Leveraging Hybrid Intelligence for Early Detection and Severity Assessment of Mental Health Disorders in Women

Abstract

Depression, anxiety and PTSD among women are commonly associated with distinct difficulties as they have gender specific symptoms and different degrees of severity. The importance of early diagnosis and proper severity evaluation is to provide interventions and better treatment results. But conventional means of diagnosis are limited to reflecting the complexity of such disorders, particularly in the context of incorporating multimodal data. This paper proposes an intelligent framework based on the proposed hybrid model that integrates machine learning and gender-sensitive decisionmaking algorithms to improve the timely diagnosis and prognosis of mental diseases in women. The framework makes use of the multimodal data of clinical records, psychological measures and behavioural data to make better classification and predicting the severity. The performance on the proposed model based on experiments demonstrates that it outperforms the current baseline methods that have 92% accuracy, 91% precision, 90% recall, and 90% F1-score with a time per instance of 0.85 seconds. The excellent classification and prediction in real time of the proposed model make it applicable in clinical scenarios, and it can be seen as a potential solution to customized care and early detection of mental disorders in women.