Authors: Mrs.T.Dheepa, Pradhakshna S, Vigiyalakshmi Muthu, Srividhya K
Abstract: Timely identification of bone tumors plays a vital role in improving diagnosis accuracy and treatment effectiveness. Although deep learning techniques have achieved remarkable success in medical image analysis, their lack of interpretability often restricts their acceptance in clinical environments. This study proposes an Explainable Artificial Intelligence (XAI)–driven framework for bone tumor detection using deep learning models. A convolutional neural network (CNN) is utilized to automatically extract relevant features from medical images such as X-rays or MRI scans to classify tumors into benign and malignant categories. To address the transparency issue, explainability techniques such as Gradient-weighted Class Activation Mapping (Grad-CAM) are incorporated to highlight critical image regions that influence the model’s predictions. These visual explanations assist clinicians in understanding and validating the system’s decisions. The experimental evaluation shows that the proposed model delivers reliable classification performance along with interpretable outputs, thereby supporting clinical decision-making and enhancing trust in AI-assisted diagnostic systems.
DOI: https://doi.org/10.5281/zenodo.18755185
