A Variable State Dimension Approach to Meal Detection and Meal Size Estimation: In Silico Evaluation Through Basal-Bolus Insulin Therapy for Type 1 Diabetes

Estimating Breakfast Characteristics Using Continuous Glucose Monitoring and Machine Learning in Adults With or at Risk of Type 2 Diabetes

BACKGROUND

Continuous glucose monitoring (CGM) systems allow detailed assessment of postprandial glucose responses (PPGR), offering new insights into food choices' impact on dysglycemia. However, current approaches to analyze PPGR using a CGM require manual meal logging, limiting the scalability of CGM-driven applications like personalized nutrition and at-home diabetes risk assessment.

OBJECTIVE

We propose a machine learning (ML) framework to automatically identify and characterize breakfast-related PPGRs from CGM profiles in adults at risk of or living with noninsulin-treated type 2 diabetes (T2D).

METHODS

Our PPGR estimation framework uses a random forest ML algorithm trained on 15 adults without diabetes who wore a CGM for up to four weeks. The algorithm performance was evaluated on a held-out subset of the participants' CGM data as well as on an external validation data set of 36 individuals at risk for or with noninsulin-treated T2D.

RESULTS

Our algorithm's estimations of breakfast PPGRs displayed no statistically significant differences to annotated PPGRs, in terms of incremental area under the curve and glucose rise (P > .05 for both data sets), while a small difference in prebreakfast glucose was found in the nondiabetes data set (P = .005) but not in the validation T2D data set (P = .18).

CONCLUSIONS

We designed an ML framework to automatically estimate the timing of meal events from CGM data in individuals without diabetes and in individuals at risk or with T2D. This could provide a more scalable approach for analyzing postprandial glycemia, increasing the feasibility of CGM-based precision nutrition and diabetes risk assessment applications.

Mealtime prediction using wearable insulin pump data to support diabetes management

Many patients with diabetes struggle with post-meal high blood glucose due to missed or untimely meal-related insulin doses. To address this challenge, our research aims to: (1) study mealtime patterns in patients with type 1 diabetes using wearable insulin pump data, and (2) develop personalized models for predicting future mealtimes to support timely insulin dose administration. Using two independent datasets with over 45,000 meal logs from 82 patients with diabetes, we find that the majority of people ( ∼ 60%) have irregular and inconsistent mealtime patterns that change notably through the course of each day and across months in their own historical data. We also show the feasibility of predicting future mealtimes with personalized LSTM-based models that achieve an average F1 score of > 95% with less than 0.25 false positives per day. Our research lays the groundwork for developing a meal prediction system that can nudge patients with diabetes to administer bolus insulin doses before meal consumption to reduce the occurrence of post-meal high blood glucose.

An ensemble machine learning approach for the detection of unannounced meals to enhance postprandial glucose control

BACKGROUND

Hybrid automated insulin delivery systems enhance postprandial glucose control in type 1 diabetes, however, meal announcements are burdensome. To overcome this, we propose a machine learning-based automated meal detection approach; METHODS:: A heterogeneous ensemble method combining an artificial neural network, random forest, and logistic regression was employed. Trained and tested on data from two in-silico cohorts comprising 20 and 47 patients. It accounted for various meal sizes (moderate to high) and glucose appearance rates (slow and rapid absorbing). To produce an optimal prediction model, three ensemble configurations were used: logical AND, majority voting, and logical OR. In addition to the in-silico data, the proposed meal detector was also trained and tested using the OhioT1DM dataset. Finally, the meal detector is combined with a bolus insulin compensation scheme; RESULTS:: The ensemble majority voting obtained the best meal detector results for both the in-silico and OhioT1DM cohorts with a sensitivity of 77%, 94%, 61%, precision of 96%, 89%, 72%, F1-score of 85%, 91%, 66%, and with false positives per day values of 0.05, 0.19, 0.17, respectively. Automatic meal detection with insulin compensation has been performed in open-loop insulin therapy using the AND ensemble, chosen for its lower false positive rate. Time-in-range has significantly increased 10.48% and 16.03%, time above range was reduced by 5.16% and 11.85%, with a minimal time below range increase of 0.35% and 2.69% for both in-silico cohorts, respectively, compared to the results without a meal detector; CONCLUSION:: To increase the overall accuracy and robustness of the predictions, this ensemble methodology aims to take advantage of each base model's strengths. All of the results point to the potential application of the proposed meal detector as a separate module for the detection of meals in automated insulin delivery systems to achieve improved glycemic control.

