Chaotic Particle Swarm Optimization Empowered Deep Convolutional Neural Network for Prostate Cancer Diagnosis
Gokulnath Chandra Babu, N. Kalaiarasi, R Yeshaswini, P Brinda, K. Jeyakarthika, S.T. Gopukumar · 2024
Prostate cancer (PCa) Classification on magnetic resonance imaging (MRI) is an innovative use of medical imaging and machine learning (ML) models that objects to precisely recognize and classify PCa lesions in MRI scans. This advanced technique permits healthcare experts to analyze PCa more efficiently, define its level and severity, and improve modified treatment tactics. By analyzing subtle differences in tissue faces and recognizing cancerous areas, this methodology provides enhanced early recognition which is vital for patient results. It allows oncologists and radiologists to make knowledgeable verdicts, improving accuracy and efficacy of PCa diagnosis and successive treatment plans. This manuscript provides a proposal for Chaotic Particle Swarm Optimization Empowered Deep Convolutional Neural Network (CPSO-DCNN) technique for PCa Diagnosis. The developed CPSO-DCNN architecture incorporates a multi-faceted approach, starting with Wiener filtering (WF) for sound reduction, improving excellence and clarity of medical images. Feature extraction accomplished by employing SqueezeNet which is a great deep learning (DL) design expert at taking complex image patterns. This method further influences CPSO for hyperparameter tuning, enhancing its performance in order to support exact dataset characteristics. For critical task of PCa diagnosis, classification performed via bi-directional gated recurrent unit (BiGRU) networks and well-known for their capability to seize temporal needs within data. The simulation outcomes highlighted that CPSO-DCNN technique characterizes a substantial development in field of medical diagnostics. It provides a strong and exact solution for early recognition and classification of PCa.