Computer Aided Diagnosis for COVID-19 with Quantum Computing and Transfer Learning

Computer Aided Diagnosis (CAD) is a promising medical technology enabling remote medical analysis and supporting diagnostics. The COVID-19 pandemic posed unprecedented challenges, prompting swift responses from academia and industry. Quantum Machine Learning (QML) addresses these challenges by extracting complex features from medical images and combining quantum algorithms with machine learning models and techniques. A set of respiratory condition images is used to implement transfer learning. This study presents a CAD system based on Quantum Deep Learning for X-ray COVID-19 binary image classification, where the ResNet50 model is first used to train the weights on a multiclass chest conditions dataset, and later implemented on a binary classification COVID-19 dataset for a performance comparison with its hybrid quantum counterpart. Our results demonstrate a favorable tradeoff performance compared to the proposed ResNet50 model, obtaining 0.8800 accuracy, 0.9388 recall, and 0.8846 F1 score. These findings emphasize the potential of transfer learning techniques for quantum hybrid methods and the effective implementation of CAD systems.