Integrated image and sensor-based food intake detection in free-living

The first step in any dietary monitoring system is the automatic detection of eating episodes. To detect eating episodes, either sensor data or images can be used, and either method can result in false-positive detection. This study aims to reduce the number of false positives in the detection of eating episodes by a wearable sensor, Automatic Ingestion Monitor v2 (AIM-2). Thirty participants wore the AIM-2 for two days each (pseudo-free-living and free-living). The eating episodes were detected by three methods: (1) recognition of solid foods and beverages in images captured by AIM-2; (2) recognition of chewing from the AIM-2 accelerometer sensor; and (3) hierarchical classification to combine confidence scores from image and accelerometer classifiers. The integration of image- and sensor-based methods achieved 94.59% sensitivity, 70.47% precision, and 80.77% F1-score in the free-living environment, which is significantly better than either of the original methods (8% higher sensitivity). The proposed method successfully reduces the number of false positives in the detection of eating episodes.

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

Objective digital data is scarce yet needed in many domains to enable research that can transform the standard of healthcare. While data from consumer-grade wearables and smartphones is more accessible, there is critical need for similar data from clinical-grade devices used by patients with a diagnosed condition. The prevalence of wearable medical devices in the diabetes domain sets the stage for unique research and development within this field and beyond. However, the scarcity of open-source datasets presents a major barrier to progress. To facilitate broader research on diabetes-relevant problems and accelerate development of robust computational solutions, we provide the DiaTrend dataset. The DiaTrend dataset is composed of intensive longitudinal data from wearable medical devices, including a total of 27,561 days of continuous glucose monitor data and 8,220 days of insulin pump data from 54 patients with diabetes. This dataset is useful for developing novel analytic solutions that can reduce the disease burden for people living with diabetes and increase knowledge on chronic condition management in outpatient settings.

A Deep Learning Framework for Automatic Meal Detection and Estimation in Artificial Pancreas Systems

Current artificial pancreas (AP) systems are hybrid closed-loop systems that require manual meal announcements to manage postprandial glucose control effectively. This poses a cognitive burden and challenge to users with T1D since this relies on frequent user engagement to maintain tight glucose control. In order to move towards fully automated closed-loop glucose control, we propose an algorithm based on a deep learning framework that performs multitask quantile regression, for both meal detection and carbohydrate estimation. Our proposed method is evaluated in silico on 10 adult subjects from the UVa/Padova simulator with a Bio-inspired Artificial Pancreas (BiAP) control algorithm over a 2 month period. Three different configurations of the AP are evaluated -BiAP without meal announcement (BiAP-NMA), BiAP with meal announcement (BiAP-MA), and BiAP with meal detection (BiAP-MD). We present results showing an improvement of BiAP-MD over BiAP-NMA, demonstrating 144.5 ± 6.8 mg/dL mean blood glucose level (-4.4 mg/dL, p< 0.01) and 77.8 ± 6.3% mean time between 70 and 180 mg/dL (+3.9%, p< 0.001). This improvement in control is realised without a significant increase in mean in hypoglycaemia (+0.1%, p= 0.4). In terms of detection of meals and snacks, the proposed method on average achieves 93% precision and 76% recall with a detection delay time of 38 ± 15 min (92% precision, 92% recall, and 37 min detection time for meals only). Furthermore, BiAP-MD handles hypoglycaemia better than BiAP-MA based on CVGA assessment with fewer control errors (10% vs. 20%). This study suggests that multitask quantile regression can improve the capability of AP systems for postprandial glucose control without increasing hypoglycaemia.

The northeast glucose drift: Stratification of post-breakfast dysglycemia among predominantly Hispanic/Latino adults at-risk or with type 2 diabetes

BACKGROUND

There is minimal experience in continuous glucose monitoring (CGM) among underserved racial/ethnic minority populations with or at risk of type 2 diabetes (T2D), and therefore a lack of CGM-driven insight for these individuals. We analyzed breakfast-related CGM profiles of free-living, predominantly Hispanic/Latino individuals at-risk of T2D, with pre-T2D, or with non-insulin treated T2D.

METHODS

Starting February 2019, 119 participants in Santa Barbara, CA, USA, (93 female, 87% Hispanic/Latino [predominantly Mexican-American], age 54·4 [±12·1] years), stratified by HbA_1c levels into (i) at-risk of T2D, (ii) with pre-T2D, and (iii) with non-insulin treated T2D, wore blinded CGMs for two weeks. We compared valid CGM profiles from 106 of these participants representing glucose response to breakfast using four parameters.

