.png){kind=icon}
Return
Quantum splitting convolutional neural network-based distributed quantum disease detection model
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2025.131410.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Quantum computing leverages the properties of superposition and entanglement to enable high-dimensional feature extraction from medical images, demonstrating its potential to improve the accuracy of disease detection. However, practical applications are still constrained by the current performance bottlenecks of quantum hardware and affected by the excessive entangling gate operations. In this work, we propose a distributed quantum disease detection (DQDD) model using quantum splitting convolutional neural network. The DQDD employs a hybrid quantum convolutional neural network (QCNN) to capture low-dimensional and high-dimensional features in medical images. We introduce the quantum circuit splitting technique to our model, reconstructing an 8-qubit QCNN with 5 qubits, while simultaneously reducing the entangling operations in each sub-circuit. The model comprises two core components, the Information Distillation mechanism designed to extract pivotal features and the Non-saliency Filtering mechanism dedicated to eliminating irrelevant patterns. Both components operate cohesively, facilitating parallel computation across single or multiple quantum computing devices. Evaluations on ISIC2017 and HAM10000 skin cancer datasets demonstrate that our model outperforms existing quantum models in skin cancer detection, doing so with fewer parameters. Superior generalization capability is further validated through cross-domain testing on the MedMNIST dataset.
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W
Organization
No organization information available
Cited Papers
.png){kind=icon}
ShareSaveShareSaveShareSaveShareSaveShareSaveShareSaveShareSaveShareSave
View more