利用神经架构搜索和深度强化学习推进 1 型糖尿病患者的血糖预测

IF 5.3 2区 医学 Q1 ENGINEERING, BIOMEDICAL
Peter Domanski , Aritra Ray , Kyle Lafata , Farshad Firouzi , Krishnendu Chakrabarty , Dirk Pflüger
{"title":"利用神经架构搜索和深度强化学习推进 1 型糖尿病患者的血糖预测","authors":"Peter Domanski ,&nbsp;Aritra Ray ,&nbsp;Kyle Lafata ,&nbsp;Farshad Firouzi ,&nbsp;Krishnendu Chakrabarty ,&nbsp;Dirk Pflüger","doi":"10.1016/j.bbe.2024.07.006","DOIUrl":null,"url":null,"abstract":"<div><p>For individuals with Type-1 diabetes mellitus, accurate prediction of future blood glucose values is crucial to aid its regulation with insulin administration, tailored to the individual’s specific needs. The authors propose a novel approach for the integration of a neural architecture search framework with deep reinforcement learning to autonomously generate and train architectures, optimized for each subject over model size and analytical prediction performance, for the blood glucose prediction task in individuals with Type-1 diabetes. The authors evaluate the proposed approach on the OhioT1DM dataset, which includes blood glucose monitoring records at 5-min intervals over 8 weeks for 12 patients with Type-1 diabetes mellitus. Prior work focused on predicting blood glucose levels in 30 and 45-min prediction horizons, equivalent to 6 and 9 data points, respectively. Compared to the previously achieved best error, the proposed method demonstrates improvements of 18.4 % and 22.5 % on average for mean absolute error in the 30-min and 45-min prediction horizons, respectively, through the proposed deep reinforcement learning framework. Using the deep reinforcement learning framework, the best-case and worst-case analytical performance measured over root mean square error and mean absolute error was obtained for subject ID 570 and subject ID 584, respectively. Models optimized for performance on the prediction task and model size were obtained after implementing neural architecture search in conjunction with deep reinforcement learning on these two extreme cases. The authors demonstrate improvements of 4.8 % using Long Short Term Memory-based architectures and 5.7 % with Gated Recurrent Units-based architectures for patient ID 570 on the analytical prediction performance by integrating neural architecture search with deep reinforcement learning framework. The patient with the lowest performance (ID 584) on the deep reinforcement learning method had an even greater performance boost, with improvements of 10.0 % and 12.6 % observed for the Long Short-Term Memory and Gated Recurrent Units, respectively. The subject-specific optimized models over performance and model size from the neural architecture search in conjunction with deep reinforcement learning had a reduction in model size which ranged from 20 to 150 times compared to the model obtained using only the deep reinforcement learning method. The smaller size, indicating a reduction in model complexity in terms of the number of trainable network parameters, was achieved without a loss in the prediction performance.</p></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"44 3","pages":"Pages 481-500"},"PeriodicalIF":5.3000,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S0208521624000536/pdfft?md5=93b6aff09e56179150aed20a868b9e84&pid=1-s2.0-S0208521624000536-main.pdf","citationCount":"0","resultStr":"{\"title\":\"Advancing blood glucose prediction with neural architecture search and deep reinforcement learning for type 1 diabetics\",\"authors\":\"Peter Domanski ,&nbsp;Aritra Ray ,&nbsp;Kyle Lafata ,&nbsp;Farshad Firouzi ,&nbsp;Krishnendu Chakrabarty ,&nbsp;Dirk Pflüger\",\"doi\":\"10.1016/j.bbe.2024.07.006\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><p>For individuals with Type-1 diabetes mellitus, accurate prediction of future blood glucose values is crucial to aid its regulation with insulin administration, tailored to the individual’s specific needs. The authors propose a novel approach for the integration of a neural architecture search framework with deep reinforcement learning to autonomously generate and train architectures, optimized for each subject over model size and analytical prediction performance, for the blood glucose prediction task in individuals with Type-1 diabetes. The authors evaluate the proposed approach on the OhioT1DM dataset, which includes blood glucose monitoring records at 5-min intervals over 8 weeks for 12 patients with Type-1 diabetes mellitus. Prior work focused on predicting blood glucose levels in 30 and 45-min prediction horizons, equivalent to 6 and 9 data points, respectively. Compared to the previously achieved best error, the proposed method demonstrates improvements of 18.4 % and 22.5 % on average for mean absolute error in the 30-min and 45-min prediction horizons, respectively, through the proposed deep reinforcement learning framework. Using the deep reinforcement learning framework, the best-case and worst-case analytical performance measured over root mean square error and mean absolute error was obtained for subject ID 570 and subject ID 584, respectively. Models optimized for performance on the prediction task and model size were obtained after implementing neural architecture search in conjunction with deep reinforcement learning on these two extreme cases. The authors demonstrate improvements of 4.8 % using Long Short Term Memory-based architectures and 5.7 % with Gated Recurrent Units-based architectures for patient ID 570 on the analytical prediction performance by integrating neural architecture search with deep reinforcement learning framework. The patient with the lowest performance (ID 584) on the deep reinforcement learning method had an even greater performance boost, with improvements of 10.0 % and 12.6 % observed for the Long Short-Term Memory and Gated Recurrent Units, respectively. The subject-specific optimized models over performance and model size from the neural architecture search in conjunction with deep reinforcement learning had a reduction in model size which ranged from 20 to 150 times compared to the model obtained using only the deep reinforcement learning method. The smaller size, indicating a reduction in model complexity in terms of the number of trainable network parameters, was achieved without a loss in the prediction performance.</p></div>\",\"PeriodicalId\":55381,\"journal\":{\"name\":\"Biocybernetics and Biomedical Engineering\",\"volume\":\"44 3\",\"pages\":\"Pages 481-500\"},\"PeriodicalIF\":5.3000,\"publicationDate\":\"2024-07-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.sciencedirect.com/science/article/pii/S0208521624000536/pdfft?md5=93b6aff09e56179150aed20a868b9e84&pid=1-s2.0-S0208521624000536-main.pdf\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Biocybernetics and Biomedical Engineering\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0208521624000536\",\"RegionNum\":2,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, BIOMEDICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Biocybernetics and Biomedical Engineering","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0208521624000536","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, BIOMEDICAL","Score":null,"Total":0}
引用次数: 0

