Assessing Mealtime Macronutrient Content: Patient Perceptions Versus Expert Analyses via a Novel Phone App

Gillingham, Melanie B.; Li, Zoey; Beck, Roy W.; Calhoun, Peter; Castle, Jessica R.; Clements, Mark A.; Dassau, Eyal; Doyle, Francis J.; Gal, Robin L.; Jacobs, Peter G.; Patton, Susana R.; Rickels, Michael R.; Riddell, Michael C.; Martin, Corby K.

doi:10.1089/dia.2020.0357

Cited by 12 publications

(7 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regional or national restrictions of other diabetes apps [4][5][6][7] hinder people in accessing assistance with T1D management. This situation has led us to seek new solutions and develop this app.…”

Section: Discussionmentioning

confidence: 99%

“…T1D management has been improved by 1) nutrition apps [4] that provide information about carbohydrates and other nutrients, 2) physical activity apps [5]; 3) glucose monitoring apps that show continuous glucose monitoring data and blood glucose measurements; 4) insulin titration apps [6] that calculate basal, prandial and correction doses, and 5) insulin delivery apps that work with smart pens or pumps [7]. However, most of these apps are only available in certain countries or regions and cannot be downloaded worldwide.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

“My Friend Diabetes Carbohydrate-Bolus Calculator”: User Experiences of a Bolus Calculating Carbohydrate Counting Mobile App for People with Type 1 Diabetes

Karakuş¹,

Gökçe²,

Muradoğlu³

et al. 2023

View full text Add to dashboard Cite

Objective: Meal management in Type 1 Diabetes (T1D) has barriers such as miscalculation of doses, insufficient carbohydrate counting and numeracy skills. "My Friend Diabetes Carbohydrate Bolus Calculator" mobile app was developed as a hybrid version of nutrition apps and insulin titration apps to calculate meal's carbs and the matching bolus dose. This study investigates facilitator role of the app on lives of people with T1D. Materials and methods: People with T1D or their caregivers were recruited from children diabetes foundation social media accounts, and they answered an online survey which examines the effects of the app on carbohydrate counting, diabetes management, and the usability of the app with 17 Likert-type questions (5-point). Descriptive and non-parametric tests were performed to analyze responses. Results: Of the 165 people who completed the survey, 58 (35.2%) had T1D, 107 (64.8%) had relatives with T1D; 87 (52.7%) participants were female, the mean age was 13.1 ± 8.9 years, and the mean duration of diabetes was 4.72 years. Of the participants, 130 used the app and agreed that the app improved the users' meal management, diabetes management, carbohydrate, and dose calculations (Mean score = 4.38 ± 0.57). They were more confident in the dose calculation, freer in the food choices, and more confident in diabetes care because of the app (Mean score = 4.46 ± 0.57). Conclusions: ''My Friend Diabetes Carbohydrate-Bolus Calculator'' mobile app improved T1D management. Diabetes teams can reach more people through mobile apps and improve their clinical outcomes.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“My Friend Diabetes Carbohydrate-Bolus Calculator”: User Experiences of a Bolus Calculating Carbohydrate Counting Mobile App for People with Type 1 Diabetes

Karakuş¹,

Gökçe²,

Muradoğlu³

et al. 2023

View full text Add to dashboard Cite

show abstract

“…The architecture of the network and other learning hyper-parameters were determined using grid search. The search space was defined as follows: {Input history length = [1, 2, and 3 hours], LSTM units = [32,64,128,256,512], hidden dense layers = [1,2,3,4,5], hidden units in the first dense layer = [512,256,128,64,32], learning rate = [1e-5, 1e-4, 1e-3], batch size = [32,64,128]}. To prevent overfitting, we used early stopping (i.e., training was stopped when the MSE of the validation dataset stopped improving or got worse, indicating that the network had started to memorize the training data).…”

Section: Lstm Glucose Forecasting Modelsmentioning

confidence: 99%

“…This may be due to unreliable or missing meal reports from Tidepool participants as well as errors in carbohydrate counting. 32…”

Section: Prediction Accuracy Of Algorithms Trained On Different Datasetsmentioning

confidence: 99%

Incorporating Glucose Variability into Glucose Forecasting Accuracy Assessment Using the New Glucose Variability Impact Index and the Prediction Consistency Index: An LSTM Case Example

