Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Leveraging Quantum-Based Architectures for Robust Diagnostics

Shabnam Sodagari, Tommy Long

TL;DR

This study addresses four-class discrimination of Normal, Stone, Cyst, and Tumor in kidney CT images using a hybrid quantum–classical framework. It combines a pretrained ResNet50 encoder for feature extraction with a quantum convolutional neural network (QCNN) that uses angle encoding to map latent features to qubits, followed by interaction layers to capture quantum feature correlations. The approach is evaluated on 8-qubit and 12-qubit QCNN configurations, with data preprocessing, augmentation, and weighted sampling to manage class imbalance; both models converge rapidly with test accuracy around 0.99, while the 12-qubit setup yields superior recall and F1 for Cyst and Tumor (including perfect Cyst recall). Overall, the results demonstrate that integrating classical feature extraction with quantum circuit processing can enhance diagnostic performance and robustness in challenging medical imaging tasks, and provide insights into the trade-offs associated with qubit count and circuit depth.

Abstract

The objective of this study is to diagnose and differentiate kidney stones, cysts, and tumors using Computed Tomography (CT) images of the kidney. This study leverages a hybrid quantum-classical framework in this regard. We combine a pretrained ResNet50 encoder, with a Quantum Convolutional Neural Network (QCNN) to explore quantum-assisted diagnosis. We pre-process the kidney images using denoising and contrast limited adaptive histogram equalization to enhance feature extraction. We address class imbalance through data augmentation and weighted sampling. Latent features extracted by the encoder are transformed into qubits via angle encoding and processed by a QCNN. The model is evaluated on both 8-qubit and 12-qubit configurations. Both architectures achieved rapid convergence with stable learning curves and high consistency between training and validation performance. The models reached a test accuracy of 0.99, with the 12-qubit configuration providing improvements in overall recall and precision, particularly for Cyst and Tumor detection, where it achieved perfect recall for Cysts and a tumor F1-score of 0.9956. Confusion matrix analysis further confirmed reliable classification behavior across all classes, with very few misclassifications. Results demonstrate that integrating classical pre-processing and deep feature extraction with quantum circuits enhances medical diagnostic performance.

Leveraging Quantum-Based Architectures for Robust Diagnostics

TL;DR

This study addresses four-class discrimination of Normal, Stone, Cyst, and Tumor in kidney CT images using a hybrid quantum–classical framework. It combines a pretrained ResNet50 encoder for feature extraction with a quantum convolutional neural network (QCNN) that uses angle encoding to map latent features to qubits, followed by interaction layers to capture quantum feature correlations. The approach is evaluated on 8-qubit and 12-qubit QCNN configurations, with data preprocessing, augmentation, and weighted sampling to manage class imbalance; both models converge rapidly with test accuracy around 0.99, while the 12-qubit setup yields superior recall and F1 for Cyst and Tumor (including perfect Cyst recall). Overall, the results demonstrate that integrating classical feature extraction with quantum circuit processing can enhance diagnostic performance and robustness in challenging medical imaging tasks, and provide insights into the trade-offs associated with qubit count and circuit depth.

Abstract

The objective of this study is to diagnose and differentiate kidney stones, cysts, and tumors using Computed Tomography (CT) images of the kidney. This study leverages a hybrid quantum-classical framework in this regard. We combine a pretrained ResNet50 encoder, with a Quantum Convolutional Neural Network (QCNN) to explore quantum-assisted diagnosis. We pre-process the kidney images using denoising and contrast limited adaptive histogram equalization to enhance feature extraction. We address class imbalance through data augmentation and weighted sampling. Latent features extracted by the encoder are transformed into qubits via angle encoding and processed by a QCNN. The model is evaluated on both 8-qubit and 12-qubit configurations. Both architectures achieved rapid convergence with stable learning curves and high consistency between training and validation performance. The models reached a test accuracy of 0.99, with the 12-qubit configuration providing improvements in overall recall and precision, particularly for Cyst and Tumor detection, where it achieved perfect recall for Cysts and a tumor F1-score of 0.9956. Confusion matrix analysis further confirmed reliable classification behavior across all classes, with very few misclassifications. Results demonstrate that integrating classical pre-processing and deep feature extraction with quantum circuits enhances medical diagnostic performance.

Paper Structure

This paper contains 4 sections, 3 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. Problem and Approach
  3. Simulation Results
  4. Conclusion

Figures (3)

  • Figure 1: An example of our pre-processing of a kidney image using denoising and CLAHE.
  • Figure 2: Comparison of confusion matrices for (a) $8$ qubits and (b) $12$ qubits, both trained with a learning rate of $0.001$, batch size of $8$, and sample size of $4000$.
  • Figure 3: Comparison of training and validation accuracy and loss for (a) $8$ qubits and (b) $12$ qubits, both trained over $100$ epochs with a learning rate of $0.001$, batch size of $8$, and sample size of $4000$.