Among the most fatal illnesses to individuals worldwide, Bone cancer has grown to be a significant socioeconomic and global health issue. A variety of detrimental effects, as well as expensive diagnostic and treatment expenses, drive researchers to create an effective and affordable identifying technique for detecting bone cancer. The most sensitive, beneficial, and widely used imaging tool for detecting bone abnormalities is MRI imaging; however, it is problematic to manually interpret MRI for bone cancer, extract the essential features, and choose high-performance classifiers. Over the past few decades, the adoption of algorithms based on deep learning in image analysis has grown exponentially. To identify bone cancer, researchers implemented deep learning techniques, specifically convolutional neural networks (CNNs), with the goal of developing a prediction model for bone cancer classification. Investigations and analyses have been done on CNN and Gabor filter efficacy. The GCNN approach employed a publicly available MRI imaging data set for Bone cancer, which is classified into normal or abnormal class labels. In terms of performance, the GCNN model performs better than the other CNN models. The outcome seems better than already developed techniques; therefore, it may be used in clinical settings as well as a medical professional's load-reducing tool.