Jiaolong Qin1, Maobin Wei1, Haiyan Liu2, Jianhuai Chen2, Rui Yan2, Zhijian Yao3, Qing Lu4. 1. Key Laboratory of Child Development and Learning Science (Ministry of Education), Research Centre for Learning Science, Southeast University, Si Pailou 2, Nanjing 210096, China. 2. Department of Psychiatry, Affiliated Nanjing Brain Hospital of Nanjing Medical University, No. 264 Guangzhou Road, Nanjing 210029, China. 3. Department of Psychiatry, Affiliated Nanjing Brain Hospital of Nanjing Medical University, No. 264 Guangzhou Road, Nanjing 210029, China; Nanjing Brain Hospital, Nanjing University Medical School, 22 Hankou Road, Nanjing 210093, China. Electronic address: zjyao@njmu.edu.cn. 4. Key Laboratory of Child Development and Learning Science (Ministry of Education), Research Centre for Learning Science, Southeast University, Si Pailou 2, Nanjing 210096, China; Suzhou Research Institute of Southeast University, 399 Linquan Street, Suzhou 215123, China. Electronic address: luq@seu.edu.cn.
Abstract
BACKGROUND: Accumulated evidence has illuminated the topological infrastructure of major depressive disorder (MDD). However, the changes of topological properties of anatomical brain networks in remitted major depressive disorder patients (rMDD) remain an open question. The present study provides an exploratory examination of pattern changes among current major depressive disorder patients (cMDD), rMDD patients and healthy controls (HC) by means of a pattern recognition analysis. METHODS: Twenty-eight cMDD patients (age range: 22-54, mean age: 39.57), 15 rMDD patients (age range: 23-53, mean age: 38.40) and 30 HC (23-54, mean age: 35.57) were enrolled. For each subject, we computed five kinds of weighted white matter (WM) networks via employing five physiological parameters (i.e. fractional anisotropy, mean diffusivity, λ1, λ2 and λ3) and then calculated three network measures of these weighted networks. We treated these measures as features and fed into a feature selection mechanism to choose the most discriminative features for linear support vector machine (SVM) classifiers. RESULTS: Linear SVM could excellently distinguish the three groups with the 100% classification accuracy of recognizing cMDD/rMDD from HC, and 97.67% classification accuracy of recognizing cMDD from rMDD. The further pattern analysis found two types of discriminative patterns among cMDD, rMDD and HC. (i) Compared with HC, both cMDD and rMDD exhibited the similar deficit patterns of node strength primarily involving the salience network (SN), default mode network (DMN) and frontoparietal network (FPN). (ii) Compared with cMDD and rMDD showed the altered pattern of intra-communicability within DMN and inter-communicability between DMN and the other sub-networks including the visual recognition network (VRN) and SN. LIMITATIONS: The present study had a limited sample size and a lack of larger independent data set to validate the methods and confirm the findings. CONCLUSIONS: These findings implied that the impairment of MDD was closely associated with the alterations of connections within SN, DMN and FPN, whereas the remission of MDD was benefitted from the network compensatory of intra-communication within DMN and inter-communication between DMN and the other sub-networks (i.e., VRN and SN).
BACKGROUND: Accumulated evidence has illuminated the topological infrastructure of major depressive disorder (MDD). However, the changes of topological properties of anatomical brain networks in remitted major depressive disorderpatients (rMDD) remain an open question. The present study provides an exploratory examination of pattern changes among current major depressive disorderpatients (cMDD), rMDDpatients and healthy controls (HC) by means of a pattern recognition analysis. METHODS: Twenty-eight cMDDpatients (age range: 22-54, mean age: 39.57), 15 rMDDpatients (age range: 23-53, mean age: 38.40) and 30 HC (23-54, mean age: 35.57) were enrolled. For each subject, we computed five kinds of weighted white matter (WM) networks via employing five physiological parameters (i.e. fractional anisotropy, mean diffusivity, λ1, λ2 and λ3) and then calculated three network measures of these weighted networks. We treated these measures as features and fed into a feature selection mechanism to choose the most discriminative features for linear support vector machine (SVM) classifiers. RESULTS: Linear SVM could excellently distinguish the three groups with the 100% classification accuracy of recognizing cMDD/rMDD from HC, and 97.67% classification accuracy of recognizing cMDD from rMDD. The further pattern analysis found two types of discriminative patterns among cMDD, rMDD and HC. (i) Compared with HC, both cMDD and rMDD exhibited the similar deficit patterns of node strength primarily involving the salience network (SN), default mode network (DMN) and frontoparietal network (FPN). (ii) Compared with cMDD and rMDD showed the altered pattern of intra-communicability within DMN and inter-communicability between DMN and the other sub-networks including the visual recognition network (VRN) and SN. LIMITATIONS: The present study had a limited sample size and a lack of larger independent data set to validate the methods and confirm the findings. CONCLUSIONS: These findings implied that the impairment of MDD was closely associated with the alterations of connections within SN, DMN and FPN, whereas the remission of MDD was benefitted from the network compensatory of intra-communication within DMN and inter-communication between DMN and the other sub-networks (i.e., VRN and SN).
Authors: Michelle N Servaas; Harriëtte Riese; Remco J Renken; Marieke Wichers; Jojanneke A Bastiaansen; Caroline A Figueroa; Hanneke Geugies; Roel Jt Mocking; Linda Geerligs; Jan-Bernard C Marsman; André Aleman; Aart H Schene; Robert A Schoevers; Henricus G Ruhé Journal: Neuropsychopharmacology Date: 2017-03-31 Impact factor: 7.853
Authors: David M Schnyer; Peter C Clasen; Christopher Gonzalez; Christopher G Beevers Journal: Psychiatry Res Neuroimaging Date: 2017-03-23 Impact factor: 2.376
Authors: Min-Hee Lee; Dong Youn Kim; Moo K Chung; Andrew L Alexander; Richard J Davidson Journal: IEEE Trans Biomed Eng Date: 2018-01-15 Impact factor: 4.538