Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin Logs
<i>Background and Objectives</i>: The daily lifestyle management of diabetes requires accurate predictions of the blood glucose level between meals. The objective of this study was to improve the accuracy achieved by previous work, especially on the mid-term, i.e., 120 to 180 min predict...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-06-01
|
| Series: | Medicina |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1648-9144/57/7/676 |
| _version_ | 1797528173875822592 |
|---|---|
| author | Rebaz A. H. Karim István Vassányi István Kósa |
| author_facet | Rebaz A. H. Karim István Vassányi István Kósa |
| author_sort | Rebaz A. H. Karim |
| collection | DOAJ |
| description | <i>Background and Objectives</i>: The daily lifestyle management of diabetes requires accurate predictions of the blood glucose level between meals. The objective of this study was to improve the accuracy achieved by previous work, especially on the mid-term, i.e., 120 to 180 min prediction horizons, for insulin-dependent patients. <i>Materials and Methods</i>: An absorption model-based method is proposed to train an artificial neural network with the bolus and basal insulin dosing and timing, the baseline blood glucose level, the maximal glucose infusion rate, and the total carbohydrate content as parameters. The approach was implemented in various algorithmic setups, and it was validated on data from a small-scale clinical trial with continuous glucose monitoring. <i>Results</i>: Root mean square error results for the mid-term horizons are 1.72 mmol/L (120 min) and 1.95 mmol/L (180 min). The accuracy of the proposed model measured on the clinical data is better than the accuracy reported by any other currently available and comparable models. <i>Conclusions</i>: A relatively short (ca. two weeks) training sample of a continuous glucose monitor and dietary/insulin log is sufficient to provide accurate predictions. For the outpatient application in practice, a hybrid model is proposed that combines the present mid-term method with the authors’ previous work for short-term predictions. |
| first_indexed | 2024-03-10T09:54:25Z |
| format | Article |
| id | doaj.art-0eb980075eca4fbda2bde0b3791041b2 |
| institution | Directory Open Access Journal |
| issn | 1010-660X 1648-9144 |
| language | English |
| last_indexed | 2024-03-10T09:54:25Z |
| publishDate | 2021-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Medicina |
| spelling | doaj.art-0eb980075eca4fbda2bde0b3791041b22023-11-22T02:25:19ZengMDPI AGMedicina1010-660X1648-91442021-06-0157767610.3390/medicina57070676Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin LogsRebaz A. H. Karim0István Vassányi1István Kósa2Medical Informatics Research & Development Center, University of Pannonia, 8200 Veszprém, HungaryMedical Informatics Research & Development Center, University of Pannonia, 8200 Veszprém, HungaryCardiac Rehabilitation Institute of the Military Hospital, 8230 Balatonfüred, Hungary<i>Background and Objectives</i>: The daily lifestyle management of diabetes requires accurate predictions of the blood glucose level between meals. The objective of this study was to improve the accuracy achieved by previous work, especially on the mid-term, i.e., 120 to 180 min prediction horizons, for insulin-dependent patients. <i>Materials and Methods</i>: An absorption model-based method is proposed to train an artificial neural network with the bolus and basal insulin dosing and timing, the baseline blood glucose level, the maximal glucose infusion rate, and the total carbohydrate content as parameters. The approach was implemented in various algorithmic setups, and it was validated on data from a small-scale clinical trial with continuous glucose monitoring. <i>Results</i>: Root mean square error results for the mid-term horizons are 1.72 mmol/L (120 min) and 1.95 mmol/L (180 min). The accuracy of the proposed model measured on the clinical data is better than the accuracy reported by any other currently available and comparable models. <i>Conclusions</i>: A relatively short (ca. two weeks) training sample of a continuous glucose monitor and dietary/insulin log is sufficient to provide accurate predictions. For the outpatient application in practice, a hybrid model is proposed that combines the present mid-term method with the authors’ previous work for short-term predictions.https://www.mdpi.com/1648-9144/57/7/676mid-term blood glucose level predictionbasal insulinlifestyle support for diabetesoutpatient careartificial neural networks |
| spellingShingle | Rebaz A. H. Karim István Vassányi István Kósa Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin Logs Medicina mid-term blood glucose level prediction basal insulin lifestyle support for diabetes outpatient care artificial neural networks |
| title | Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin Logs |
| title_full | Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin Logs |
| title_fullStr | Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin Logs |
| title_full_unstemmed | Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin Logs |
| title_short | Improved Methods for Mid-Term Blood Glucose Level Prediction Using Dietary and Insulin Logs |
| title_sort | improved methods for mid term blood glucose level prediction using dietary and insulin logs |
| topic | mid-term blood glucose level prediction basal insulin lifestyle support for diabetes outpatient care artificial neural networks |
| url | https://www.mdpi.com/1648-9144/57/7/676 |
| work_keys_str_mv | AT rebazahkarim improvedmethodsformidtermbloodglucoselevelpredictionusingdietaryandinsulinlogs AT istvanvassanyi improvedmethodsformidtermbloodglucoselevelpredictionusingdietaryandinsulinlogs AT istvankosa improvedmethodsformidtermbloodglucoselevelpredictionusingdietaryandinsulinlogs |