Effectiveness of Machine Learning for Mental Health

Abstract

Analyze and finding the most used AI applications or methods in the Mental Health sector and suggesting appropriate directions for advanced research is the primary objective of this research. With this purpose, the author selects papers reviewing to analyze thirty-one articles included two books, two conferences, six generics, one report, and 16 journals. The author identified that most of the research followed the same theories although, some are different or merge. The researcher found the most used recognizing technologies in real life; find out from a person’s voice or speech and eyetracking and neuroimaging technologies for checking brain abnormalities. Besides, the chatbot is the most used AI assistant in digital health or virtual life. As the main aim is to analyze the Bangladeshi youth’s mental health, the researcher has done a survey on 19-29 years people. With the result of this survey, the researcher becomes clear about the upcoming danger that will only for mental disorders like as- Anxiety, Depression, and PTSD. Not only that, this will cause a huge loss for people's unaware attitude. This research also offers an implication for future research directions and also identified existing knowledge gaps. Besides, a three-phase model proposed in the study is a combination of virtual and real-life AI technologies. Moreover, this review did not include any psychiatrist techniques of treatment or, any real-life dataset from patients. So, it was a limitation and a future working scope also. Applying deep learning, machine learning algorithms (e.g. support vector machine (SVM) random forest), BCI, virtual reality in mental illness detection, and treatments are also used effectively as founded by reviewing those articles.