Prediction of Post-Traumatic Stress Disorder (PTSD) in Healthcare Workers Based on Machine Learning

Machine learning has been used in various fields like mental health. One area explored is post-traumaticstress disorder (PTSD). The condition known as PTSD may manifest after an individual hasencountered or observed a traumatic event, affecting their mental wellbeing.Background and aims: The aim of this study was to investigate the factors that affect on post-traumaticstress disorder using machine learning methods.Method: In this cross-sectional study, ۶۳۷ Mazandaran University of Medical Sciences healthcareworkers were studied from ۲۷ June to ۲ September ۲۰۲۱. Participants included nurses, medical,midwifes, technicians, and support staffs. The Demographic Information Checklist and PTSDand DASS۲۱ Questionnaire was completed. Data analysis was performed by machine learningalgorithm using neural network the python.Results: After performance model, test loss equals ۰.۰۳ and test accuracy equals ۰.۹۹. The mostimportant features in model was depression, stress, anxiety, history of psychiatric disorders, workinghours with COVID Patient, contact group with patient and occupation.Conclusion: It seems the performance of model is quite good with a low test loss and high accuracy.The features that were found to be most important in the model suggests that these factorsmay play a significant role in predicting outcomes related to PTSD in individuals who have beenexposed to COVID-۱۹ patients. It may be helpful to further investigate these factors and howthey interact with each other to gain a deeper understanding of their impact on mental health andwell-being.

کلیدواژه ها:

Machine learning ، neural network ، post-traumatic stress disorder ، COVID ، health workers