Subject-Conditioned Glucose Forecasting in Type-1 Diabetes

Davide Marelli; Giorgia Rigamonti; Mirko Paolo Barbato; Paolo Napoletano

REVIEW 5 major objections 6 minor 31 references

Subject-Conditioned Glucose Forecasting in Type-1 Diabetes

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that explicit subject conditioning improves glucose forecasts and adverse-event detection in type 1 diabetes.

desk verdict New SC-RoPE conditioning idea, but the LOSOCV protocol has a load-bearing ambiguity that could make the reported gains leakage artifacts; the abstract also overstates the results. read the letter →

arxiv 2607.19006 v1 pith:35LBN67H submitted 2026-07-21 cs.LG q-bio.QM

Giorgia Rigamonti , Mirko Paolo Barbato , Davide Marelli , Paolo Napoletano This is my paper · ORCID

classification cs.LGq-bio.QM

keywords Type1diabetesbloodglucosepredictionsubjectconditioningpersonalizedforecastingcontinuousmonitoringadverseeventdetectionmultimodaldeeplearningrotarypositionalembedding

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Type 1 diabetes management depends on anticipating dangerously low and high blood glucose before they occur. This paper proposes SCGP, an architecture that learns a compact subject signature from each patient's basal insulin, bolus insulin, and carbohydrate records, then uses that signature to condition a glucose-forecasting network so predictions are tailored to the individual. The paper claims SCGP consistently improves adverse-event detection, especially hypoglycemia sensitivity, and yields earlier predictions (higher time gain) while keeping RMSE on par with established personalized baselines on two benchmark datasets at 30- and 60-minute horizons. If the claim holds, personalization can be achieved with only CGM data at inference time, once the subject signature is fixed.

What carries the argument

The load-bearing mechanism is SC-RoPE (subject-conditioned rotary positional embedding), which modulates the temporal glucose representation by subject identity: each time step's features are split into even and odd components and rotated by angle $\theta_{i,t} = m_t \cdot e_i$, where $m_t$ is the temporal position and $e_i$ is the subject embedding. Alongside it, a Bayesian Gaussian mixture model (BGMM) aggregates per-subject embeddings into a Gaussian descriptor $\mathbf{g}_i = [\mu_i \parallel \text{diag}(\Sigma_i)]$, so each person is summarized by both a central tendency and a variability profile. Together they let the same glucose encoder produce subject-adapted forecasts while keeping glucose dynamics themselves subject-independent.

What would settle it

Re-run the leave-one-subject-out evaluation with held-out subjects' signatures computed only from their training-period basal, bolus, and carbohydrate data (or from other subjects only), and check whether the hypoglycemia-sensitivity and time-gain improvements survive; if they do not, the central claim would be refuted.

Watch

Extended reading notes

Core claim

SCGP's central claim is that separating subject characterization from glucose dynamics modeling improves personalized blood glucose forecasting. A recurrent encoder maps each subject's contextual signals (basal insulin, bolus insulin, carbohydrates) to a 64-dimensional embedding trained with ArcFace loss; a Bayesian Gaussian mixture model (BGMM) converts those embeddings into a probabilistic subject descriptor; and a subject-conditioned rotary positional embedding (SC-RoPE) injects the descriptor into the temporal glucose representation by rotating even and odd feature components by angles that depend on both subject identity and temporal position. The paper reports that this design outperforms patient-independent models and two patient-specific baselines in hypo- and hyperglycemia sensitivity at 30- and 60-minute prediction horizons, with better time gain and comparable RMSE, and that predictions remain clinically safe under Clarke Error Grid analysis. The paper reads this as evidence that explicit subject conditioning, not implicit fine-tuning, is the effective ingredient.

Load-bearing premise

The load-bearing assumption is that the subject signature for a held-out patient is built without using that patient's future test-period records; if test-period basal, bolus, or carbohydrate data leak into the signature, the reported gains could come from seeing the future rather than from personalization.