摘要

对于 1 型糖尿病患者来说,准确预测未来的血糖值至关重要,这有助于根据患者的具体需求使用胰岛素进行调节。作者提出了一种将神经架构搜索框架与深度强化学习相结合的新方法,用于自主生成和训练架构,针对每个受试者的模型大小和分析预测性能进行优化,以完成 1 型糖尿病患者的血糖预测任务。作者在 OhioT1DM 数据集上对所提出的方法进行了评估,该数据集包括 12 名 1 型糖尿病患者 8 周内 5 分钟间隔的血糖监测记录。之前的工作侧重于预测 30 分钟和 45 分钟预测范围内的血糖水平,分别相当于 6 个和 9 个数据点。与之前达到的最佳误差相比,通过所提出的深度强化学习框架,所提出的方法在 30 分钟和 45 分钟预测范围内的平均绝对误差分别平均提高了 18.4% 和 22.5%。利用深度强化学习框架,ID 570 和 ID 584 分别获得了以均方根误差和平均绝对误差衡量的最佳和最差分析性能。在这两个极端案例上结合深度强化学习实施神经架构搜索后,获得了预测任务性能和模型大小的优化模型。作者证明,通过将神经架构搜索与深度强化学习框架相结合,使用基于长短期记忆的架构和基于门控递归单元的架构,ID 570 患者的分析预测性能分别提高了 4.8% 和 5.7%。深度强化学习方法性能最低的患者(ID 584)的性能提升幅度更大,长短期记忆和门控递归单元的性能分别提高了 10.0% 和 12.6%。与仅使用深度强化学习方法获得的模型相比,通过神经架构搜索和深度强化学习获得的特定主题优化模型在性能和模型大小上减少了 20 到 150 倍。模型规模的缩小表明,就可训练网络参数的数量而言,模型的复杂性有所降低,但预测性能并没有降低。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Advancing blood glucose prediction with neural architecture search and deep reinforcement learning for type 1 diabetics

Advancing blood glucose prediction with neural architecture search and deep reinforcement learning for type 1 diabetics

For individuals with Type-1 diabetes mellitus, accurate prediction of future blood glucose values is crucial to aid its regulation with insulin administration, tailored to the individual’s specific needs. The authors propose a novel approach for the integration of a neural architecture search framework with deep reinforcement learning to autonomously generate and train architectures, optimized for each subject over model size and analytical prediction performance, for the blood glucose prediction task in individuals with Type-1 diabetes. The authors evaluate the proposed approach on the OhioT1DM dataset, which includes blood glucose monitoring records at 5-min intervals over 8 weeks for 12 patients with Type-1 diabetes mellitus. Prior work focused on predicting blood glucose levels in 30 and 45-min prediction horizons, equivalent to 6 and 9 data points, respectively. Compared to the previously achieved best error, the proposed method demonstrates improvements of 18.4 % and 22.5 % on average for mean absolute error in the 30-min and 45-min prediction horizons, respectively, through the proposed deep reinforcement learning framework. Using the deep reinforcement learning framework, the best-case and worst-case analytical performance measured over root mean square error and mean absolute error was obtained for subject ID 570 and subject ID 584, respectively. Models optimized for performance on the prediction task and model size were obtained after implementing neural architecture search in conjunction with deep reinforcement learning on these two extreme cases. The authors demonstrate improvements of 4.8 % using Long Short Term Memory-based architectures and 5.7 % with Gated Recurrent Units-based architectures for patient ID 570 on the analytical prediction performance by integrating neural architecture search with deep reinforcement learning framework. The patient with the lowest performance (ID 584) on the deep reinforcement learning method had an even greater performance boost, with improvements of 10.0 % and 12.6 % observed for the Long Short-Term Memory and Gated Recurrent Units, respectively. The subject-specific optimized models over performance and model size from the neural architecture search in conjunction with deep reinforcement learning had a reduction in model size which ranged from 20 to 150 times compared to the model obtained using only the deep reinforcement learning method. The smaller size, indicating a reduction in model complexity in terms of the number of trainable network parameters, was achieved without a loss in the prediction performance.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
CiteScore
16.50
自引率
6.20%
发文量
77
审稿时长
38 days
期刊介绍: Biocybernetics and Biomedical Engineering is a quarterly journal, founded in 1981, devoted to publishing the results of original, innovative and creative research investigations in the field of Biocybernetics and biomedical engineering, which bridges mathematical, physical, chemical and engineering methods and technology to analyse physiological processes in living organisms as well as to develop methods, devices and systems used in biology and medicine, mainly in medical diagnosis, monitoring systems and therapy. The Journal''s mission is to advance scientific discovery into new or improved standards of care, and promotion a wide-ranging exchange between science and its application to humans.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信