Neural network-based decision support system for pre-diagnosis of psychiatric disorders

Psychiatric disorders are mental conditions affecting emotional, cognitive, affective and behavioral states and causing impairment and suffering. The early and accurate diagnosis of such disorders is crucial for recovery and improvement. Artificial Intelligence is extremely implicated in medical and clinical fields bringing efficient results and solutions. This paper introduces a psychiatric pre-diagnosis approach to simplify the modeling of a decision support system using neural networks. The choice of neural network as a decisional tool is made after a comparative study with Case-Based Reasoning. The efficiency of the pre-diagnosis neural network in our experiment reaches the accuracy of 90% in identifying some categories like psychotic disorders category.