Quantum Machine Learning for Digital Health?

Abstract

Machine learning for analysing health data is booming with the increasing use of electronic health and medical records. While classical machine learning techniques for health data approach commercialisation, there is not yet clear evidence whether quantum machine learning (QML) will provide any empirical advantage for digital health data processing.

In this talk I will present the results of a recent systematic literature review we conducted, which assesses whether QML algorithms have the potential to outperform existing classical methods in efficacy or efficiency. I will begin by outlining the strict systematic review framework in which the work was conducted, largely unheard of in physics research, and then discuss our findings. Of 169 eligible studies included in the review, we found that most contained widespread technical misconceptions about QML, with 123 excluded for insufficient rigor in analysis. Of the remaining 46 studies, we found that only 16 studies consider realistic QML operating conditions, either by testing algorithms on quantum hardware, or using noisy quantum circuits when assessing QML algorithms. We noted a focus on clinical decision support over health service delivery or public health, and that nearly all the QML models used are linear quantum models. Scalability was not discussed, casting doubt on the models utility for very large health data sets. Our investigation paves the way for meaningful dialogue about QML use-case discovery in digital health, and highlights the challenges of interdisciplinary collaborative research.

Carolyn Wood

Postdoctoral Scientist

Carolyn Wood is a postdoctoral researcher at the University of Queensland, in Brisbane, Australia focusing on quantum machine learning and physics at the interface between quantum mechanics and general relativity.