Abstract
The paucity of readily available medical data poses a major challenge for the development of AI (artificial intelligence)-based healthcare applications and devices. To aid in overcoming this challenge, we propose a sensor-based medical time series data synthesis system especially designed for the training of diabetic foot diagnosis models. The proposed system utilizes statistical methods, augmentation techniques, and the NeuralProphet model to accomplish its purpose while still maintaining medical validity. Our results show that the generated synthetic time series data follow the trends and tendencies of real data. We also verify our work using machine learning-based clustering. By successfully clustering the synthetic data generated by our proposed system, we prove that our system is capable of meeting its objectives.
Similar content being viewed by others
References
“Diabetes”. World Health Organization; 2020. https://www.who.int/news-room/fact-sheets/detail/diabetes. Accessed 1 Apr 2021.
American Podiatric Medical Association. Diabetic wound care.
Nobuyoshi A. The diagnostic classification of critical limb ischemia. Ann Vasc Dis. 2018;ra-18.
Kalani M, Brismar K, Fagrell B, Ostergren J, Jörneskog G. Transcutaneous oxygen tension and toe blood pressure as predictors for outcome of diabetic foot ulcers. Diabetes Care. 1999;22(1):147–51.
de Meijer VE, Van’t Sant HP, Spronk S, Kusters FJ, den Hoed PT. Reference value of transcutaneous oxygen measurement in diabetic patients compared with nondiabetic patients. J Vasc Surg. 2008;48(2):382–8.
Murch C. Tramul: transcutaneous oxygen measurement and diabetic foot ulceration. Diabet Foot J. 2019;22(2):48–54.
Wyss CR, Matsen FA 3rd, Simmons CW, Burgess EM. Transcutaneous oxygen tension measurements on limbs of diabetic and nondiabetic patients with peripheral vascular disease. Surgery. 1984;95(3):339–46.
Yang C, Weng H, Chen L, Yang H, Luo G, Mai L, Jin G, Yan L. Transcutaneous oxygen pressure measurement in diabetic foot ulcers: mean values and cut-point for wound healing. J Wound Ostomy Cont Nurs. 2013;40(6):585–9.
Jayun H, Lee SH, Son HM, Park J-U, Chung T-M. A synthetic data generation model for diabetic foot treatment. In: International conference on future data and security engineering. Springer; 2020. p. 249–64.
Ming LJZ, Ng LNS, Thomas C. Prevention and treatment of diabetic foot ulcers. J R Soc Med. 2017;110(3):104–9.
Pendsey S. Understanding diabetic foot. Int J Diabetes Dev Ctries. 2010;30(2):75.
Singh S, Pai DR, Yuhhui C. Diabetic foot ulcer-diagnosis and management. Clin Res Foot Ankle. 2013;1(3):120.
Yale A, Dash S, Dutta R, Guyon I, Pavao A, Bennett KP. Generation and evaluation of privacy preserving synthetic health data. Neurocomputing. 2020;416:244–55.
Dahmen J, Cook D. Synsys: a synthetic data generation system for healthcare applications. Sensors. 2019;19(5):1181.
Walonoski J, Kramer M, Nichols J, Quina A, Moesel C, Hall D, Duffett C, Dube K, Gallagher T, McLachlan S. Synthea: an approach, method, and software mechanism for generating synthetic patients and the synthetic electronic health care record. J Am Med Inform Assoc. 2018;25(3):230–8.
Esteban C, Hyland SL, Rätsch G. Real-valued (medical) time series generation with recurrent conditional gans (2017). arXiv:1706.02633.
Yale A, Dash S, Dutta R, Guyon I, Pavao A, Bennett KP. Generation and evaluation of privacy preserving synthetic health data. Neurocomputing. 2020;416:244–55.
Mata AG. A comparison between lstm and facebook prophet models: a financial forecasting case study.
Liu W, Qin C, Gao K, Li H, Qin Z, Cao Y, Si W. Research on medical data feature extraction and intelligent recognition technology based on convolutional neural network. IEEE Access. 2019;7:150157–67.
Um TT, Pfister FMJ, Pichler D, Endo S, Lang M, Hirche S, Fietzek U, Kulić D. Data augmentation of wearable sensor data for Parkinson’s disease monitoring using convolutional neural networks. In: Proceedings of the 19th ACM international conference on multimodal interaction. 2017. p. 216–20.
Deng S, Hua W, Wang B, Wang G, Zhou X. Few-shot human activity recognition on noisy wearable sensor data. In: International conference on database systems for advanced applications. Springer; 2020. p. 54–72.
Wen Q, Sun L, Song X, Gao J, Wang X, Xu H. Time series data augmentation for deep learning: a survey. 2020. arXiv:2002.12478.
Triebe O, Laptev N, Rajagopal R. Ar-net: a simple auto-regressive neural network for time-series. 2019. CoRR. arXiv:1911.12436.
Benhamou Y, Begarin L, David N, Cailleux N, Bessin C, Lévesque H, Edet S. Detection of microcirculatory impairment by transcutaneous oxymetry monitoring during hemodialysis: an observational study. BMC Nephrol. 2014;15(1):1–8.
Makris K, Spanou L. Is there a relationship between mean blood glucose and glycated hemoglobin? J Diabetes Sci Technol. 2011;5(6):1572–83.
Elaine Biostatistics. An introduction to medical statistics for health care professionals: describing and presenting data. Musculoskelet Care. 2004;2:218–28.
Dua D, Graff C, et al. Uci machine learning repository. 2017.
National diabetes statistics report. 2020.
Selvin E, Zhu H, Brancati FL. Elevated a1c in adults without a history of diabetes in the U.S. Diabetes Care. 2009;32(5):828–33.
Lund E. Comparison of additive and multiplicative models for reproductive risk factors and post-menopausal breast cancer. Stat Med. 1995;14(3):267–74.
Tian Y, Huffman GJ, Adler RF, Tang L, Sapiano M, Maggioni V, Huan W. Modeling errors in daily precipitation measurements: additive or multiplicative? Geophys Res Lett. 2013;40(10):2060–5.
Yan X, Xie H, Tong W. A multiple linear regression data predicting method using correlation analysis for wireless sensor networks. In: Proceedings of 2011 cross strait quad-regional radio science and wireless technology conference. 2011.
Wang W, Lyu G, Shi Y, Liang X. Time series clustering based on dynamic time warping. In: 2018 IEEE 9th international conference on software engineering and service science (ICSESS). 2018. p. 487–90.
Tavenard R, Faouzi J, Vandewiele G, Divo F, Androz G, Holtz C, Payne M, Yurchak R, Rußwurm M, Kolar K, Woods E. Tslearn, a machine learning toolkit for time series data. J Mach Learn Res. 2020;21(118):1–6.
Funding
This work was supported by Institute of Information & communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (No. 2020-0-00990, Platform Development and Proof of High Trust & Low Latency Processing for Heterogeneous.Atypical.Large Scaled Data in 5G-IoT Environment).
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval
In this study, the original research has been carried out by following the ethical principles.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This article is part of the topical collection “Future Data and Security Engineering 2020” guest edited by Tran Khanh Dang.
Rights and permissions
About this article
Cite this article
Hyun, J., Lee, Y., Son, H.M. et al. Synthetic Data Generation System for AI-Based Diabetic Foot Diagnosis. SN COMPUT. SCI. 2, 345 (2021). https://doi.org/10.1007/s42979-021-00667-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s42979-021-00667-9