Detection of Depression and Suicide Risk Based on Text From Clinical Interviews Using Machine Learning: Possibility of a New Objective Diagnostic Marker

• DOI: 10.3389/fpsyt.2022.801301
• Publication Date: 2022-05-24T05:45:04Z
• Authors: Daun Shin; Kyungdo Kim; Seung-Bo Lee; Changwoo Lee; Ye Seul Bae; Won Ik Cho; Min Ji Kim; C. Hyung Keun Park; Eui Kyu Chie; Nam Soo Kim; Yong Min Ahn
• Keywords: Depression; Mini-international neuropsychiatric interview

Depression and suicide are critical social problems worldwide, but tools to objectively diagnose them lacking. Therefore, this study aimed depression through machine learning determine whether it is possible identify groups at high risk of words spoken by the participants in a semi-structured interview.A total 83 healthy depressed patients were recruited. All recorded during Mini-International Neuropsychiatric Interview. Through assessment from interview items, with classified into high-suicide-risk (31 participants) low-suicide-risk (52 groups. The recording was transcribed text after only uttered participant extracted. In addition, all evaluated for depression, anxiety, suicidal ideation, impulsivity. chi-square test student's T-test used compare clinical variables, Naive Bayes classifier model.A 21,376 extracted model diagnosing based on confirmed an area under curve (AUC) 0.905, sensitivity 0.699, specificity 0.964. that distinguished two using statistically significant demographic AUC 0.761. DeLong result (p-value 0.001) text-based classification superior model. When predicting group, demographics-based 0.499, while one 0.632. However, ensemble incorporating variables 0.800.The possibility confirmed; regarding risk, diagnosis accuracy increased when incorporated. participants' show potential as objective diagnostic marker learning.