Optimization of Diagnosis and Treatment of Depression Based on Multidimensional Clinical Assessment Classification

Major depressive disorder (MDD) characterized by high prevalence, high recurrence rate and high disability rate is a mental illness with the heaviest burden and has become a major public health issue in China and the world. Great challenges in diagnosis and treatment of depression consist of the complicated pathogenesis, a lack of objective diagnostic criteria, unsatisfactory treatment outcomes and poor treatment compliance. The previous studies of our research team showed that depression is affected by multiple factors. We could explore important markers for the diagnosis, treatment and prediction of treatment efficacy in depressed patients' data collected from different dimensions including immunometabolism, brain electrophysiology, brain structure and functional neural circuits, neuropsychology and psychophysiology. Our completed studies in the National Science and Technology Support Program and National Key Research and Development (R\&D) Program of the 12th and 13th Five-Year Plan in China found that treatment designed for specific clinical subtypes can improve the treatment effect, and meanwhile, the application of electronic-measurement based care (e-MBC) combined with smart mobile terminals can effectively provide whole-course medical management for patients, improve patient compliance and increase the efficacy of clinical diagnosis and treatment. However, due to disease diagnosis based on clinical symptomatology without subtype distinction and lack of multi-scale biological data mining, multidimensional assessment and deep integration, the results of most previous studies can hardly be used in clinical practice. Therefore, there is a strong urge to carry out a systematic research in which multidimensional evaluation of clinical characteristics and a large scale of data collection and mining are needed to form clinical diagnosis and optimal treatment regimens for depression subtypes. To achieve the goal, patients with depression will be our research subject in this study. First, on the basis of the previous cohort study and the whole-course e-MBC, patients' data of movement, respiratory rate, heart rate and sleep will be further collected. With the help of artificial intelligence (AI) technology such as deep machine learning, the data integrated with EEG imaging and specific immunometabolic markers in blood will be analyzed with clinical characteristics. The model of diagnosis and classification will be established based on multidimensional clinical assessment and verified. Second, through a prospective multicenter randomized controlled trial, optimal treatment regimens for different depression subtypes and individualized magnetic stimulation physical intervention technology navigated by AI will be explored so as to establish a predictive model of curative effect. Finally, long-term follow-up and its regular data collection can be completed on the patient diagnosis and treatment platform which is linked to the e-MBC. Thus, a stable clinical cohort and an advanced database containing multidimensional information of depression will be set up. The whole course e-MBC management platform will be optimized and promoted to improve patient compliance, treatment efficiency and prognosis. This study can provide evidence for precise diagnosis and classification of depression and optimal treatment regimens for different subtypes.