Archimedes optimization-based Elman Recurrent Neural Network for detection of post-traumatic stress disorder

Abstract

Post-traumatic stress disorder (PTSD) is a stress-based disorder that occurs when a person is vulnerable to undesirable traumatic events like injury or harm. Numerous research works reported dissimilar complex issues for detecting PTSD severity in patients which makes the PTSD diagnosis process a complicated task. So, a novel Multi-Strategy Seeker Archimedes Optimization-based Elman Recurrent Neural Network (MSSAO-ERNN) technique to detect PTSD symptoms in earlier stages using speech samples. Our proposed MSSAO-ERNN technique efficiently identifies PTSD symptoms from the dataset and helps the physician to diagnose accurately. The proposed approach constitutes various significant phases that include the pre-processing phase, feature extraction phase, and classification phase to detect the presence of PTSD in the datasets. The performance of the proposed method is determined by computing various metrics for the accurate identification of PTSD. Finally, three different datasets namely the NNE dataset, the FME hospital dataset as well as TIMIT dataset are employed in this paper for PTSD diagnosis. The experimental results revealed that the proposed method attained a superior accuracy of 97% under different PTSD techniques.