A Precise Classification of Medical Images with a Cost-Efficient Quantum-Classical Hybrid Neural Network

This work discusses a cost-efficient hybrid quantum-classical neural network for classification of medical images. By embedding a 4-qubit variational quantum circuit (VQC) within a compact convolution neural network (CNN), the hybrid model achieves the accuracy of deep classical networks (99.89%) with much less parameters and faster training than ResNet-18 and VGG-16 model. Compared to a conventional CNN that shows the best training speed, the hybrid still remains even more parameter-efficient with similar speed, confirming its potential for real-time medical diagnostics in resource-constrained conditions.