Predictive Modeling Of Real-Time Colorectal Cancer Via Hyperparameter Configuration With Deep Learning Using Public Health Indicator Analysis

Public health indicator analysis plays a crucial part in colorectal cancer (CRC) recognition by leveraging epidemiological data and health metrics to recognize trends, disparities, and risk factors in diagnosis and screening. Public health officials use screening, incidence, and mortality rates to analyze current screening programs, target interventions for underserved populations, and improve early recognition and treatment results. With growing case numbers, CRC ranks second in women and third in men. The growing diagnostic workload and variability in biomarker assessment have prompted the exploration of robust machine-based models for pathology labs. Diagnosing CRC from histopathological images involves analyzing tissue samples for cancerous cellular changes. These images offer elaborate insights into tissue microstructure, aiding pathologists in examining cellular characteristics such as size, shape, and organization. Artificial Intelligence (AI) and fractals algorithms have proven advanced in the healthcare field, having the inherent ability for medical applications. The incorporation of AI in CRC diagnosis aims to improve the efficiency of pathologists, reduce diagnostic errors, and ultimately contribute to better patient outcomes through early and accurate detection of cancerous lesions. The study proposes a Golden Jackal Fractal Optimization with Deep Learning Assisted CRC Detection (GJODL-CRCD) technique on HIs. The GJODL-CRCD technique mainly examines the HIs to recognize and identify the CRC. The GJODL-CRCD method applies the Gabor filtering (GF) approach as a noise removal procedure to achieve this. Furthermore, the GJODL-CRCD technique employs the DenseNet model for learning and capturing intrinsic feature patterns from the preprocessed imageries. For hyperparameter selection of the DenseNet model, the GJODL-CRCD approach uses the GJO fractal algorithm. Finally, the GJODL-CRCD method uses the Elman Neural Network (ENN) model to identify and classify CRC. The performance evaluation of the GJODL-CRCD technique takes place on benchmark medical image data. The experimental results pointed out the betterment of the GJODL-CRCD model over other methods.