shen Hui1 Kang Yan2

（1.National University of Defense Technology；2. Wuhan Institute of Physics and Mathematics （WIPM） of Chinese Academy of Sciences ）

Abstract：This report aims at introducing the recent advances in study of brain networks， supported by the National Basic Research Program of China （2011CB707802）. Main research scopes of this project include theory and methodology applied in brain networks， experimental study of small animals， and the potential clinical applications. First， in the study for theory and methods of brain networks， we concentrated on the dynamic causal mode （DCM）， and mainly researched a new parameter estimation approach of DCM based stochastic filtering theory. We also applied multiple voxel pattern analysis approach based statistic learning theory into pattern classification of resting-state functional brain networks， which were suggested to provide the potential biomarkers for clinical diagnoses of various mental disorders. Second， in the experimental studies of animals， we mainly analyzed the anatomical networks based on DTI in Tupaia belangeri chinensis， which was expected to provide an effective animal mode for brain networks studies in psychiatry. Finally， we used the theory and methods of brain network to investigate pathological modes of a variety of mental disorder， including major depression， behavioral addiction， and smoking addiction. These results not only validate the reliability and robustness of proposed methods for brain network analyses， but shed some new lights on the pathological causes and neural mechanism of these disorders.

Key Words：Brain networks； Magnetic resonance imaging； Psychiatry； Animal mode