FINDINGS

A "northeast drift" was observed in breakfast glucose responses comparing at-risk to pre-T2D to T2D participants. T2D participants had a significantly higher pre-breakfast glucose level, glucose rise, glucose incremental area under the curve (all p < 0·0001), and time to glucose peak (p < 0·05) compared to pre-T2D and at-risk participants. After adjusting for demographic and clinical covariates, pre-breakfast glucose and time to peak (p < 0·0001) were significantly associated with HbA_1c. The model predicted HbA_1c within (0·55 ± 0·67)% of true laboratory HbA_1c values.

INTERPRETATION

For predominantly Hispanic/Latino adults, the average two-week breakfast glucose response shows a progression of dysglycemia from at-risk of T2D to pre-T2D to T2D. CGM-based breakfast metrics have the potential to predict HbA_1c levels and monitor diabetes progression.

FUNDING

US Department of Agriculture (Grant #2018-33800-28404), a seed grant from the industry board fees of the NSF Engineering Research Center for Precise Advanced Technologies and Health Systems for Underserved Populations (PATHS-UP) (Award #1648451), and the Elsevier foundation.

A New Meal Absorption Model for Artificial Pancreas Systems

BACKGROUND

Artificial pancreas (AP) systems reduce the treatment burden of Type 1 Diabetes by automatically regulating blood glucose (BG) levels. While many disturbances stand in the way of fully closed-loop (automated) control, unannounced meals remain the greatest challenge. Furthermore, different types of meals can have significantly different glucose responses, further increasing the uncertainty surrounding the meal.

METHODS

Effective attenuation of a meal requires quick and accurate insulin delivery because of slow insulin action relative to meal effects on BG. The proposed Variable Hump (VH) model adapts to meals of varying compositions by inferring both meal size and shape. To appropriately address the uncertainty of meal size, the model divides meal absorption into two disjoint regions: a region with coarse meal size predictions followed by a fine-grain region where predictions are fine-tuned by adapting to the meal shape.

RESULTS

Using gold-standard triple tracer meal data, the proposed VH model is compared to three simpler second-order response models. The proposed VH model increased model fit capacity by 22% and prediction accuracy by 12% relative to the next best models. A 47% increase in the accuracy of uncertainty predictions was also found. In a simple control scenario, the controller governed by the proposed VH model provided insulin just as fast or faster than the controller governed by the other models in four out of the six meals. While the controllers governed by the other models all delivered at least a 25% excess of insulin at their worst, the VH model controller only delivered 9% excess at its worst.

CONCLUSIONS

The VH Model performed best in accuracy metrics and succeeded over the other models in providing insulin quickly and accurately in a simple implementation. Use in an AP system may improve prediction accuracy and lead to better control around mealtimes.

A New Stochastic Approach for Modeling Glycemic Disturbances in Type 2 Diabetes

OBJECTIVE

To improve insulin treatment in type 2 diabetes (T2D) using model-based control techniques, the underlying model needs to be individualized to each patient. Due to the impact of unknown meals, exercise and other factors on the blood glucose, it is difficult to utilize available data from continuous glucose monitors (CGMs) for model fitting and parameter estimation purposes.

METHODS

To overcome this problem, we propose a novel method for modeling the glycemic disturbances as a stochastic process. To differentiate meals from other glycemic disturbances, we model the meal intake as a separate stochastic process while encompassing all other disturbances in another stochastic process. Using particle filtering, we validate the model on simulations as well as on clinical data.

RESULTS

Based on simulated CGM data, the residuals generated by the particle filter are white, indicating a good model fit. For the clinical data, we use parameter values estimated based on fasting glucose data. The residuals obtained from clinical CGM data contain correlations up to lag 5.

CONCLUSION

The proposed model is shown to adequately describe the meal-induced glucose fluctuations in simulated CGM data while validations on clinical CGM data show promising results as well.

SIGNIFICANCE

The proposed model may lay the grounds for new ways of utilizing available CGM data, including CGM-based parameter estimation and stochastic optimal control.

Automated meal detection from continuous glucose monitor data through simulation and explanation

Background

Artificial pancreas systems aim to reduce the burden of type 1 diabetes by automating insulin dosing. These systems link a continuous glucose monitor (CGM) and insulin pump with a control algorithm, but require users to announce meals, without which the system can only react to the rise in blood glucose.

