<![CDATA[The increasing prevalence of psychiatric disorders such as depression, bipolar disorder, and post-traumatic stress disorder has drawn attention to the need for more efficient and accurate diagnostic tools. In this context, machine learning offers promising solutions by enabling the analysis of complex and high-dimensional data. This study aims to evaluate the diagnostic performance of ML models applied to various psychiatric disorders by comparing the effectiveness of different data types such as EEG, MRI, video and audio recordings, photographs, survey responses, and clinical data. A total of 44 scientific studies published between 2015 and 2024 were systematically reviewed in accordance with PRISMA 2020 guidelines. The studies included applied ML or deep learning models to adult participants. The results show that the most successful data types varied by disorder. In conclusion, the choice of data type significantly influences the performance of ML models in psychiatric diagnosis. EEG, survey, and clinical data emerged as the most reliable across different conditions, while SVM, Random Forest, and CNN-based models provided the best classification results. These findings offer a valuable reference point for future research and the development of AI-assisted diagnostic tools.]]>
Copyrights © 2025
