In this research work, we present a compact, end-to-end framework for binary cancer detection (cancer vs. non-cancer) on two imaging problems, oral lesions and peripheral blood-smear leukemia, by combining classical deep convolutional networks with shallow variational quantum circuits. Public datasets from Kaggle were segmented with U-Net to isolate regions of interest, augmented to improve generalization, and split into train/validation/test (70/20/10). Proposed research work benchmarks the ResNet family (18/34/50/101/152) and introduces Quantum-ResNet variants by inserting data-reuploading variational quantum layers (1-4 qubits) into the classification head. Across 50 epochs on Google Colab (T4 GPU), the best hybrid models improve over classical baselines. For oral cancer, Q-ResNet50 (3-qubit) attains accuracy 98.89 % and F1 97.98 %; for leukemia, Q-ResNet50 (4-qubit) reaches accuracy 99.42 % and F1 98.95 %. These results suggest that small quantum layers can serve as expressive nonlinearities that enhance separability on medical image features without substantial parameter growth or training instability.
Early detection of oral cancer markedly improves clinical outcomes, yet specialized diagnostic tools remain scarce in low-resource settings. Smartphone-based screening is a scalable alternative but needs lightweight models that run within edge-hardware constraints. Hybrid classical-quantum architectures are emerging candidates for parameter-efficient learning, yet most rely on qubit hardware that needs cryogenic operation, unsuitable for edge deployment. Continuous-variable (CV) photonic quantum computing, which operates at room temperature, offers a complementary route. We investigate a hybrid classical-CV quantum classifier for oral cancer detection from smartphone images. The pipeline combines a MobileNetV1 feature extractor, principal component analysis to 16 dimensions, and a parameterized CV-QNN of displacement, interferometric, and Kerr gates on a photonic backend. We propose a simplified
Φ∘D∘U1 CV-QNN architecture that cuts trainable parameters 40-45% relative to the standard CV-QNN layer of Killoran et al. (2019a), and identify dimensionality-reduction and encoding-restriction strategies that mitigate barren plateaus, raising loss-gradient variance by roughly 58 orders of magnitude. Whether the simplified layer beats the full layer is width-dependent: the full layer holds a small but significant edge at two qumodes, whereas the simplified layer is significantly better at four qumodes using 44% fewer parameters. The strongest model, a four-qumode simplified CV-QNN with only 18 parameters, attains the highest validation AUC of all models, exceeds a 55-parameter classical baseline using 67% fewer parameters, and reaches 100% calibrated test accuracy across all seeds. These results support CV photonic quantum machine learning for parameter-efficient, room-temperature medical image classification and motivate progress toward edge quantum AI.