The rise in the data and its complexity in the healthcare sector demands accurate and precise diagnosis of patients, most of which are computer aided. However, the challenge arises when the outputs solely depend on the medical practitioners to look precisely into, which could have the potential of getting errors. One of the major error-prone areas is the detection of bone fractures from X-rays and CT reports which could be due to its lengthy process of detection and high workload. Thus examination of bone fractures solely by radiologists may result in inappropriate treatment of the patients. To overcome it we can create systems that utilize X-ray images to assist radiologists in identifying bone fractures, enhancing diagnostic accuracy, and streamlining treatment planning for patients with fractures. The paper focuses on deep learning-assisted detection of bone fractures as we proceeded with the CNN technique, which is effective for image-based tasks. It’s a multi-layer structure of convolutional, max-pooling, and fully connected layers. An X-ray image dataset consisting of two directories of train and test data was used. The analysis of the CNN model gives the outcome of 96% accuracy over the trained model. The evaluation metrics of the model reflect the model’s ability to differentiate between fractured bones and non-fractured bones. Thus the proposed work brings out notable results in comparison to others’ work in terms of accuracy that indicate that the model accurately diagnoses the bone fractures which indeed reduces the chances of error via self-examining and benefits radiologists.