Diabetic retinopathy is one of the leading causes of vision loss worldwide, and early automated detection from retinal images is critical to timely treatment. This paper explores a quantum machine learning approach that integrates classical deep feature extraction and dimensionality reduction with a variational quantum circuit inspired by Grover’s algorithm. The model encodes retinal image features into a compact quantum representation and employs iterative quantum amplitude amplification to improve classification accuracy. Experiments conducted on the APTOS 2019 dataset demonstrate that this hybrid quantum-classical method achieves performance comparable to classical Support Vector Machine classifiers, a positive indicator given the early stage of quantum machine learning. These findings highlight the promise of quantum-enhanced techniques for advancing medical image analysis.