Medical Imaging Classification with Cold-Atom Reservoir Computing using Auto-Encoders and Surrogate-Driven Training

2026-05-07

AI Summary

This work addresses the challenge of effectively training quantum reservoir computing models for high-dimensional medical images—such as in polyp detection—where non-differentiability of quantum measurements impedes end-to-end optimization. To overcome this, the authors propose a hybrid quantum-classical architecture that integrates a guided autoencoder with a neutral-atom quantum reservoir. By employing a differentiable surrogate model, the framework enables end-to-end training, compressing input images into low-dimensional representations that are encoded into the detuning parameters of a Rydberg Hamiltonian for quantum processing. A linear classifier then performs binary classification. This approach, the first to combine guided autoencoding with cold-atom reservoir computing, simultaneously enhances classification accuracy and reconstruction fidelity while circumventing the non-differentiability barrier of quantum measurements. Experiments demonstrate superior performance over conventional PCA and unguided autoencoder baselines, with robustness and practical promise on NISQ-era hardware.

Abstract

We introduce a hybrid quantum-classical pipeline, based on neutral-atom reservoir computing, for medical image classification, focusing on the binary classification task of polyp detection. To deal effectively with the high dimensionality, we integrate a guided auto-encoder. This pipeline learns compact and discriminative representations of image data that are also well-suited for quantum reservoir computing. A key challenge in such systems is the non-differentiable nature of quantum measurements, which creates a 'gradient barrier' for standard training. We overcome this barrier by incorporating a differentiable surrogate model that emulates the quantum layer, enabling end-to-end backpropagation through the entire system. This guided training process is jointly optimized for classification accuracy and for faithful image recovery from the auto-encoder. The learned latent representations are encoded as pulse detuning parameters within a Rydberg Hamiltonian, and quantum embeddings are subsequently obtained through expectation values. These embeddings are then passed to a linear classifier. Our simulations show that this method outperforms some traditional approaches that use PCA or unguided autoencoders. We also conduct ablation studies to assess the impact of various quantum and training parameters, demonstrating the robustness and flexibility of our proposed pipeline for real-world medical imaging applications, even in the current NISQ era.

Problem

• medical imaging classification
• quantum reservoir computing
• gradient barrier
• high-dimensional data
• polyp detection

Innovation

• quantum reservoir computing
• guided auto-encoder
• surrogate model
• differentiable training
• Rydberg Hamiltonian