International Journal of Science, Engineering and TechnologyAuthors: Raunak Sawa, Sharad Menon, Rishabh Gupta, Umamaheshwari M
Abstract: With the advent of artificial intelligence (AI), it is now pos-sible to assess mental health conditions in new ways using data from dig-ital sources. Traditional diagnosis is based on patients’ statements and a few tests, along with clinical observations, which can result in a delay in diagnosing and treating issues early. In recent years, research on the application of AI and machine learning in mental health assessment has increased. This review is a summary of literature published from 2015 to present date. The studies relied on the data collected from smartphones, wearable devices and social media. Sophisticated temporal and behav-ioral patterns in mental health-related information have been captured with the help of deep learning techniques. Issues with the problems are however that the data is not standardized, there is a lack of understand-ing of the models, privacy concerns and clinical validation of the models is poor. This review identifies gaps that need to be addressed in research on technologies, including real-time monitoring systems, explainable AI methods, personalized prediction models, and privacy-preserving learn-ing. Such advancements are crucial for enhancing the reliability, equity, and trustworthiness of the system and unlocking the potential of AI in mental health diagnostics and applications.
DOI: http://doi.org/10.5281/zenodo.22108545