Editorial extensions

If this is right

At inference time only CGM readings are required, so forecasts could be generated even when insulin and meal records are unavailable or unreliable, provided a signature already exists.
The reported improvement in hypoglycemia sensitivity at the 60-minute horizon suggests earlier warnings of dangerous low-glucose events, with only a small increase in false alarms.
Separating subject representation from temporal modeling is a general design for personalized time-series forecasting; the same two-module pattern could apply to other physiological or behavioral signals.
The probabilistic subject descriptor expresses uncertainty about which patient group a person resembles, which could feed into confidence-aware clinical alarms.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A decisive unstated condition is whether the subject signature for a new patient can be built from a short observation period; if yes, the method is deployable without historical context, and if no, a data-collection phase is required before personalized forecasts begin.
Replacing the BGMM with the raw 64-dimensional embedding in an ablation would reveal whether the probabilistic aggregation is contributing the improvement or whether SC-RoPE alone carries it.
The small specificity drop under personalization implies the clinical benefit depends on alarm-fatigue tolerance; the paper reports false alarm rates but does not compute a net utility across that trade-off.
The same conditioning mechanism could be tested with other subject signatures, such as summaries of heart rate or physical activity, which the paper does not explore.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The genuinely new thing here is SC-RoPE: conditioning the rotation angle of a rotary positional encoding on a subject embedding. That is a real contribution, and the idea of separating subject characterization from glucose dynamics is sensible. The GitHub link, two datasets, an ablation study, and a clinical error-grid analysis are all assets.

The soft spots are substantial. The biggest one is the LOSOCV protocol. Section II-F says subject embeddings are obtained from the external-feature encoder, aggregated via BGMM, and injected through SC-RoPE, but it never states which portion of a held-out subject's basal, bolus, and carbohydrate data is used to compute that embedding. The paper says inference relies exclusively on CGM data and also mentions an "initial observation phase," which suggests the training portion, but it is not explicit. If the external-feature encoder runs on the held-out subject's full record including test windows, then the conditioning signal contains future information, and the reported detection gains are leakage artifacts. This is load-bearing because the entire evaluation rests on the separation between observation and test phases.

Second, the abstract says SCGP "consistently improves forecasting performance." Table I contradicts that. On RMSE, SCGP-specific often ties or loses to the CNN and Bi-GRU baselines (e.g., OhioT1DM PH=30: 12.86 vs 12.85 for CNN; DiaTrend PH=60: 24.10 vs 23.74). The real advantages are in hypoglycemia sensitivity and time gain, not overall accuracy. The abstract needs to be tempered.

Third, there are no significance tests. Given the large per-patient variance, many of these differences could be noise. Fourth, the comparison set omits the SOTA models cited in the introduction (GluFormer, TimesNet, BG-BERT), so the claim of outperforming existing methods is not properly tested. Fifth, the composite loss includes terms that directly penalize clinical-zone errors and temporal lag, while Table I reports event sensitivity and Time Gain. That means part of the reported advantage is baked into the objective. A fair comparison would give baselines the same loss or ablate the loss terms.

None of this is fatal. The SC-RoPE idea is worth pursuing, and the evaluation issues are fixable. A clear protocol statement, matched-loss baselines, significance testing, and a corrected abstract would turn this into a solid contribution. The paper deserves peer review, but the referee should demand that clarification and re-analysis before acceptance.

Referee Report

5 major / 6 minor

Summary. The paper proposes SCGP, a multimodal deep learning architecture for personalized blood glucose forecasting in Type 1 Diabetes. Subject-specific embeddings are learned from basal insulin, bolus insulin, and carbohydrate data using a GRU encoder with ArcFace loss, then aggregated via a Bayesian Gaussian Mixture Model and injected into a glucose regression network through a subject-conditioned rotary positional embedding (SC-RoPE). The architecture is evaluated on OhioT1DM and DiaTrend under LOSOCV and fine-tuning, comparing patient-independent and patient-specific configurations against adapted CNN and Bi-GRU baselines. The central claim is that explicit subject conditioning improves adverse-event detection and time gain while maintaining accurate predictions across 30- and 60-minute horizons, using only CGM data at inference.

