KDA
- Current
- Browse
- Collections
-
For contributors
- For Authors
- Instructions to authors
- Article processing charge
- e-submission
- For Reviewers
- Instructions for reviewers
- How to become a reviewer
- Best reviewers
- For Readers
- Readership
- Subscription
- Permission guidelines
- About
- Editorial policy
Search
- Page Path
- HOME > Search
Original Article
- Technology/Device
- Development and Validation of a Deep Learning Based Diabetes Prediction System Using a Nationwide Population-Based Cohort
- Sang Youl Rhee, Ji Min Sung, Sunhee Kim, In-Jeong Cho, Sang-Eun Lee, Hyuk-Jae Chang
- Diabetes Metab J. 2021;45(4):515-525. Published online February 25, 2021
- DOI: https://doi.org/10.4093/dmj.2020.0081
- 8,464 View
- 256 Download
- 6 Web of Science
- 5 Crossref
-
Graphical Abstract
Abstract
PDF
Supplementary Material
PubReader 
ePub - Background
Previously developed prediction models for type 2 diabetes mellitus (T2DM) have limited performance. We developed a deep learning (DL) based model using a cohort representative of the Korean population.
Methods
This study was conducted on the basis of the National Health Insurance Service-Health Screening (NHIS-HEALS) cohort of Korea. Overall, 335,302 subjects without T2DM at baseline were included. We developed the model based on 80% of the subjects, and verified the power in the remainder. Predictive models for T2DM were constructed using the recurrent neural network long short-term memory (RNN-LSTM) network and the Cox longitudinal summary model. The performance of both models over a 10-year period was compared using a time dependent area under the curve.
Results
During a mean follow-up of 10.4±1.7 years, the mean frequency of periodic health check-ups was 2.9±1.0 per subject. During the observation period, T2DM was newly observed in 8.7% of the subjects. The annual performance of the model created using the RNN-LSTM network was superior to that of the Cox model, and the risk factors for T2DM, derived using the two models were similar; however, certain results differed.
Conclusion
The DL-based T2DM prediction model, constructed using a cohort representative of the population, performs better than the conventional model. After pilot tests, this model will be provided to all Korean national health screening recipients in the future. -
Citations
Citations to this article as recorded by
- Remnant Cholesterol Is an Independent Predictor of Type 2 Diabetes: A Nationwide Population-Based Cohort Study
Ji Hye Huh, Eun Roh, Seong Jin Lee, Sung-Hee Ihm, Kyung-Do Han, Jun Goo Kang
Diabetes Care.2023; 46(2): 305. CrossRef - A scoping review of artificial intelligence-based methods for diabetes risk prediction
Farida Mohsen, Hamada R. H. Al-Absi, Noha A. Yousri, Nady El Hajj, Zubair Shah
npj Digital Medicine.2023;[Epub] CrossRef - Development of Various Diabetes Prediction Models Using Machine Learning Techniques
Juyoung Shin, Jaewon Kim, Chanjung Lee, Joon Young Yoon, Seyeon Kim, Seungjae Song, Hun-Sung Kim
Diabetes & Metabolism Journal.2022; 46(4): 650. CrossRef - Prediction Model for Hypertension and Diabetes Mellitus Using Korean Public Health Examination Data (2002–2017)
Yong Whi Jeong, Yeojin Jung, Hoyeon Jeong, Ji Hye Huh, Ki-Chul Sung, Jeong-Hun Shin, Hyeon Chang Kim, Jang Young Kim, Dae Ryong Kang
Diagnostics.2022; 12(8): 1967. CrossRef - Improving Machine Learning Diabetes Prediction Models for the Utmost Clinical Effectiveness
Juyoung Shin, Joonyub Lee, Taehoon Ko, Kanghyuck Lee, Yera Choi, Hun-Sung Kim
Journal of Personalized Medicine.2022; 12(11): 1899. CrossRef
- Remnant Cholesterol Is an Independent Predictor of Type 2 Diabetes: A Nationwide Population-Based Cohort Study
First
Prev
