Forecasting of Glucose Levels and Hypoglycemic Events: Head-to-Head Comparison of Linear and Nonlinear Data-Driven Algorithms Based on Continuous Glucose Monitoring Data Only
Open accessThe DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
Abstract
In type 1 diabetes management, the availability of algorithms capable of accurately forecasting future blood glucose (BG) concentrations and hypoglycemic episodes could enable proactive therapeutic actions, e.g., the consumption of carbohydrates to mitigate, or even avoid, an impending critical event. The only input of this kind of algorithm is often continuous glucose monitoring (CGM) sensor data, because other signals (such as injected insulin, ingested carbs, and physical activity) are frequently unavailable. Several predictive algorithms fed by CGM data only have been proposed in the literature, but they were assessed using datasets originated by different experimental protocols, making a comparison of their relative merits difficult. The aim of the present work was to perform a head-to-head comparison of thirty different linear and nonlinear predictive algorithms using the same dataset, given by 124 CGM traces collected over 10 days with the newest Dexcom G6 sensor available on the market and considering a 30-min prediction horizon. We considered the state-of-the art methods, investigating, in particular, linear black-box methods (autoregressive; autoregressive moving-average; and autoregressive integrated moving-average, ARIMA) and nonlinear machine-learning methods (support vector regression, SVR; regression random forest; feed-forward neural network, fNN; and long short-term memory neural network). For each method, the prediction accuracy and hypoglycemia detection capabilities were assessed using either population or individualized model parameters. As far as prediction accuracy is concerned, the results show that the best linear algorithm (individualized ARIMA) provides accuracy comparable to that of the best nonlinear algorithm (individualized fNN), with root mean square errors of 22.15 and 21.52 mg/dL, respectively. As far as hypoglycemia detection is concerned, the best linear algorithm (individualized ARIMA) provided precision = 64%, recall = 82%, and one false alarm/day, comparable to the best nonlinear technique (population SVR): precision = 63%, recall = 69%, and 0.5 false alarms/day. In general, the head-to-head comparison of the thirty algorithms fed by CGM data only made using a wide dataset shows that individualized linear models are more effective than population ones, while no significant advantages seem to emerge when employing nonlinear methodologies.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Saito, T., & Rehmsmeier, M. (2015) · PLOS ONE · doi:10.1371/journal.pone.0118432 · Cited by 5,309
Kernel methods in system identification, machine learning and function estimation: A survey
Pillonetto, G., Dinuzzo, F., Chen, T., De Nicolao, G., & Ljung, L. (2014) · Automatica · doi:10.1016/j.automatica.2014.01.001 · Cited by 741 · cited 3×
“, 100) and adds to the standard PEM cost function a regularization term representing a suitable prior on the unknown coefficients, which allows avoiding overfitting [32].”
BackgroundMethodResultInfluentialNon-Invasive Blood Glucose Monitoring Technology: A Review
Tang, L., Chang, S. J., Chen, C.-J., & Liu, J.-T. (2020) · Sensors · doi:10.3390/s20236925 · Cited by 481
Cappon, G., Vettoretti, M., Sparacino, G., & Facchinetti, A. (2019) · Diabetes & Metabolism Journal · doi:10.4093/dmj.2019.0121 · Cited by 477
Continuous glucose monitoring: A review of the technology and clinical use
Klonoff, D. C., Ahn, D., & Drincic, A. (2017) · Diabetes Research and Clinical Practice · doi:10.1016/j.diabres.2017.08.005 · Cited by 320
A review of personalized blood glucose prediction strategies for T1DM patients
Oviedo, S., Vehí, J., Calm, R., & Armengol, J. (2017) · International Journal for Numerical Methods in Biomedical Engineering · doi:10.1002/cnm.2833 · Cited by 279 · cited 1×
“Nonetheless, the focus of [19] was on a methodological review rather than on performing a head-to-head comparison on the same dataset.”
BackgroundPrevention of Nocturnal Hypoglycemia Using Predictive Alarm Algorithms and Insulin Pump Suspension
Buckingham, B., Chase, H. P., Dassau, E., Cobry, E., Clinton, P., Gage, V., Caswell, K., Wilkinson, J., Cameron, F., Lee, H., et al. (2010) · Diabetes Care · doi:10.2337/dc09-2303 · Cited by 169 · cited 3×
“, hypotreatments or pump suspension [25,37]) triggered 20–30 min before hypoglycemia are effective in avoiding/mitigating the episodes.”
BackgroundZecchin, C., Facchinetti, A., Sparacino, G., & Cobelli, C. (2016) · Journal of Diabetes Science and Technology · doi:10.1177/1932296816654161 · Cited by 64 · cited 3×
“Moreover, given the different nature of glucose fluctuations during the day and night (larger in the former case due to meal ingestion and less pronounced in the latter case) [14,20], we designed specific versions for these two time periods.”
BackgroundMethodSelf-citationCamerlingo, N., Vettoretti, M., Del Favero, S., Cappon, G., Sparacino, G., & Facchinetti, A. (2019) · Diabetes Technology & Therapeutics · doi:10.1089/dia.2019.0139 · Cited by 22 · cited 2×
“Furthermore, the employment of CGM to provide short-term predictions of future glucose values or to forecast forthcoming hypo/hyperglycemic episodes could lead to a further improvement, since targeted preventive measures—such as preventive hypotreatments (fast-acting carbohydrate consumption) [10] or correction insuli…”
BackgroundSelf-citation
Cited by
Enhancing the Capabilities of Continuous Glucose Monitoring With a Predictive App
Herrero, P., Andorrà, M., Babion, N., Bos, H., Koehler, M., Klopfenstein, Y., Leppäaho, E., Lustenberger, P., Peak, A., Ringemann, C., et al. (2024) · Journal of Diabetes Science and Technology · doi:10.1177/19322968241267818 · Cited by 31
Digital Solutions to Diagnose and Manage Postbariatric Hypoglycemia
Schönenberger, K. A., Cossu, L., Prendin, F., Cappon, G., Wu, J., Fuchs, K. L., Mayer, S., Herzig, D., Facchinetti, A., & Bally, L. (2022) · Frontiers in Nutrition · doi:10.3389/fnut.2022.855223 · Cited by 16
Self-citationPiersanti, A., Littero, M., Del Giudice, L. L., Marcantoni, I., Burattini, L., Tura, A., & Morettini, M. (2025) · Sensors · doi:10.3390/s25134038 · Cited by 5
Metadata
- Authors
- Prendin, F., Del Favero, S., Vettoretti, M., Sparacino, G., & Facchinetti, A.
- Venue
- Sensors
- Publisher
- MDPI AG
- Published
- 2021-02-27
- Volume
- 21
- Issue
- 5
- Pages
- 1647
- ISSN
- 1424-8220 (ISSN-L)
- OpenAlex
- W3135262141
- Cited by
- 72
- References
- 9
- FWCI
- 6.23
- Type
- journal article
- Language
- en
- Open access
- gold
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.