Significance. If the validity of the evaluation is established, SCGP would contribute a novel and plausible mechanism for explicit subject conditioning in glucose forecasting, with source code and two benchmark datasets supporting reproducibility. The disentanglement of subject characterization from glucose dynamics, combined with the probabilistic BGMM and SC-RoPE modulation, is a technically interesting design. However, the current evidence for the central claims is weakened by an under-specified protocol for computing held-out subject embeddings, a composite loss that directly optimizes the reported clinical metrics, and a mismatch between the abstract's "consistently improves" and the RMSE rows of Table I. These issues prevent the results from being taken at face value and require substantial revision.

major comments (5)

[Section II-F and III-D] The LOSOCV protocol does not specify whether the subject embedding for a held-out subject is computed from the entire record, including the future test-period basal, bolus, and carbohydrate data, or only from data strictly preceding the test window. This is load-bearing because the paper's inference-time claim 'relies exclusively on CGM measurements' and the causal interpretation of subject conditioning both require that no test-period contextual information enters the embedding. If the external-feature encoder processes the held-out subject's full contextual record, the reported gains in event sensitivity and time gain could be leakage artifacts. The authors must state exactly which data are used to compute each held-out subject's embedding and, if the full record is currently used, re-run the evaluation with embeddings derived only from the observation phase (e.g., the training portion before the first test timestamp).
[Section II-G and Section III-B (Table I)] The composite loss in Eq. (10) includes L_e and L_l, which directly penalize clinical-zone prediction errors and expected temporal lag, while Table I reports Hypo/Hyper Sen and Time Gain as outcomes. The baselines (CNN and Bi-GRU) are trained with their original losses, which do not contain these terms. Consequently, the observed improvements in event sensitivity and time gain are partly built into the SCGP objective rather than arising solely from subject conditioning. To support the claim that the architecture is responsible for the gains, the authors should either train all models with the same composite loss or ablate L_e and L_l from SCGP and show that the event-detection advantage persists.
[Section III-B, Table I and Abstract] The statement that 'personalized models achieve lower RMSE and higher Time Gain than patient-independent counterparts' is contradicted by the table itself. For example, on OhioT1DM at PH=30, SCGP+ has RMSE 12.86 versus SCGP- 12.46, and Bi-GRU+ has 13.66 versus 12.85; on DiaTrend at PH=30, SCGP+ RMSE is 14.76 versus 14.53. The abstract's claim that SCGP 'consistently improves forecasting performance' is therefore an overstatement. The authors should either recalibrate the claims to focus on the event-detection and time-gain trade-off or provide a clinical utility analysis justifying RMSE increases in exchange for improved sensitivity.
[Section III-B] No statistical significance tests are reported for any metric. The reported standard deviations across subjects are large (e.g., OhioT1DM PH=60 Hypo Sen 63.98±22.20 for SCGP+ vs 57.97±20.79 for SCGP-), and many differences are plausibly within noise. The authors should report paired per-subject differences and appropriate non-parametric tests (e.g., Wilcoxon signed-rank test across LOSOCV folds) for the key metrics, and temper qualitative claims such as 'consistently improving' and 'marked improvement' accordingly.
[Introduction and Section III-B] The introduction states that SCGP 'outperforms existing methods in the early detection of hypo- and hyper-glycemic events,' but the experimental comparison is limited to two adapted baselines (a CNN and a Bi-GRU). No comparison is provided against the cited state-of-the-art models such as GluFormer, BG-BERT, LST, Informer, or TimesNet, nor are published results on OhioT1DM used as reference points. The claim should be scoped to the two baselines actually evaluated, or the benchmark should be extended to include at least one recent strong baseline with comparable experimental conditions.

