Accurate differentiation of pulmonary emphysema from chest X-rays remains challenging due to imbalanced data, image noise, and the limitations of traditional deep networks in managing diagnostic uncertainty. In this study, emphysema detection is operationally defined based on radiographic irregular radiolucency, which is a validated marker of emphysematous changes so introduces the quantum–Fuzzy Neural Network (QFNN), augmented by the Giza Pyramids Construction(GPC) algorithm, to facilitate interpretable and robust emphysema classification. The proposed methodology incorporates a U-Net encoder–decoder for multi-scale feature extraction, a four-qubit variational quantum layer for nonlinear feature encoding, and a Mamdani fuzzy inference module with five Gaussian rules to model diagnostic uncertainty. The GPC algorithm adaptively adjusts convolutional, quantum, and fuzzy parameters while balancing the trade-off between exploration and convergence. Experiments were conducted using the emphysema dataset from Çallı et al., which includes 2418 chest radiographs for training and 422 for testing. The experimental results indicated that the QFNN–GPC model achieved 93.21% accuracy, 0.8102 recall, and an AUC of 0.9011, surpassing both classical and quantum-only baselines. These findings suggest that the hybrid quantum–fuzzy framework significantly reduces overfitting, enhances interpretability, and improves the reliability of uncertainty estimation, offering a promising pathway toward transparent quantum-enhanced medical diagnosis.
Oviesi et al. (2026) studied this question.