Enhanced Bone Cancer Detection in MRI Scans via a Hybrid Quantum-Classical Convolutional Neural Network and War Strategy Optimization

Bone cancer is a relatively rare and aggressive form of cancer that originates in the bones and the adjacent soft tissues; the disease is characterized by marked morbidity and mortality. Frequently, early detection is possible, and such medical conditions should be treated to enhance the result of the therapy. MRI is very useful in diagnosing bone cancer since they give clear images of bones and soft tissues. In this work, a new method of bone cancer detection from MRI images is presented with unsharp structure guided filtering for preprocessing and a CSwin transformer to extract features from the preprocessed images. This method further improves visualization and selection of MRI images by improving the representation of features in the image. The main part of the designed method is the Hybrid Quantum-Classical Convolutional Neural Network with War Strategy Optimization (HQCCNN-WSO), which combines the fundamentals of quantum computing and traditional neural networks to enhance the classification performance. Based on the parameters of the War Strategy Optimization algorithm, the model distinguishes between cancerous lesions and healthy bone tissue, and the parameters are further refined for improved performance. The suggested approach demonstrates remarkable accuracy and dependability in the identification of bone cancer, achieving over 99% across all performance criteria.