minor comments (6)

[Table I footnote] The footnote reads '* Inboldthe best score'; it should be '* In bold the best score.'
[Footnote 1] The source-code footnote is malformed: '1/githubSource code available at https://github.com/unimib-islab/SCGP.' should be cleaned up to read 'Source code available at ...' as a proper footnote.
[Section II-E, Eq. (7)] In Eq. (7), θ_i,t = m_t · e_i is ambiguous because e_i is a vector and the subsequent element-wise operations in Eqs. (8)-(9) require θ_i,t to be a vector; the notation '·' should be clarified as element-wise multiplication or a broadcast scalar product.
[Section II-E, Eqs. (4)-(6)] The concatenation symbol '∥' in Eq. (6) is not defined; it should be introduced in the text as vector concatenation.
[Section III-A, Time Gain definition] The delay formula uses an undefined loss L in 'delay(y, ŷ) = argmin_k Σ_i L(y_i - ŷ_{i-k})²'; the loss should be specified as squared error.
[Section III-B, FAR discussion] The text states that SCGP patient-specific configurations show only marginal false-alarm-rate increases (+0.1% at PH=30, +0.4% at PH=60), but FAR values are not reported in Table I; either include the FAR numbers or omit the quantitative claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the claimed improvements are empirical comparisons on held-out data, and the composite loss terms are surrogates, not identical to the reported metrics.

full rationale

The derivation chain is empirical rather than formal: subject embeddings are learned from contextual data with an ArcFace classification objective, aggregated through a BGMM, injected via SC-RoPE, and the glucose forecaster is trained, validated, and tested on held-out temporal and subject splits of OhioT1DM and DiaTrend. The reported gains are measured on held-out subjects under LOSOCV against external benchmarks and baselines, so the central claim is not an identity or a renamed input. The composite loss L contains event-aware and lag-aware terms, and Table I reports event sensitivity and Time Gain; however, L_e and L_l are surrogate training objectives, not the same functions as the evaluation metrics: the lag loss is a softmin-expected shift rather than Time Gain's argmin delay, and the event penalty is a margin-based asymmetric error, not the thresholded sensitivity score. This makes the comparison to baselines without those loss terms meaningful rather than circular. Self-citations [12], [13], and [23] are used for preprocessing, baselines, and prior architecture choices, not as the evidence for the central claim, so they are not load-bearing. The LOSOCV ambiguity about whether held-out subject embeddings use test-period contextual data is a possible leakage or correctness issue, not a circularity of definition, and is outside this pass.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on several unverified modeling assumptions. The most load-bearing is that the subject descriptor for a held-out subject is computed without future leakage. The paper also assumes the ArcFace embedding transfers to unseen subjects and that SMOTE does not corrupt temporal structure. Hyperparameters such as loss weights, the BGMM component count, and the safety margin are set by hand, and some are not reported.

free parameters (5)

loss weights lambda_s, lambda_e, lambda_l = 0.4, 0.6, 0.07
Manual weighting of the composite loss; affects the balance between accuracy, robustness, event penalty, and lag alignment.
BGMM component count K = not reported
The number of Gaussian components in Eq. (1) is a free modeling choice and is never stated in the paper.
max lag K in L_l = 3 time steps
Sets the temporal tolerance in the lag-aware loss; directly influences Time Gain.
safety margin delta in L_e = not reported
Controls how much error is tolerated before the clinical penalty is applied; not specified in the paper.
subject embedding dimension d = 64
Dimensionality of the subject signature; an architectural choice that affects the capacity of the conditioning signal.

