Quantum machine learning for healthcare - progress

Principal Investigator: Miguel Ángel González Ballester

MdM-supported staff: Francesco Venturelli

External collaborators: Alba Cervera Lierta (BSC)

Objectives: The project explores the interface between machine learning and quantum computing, particularly for the analysis of medical imaging data. We will explore variational quantum circuits as approaches for data classification and regression, analysing their theoretical properties and developing proof of concept prototypes. Data encoding, i.e. the process of encoding classical data into a quantum state amenable for further quantum processing, will be approached by devising novel methods tuned for imaging data. Alternative quantum computation paradigms, such as quantum annealing, will also be explored for applications in image segmentation and analysis.

Results to date: An in-depth study of the state of the art of quantum machine learning, with a particular focus on image analysis, has been carried out. Initial prototypes have been developed for image encoding and classification. Furthermore, first implementations of a quantum annealing approach for image processing are being developed, by casting the image segmentation problem into a quadratic unconstrained binary optimization (QUBO), amenable for analysis with annealers.

Activities:

- Quantum Techniques in Machine Learning (QTML) 2024, Melbourne, Australia

- AI4Quantum 2025, Favrholm, Denmark

Next steps:

- Develop quantum encoding techniques for medical imaging data

- Further explore QUBO-type problems for optimization in a quantum annealing setting

- Analyse convergence properties of quantum variational circuits

- Application to healthcare data