A Brief Review of The Different Approaches for Mood Disorders Detection and Diagnosis

Abstract

Mood disorders are a mental disease that affects seriously the feelings and the thinking manner of people. They consist of severely emotional state fluctuations which is different from normal fluctuations. It seems that early detection is the best way to deal with this problem and take the adequate treatment for these disorders. The emergence of computer aided diagnosis applications, in recent years, has given an efficient help for psychiatrist to detect mood fluctuations and diagnose the pathological state in the adequate time. In this context, recent studies apply artificial intelligence in order to detect emotional states from one or more combined attributes. Four attributes are commonly used to detect a mood disorder: facial expressions, speech voice signals, body movements and texts shared in social media. This paper reviews recent research studies conducted for the detection and diagnosis of mood disorders using these attributes. This paper firstly introduces CAD systems in medicine and specifically in mental health domain. Then, it presents a brief explanation of the most used artificial intelligence approaches in CAD systems. After that, It reviews the recent studies performed to detect mood disorders from the four attributes mentioned above. The last section discusses results from the presented approaches.