Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients

Abstract Accurate prediction of blood glucose variations in type 2 diabetes (T2D) will facilitate better glycemic control and decrease the occurrence of hypoglycemic episodes as well as the morbidity and mortality associated with T2D, hence increasing the quality of life of patients. Owing to the co...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Yixiang Deng, Lu Lu, Laura Aponte, Angeliki M. Angelidi, Vera Novak, George Em Karniadakis, Christos S. Mantzoros
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Computer applications to medicine. Medical informatics R858-859.7
Acceso en línea:	https://doaj.org/article/880a0997af744705abe7461726f7f548
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:880a0997af744705abe7461726f7f548
record_format	dspace
spelling	oai:doaj.org-article:880a0997af744705abe7461726f7f5482021-12-02T16:09:40ZDeep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients10.1038/s41746-021-00480-x2398-6352https://doaj.org/article/880a0997af744705abe7461726f7f5482021-07-01T00:00:00Zhttps://doi.org/10.1038/s41746-021-00480-xhttps://doaj.org/toc/2398-6352Abstract Accurate prediction of blood glucose variations in type 2 diabetes (T2D) will facilitate better glycemic control and decrease the occurrence of hypoglycemic episodes as well as the morbidity and mortality associated with T2D, hence increasing the quality of life of patients. Owing to the complexity of the blood glucose dynamics, it is difficult to design accurate predictive models in every circumstance, i.e., hypo/normo/hyperglycemic events. We developed deep-learning methods to predict patient-specific blood glucose during various time horizons in the immediate future using patient-specific every 30-min long glucose measurements by the continuous glucose monitoring (CGM) to predict future glucose levels in 5 min to 1 h. In general, the major challenges to address are (1) the dataset of each patient is often too small to train a patient-specific deep-learning model, and (2) the dataset is usually highly imbalanced given that hypo- and hyperglycemic episodes are usually much less common than normoglycemia. We tackle these two challenges using transfer learning and data augmentation, respectively. We systematically examined three neural network architectures, different loss functions, four transfer-learning strategies, and four data augmentation techniques, including mixup and generative models. Taken together, utilizing these methodologies we achieved over 95% prediction accuracy and 90% sensitivity for a time period within the clinically useful 1 h prediction horizon that would allow a patient to react and correct either hypoglycemia and/or hyperglycemia. We have also demonstrated that the same network architecture and transfer-learning methods perform well for the type 1 diabetes OhioT1DM public dataset.Yixiang DengLu LuLaura AponteAngeliki M. AngelidiVera NovakGeorge Em KarniadakisChristos S. MantzorosNature PortfolioarticleComputer applications to medicine. Medical informaticsR858-859.7ENnpj Digital Medicine, Vol 4, Iss 1, Pp 1-13 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Computer applications to medicine. Medical informatics R858-859.7
spellingShingle	Computer applications to medicine. Medical informatics R858-859.7 Yixiang Deng Lu Lu Laura Aponte Angeliki M. Angelidi Vera Novak George Em Karniadakis Christos S. Mantzoros Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients
description	Abstract Accurate prediction of blood glucose variations in type 2 diabetes (T2D) will facilitate better glycemic control and decrease the occurrence of hypoglycemic episodes as well as the morbidity and mortality associated with T2D, hence increasing the quality of life of patients. Owing to the complexity of the blood glucose dynamics, it is difficult to design accurate predictive models in every circumstance, i.e., hypo/normo/hyperglycemic events. We developed deep-learning methods to predict patient-specific blood glucose during various time horizons in the immediate future using patient-specific every 30-min long glucose measurements by the continuous glucose monitoring (CGM) to predict future glucose levels in 5 min to 1 h. In general, the major challenges to address are (1) the dataset of each patient is often too small to train a patient-specific deep-learning model, and (2) the dataset is usually highly imbalanced given that hypo- and hyperglycemic episodes are usually much less common than normoglycemia. We tackle these two challenges using transfer learning and data augmentation, respectively. We systematically examined three neural network architectures, different loss functions, four transfer-learning strategies, and four data augmentation techniques, including mixup and generative models. Taken together, utilizing these methodologies we achieved over 95% prediction accuracy and 90% sensitivity for a time period within the clinically useful 1 h prediction horizon that would allow a patient to react and correct either hypoglycemia and/or hyperglycemia. We have also demonstrated that the same network architecture and transfer-learning methods perform well for the type 1 diabetes OhioT1DM public dataset.
format	article
author	Yixiang Deng Lu Lu Laura Aponte Angeliki M. Angelidi Vera Novak George Em Karniadakis Christos S. Mantzoros
author_facet	Yixiang Deng Lu Lu Laura Aponte Angeliki M. Angelidi Vera Novak George Em Karniadakis Christos S. Mantzoros
author_sort	Yixiang Deng
title	Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients
title_short	Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients
title_full	Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients
title_fullStr	Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients
title_full_unstemmed	Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients
title_sort	deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/880a0997af744705abe7461726f7f548
work_keys_str_mv	AT yixiangdeng deeptransferlearninganddataaugmentationimproveglucoselevelspredictionintype2diabetespatients AT lulu deeptransferlearninganddataaugmentationimproveglucoselevelspredictionintype2diabetespatients AT lauraaponte deeptransferlearninganddataaugmentationimproveglucoselevelspredictionintype2diabetespatients AT angelikimangelidi deeptransferlearninganddataaugmentationimproveglucoselevelspredictionintype2diabetespatients AT veranovak deeptransferlearninganddataaugmentationimproveglucoselevelspredictionintype2diabetespatients AT georgeemkarniadakis deeptransferlearninganddataaugmentationimproveglucoselevelspredictionintype2diabetespatients AT christossmantzoros deeptransferlearninganddataaugmentationimproveglucoselevelspredictionintype2diabetespatients
_version_	1718384421031116800

Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients

Ejemplares similares