Approved Research

Combined analysis of multi-source heterogeneous data for depression detection and diagnosis

Lay summary

More than 350 million people in the world suffer from depression, and depression has become the fourth largest disease in the world. It is predicted that depression will become the first disease in the global disease burden in 2030. However, diagnostic biomarkers for depression are lacking, and previous studies on depression have reported many inconsistent results, which may be attributed to the small sample size and within-sample demographic and clinical heterogeneity. Using the UK Biobank database, this study aims to integrate clinical data, psychosocial factors, neuroimaging, and genetic data of depression, and use large sample data to explore precise objective biomarkers related to depression.

Combining multiple computer methods such as machine learning and deep learning, and integrating genetic, brain imaging and cognitive/emotional data, and demographic data from the UK Biobank, the onset of depression is predicted through cross-sectional and prospective follow-up data. We will  examine associations between genes, brain imaging and cognitive/mood data, demographic data, and explore how these associations are mediated by brain MRI or mediated by biomarker variables (inflammatory markers, vitamins, etc.) of. It is expected to screen out 2-3 accurate, sensitive and effective monitoring biomarkers for depression, and build a prevention and treatment network for depression research. Provide active and effective scientific and technological support for accelerating technological breakthroughs in the prevention and control of depression, controlling the growth of medical expenses, promoting the rational and standardized application of technology, reducing medical and social burdens, and curbing the high incidence and mortality of major chronic diseases.