Quantum Computing for Medical Imaging
QCNN for medical imaging - Access to real quantum hardware via Helsinki
Overview
QCNN for medical imaging - Access to real quantum hardware via Helsinki
Quantum Computing for Medical Imaging
To overcome the limitations of classical neural networks, often subject to overfitting with small medical image datasets, this DOPE Health project explores the use of Quantum Convolutional Neural Networks (QCNN).
The approach combines quantum and classical models to improve diagnostic capabilities, leveraging the computational advantages of quantum computing.
The team has already achieved a significant result: access to real quantum hardware through an international internship in Helsinki, Finland.
Video
Technical Specs
QCNN - Quantum Convolutional Neural Networks
QCNNs are a quantum variant of classical CNNs, which exploit phenomena such as superposition and entanglement to process images more efficiently.
Overfitting Problem
Medical image datasets are often limited due to the complexity of acquisition and annotation. Classical networks tend to memorize rather than generalize.
Hybrid Approach
The project combines quantum and classical layers to maximize the advantages of both technologies.
Project Partners
Università di GenovaTimeline
Timeline TBD
Gallery
No media available.