assumptions (4)

ad hoc to paper The subject embedding for a held-out subject is computed from contextual data that do not include the test-period signal.
Central to the LOSOCV protocol; if violated, personalization leaks future information into forecasts.
domain assumption The ArcFace-trained embedding space transfers to subjects not in the classification training set.
Stage 1 trains a classifier on known subjects; LOSOCV requires meaningful embeddings for unseen subjects.
domain assumption SMOTE oversampling does not break temporal dependencies or create artificial trajectories that distort evaluation.
SMOTE is applied to time-series training sets, but no safeguards or analysis are provided for temporal validity.
standard math Variational BGMM inference with conjugate priors converges to a descriptor that is stable across folds.
The model relies on standard variational Bayesian inference, but hyperparameters such as W0, nu0, and beta0 are not reported.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Subject-Conditioned Glucose Forecasting in Type-1 Diabetes." pith.science (2026). https://pith.science/paper/35LBN67H

@misc{pith2026260719006,
  author       = {Pith},
  title        = {Pith review of: Subject-Conditioned Glucose Forecasting in Type-1 Diabetes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/35LBN67H}},
  note         = {Machine review of arXiv:2607.19006}
}

read the original abstract

Accurate forecasting of blood glucose concentration is key in the management of Type 1 Diabetes, facilitating early detection of adverse glycemic events and supporting timely therapeutic interventions. Despite recent advances in glucose prediction, most existing approaches rely on population-level representations or implicit personalization strategies that fail to deliver effective subject-specific forecasts. In this work, we propose Subject-Conditioned Glucose Prediction (SCGP), a novel multimodal deep learning architecture conceived for personalized blood glucose prediction. SCGP conditions glucose predictions based on observed glucose data and a compact subject-specific representation learned from contextual information. By explicitly separating subject characterization from glucose dynamics modeling and avoiding early fusion of heterogeneous inputs, the proposed framework effectively captures inter-subject variability while preserving robust and reliable temporal modeling. Experiments on two state-of-the-art benchmark datasets demonstrate that SCGP consistently improves forecasting performance, enabling reliable detection of adverse glycemic events across multiple prediction horizons, highlighting the benefits of explicit subject conditioning for personalized diabetes management.

Figures

Figures reproduced from arXiv: 2607.19006 by the authors.

**Figure 1.** Architecture of the proposed Subject-Conditioned Glucose Prediction framework. Subject embeddings learned from contextual signals are probabilistically modeled via a Bayesian GMM and injected into the glucose forecasting network through SC-RoPE. constructed time series from continuous glucose monitoring intervals. Linear interpolation enforced a uniform five-minute sampling interval and filled occasional CGM gaps. … view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 29 canonical work pages

[1]

Type 1 diabetes,

M. A. Atkinson, G. S. Eisenbarth, and A. W. Michels, “Type 1 diabetes,” The lancet, vol. 383, no. 9911, pp. 69–82, 2014

work page 2014
[2]

Continuous glucose monitoring: a review of successes, challenges, and opportunities,

D. Rodbard, “Continuous glucose monitoring: a review of successes, challenges, and opportunities,”Diabetes technology & therapeutics, vol. 18, no. S2, pp. S2–3, 2016

work page 2016
[3]

Continuous subcutaneous insulin infusion: a comprehensive review of insulin pump therapy,

M. J. Lenhard and G. D. Reeves, “Continuous subcutaneous insulin infusion: a comprehensive review of insulin pump therapy,”Archives of Internal Medicine, vol. 161, no. 19, pp. 2293–2300, 2001

work page 2001
[4]

Engineering digital biomarkers of inter- stitial glucose from noninvasive smartwatches,

B. Bent, P. J. Cho, M. Henriquez, A. Wittmann, C. Thacker, M. Feinglos, M. J. Crowley, and J. P. Dunn, “Engineering digital biomarkers of inter- stitial glucose from noninvasive smartwatches,”NPJ Digital Medicine, vol. 4, no. 1, p. 89, 2021

work page 2021
[5]

Arima models,

R. H. Shumway, D. S. Stoffer, R. H. Shumway, and D. S. Stoffer, “Arima models,”Time series analysis and its applications: with R examples, pp. 75–163, 2017

work page 2017
[6]

