Convolutional slime mold deep learning model for diagnosis of PD

The proposed study aims to develop an efficient PD detection scheme using a novel optimized deep learning mechanism. Initially, the input multiple human voice recordings are pre-processed to lessen the unwanted noises. Then, the relevant features are selected to reduce the complexity problems in the feature selection stage using chi-square feature statistical model. Finally, an Enhanced Convolutional Slime Mold Attention (ECSMA) model is proposed for categorizing the input voice recordings. The simulation results portray that the proposed PD detection model attains higher performance than other existing methods and mitigate the costs of healthcare in identifying upcoming disease stages.