Speech signals serve as a crucial medium for information exchange, reflecting various aspects of human being as sentiments, age, regional accent, gender, and health status. This study focuses on a gender-wise comparative analysis for the accurate detection of Psychogene Dysphonie using voice signals. Mel Frequency Cepstral Coefficients (MFCCs) play pivotal role as key features in computing gender-specific disease identification accuracy. Employing a machine learning algorithm, specifically Recurrent Neural Network Bidirectional Long Short-Term Memory (RNN_BiLSTM) with ADAM optimization, the study achieves significant results. Disease identification recall values peaks at 87.23% for disease-base male speakers and 99.20% for the healthy male speakers, while female speakers attain a recall values of 76.88% for the disease-based speakers and 85.63% for the healthy speakers. Overall, the gender-wise disease identification accuracy stands at an impressive 95.69% during the analysis of male voice signals and 76.88% for the female speakers. The evaluation employs a confusion matrix, offering a detailed assessment of the model’s performance in accurately detecting Psychogenic Dysphonia.