Deep learning in predicting the treatment outcomes of depressed patients
10.3760/cma.j.cn371468-20220301-00085
- VernacularTitle:抑郁症疗效的深度学习预测模型
- Author:
Xiaoyu CHEN
1
;
Yiru HU
;
Bin ZHANG
Author Information
1. 广州医科大学附属脑科医院精神心理脑功能实验室，广州 510370
- Keywords:
Deep learning;
Prediction model;
Depression;
Genomics;
Functional magnetic resonance imaging
- From:
Chinese Journal of Behavioral Medicine and Brain Science
2022;31(11):1041-1045
- CountryChina
- Language:Chinese
-
Abstract:
The optimal antidepressant therapies for different patients have been identified mostly by trial and error. Selecting an effective treatment based on the specific biomarkers may be an important step toward personalized treatment of depression. Deep learning is a branch of machine learning, that is capable of processing high-dimensional and complex data.It automatically extracts and learns from large amounts of demographic, clinical symptoms, genomics and neuroimaging data. Researchers have been using deep learning algorithms to develop prediction model of anti-depressant response in order to guide clinicians to make a precise prescription for depression and further advance personalized treatment globally. This article reviews the application of deep learning in predicting the treatment outcomes of depression. Additionally, deep learning based on multi-omics data applied in treatment outcome's prediction is direction with prospects in the future.
