Hybrid Classical-Quantum (HCQ) Alzheimer's Classification via Supervised $β$-VAE and Quantum Kernels
Researchers have developed a novel Hybrid Classical-Quantum (HCQ) pipeline for classifying Alzheimer's disease (AD) using 3D structural MRI scans. The system employs a supervised 3D $\beta$-variational autoencoder (VAE) to extract disease-aware features, which are then encoded into quantum states. These states are used to construct a quantum kernel for a Support Vector Machine (SVM), achieving up to 72.1% accuracy and an AUC of 0.799 on the ADNI-1 dataset. This approach integrates classical deep learning with quantum computing for enhanced diagnostic capabilities in biomedical imaging. AI
IMPACT This hybrid approach could offer a new framework for diagnostic classification in biomedical imaging, potentially improving accuracy and efficiency.