An algorithm mimicking pancreas pulsatile behavior improves artificial pancreas performance

2021 ◽  
pp. 039139882110271
Author(s):  
Guillermo Cocha ◽  
Victor Tedesco ◽  
Carlos D’Attellis ◽  
Carlos Amorena
Keyword(s):  
Blood Glucose ◽  
Linear System ◽  
Insulin Infusion ◽  
Insulin Pump ◽  
Artificial Pancreas ◽  
Insulin Concentration ◽  
Postprandial Blood Glucose ◽  
Diabetic Patients ◽  
Control Input ◽  
Blood Insulin

Background: Artificial pancreas design using subcutaneous insulin infusion without pre-meal feed-forward boluses often induces an over-response leading to hypoglycemia due to the increase of blood insulin concentration sustained in time. The objective of this work was to create an algorithm for controlling the function of insulin pumps in closed-loop systems to improve blood glucose management in type 1 diabetic patients by mimicking the pulsatile behaviour of the pancreas. Methods: A controller tuned in a pulsatile way promotes damped oscillations of blood insulin concentration injected through an insulin pump. We tested it in a simulated environment, using nine ‘in silica’ subjects. The control algorithm is founded on feedback linearization where through a change of variables, the nonlinear system turns into an equivalent linear system, suitable for implementing through a PID controller. We compared the results obtained ‘in silica’ with the volume injected by an insulin pump controlled by this algorithm. Results: The use of this algorithm resulted in a pulsatile control of postprandial blood glucose concentration, avoiding hypoglycaemic episodes. The results obtained ‘in silica’ were replicated in a real pump ‘in vitro’. Conclusions: With this proposed linear system, an appropriate control input can be designed. The controller works with a damped pulsatile pattern making the insulin infusion from the pump and blood insulin concentration pulsatile. This operational would improve the performance of an artificial pancreas.

Diurnal plasma profiles of metabolite and hormone concentration in insulin-dependent diabetic patients during conventional insulin treatment and continuous subcutaneous insulin infusion

1987 ◽  
Vol 114 (3) ◽  
pp. 433-439 ◽  
Author(s):  
K. Hermansen ◽  
A. Møller ◽  
C. K. Christensen ◽  
J. S. Christiansen ◽  
O. Schmitz ◽  
...  
Keyword(s):  
Growth Hormone ◽  
Blood Glucose ◽  
Continuous Subcutaneous Insulin Infusion ◽  
Insulin Infusion ◽  
Insulin Treatment ◽  
Insulin Pump ◽  
Hormonal Control ◽  
Diabetic Patients ◽  
Subcutaneous Insulin ◽  
Insulin Dependent

Abstract. In addition to hyperglycaemia, derangement of metabolic and hormonal control may play an important role in the development of microvascular complications in diabetes. Little, however, is known about the impact of insulin pump treatment on metabolic and hormonal parameters. In a 6-month prospective randomized study in insulin-dependent diabetics we therefore investigated the effects of continuous subcutaneous insulin infusion by pump (10 patients) and conventional insulin treatment (10 patients) on the 24-h profiles of blood glucose, glycerol, lactate, 3-hydroxybutyrate, insulin, glucagon and growth hormone by measuring the respective concentrations every 2 h. We found that average blood glucose levels and HbA1c were significantly lower in the group treated by continuous subcutaneous insulin infusion as compared with the group on conventional insulin treatment. Furthermore, we observed an improvement in diurnal levels of lactate and 3-hydroxybutyrate in the pumptreated group which was not seen in the conventionally treated group. A slight increment in alanine was seen in the group treated with insulin pump. Serum growth hormone, glycerol, plasma free insulin as well as the daily insulin supply were unchanged and identical in the two groups. It is noteworthy that in the pump group, the decrease in blood glucose and 3-hydroxybutyrate takes place concomitantly with a significant suppression of glucagon.

