Enhanced Multiple Convolutional Neural Networks Based CCS P System for Precise Classification in Membrane System-Based Applications

In the realm of pseudo-random simultaneous physiological computing models, exemplified by the innovative cell-like P systems (CLPS) with channel sections and symport ad antiport rules known as CCS P systems, this paper introduces a progressive variant—CCSs P systems. This augmentation incorporates a synchronization rule, adding a nuanced layer to the original CCS P system's channel sections and symport/antiport rules. The study delves into the computational analysis of drug sensitivity data, meticulously generated by Rising for various cancer cell lines and an array of chemicals. With a specific focus on advancing anticancer treatments for brain tumours, the research confronts the inherent challenges of accurate segmentation in the clinical context. Given the diverse sizes, shapes, and individual variations of brain tumours, achieving precise segmentation is a formidable task. In response, this study pioneers a distinctive fully convolutional neural network, seamlessly integrating a feature reuse module for iterative structure utilization and a feature conformity module to enhance the synthesis of multiple feature maps while minimizing noise. The ubiquity of convolutional neural networks (CNNs) is undeniable, yet their time-intensive training, resource demands, and data complexity necessitate novel approaches. To tackle these challenges, the research introduces a groundbreaking hypergraph membrane system—an innovative distributed and parallel computation paradigm poised to implement the Multiple Convolutional Neural Networks (MCNN).