Method for determining mood disorder onset risk based on regression model

Abstract

The invention relates to the technical field of electronic informatization detection of biochemical indexes, and provides a regression model-based onset risk determination method for mood disorder patients. The method comprises: extracting biochemical index data of a mood disorder patient; cleaning data to obtain valid data packets; carrying out gender grouping and detection; establishing a male regression model and / or a female regression model, and carrying out regression processing to obtain an ROC curve graph and an AUC result; and finally importing the related parameters into an SPSS decision tree to obtain a male decision tree model and / or a female decision tree model. The invention also designs a system for the determination method. According to the invention, clinical biochemical indexes are subjected to data analysis through an informatization system, a multi-distance linear regression model and a decision tree model are established, accurate evaluation and risk probability prediction can be performed on patients with mental examination of MDD or BD, the diagnosis speed and accuracy of treated doctors are improved, and the prognosis of diseases is further improved.

Description

technical field [0001] The invention relates to the technical field of electronic information detection of biochemical indicators, in particular to a method for determining the risk probability of a mood disorder based on a regression model. Background technique [0002] The diagnosis of mood disorders mainly relies on clinical phenomenology and lacks the utilization of biochemical parameters. Therefore, the current dilemma is that the ten-year diagnostic consistency of major depressive disorder (MDD) is only 45.5%, while the long-term consistency of bipolar disorder (major depressive disorder, BD) seven-year cohort study is only 45.5%. The consistency rate is also only 71.9%. Many researchers are trying to find biological markers to predict these diagnoses. A research group found that some independent conventional biochemical markers associated with MDD, such as uric acid, can be used as predictors of future development of BD in MDD individuals, and the factor has good ac...

AI Technical Summary

Problems solved by technology

[0004] At present, research on the development of biological markers for the early diagnosis of mood disorders mainly focuses on basic research such as hematology, genetics, and imaging. However, the research results are inconsistent and even contradictory.

As a result, these studies have encountered many difficulties under the current medical conditions, and practical clinical applications are rarely seen.

Even if a good biomarker is discovered, it cannot be effectively promoted due to the limitations of medical staff awareness, medical insurance, materials and equipment, etc.

[0005] But there is now a very valid implication: there are significant sex differences in the prevalence of MDD that do not exist in BD

Images