<p>Liver Cancer (LC), specifically Hepato Cellular Carcinoma (HCC), poses a global health threat. Biomedical research is vital, but its inadequate access to superior treatments is a significant challenge. A comprehensive, collaborative healthcare approach is essential to address the complexity of HCC pathophysiology and unequal intervention access. This study’s objective is to develop a novel framework for LC prediction that integrates advanced imaging techniques, Machine Learning (ML) algorithms, and clinical data analysis to optimize diagnostic accuracy and enhance patient outcomes. The study gathers a dataset of CT medical images from Kaggle. To develop data quality, preprocessing involved median filtering and Histogram Equalization (HE). Mean-shift segmentation was applied for effective image segmentation, ensuring precise delineation of liver regions. Feature extraction now utilizes Histogram of Oriented Gradients (HOG), capturing relevant information. The study represents a classification technique, such as the Integrated Randomized Gated Layer Recurrent NeuroNet (IRGLRN) approach, which integrates the Recurrent Neural Network (RNN) and Random Forest (RF) for LC diagnosis. By feature regularizing and integrating RNN and RF, through the fusion of these two architectures, the investigated model optimizes feature weights, leading to enhanced and robust classification of LC. A Python tool was utilized as the structure for implementing the proposed IRGLRN algorithm for LC classification. The algorithm, a fusion of RNN and RF, aimed to regularize the feature to achieve improved and robust classification of LC. The proposed IRGLRN model obtains an accuracy of 99.49%, precision of 99%, recall of 99%, F1 score of 99%, specificity of 100%, sensitivity of 99.29%, and a Dice score of 99.62%. This comprehensive methodology surpassed the potential for an enhanced and robust LC model compared to existing methods.</p>