摘要

Depression is a psychological disorder, which affects one's quality of life due to its chronic course. Its early screening is the key to curb the overall disease load. This paper is an attempt to develop a mathematical model that captures several common symptoms of depression using a Concept Hierarchical Tree (CHT) and based on the answers obtained, computes the Depression Load (DL) by a Connected Graph-based Approach (CGA) and multiple linear regressions. The proposed CHT-CGA model is implemented as a tool in a Java Script and has been validated with 123 real-world depression cases. The prototype tool is able to diagnose 'mild,' moderate; and 'severe' depression with 90%, 95%, and 94% accuracies, respectively with the average accuracy of 93%. It has been proposed that the tool might be deployed in rural health centres for initial screening and taking referral decision.

  • 出版日期2012-9