Mosquera-Lopez

Jacobs

2021

J Diabetes Sci Technol

View full text Add to dashboard Cite

Background: In this work, we developed glucose forecasting algorithms trained and evaluated on a large dataset of free-living people with type 1 diabetes (T1D) using closed-loop (CL) and sensor-augmented pump (SAP) therapies; and we demonstrate how glucose variability impacts accuracy. We introduce the glucose variability impact index (GVII) and the glucose prediction consistency index (GPCI) to assess the accuracy of prediction algorithms. Methods: A long-short-term-memory (LSTM) neural network was designed to predict glucose up to 60 minutes in the future using continuous glucose measurements and insulin data collected from 175 people with T1D (41,318 days) and evaluated on 75 people (11,333 days) from the Tidepool Big Data Donation Dataset. LSTM was compared with two naïve forecasting algorithms as well as Ridge linear regression and a random forest using root-mean-square error (RMSE). Parkes error grid quantified clinical accuracy. Regression analysis was used to derive the GVII and GPCI. Results: The LSTM had highest accuracy and best GVII and GPCI. RMSE for CL was 19.8 ± 3.2 and 33.2 ± 5.4 mg/dL for 30- and 60-minute prediction horizons, respectively. RMSE for SAP was 19.6 ± 3.8 and 33.1 ± 7.3 mg/dL for 30- and 60-minute prediction horizons, respectively; 99.6% and 97.6% of predictions were within zones A+B of the Parkes error grid at 30- and 60-minute prediction horizons, respectively. Glucose variability was strongly correlated with RMSE (R≥0.64, P < 0.001); GVII and GPCI demonstrated a means to compare algorithms across datasets with different glucose variability. Conclusions: The LSTM model was accurate on a large real-world free-living dataset. Glucose variability should be considered when assessing prediction accuracy using indices such as GVII and GPCI.

show abstract

“…Current carbohydrate counting methods require a level of numeracy and literacy that might be a barrier for some people with diabetes 3 . Gillingham et al showed that 49% percent of meals with <30 g of carbohydrates are overestimated while the majority (64%) of large carbohydrate meals (≥60 g) are underestimated 4 . Inaccurate carbohydrate estimations that are used for calculation of prandial insulin are associated with high prevalence of postprandial hyper-and hypoglycemia, even with hybrid insulin delivery systems 5,6 .…”

Section: Introductionmentioning

confidence: 99%

Enabling fully automated insulin delivery through meal detection and size estimation using Artificial Intelligence

et al. 2023

Self Cite

View full text Add to dashboard Cite

We present a robust insulin delivery system that includes automated meal detection and carbohydrate content estimation using machine learning for meal insulin dosing called robust artificial pancreas (RAP). We conducted a randomized, single-center crossover trial to compare postprandial glucose control in the four hours following unannounced meals using a hybrid model predictive control (MPC) algorithm and the RAP system. The RAP system includes a neural network model to automatically detect meals and deliver a recommended meal insulin dose. The meal detection algorithm has a sensitivity of 83.3%, false discovery rate of 16.6%, and mean detection time of 25.9 minutes. While there is no significant difference in incremental area under the curve of glucose, RAP significantly reduces time above range (glucose >180 mg/dL) by 10.8% (P = 0.04) and trends toward increasing time in range (70–180 mg/dL) by 9.1% compared with MPC. Time below range (glucose <70 mg/dL) is not significantly different between RAP and MPC.

show abstract

Assessing Mealtime Macronutrient Content: Patient Perceptions Versus Expert Analyses via a Novel Phone App

Cited by 12 publications

References 35 publications

“My Friend Diabetes Carbohydrate-Bolus Calculator”: User Experiences of a Bolus Calculating Carbohydrate Counting Mobile App for People with Type 1 Diabetes

“My Friend Diabetes Carbohydrate-Bolus Calculator”: User Experiences of a Bolus Calculating Carbohydrate Counting Mobile App for People with Type 1 Diabetes

Incorporating Glucose Variability into Glucose Forecasting Accuracy Assessment Using the New Glucose Variability Impact Index and the Prediction Consistency Index: An LSTM Case Example

Enabling fully automated insulin delivery through meal detection and size estimation using Artificial Intelligence

Contact Info

Product

Resources

About