Quantum Machine Learning and Data Re-Uploading: Evaluation on Benchmark and Laboratory Medicine Datasets

Background Quantum machine learning is an emerging field that may offer unique advantages over classical machine learning but has not been extensively studied with real-world healthcare data of practical size. This study evaluates the performance of a recently proposed quantum machine learning algorithm—quantum circuits with data re-uploading—compared to classical machine learning and other common quantum algorithms, using both benchmarking data and real-world laboratory medicine data.

Four datasets containing between 2 and 30 features were selected for evaluation. The quantum re-uploading algorithm was compared against variational quantum classifiers, quantum neural networks and four, commonly used classical machine learning algorithms. Initial, baseline comparisons of classification performance (F1 score) were conducted using all algorithms across the four datasets. Configuration parameters were then optimized for the quantum data re-uploading algorithm using a previously published dataset of plasma amino acid profiles to determine the impact of optimization on classification performance, followed by a final comparison against classical ML algorithms.

Baseline comparisons of quantum re-uploading showed superior classification performance on lower-dimensional datasets compared to quantum and linear ML algorithms. However, performance declined significantly as the number of input features increased. While optimization improved classification performance, marked variability was observed. Following optimization, the quantum re- uploading algorithm again performed comparably to linear algorithms but had lower classification performance on the plasma amino acid dataset compared to non-linear classical algorithms.

This study suggests that quantum data-reuploading algorithms can achieve classification performance comparable to some classical methods on lower-dimensional datasets, presenting opportunities for early research applications in laboratory medicine. However, while optimizing configuration parameters can improve classification performance, further advancements in quantum hardware and algorithm design will likely be necessary for quantum machine learning to become practically viable in laboratory medicine, and biomedical research more broadly. Competing Interest Statement W.L.S. was a technical consultant to HugoHealth, a personal health information platform (equity, fees); is a cofounder of Refactor Health, an AI-augmented data management platform for healthcare (equity); is a consultant for Detect, a point-of-care diagnostics company (equity, fees). B.N. is a consultant for Refactor Health, an AI-augmented data management platform for healthcare (equity). H.P.Y. is a consultant for Refactor Health, an AI-augmented data management platform for healthcare (fees). S.D. is now an employee at Royalty Pharma following the conclusion of study analysis. Funding Statement This work was partially supported by National Institutes of Health (NIH) grant number 1OT2OD032742-01. Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Data is available at: Edmund H Wilkes, Erin Emmett, Luisa Beltran, Gary M Woodward, Rachel S Carling, A Machine Learning Approach for the Automated Interpretation of Plasma Amino Acid Profiles, Clinical Chemistry, Volume 66, Issue 9, September 2020, Pages 1210-1218, https://doi.org/10.1093/clinchem/hvaa134 I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data Availability All data produced are available online at: Edmund H Wilkes, Erin Emmett, Luisa Beltran, Gary M Woodward, Rachel S Carling, A Machine Learning Approach for the Automated Interpretation of Plasma Amino Acid Profiles, Clinical Chemistry, Volume 66, Issue 9, September 2020, Pages 1210-1218, https://doi.org/10.1093/clinchem/hvaa134