Feasibility Study of Federated Learning on the Distributed Research Network of OMOP Common Data Model
0
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
Objectives: Since protecting patients’ privacy is a major concern in clinical research, there has been a growing need for privacy-preserving data analysis platforms. For this purpose, a federated learning (FL) method based on the Observational MedicalOutcomes Partnership (OMOP) common data model (CDM) was implemented, and its feasibility was demonstrated. Methods: We implemented an FL platform on FeederNet, which is a distributed clinical data analysis platform based on theOMOP CDM in Korea. We trained it through an artificial neural network (ANN) using data from patients who received steroidprescriptions or injections, with the aim of predicting the occurrence of side effects depending on the prescribed dose. The ANN was trained using the FL platform with the OMOP CDMs of Kyung Hee University Medical Center (KHMC) andAjou University Hospital (AUH). Results: The area under the receiver operating characteristic curves (AUROCs) for predictingbone fracture, osteonecrosis, and osteoporosis using only data from each hospital were 0.8426, 0.6920, and 0.7727 forKHMC and 0.7891, 0.7049, and 0.7544 for AUH, respectively. In contrast, when using FL, the corresponding AUROCs were0.8260, 0.7001, and 0.7928 for KHMC and 0.7912, 0.8076, and 0.7441 for AUH, respectively. In particular, FL led to a 14%improvement in performance for osteonecrosis at AUH. Conclusions: FL can be performed with the OMOP CDM, and FLoften shows better performance than using only a single institution's data. Therefore, research using OMOP CDM has beenexpanded from statistical analysis to machine learning so that researchers can conduct more diverse research.
Abstract
Objectives: Since protecting patients’ privacy is a major concern in clinical research, there has been a growing need for privacy-preserving data analysis platforms. For this purpose, a federated learning (FL) method based on the Observational MedicalOutcomes Partnership (OMOP) common data model (CDM) was implemented, and its feasibility was demonstrated. Methods: We implemented an FL platform on FeederNet, which is a distributed clinical data analysis platform based on theOMOP CDM in Korea. We trained it through an artificial neural network (ANN) using data from patients who received steroidprescriptions or injections, with the aim of predicting the occurrence of side effects depending on the prescribed dose. The ANN was trained using the FL platform with the OMOP CDMs of Kyung Hee University Medical Center (KHMC) andAjou University Hospital (AUH). Results: The area under the receiver operating characteristic curves (AUROCs) for predictingbone fracture, osteonecrosis, and osteoporosis using only data from each hospital were 0.8426, 0.6920, and 0.7727 forKHMC and 0.7891, 0.7049, and 0.7544 for AUH, respectively. In contrast, when using FL, the corresponding AUROCs were0.8260, 0.7001, and 0.7928 for KHMC and 0.7912, 0.8076, and 0.7441 for AUH, respectively. In particular, FL led to a 14%improvement in performance for osteonecrosis at AUH. Conclusions: FL can be performed with the OMOP CDM, and FLoften shows better performance than using only a single institution's data. Therefore, research using OMOP CDM has beenexpanded from statistical analysis to machine learning so that researchers can conduct more diverse research.
참고문헌 (16)
- McMahan B, "Communication-efficient learning of deep networks from decentralized data", Proc Mach Learn Res, 54, pp.1273~1282 (2017)
- Bonawitz K, "Towards federated learning at scale : System design", Proceedings of Machine Learning and Systems(MLSys), pp.374~388 (2019)
- Konecny J, "Federated learning: strategies for improving communication efficiency" (2016)[DOI]
- Xu J, "Federated learning for healthcare informatics", J Healthc Inform Res, 5(1), pp.1~19 (2021)[DOI]
- Rieke N, "The future of digital health with federated learning", NPJ Digit Med, 3, pp.119~ (2020)[DOI]
- Lee GH, "Federated learning on clinical benchmark data : performance assessment", J Med Internet Res, 22(10), pp.e20891~ (2020)[DOI]
- Vaid A, "Federated learning of electronic health records to improve mortality prediction in hospitalized patients with COVID-19 : machine learning approach", JMIR Med Inform, 9(1), pp.e24207~ (2021)[DOI]
- Li W, "Machine learning in medical imaging", pp.133~141 (2019), Springer
- Sheller MJ, "Federated learning in medicine : facilitating multi-institutional collaborations without sharing patient data", Sci Rep, 10(1), pp.12598~ (2020)[DOI]
- Brown JS, "Distributed health data networks : a practical and preferred approach to multi-institutional evaluations of comparative effectiveness, safety, and quality of care", Med Care, 48(6 Suppl), pp.S45~S51 (2010)[DOI]
- Hansen RA, "Exploration of heterogeneity in distributed research network drug safety analyses", Res Synth Methods, 5(4), pp.352~370 (2014)[DOI]
- Toh S, "Confounding adjustment in comparative effectiveness research conducted within distributed research networks", Med Care, 51(8 Suppl 3), pp.S4~S10 (2013)[DOI]
- "Observational Health Data Sciences and Informatics (OHDSI)" (2022)[URL]
- "FeederNet: a distributed clinical data analysis platform in Korea" (2022)[URL]
- Kim Y, "Comparative safety and effectiveness of alendronate versus raloxifene in women with osteoporosis", Sci Rep, 10(1), pp.11115~ (2020)[DOI]
- Suchard MA, "Comprehensive comparative effectiveness and safety of first-line antihypertensive drug classes : a systematic, multinational, large-scale analysis", Lancet, 394(10211), pp.1816~1826 (2019)[DOI]