Abstract
In this chapter we explain a quantum classical model, for automated image diagnosis. The hybrid quantum classical model mixes processing with a classical deep learning framework. The main goal of the work is to see whether quantum enhanced models can improve image feature extraction and classification accuracy, in imaging. The model looks at non-autistic brain MRwe images. The model must separate autistic brain MRwe images from non-autistic brain MRwe images. That task requires precision and interpretability. We think the model could help doctors. The proposed work uses PennyLane to simulate circuits. Uses TensorFlow / Keras to build classical neural networks. The quantum part and the classical part can work together. We compare the approach with a classical convolutional neural network. The comparison checks accuracy, computational efficiency and model interpretability. Experimental results show that the hybrid model improves performance metrics. The experimental results suggest that quantum layers add clearer feature representations. We see that the hybrid model gives accuracy and better computational efficiency. In our view the hybrid model offers an advantage, over the classical convolutional neural network. We think the proposed work provides an analysis of convergence behavior, parameter optimization and resource utilization, inside the architecture. we think the findings give insight into the benefits and present limits of quantum-classical systems. we think the findings also show how quantum computing can be added to medical imaging solutions.