Nocturnal versus diurnal hormonal counterregulation to hypoglycemia in Type 1 (insulin-dependent) diabetic patients

1993 ◽  
Vol 128 (2) ◽  
pp. 109-115 ◽  
Author(s):  
Inger Bendtson ◽  
Anne Mette Rosenfalck ◽  
Christian Binder
Keyword(s):  
Growth Hormone ◽  
Blood Glucose ◽  
Insulin Infusion ◽  
Hormone Secretion ◽  
Growth Hormone Secretion ◽  
Diabetic Patients ◽  
Hormonal Responses ◽  
Nocturnal Hypoglycemia ◽  
Insulin Dependent ◽  
Insulin Dependent Diabetic

Asymptomatic hypoglycemia in IDDM patients seems to be more frequent during the night than during the day, with reported frequencies as high as 56%. Hormonal counterregulation to diurnal and nocturnal hypoglycemia was studied in 10 insulin-dependent diabetic patients without diabetic complications in order to test whether hormonal responses were lower at night than during daytime. A lower catecholamine response might imply less marked symptoms and therefore one reason why patients are not awakened by hypoglycemia. Blood glucose was stabilized to around 6 mmol/1 by iv insulin infusion and hypoglycemia was induced by increasing the insulin infusion rate—in the night studies at 01.30, in the day studies at 08.00. Blood glucose nadirs were 1.5±0.4 (1.2–1.9) mmol/1 at night and 1.9±0.3 (1.3–2.2) mmol/l during the day; in three patients the nadirs were identical during both the night and day. One patient had no adrenaline response to daytime hypoglycemia. In general, nocturnal hypoglycemia elicited greater catecholamine responses correlated to the duration of hypoglycemia. Glucagon responses showed a great heterogeneity independently of diabetes duration and hypoglycemic level. Growth hormone secretion was reduced during the night study; however, no refractory periods were found after sleep-related growth hormone secretion. In conclusion: counter-regulatory hormonal responses tend to be greater at night than during the day and do not explain why patients are not awakened by nocturnal hypoglycemia.

scholarly journals The Application of Plasma 1,5-Anhydro-D-glucitol for Monitoring Type 2 Diabetic Patients

2005 ◽  
Vol 21 (3) ◽  
pp. 127-132 ◽  
Author(s):  
Marzena Dworacka ◽  
Hanna Winiarska
Keyword(s):  
Blood Glucose ◽  
Plasma Level ◽  
Multivariate Regression Analysis ◽  
Postprandial Blood Glucose ◽  
Effective Control ◽  
Diabetic Patients ◽  
Short Term ◽  
Type 2 Diabetic Patients ◽  
Type 2 Diabetic

Aim: Recent data have suggested that effective control of postprandial blood glucose can reduce the risk of macroangiopathic complications of diabetes, especially cardiovascular risk. 1,5-Anhydro-D-glucitol (1,5-AG) has been proposed as a marker of short-term hyperglycaemic excursions. We aimed to evaluate its usefulness in patients with type 2 diabetes and have attempted to indicate when 1,5-AG monitoring should be used in ordinary diabetes care settings. Methods: The study group consisted of 130 type 2 diabetic patients aged 36–69 years. 1,5-AG plasma level, HbA1c concentrations and daily glucose profile were measured. Mean blood glucose (MBG), M-value were calculated and maximal daily glycaemia (MxG) was established as indicators of short-term hyperglycaemic episodes. Results: 1,5-AG plasma level was negatively and HbA1c was positively correlated with fasting glycaemia (FG), MBG, M-value and MxG. Multivariate regression analysis revealed that 1,5-AG plasma level is determined by MxG only, while FG determined HbA1c concentration in blood. The analysis of 1,5-AG level and HbA1c distributions in well and poorly controlled patients revealed that persons with low HbA1c values may have decreased 1,5-AG plasma level. Conclusion: 1,5-AG plasma level monitoring is the useful method to identify well controlled, exclusively based on HbA1c levels type 2 diabetic patients with transient hyperglycaemia, accordingly patients at high risk of macroangiopathic complications.