Objective

We investigate whether CGM data can be used to automatically infer meals in daily life even in the presence of physical activity, which can raise or lower blood glucose.

Materials and Methods

We propose a novel meal detection algorithm that combines simulations with CGM, insulin pump, and heart rate monitor data. When observed and predicted glucose differ, our algorithm uses simulations to test whether a meal may explain this difference. We evaluated our method on simulated data and real-world data from individuals with type 1 diabetes.

Results

In simulated data, we detected meals earlier and with higher accuracy than was found in prior work (25.7 minutes, 1.2 g error; compared with 48.3 minutes, 17.2 g error). In real-world data, we discovered a larger number of plausible meals than was found in prior work (30 meals, 76.7% accepted; compared with 33 meals, 39.4% accepted).

Discussion

Prior research attempted meal detection from CGM, but had delays and lower accuracy in real data or did not allow for physical activity. Our approach can be used to improve insulin dosing in an artificial pancreas and trigger reminders for missed meal boluses.

Conclusions

We demonstrate that meal information can be robustly inferred from CGM and body-worn sensor data, even in challenging environments of daily life.

Toward a Fully Automated Artificial Pancreas System Using a Bioinspired Reinforcement Learning Design: In Silico Validation

OBJECTIVE

The automation of insulin treatment is the most challenge aspect of glucose management for type 1 diabetes owing to unexpected exogenous events (e.g., meal intake). In this article, we propose a novel reinforcement learning (RL) based artificial intelligence (AI) algorithm for a fully automated artificial pancreas (AP) system.

METHODS

A bioinspired RL designing method was developed for automated insulin infusion. This method has reward functions that imply the temporal homeostatic objective and discount factors that reflect an individual specific pharmacological characteristic. The proposed method was applied to a training method using an RL algorithm and was evaluated in virtual patients from the FDA approved UVA/Padova simulator with unannounced meal intakes.

RESULTS

For a single-meal experiment with preprandial fasting, the trained policy demonstrated fully automated regulation in both the basal and postprandial phases. In the in silico trial with a variation of insulin sensitivity and dawn phenomenon, the policy achieved a mean glucose of 124.72 mg/dL and percentage time in the normal range of 89.56%. The layer-wise relevance propagation provides interpretable information on AI-driven decision for robustness to sensor noise, automated postprandial regulation, and insulin stacking avoidance.

CONCLUSION

The AP algorithm based on the bioinspired RL approach enables fully automated blood glucose control with unannounced meal intake.

SIGNIFICANCE

The proposed framework can be extended to other drug-based treatments for systems with significant uncertainties.

Unannounced Meal Detection for Artificial Pancreas Systems Using Extended Isolation Forest

This study aims at developing an unannounced meal detection method for artificial pancreas, based on a recent extension of Isolation Forest. The proposed method makes use of features accounting for individual Continuous Glucose Monitoring (CGM) profiles and benefits from a two-threshold decision rule detection. The advantage of using Extended Isolation Forest (EIF) instead of the standard one is supported by experiments on data from virtual diabetic patients, showing good detection accuracy with acceptable detection delays.

Automatic Detection and Estimation of Unannounced Meals for Multivariable Artificial Pancreas System

BACKGROUND

Automatically attenuating the postprandial rise in the blood glucose concentration without manual meal announcement is a significant challenge for artificial pancreas (AP) systems. In this study, a meal module is proposed to detect the consumption of a meal and to estimate the amount of carbohydrate (CHO) intake.

METHODS

The meals are detected based on qualitative variables describing variation of continuous glucose monitoring (CGM) readings. The CHO content of the meals/snacks is estimated by a fuzzy system using CGM and subcutaneous insulin delivery data. The meal bolus amount is computed according to the patient's insulin to CHO ratio. Integration of the meal module into a multivariable AP system allows revision of estimated CHO based on knowledge about physical activity, sleep, and the risk of hypoglycemia before the final decision for a meal bolus is made.

RESULTS

The algorithm is evaluated by using 117 meals/snacks in retrospective data from 11 subjects with type 1 diabetes. Sensitivity, defined as the percentage of correctly detected meals and snacks, is 93.5% for meals and 68.0% for snacks. The percentage of false positives, defined as the proportion of false detections relative to the total number of detected meals and snacks, is 20.8%.

CONCLUSIONS

Integration of a meal detection module in an AP system is a further step toward an automated AP without manual entries. Detection of a consumed meal/snack and infusion of insulin boluses using an estimate of CHO enables the AP system to automatically prevent postprandial hyperglycemia.