The diagnostic performance of MR psychoradiology in identifying patients with mental disorders / 中华放射学杂志

ABSTRACT

Objective:To explore the efficiency of MR psychoradiology examination in screening and classification of psychiatric disorders as well as its potential clinical application.Methods:Retrospective study was conducted for 144 patients with MR psychoradiology examination, who were diagnosed mental disorders based on International Statistical Classification of Diseases and Related Health Problems Tenth Revision (ICD-10) from September 2018 to July 2020 in West China Hospital of Sichuan University. As our previous studies, imaging diagnostic models were constructed based on gray matter volume (GMV) analysis for four kinds of psychiatric disorders, including depression, anxiety disorder, bipolar disorder and schizophrenia. For 120 patients with psychiatric disorders, the GMV change pattern of brain regions was detected and subtyped by MR psychoradiology examination with 50% similarity threshold of the above imaging diagnostic models. The diagnostic efficiency of MR psychoradiology examination was evaluated with references of clinical diagnose. For 24 patients with dementia, brain atrophy was additionally measured by conventional MR examination and MR psychoradiology examination respectively.Results:The sensitivity, specificity, Youden index and accuracy of MR psychoradiology in identifying the four psychiatric disorders were 86.6% (84/97), 69.6% (16/23), 0.56 and 83.3% (100/120), including depression [77.8% (28/36), 69.0% (58/84), 0.47, 71.7% (86/120)], schizophrenia [58.8% (10/17), 91.3% (94/103), 0.50, 86.7% (104/120)], bipolar disorder [42.1% (8/19), 96.0% (97/101), 0.38, 87.5% (105/120)] and anxiety disorder [24.0% (6/25), 100% (95/95), 0.24, 84.2% (101/120)]. As for 24 patients with dementia, 58.3% (14/24) were detected with brain atrophy by conventional MR examination, while 91.7% (22/24) were detected by MR psychoradiology examination.Conclusions:MR psychoradiology can detect the subtle brain structural abnormalities of patients with mental disorders. Although it can′t be used as an independent biomarker for disease diagnosis till now, it is of great help in improving the accuracy and objectivity of diagnosis.