Meal detection based on non-individualized moving horizon estimation and classification

Kölle, Konstanze; Fougner, Anders Lyngvi; Stavdahl, Øyvind

doi:10.1109/ccta.2017.8062516

Cited by 11 publications

(13 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An augmented version of the Bergman model was also used in a method that applies invariant statistics to differentiate between effects that can be explained by the model with previously detected meals as input and those that must result from a more recent meal [96]. Another approach applies linear discriminant analysis to state horizons that were generated by moving horizon estimation using a version of the Bergman minimal model [97].…”

Section: Detection Of Mealsmentioning

confidence: 99%

Risk analysis for the design of a safe artificial pancreas control system

et al. 2018

Self Cite

View full text Add to dashboard Cite

Closed-loop glucose control has the potential to improve the glycemic control in patients with diabetes mellitus type 1. Such an artificial pancreas (AP) should keep the user safe despite all disturbances and faults. The objective of this paper is to analyze those perturbations according to their effects on the glycemic status, and thereby supporting an informed design process of the control system. As suggested by the international standard ISO 14971 for risk management of medical devices, the well proven failure modes and effects analysis (FMEA) was chosen as instrument. An FMEA scheme was modified for this purpose and applied to a single-hormone system with subcutaneous and intraperitoneal routes for glucose sensing and insulin administration. Faults that imply urgent danger and thus require fast detection and diagnosis were identified and distinguished from disturbances that can be sufficiently addressed by basic control functions, e.g. by adaptive control algorithms. Requirements and testing criteria for basic control functions as well as fault detection and diagnosis functions can be derived from the provided overview.Safety is the primary requirement on a control system that doses insulin to regulate the blood glucose level (BGL) in diabetes mellitus type 1 (DM1). People with DM1 have an insufficient pancreatic insulin production and depend on lifelong insulin administration to avoid hyperglycemia. The chronically elevated BGL would otherwise lead to several long-term complications. Hypoglycemia caused by overdosing of insulin, on the other hand, may lead to unconsciousness and death in the F.

show abstract

Section: Detection Of Mealsmentioning

confidence: 99%

Risk analysis for the design of a safe artificial pancreas control system

et al. 2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…The aim is to extract the characteristic changes in CGM caused by the onset of meals. In particular, moving horizon estimation (MHE) is used to estimate the rate of glucose appearance [15]. Other estimators such as the Kalman filter update the current estimate based on a single, most recent measurement.…”

Section: Introductionmentioning

confidence: 99%

“…At each time step, not only the most recent estimate, but the whole estimated backward horizon is updated when more recent measurements are available. The dynamical changes within the estimated horizons are exploited in the pattern recognition [15]. Both the rate of appearance of glucose in blood estimated with a moving horizon estimation (MHE) and the CGM data directly are used as inputs to the pattern recognition method.…”

Section: Introductionmentioning

confidence: 99%

Pattern Recognition Reveals Characteristic Postprandial Glucose Changes: Non-Individualized Meal Detection in Diabetes Mellitus Type 1

Kölle

Biester

Christiansen

et al. 2020

IEEE J. Biomed. Health Inform.

Self Cite

View full text Add to dashboard Cite

Accurate continuous glucose monitoring (CGM) is essential for fully automated glucose control in diabetes mellitus type 1. State-of-the-art glucose control systems automatically regulate the basal insulin infusion. Users still need to manually announce meals to dose the prandial insulin boluses. An automated meal detection could release the user and improve the glucose regulation.In this study, patterns in the postprandial CGM data are exploited for meal detection. Binary classifiers are trained to recognize the postprandial pattern in horizons of the estimated glucose rate of appearance and in CGM data. The appearance rate is determined by moving horizon estimation (MHE) based on a simple model. Linear discriminant analysis (LDA) is used for classification. The proposed method is compared to methods that detect meals when thresholds are violated.Diabetes care data from twelve free-living pediatric patients was downloaded during regular screening. Experts identified meals and their start by retrospective evaluation.The classification was tested by cross-validation. Compared to the threshold-based methods, LDA showed higher sensitivity to meals with a low rate of false alarms. Classifying horizons outperformed the other methods also with respect to time of detection.The onset of meals can be detected by pattern recognition based on estimated model states and consecutive CGM measurements. No individual tuning is necessary. This makes the method easily adopted in the clinical practice.

show abstract

“…[3], [4]. The delay between meal start and reliable meal detection can easily reach 40 min [5]- [7] although our recent studies show that it may be possible to reduce this detection time significantly [8], [9]. The fusion of measurements from two redundant SC sensors was proposed to enhance the reliability [10], but alternative sensing modalities have not been exploited.…”

Section: Introduction a Meal Detection In An Artificial Pancreasmentioning

confidence: 99%

Feasibility of Early Meal Detection Based on Abdominal Sound

Kölle

Fougner

Ellingsen

et al. 2019

IEEE J. Transl. Eng. Health Med.

Self Cite

View full text Add to dashboard Cite

In classical approaches for an artificial pancreas, continuous glucose monitoring (CGM) is the only measured variable used for insulin dosing and additional control functions. The CGM values are subject to time delays and slow dynamics between blood and the sensing location. These time lags compromise the controller's performance in maintaining (near to) normal glucose levels. Meal information could enhance the control outcome. However, meal announcement by the user is not reliable, and it takes 30 min to 40 min from meal onset until a meal is detected by methods based on CGM. In this pilot study, the use of bowel sounds for meal detection was investigated. In particular, we focused on whether bowel sounds change qualitatively during or shortly after meal ingestion. After fasting for at least 4 h, 11 healthy volunteers ingested a lunch meal at their usual time. Abdominal sound was recorded by a condenser microphone that was attached to the right upper quadrant of the abdomen by medical tape. Features that describe the power distribution over the frequency spectrum were extracted and used for classification by support vector machines. These classifiers were trained in a leave-one-out cross-validation scheme. Meals could be detected on average 10 min (std: 4.4 min) after they had started. Half of these were detected without false alarms. This shows that abdominal sound monitoring could provide an early meal detection. Further studies should investigate this possibility on a larger population in more general settings.

show abstract

Meal detection based on non-individualized moving horizon estimation and classification

Cited by 11 publications

References 19 publications

Risk analysis for the design of a safe artificial pancreas control system

Risk analysis for the design of a safe artificial pancreas control system

Pattern Recognition Reveals Characteristic Postprandial Glucose Changes: Non-Individualized Meal Detection in Diabetes Mellitus Type 1

Feasibility of Early Meal Detection Based on Abdominal Sound

Contact Info

Product

Resources

About