scholarly journals Effects of interactive education with conversation map on foot care behaviors in elderly diabetic patients with high-risk diabetic foot

2021 ◽  
Author(s):  
Zhang Yingli
Keyword(s):  
Blood Glucose ◽  
High Risk ◽  
Diabetic Foot ◽  
Postprandial Blood Glucose ◽  
Control Group ◽  
Diabetic Patients ◽  
Foot Care ◽  
Interactive Education ◽  
Conversation Map ◽  
Experimental Group

Background: Diabetic foot is one of the main reasons that seriously affect the quality of life and disability. Good foot care can prevent 85% amputation of diabetic foot patients. The purpose of this study is to explore the effects of interactive education with conversation map on foot care behaviors among elderly diabetic patients with high-risk diabetic foot.Methods: 96 elderly diabetic patients in Endocrinology Department of the Second Affiliated Hospital of Shandong First Medical University from January to December in 2020 were selected as the research objects. According to the parity of the last two digits of the hospitalization number, the odd number was divided into the control group and the even number was divided into the experimental group. The control group was given traditional health education and interactive education with conversation map for diabetic foot launched by the international diabetes federation was added to the experimental group. After 3 months of intervention, fasting blood glucose, 2 hours postprandial blood glucose, glycosylated hemoglobin and foot care behaviors of the two groups were observed.Results: The fasting blood glucose, 2 hours postprandial blood glucose, glycosylated hemoglobin and foot care behaviors of the experimental group were significantly better than those of the control group and the difference was statistically significant (p<0.05).Conclusions: The interactive education with conversation map can effectively improve the foot care behaviors of elderly diabetic patients with high-risk diabetic foot, and then prevent the occurrence of diabetic foot.

Conventional approach to glucose management for diabetic patients undergoing coronary artery bypass surgery

Perfusion ◽  
2002 ◽  
Vol 17 (2) ◽  
pp. 141-144 ◽  
Author(s):  
Patricia A Gustafson ◽  
Debra L Zarro ◽  
David A Palanzo ◽  
Norman J Manley ◽  
Ralph M Montesano ◽  
...  
Keyword(s):  
Blood Glucose ◽  
Insulin Infusion ◽  
Diabetic Patients ◽  
Type I ◽  
Cardioplegic Solution ◽  
Glucose Levels ◽  
Blood Glucose Levels ◽  
Group I ◽  
Significant Difference ◽  
Group Ii

Continuous insulin infusion was not an effective mode of treatment in maintaining safe blood glucose levels (<200 mg/dl) during the intraoperative period of diabetic patients requiring open-heart surgery. The two modifications investigated to gain better control of the blood glucose were a change in the base solution of the cardioplegia and the use of a sliding insulin scale. Fifty patients including Type I and Type II diabetics were selected for the purpose of this study. The patients were then randomly divided into two groups categorized by the type of cardioplegic solution administered and the mode of insulin treatment. Group I patients received a dextrose 5%-based cardioplegic solution and blood glucose was treated via continuous intravenous insulin infusion. Group II patients received normal saline 0.9%-based cardioplegic solution and blood glucose was treated via sliding scale. Blood glucose levels were monitored pre- and postcardio- pulmonary bypass (CPB) and every 30 min while on CPB. Glucose values were analyzed by group t test. A p value of < 0.05 was considered statistically significant. When comparing Group I (mean=258 mg/dl) with Group II (mean= 158 mg/dl), there was a statistically significant difference between the glucose values at each of the time intervals when the glucose values were recorded. In conclusion, Group II maintained an acceptable blood glucose level (<200 mg/dl) throughout the entire intra- operative period, which suggests that the combination of the sliding insulin scale and modification of the base cardioplegic solution was an effective mode of treatment.