Blood glucose prediction model for type 1 diabetes based on extreme gradient boosting,

G. Alfian, M. Syafrudin, J. Rhee, M. Anshari, M. Mustakim, and I. Fahrurrozi, “Blood glucose prediction model for type 1 diabetes based on extreme gradient boosting,” inIOP Conference Series: Materials Science and Engineering, vol. 803, no. 1. IOP Publishing, 2020, p. 012012

work page 2020
[7]

Blood glucose level prediction as time-series modeling using sequence-to-sequence neural networks,

A. Bhimireddy, P. Sinha, B. Oluwalade, J. W. Gichoya, and S. Purkayastha, “Blood glucose level prediction as time-series modeling using sequence-to-sequence neural networks,” inKDH@ECAI, 2020, pp. 125–130

work page 2020
[8]

Bgformer: An improved informer model to enhance blood glucose prediction,

Y . Xue, S. Guan, and W. Jia, “Bgformer: An improved informer model to enhance blood glucose prediction,”Journal of Biomedical Informatics, vol. 157, p. 104715, 2024

work page 2024

Show all 31 references

[9]

Gluformer: Transformer-based personalized glucose forecasting with uncertainty quantification,

R. Sergazinov, M. Armandpour, and I. Gaynanova, “Gluformer: Transformer-based personalized glucose forecasting with uncertainty quantification,” inICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing. IEEE, 2023, pp. 1–5

2023
[10]

Blood glucose prediction algorithm via timesnet-mpinn,

H. Zhang, M. Wang, and R. Song, “Blood glucose prediction algorithm via timesnet-mpinn,” in2025 IEEE International Annual Conference on Complex Systems and Intelligent Science (CSIS-IAC). IEEE, 2025, pp. 763–768

2025
[11]

Predicting adverse events for patients with type-1 diabetes via self-supervised learning,

X. Zheng, S. Ji, and C. Wu, “Predicting adverse events for patients with type-1 diabetes via self-supervised learning,” inICASSP 2024- 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2024, pp. 1526–1530. 1https://fondazioneanthem.it/

2024
[12]

Improving detection of type-1 diabetes adverse events using gru networks,

G. Rigamonti, M. P. Barbato, D. Marelli, and P. Napoletano, “Improving detection of type-1 diabetes adverse events using gru networks,” in2024 IEEE 8th Forum on Research and Technologies for Society and Industry Innovation (RTSI). IEEE, 2024, pp. 79–84

2024
[13]

Lightweight sequential transformers for blood glucose level prediction in type-1 diabetes,

M. P. Barbato, G. Rigamonti, D. Marelli, and P. Napoletano, “Lightweight sequential transformers for blood glucose level prediction in type-1 diabetes,”IEEE Journal of Biomedical and Health Informatics, pp. 1–11, 2025

2025
[14]

Neural network incorporating meal information improves accuracy of short-time prediction of glucose concentration,

C. Zecchin, A. Facchinetti, G. Sparacino, G. De Nicolao, and C. Cobelli, “Neural network incorporating meal information improves accuracy of short-time prediction of glucose concentration,”IEEE Transactions on Biomedical Engineering, vol. 59, no. 6, pp. 1550–1560, 2012

2012
[15]

A review of personalized blood glucose prediction strategies for t1dm patients,

S. Oviedo, J. Veh ´ı, R. Calm, and J. Armengol, “A review of personalized blood glucose prediction strategies for t1dm patients,”International journal for numerical methods in biomedical engineering, vol. 33, no. 6, p. e2833, 2017

2017
[16]

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

Y . Deng, L. Lu, L. Aponte, A. M. Angelidi, V . Novak, G. E. Karniadakis, and C. S. Mantzoros, “Deep transfer learning and data augmentation improve glucose levels prediction in type 2 diabetes patients,”NPJ Digital Medicine, vol. 4, no. 1, p. 109, 2021

