Revolutionizing Osteosarcoma Diagnosis: A Comparative Analysis of Deep Learning Models for Precise Bone Cancer Detection Using Multi-Modal Medical Imaging

Osteosarcoma, a highly aggressive primary bone malignancy, poses a significant challenge to public health, necessitating precise and timely diagnostic solutions. This study delves deep into the realm of sophisticated deep learning models tailored specifically for the meticulous detection of osteosarcoma, leveraging diverse multi-modal medical imaging datasets. Through a meticulous comparative analysis, we rigorously scrutinize the performance of cutting-edge algorithms: Support Vector Machine (SVM), the MobileNetV2 architecture, a customized Convolutional Neural Network (CNN), and the U-Net model. Our evaluation metrics, meticulously examining accuracy, precision, recall, and F1-score, offer a detailed insight into the diagnostic capabilities of each model. Among these, our custom CNN emerges as a frontrunner, showcasing the transformative potential of artificial intelligence in reshaping the landscape of osteosarcoma diagnosis. This research not only propels the discourse on the application of deep learning but also underscores the pivotal role of state-of-the-art technology in enhancing the precision and efficacy of bone cancer detection methodologies. With an accuracy of 72%, our CNN surpasses other models, including MobileNetV2 (76%), U-Net (85%), and an enhanced CNN designed for binary classification (95%). This highlights the significant strides made by deep learning, particularly in advancing the accuracy and reliability of osteosarcoma diagnosis.