scholarly journals Stacked LSTM based deep recurrent neural network with kalman smoothing for blood glucose prediction

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Md Fazle Rabby ◽  
Yazhou Tu ◽  
Md Imran Hossen ◽  
Insup Lee ◽  
Anthony S. Maida ◽  
...  
Keyword(s):  
Neural Network ◽  
Blood Glucose ◽  
Recurrent Neural Network ◽  
Insulin Infusion ◽  
Diabetes Management ◽  
Short Term Memory ◽  
Artificial Pancreas ◽  
Time Interval ◽  
Kalman Smoothing ◽  
Deep Recurrent Neural Network

Abstract Background Blood glucose (BG) management is crucial for type-1 diabetes patients resulting in the necessity of reliable artificial pancreas or insulin infusion systems. In recent years, deep learning techniques have been utilized for a more accurate BG level prediction system. However, continuous glucose monitoring (CGM) readings are susceptible to sensor errors. As a result, inaccurate CGM readings would affect BG prediction and make it unreliable, even if the most optimal machine learning model is used. Methods In this work, we propose a novel approach to predicting blood glucose level with a stacked Long short-term memory (LSTM) based deep recurrent neural network (RNN) model considering sensor fault. We use the Kalman smoothing technique for the correction of the inaccurate CGM readings due to sensor error. Results For the OhioT1DM (2018) dataset, containing eight weeks’ data from six different patients, we achieve an average RMSE of 6.45 and 17.24 mg/dl for 30 min and 60 min of prediction horizon (PH), respectively. Conclusions To the best of our knowledge, this is the leading average prediction accuracy for the ohioT1DM dataset. Different physiological information, e.g., Kalman smoothed CGM data, carbohydrates from the meal, bolus insulin, and cumulative step counts in a fixed time interval, are crafted to represent meaningful features used as input to the model. The goal of our approach is to lower the difference between the predicted CGM values and the fingerstick blood glucose readings—the ground truth. Our results indicate that the proposed approach is feasible for more reliable BG forecasting that might improve the performance of the artificial pancreas and insulin infusion system for T1D diabetes management.

scholarly journals Intraperitoneal insulin infusion: on the way to the artificial pancreas

2015 ◽  
Vol 18 (3) ◽  
pp. 32-45 ◽  
Author(s):  
Vladimir Aleksandrovich Karpel'ev ◽  
Elena Anatol'evna Fedorova ◽  
Yury Ivanovich Philippov ◽  
Aleksandr Yur'evich Mayorov ◽  
Marina Vladimirovna Shestakova
Keyword(s):  
Insulin Infusion ◽  
Insulin Pump ◽  
Artificial Pancreas ◽  
Glucose Monitoring ◽  
Promising Alternative ◽  
Insulin Pharmacokinetics ◽  
Intravenous Insulin ◽  
Modern Engineering ◽  
Intraperitoneal Insulin ◽  
Fast Insulin

Creating an "artificial pancreas" (a "closed loop" insulin pump, with self-adjusting insulin abilities, based on real time continuous glucose monitoring data) – is one of the most actual medical challenges of modern engineering and cybernetics.Artificial pancreas (AP) prototypes based on wearable insulin pump with subcutaneous insulin delivery are still problematic, mainly because of slow insulin pharmacokinetics. Intravenous insulin infusion via AP allows effectively maintain euglycaemia for inpatients, due to insulin pharmacokinetics and pharmacodynamics advantages. Unfortunately, it can’t be used for outpatients. Intraperitoneal insulin infusion is still relatively infrequently used in the world, but it is a promising alternative, compared to both previous methods due to a physiological action profile, fast insulin pharmacokinetics, relatively better safety and availability for outpatient usage.The purpose of this review is to describe the intraperitoneal insulin infusion features for diabetes patients at a point of AP creation perspectives. 

Sign in / Sign up

Export Citation Format

Share Document