2021
[17]

Metabolic interventions as adjunctive therapies to insulin in type 1 diabetes: Current clinical landscape and perspectives,

J. Podobnik and K. J. Prentice, “Metabolic interventions as adjunctive therapies to insulin in type 1 diabetes: Current clinical landscape and perspectives,”Diabetes, Obesity and Metabolism, 2025

2025
[18]

A per- sonalized blood glucose level prediction model with a fine-tuning strategy: A proof-of-concept study,

W. Seo, S.-W. Park, N. Kim, S.-M. Jin, and S.-M. Park, “A per- sonalized blood glucose level prediction model with a fine-tuning strategy: A proof-of-concept study,”Computer methods and programs in biomedicine, vol. 211, p. 106424, 2021

2021
[19]

A meta-learning approach to personalized blood glucose prediction in type 1 diabetes,

S. Langarica, M. Rodriguez-Fernandez, F. N ´u˜nez, and F. J. Doyle III, “A meta-learning approach to personalized blood glucose prediction in type 1 diabetes,”Control Engineering Practice, vol. 135, p. 105498, 2023

2023
[20]

Personalized blood glucose prediction for type 1 diabetes using evidential deep learning and meta- learning,

T. Zhu, K. Li, P. Herrero, and P. Georgiou, “Personalized blood glucose prediction for type 1 diabetes using evidential deep learning and meta- learning,”IEEE Transactions on Biomedical Engineering, vol. 70, no. 1, pp. 193–204, 2023

2023
[21]

Deep learning-based glucose prediction models: a guide for practitioners and a curated dataset for improved diabetes management,

S. Langarica, D. de la Vega, N. Cariman, M. Miranda, D. C. Andrade, F. N ´u˜nez, and M. Rodriguez-Fernandez, “Deep learning-based glucose prediction models: a guide for practitioners and a curated dataset for improved diabetes management,”IEEE Open Journal of Engineering in Me...

2024
[22]

Personalized glucose fore- casting for people with type 1 diabetes using large language models,

F. J. Lara-Abelenda, D. Chushig-Muzo, P. Peiro-Corbacho, A. M. W¨agner, C. Granja, and C. Soguero-Ruiz, “Personalized glucose fore- casting for people with type 1 diabetes using large language models,” Computer Methods and Programs in Biomedicine, vol. 265, p. 108737, 2025

2025
[23]

Tailoring adverse event prediction in type 1 diabetes with patient-specific deep learning models,

G. Rigamonti, M. P. Barbato, D. Marelli, and P. Napoletano, “Tailoring adverse event prediction in type 1 diabetes with patient-specific deep learning models,”arXiv preprint arXiv:2601.14917, 2026

2026
[24]

Multimodal deep learning for biomedical data fusion: a review,

S. R. Stahlschmidt, B. Ulfenborg, and J. Synnergren, “Multimodal deep learning for biomedical data fusion: a review,”Briefings in bioinformat- ics, vol. 23, no. 2, p. bbab569, 2022

2022
[25]

The OhioT1DM Dataset for Blood Glucose Level Prediction: Update 2020,

B. R. Marling C., “The OhioT1DM Dataset for Blood Glucose Level Prediction: Update 2020,”CEUR work. proceedings, pp. 71–74, 2020

2020
[26]

Diatrend: A dataset from advanced diabetes technology to enable development of novel analytic solutions,

T. Prioleau, A. Bartolome, R. Comi, and C. Stanger, “Diatrend: A dataset from advanced diabetes technology to enable development of novel analytic solutions,”Scientific Data, vol. 10, no. 1, p. 556, 2023

2023
[27]

K. Bach, R. C. Bunescu, C. Marling, and N. Wiratunga, Eds.,Pro- ceedings of the 5th International Workshop on Knowledge Discovery in Healthcare Data co-located with 24th European Conference on Artificial Intelligence, KDH@ECAI 2020, Santiago de Compostela, Spain & Virtually, A...

