Synthetic data as an enabler for machine learning applications in medicine

Rajotte, Jean-François; Bergen, Robert V.; Buckeridge, David L.; Emam, Khaled El; Ng, Raymond T.; Strome, Elissa

doi:10.1016/j.isci.2022.105331

Cited by 47 publications

(25 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While there is a body of work on the synthesis of medical images and other data types [ 111 ], our focus in this paper was on structured longitudinal data. The synthesis of multi-modal data would be an important direction for future research.…”

Section: Discussionmentioning

confidence: 99%

A method for generating synthetic longitudinal health data

Mosquera

Emam

Ding

et al. 2023

BMC Med Res Methodol

Self Cite

View full text Add to dashboard Cite

Getting access to administrative health data for research purposes is a difficult and time-consuming process due to increasingly demanding privacy regulations. An alternative method for sharing administrative health data would be to share synthetic datasets where the records do not correspond to real individuals, but the patterns and relationships seen in the data are reproduced. This paper assesses the feasibility of generating synthetic administrative health data using a recurrent deep learning model. Our data comes from 120,000 individuals from Alberta Health’s administrative health database. We assess how similar our synthetic data is to the real data using utility assessments that assess the structure and general patterns in the data as well as by recreating a specific analysis in the real data commonly applied to this type of administrative health data. We also assess the privacy risks associated with the use of this synthetic dataset. Generic utility assessments that used Hellinger distance to quantify the difference in distributions between real and synthetic datasets for event types (0.027), attributes (mean 0.0417), Markov transition matrices (order 1 mean absolute difference: 0.0896, sd: 0.159; order 2: mean Hellinger distance 0.2195, sd: 0.2724), the Hellinger distance between the joint distributions was 0.352, and the similarity of random cohorts generated from real and synthetic data had a mean Hellinger distance of 0.3 and mean Euclidean distance of 0.064, indicating small differences between the distributions in the real data and the synthetic data. By applying a realistic analysis to both real and synthetic datasets, Cox regression hazard ratios achieved a mean confidence interval overlap of 68% for adjusted hazard ratios among 5 key outcomes of interest, indicating synthetic data produces similar analytic results to real data. The privacy assessment concluded that the attribution disclosure risk associated with this synthetic dataset was substantially less than the typical 0.09 acceptable risk threshold. Based on these metrics our results show that our synthetic data is suitably similar to the real data and could be shared for research purposes thereby alleviating concerns associated with the sharing of real data in some circumstances.

show abstract

Section: Discussionmentioning

confidence: 99%

A method for generating synthetic longitudinal health data

Mosquera

Emam

Ding

et al. 2023

BMC Med Res Methodol

Self Cite

View full text Add to dashboard Cite

show abstract

“…This data set can be used with any machine‐learning module, thus allowing other researchers to utilize it in their research. Synthetic data generation for machine‐learning is an active area of research, especially in the domain of medicine and health care, where the privacy of patient data is a big concern [3, 14, 22, 29, 40, 46]. To the best of our knowledge, no existing open‐source contribution automate the diagnosis of the three hearing loss attributes: type, degree, and configuration.…”

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Automated hearing impairment diagnosis using machine‐learning: An open‐source software development undergraduate research project

Taylor,

Sheikh

2024

Comp Applic In Engineering

View full text Add to dashboard Cite

Approximately 700 million people will have disabling hearing loss by 2050. Underdeveloped and developing countries, which encompass a considerable proportion of people with disabling hearing impairment, have a sparse number of audiologists and otolaryngologists. The lack of specialists leaves most hearing impairments undiagnosed for a long time, resulting in negative societal and economic impacts. In this article, we propose an automated hearing impairment diagnosis software—based on machine learning—to support audiologists and otolaryngologists in accurately and efficiently diagnosing and classifying hearing loss. We present the design, implementation, and performance analysis of an open‐source automated hearing impairment diagnosis software, which consists of two modules: a hearing test data‐generation module and a machine‐learning model. The data‐generation module produces a diverse and exhaustive data set for training and evaluating the machine‐learning model. By employing multiclass and ultilabel classification techniques to learn from the hearing test data, the model can swiftly predict the type, degree, and configuration of hearing loss with high reliability. Our proposed machine‐learning model demonstrates promising results with a prediction time of 634 ms, a log‐loss reduction rate of 0.9848 and accuracy, precision, recall, and f1‐score of 1.0000—showing the model's applicability to assist audiologists and otolaryngologists in rapidly and accurately classifying the type, degree, and configuration of hearing loss. In addition to the technical contributions, this article also highlights the importance of involving undergraduate students in open‐source software development projects which have a direct impact on improving the quality of human life.

show abstract

“…These limitations represent a unique opportunity to develop in silico trials that are completed as planned, safely, and that include digital cohorts with a representative distribution of subject characteristics and numbers large enough for appropriate statistical power. As pointed out in [9], in silico data has the potential to address lack of data availability, sharing mechanisms and privacy challenges associated with the use of medical information.…”

Section: Introductionmentioning

confidence: 99%

The stochastic digital human is now enrolling for in silico imaging trials—methods and tools for generating digital cohorts

Badano,

Lago,

Sizikova

et al. 2023

Prog. Biomed. Eng.

View full text Add to dashboard Cite

Randomized clinical trials, while often viewed as the highest evidentiary bar by which to judge the quality of a medical intervention, are far from perfect. In silico imaging trials are computational studies that seek to ascertain the performance of a medical device by collecting this information entirely via computer simulations. The benefits of in silico trials for evaluating new technology include significant resource and time savings, minimization of subject risk, the ability to study devices that are not achievable in the physical world, allow for the rapid and effective investigation of new technologies and ensure representation from all relevant subgroups. To conduct in silico trials, digital representations of humans are needed. We review the latest developments in methods and tools for obtaining digital humans for in silico imaging studies. First, we introduce terminology and a classification of digital human models. Second, we survey available methodologies for generating digital humans with healthy status and for generating diseased cases and discuss briefly the role of augmentation methods. Finally, we discuss approaches for sampling digital cohorts and understanding the trade-offs and potential for study bias associated with selecting specific patient distributions.

show abstract

Synthetic data as an enabler for machine learning applications in medicine

Cited by 47 publications

References 42 publications

A method for generating synthetic longitudinal health data

A method for generating synthetic longitudinal health data

Automated hearing impairment diagnosis using machine‐learning: An open‐source software development undergraduate research project

The stochastic digital human is now enrolling for in silico imaging trials—methods and tools for generating digital cohorts

Contact Info

Product

Resources

About