2020
[28]

Feature transformation for efficient blood glucose prediction in type 1 diabetes mellitus patients,

H. Butt, I. Khosa, and M. A. Iftikhar, “Feature transformation for efficient blood glucose prediction in type 1 diabetes mellitus patients,” Diagnostics, vol. 13, no. 3, p. 340, 2023

2023
[29]

Smote: synthetic minority over-sampling technique,

N. V . Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer, “Smote: synthetic minority over-sampling technique,”Journal of artificial intel- ligence research, vol. 16, pp. 321–357, 2002

2002
[30]

Arcface: Additive angular margin loss for deep face recognition,

J. Deng, J. Guo, N. Xue, and S. Zafeiriou, “Arcface: Additive angular margin loss for deep face recognition,” inProc. of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4690– 4699

2019
[31]

Evaluating clinical accuracy of systems for self-monitoring of blood glucose,

W. Clarke, C. DC, L. Gonder-Frederick, W. Carter, and S. Pohl, “Evaluating clinical accuracy of systems for self-monitoring of blood glucose,”Diabetes care, vol. 10, pp. 622–8, 09 1987

1987

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Type 1 diabetes,

M. A. Atkinson, G. S. Eisenbarth, and A. W. Michels, “Type 1 diabetes,” The lancet, vol. 383, no. 9911, pp. 69–82, 2014

work page 2014

[2] [2]

Continuous glucose monitoring: a review of successes, challenges, and opportunities,

D. Rodbard, “Continuous glucose monitoring: a review of successes, challenges, and opportunities,”Diabetes technology & therapeutics, vol. 18, no. S2, pp. S2–3, 2016

work page 2016

[3] [3]

Continuous subcutaneous insulin infusion: a comprehensive review of insulin pump therapy,

M. J. Lenhard and G. D. Reeves, “Continuous subcutaneous insulin infusion: a comprehensive review of insulin pump therapy,”Archives of Internal Medicine, vol. 161, no. 19, pp. 2293–2300, 2001

work page 2001

[4] [4]

Engineering digital biomarkers of inter- stitial glucose from noninvasive smartwatches,

B. Bent, P. J. Cho, M. Henriquez, A. Wittmann, C. Thacker, M. Feinglos, M. J. Crowley, and J. P. Dunn, “Engineering digital biomarkers of inter- stitial glucose from noninvasive smartwatches,”NPJ Digital Medicine, vol. 4, no. 1, p. 89, 2021

work page 2021

[5] [5]

Arima models,

R. H. Shumway, D. S. Stoffer, R. H. Shumway, and D. S. Stoffer, “Arima models,”Time series analysis and its applications: with R examples, pp. 75–163, 2017

work page 2017

[6] [6]

Blood glucose prediction model for type 1 diabetes based on extreme gradient boosting,

G. Alfian, M. Syafrudin, J. Rhee, M. Anshari, M. Mustakim, and I. Fahrurrozi, “Blood glucose prediction model for type 1 diabetes based on extreme gradient boosting,” inIOP Conference Series: Materials Science and Engineering, vol. 803, no. 1. IOP Publishing, 2020, p. 012012

work page 2020

[7] [7]

Blood glucose level prediction as time-series modeling using sequence-to-sequence neural networks,

A. Bhimireddy, P. Sinha, B. Oluwalade, J. W. Gichoya, and S. Purkayastha, “Blood glucose level prediction as time-series modeling using sequence-to-sequence neural networks,” inKDH@ECAI, 2020, pp. 125–130

work page 2020

[8] [8]

Bgformer: An improved informer model to enhance blood glucose prediction,

Y . Xue, S. Guan, and W. Jia, “Bgformer: An improved informer model to enhance blood glucose prediction,”Journal of Biomedical Informatics, vol. 157, p. 104715, 2024

work page 2024