Youqing Wang1, Jinping Zhang2, Fanmao Zeng1, Na Wang2, Xiaoping Chen2, Bo Zhang2, Dong Zhao1, Wenying Yang2, Claudio Cobelli3. 1. 1 College of Information Science and Technology, Beijing University of Chemical Technology , Beijing, China . 2. 2 Department of Endocrinology, China-Japan Friendship Hospital , Beijing, China . 3. 3 Department of Information Engineering, University of Padova , Padova, Italy .
Abstract
BACKGROUND: A learning-type artificial pancreas has been proposed to exploit the repetitive nature in the blood glucose dynamics. We clinically evaluated the efficacy of the learning-type artificial pancreas. METHODS: We conducted a pilot clinical study in 10 participants of mean age 36.1 years (standard deviation [SD] 12.7; range 16-58) with type 1 diabetes. Each trial was conducted for eight consecutive mornings. The first two mornings were open-loop to obtain the individualized parameters. Then, the following six mornings were closed-loop, during which a learning-type model predictive control algorithm was employed to calculate the insulin infusion rate. To evaluate the algorithm's robustness, each participant took exercise or consumed alcohol on the fourth or sixth closed-loop day and the order was determined randomly. The primary outcome was the percentage of time spent in the target glucose range of 3.9-8.0 mmol/L between 0900 and 1200 h. RESULTS: The percentage of time with glucose spent in target range was significantly improved from 51.6% on day 1 to 71.6% on day 3 (mean difference between groups 17.9%, confidence interval [95% CI] 3.6-32.1; P = 0.020). There were no hypoglycemic episodes developed on day 3 compared with two episodes on day 1. There was no difference in the percentage of time with glucose spent in target range between exercise day versus day 5 and alcohol day versus day 5. CONCLUSIONS: The learning-type artificial pancreas system achieved good glycemic regulation and provided increased effectiveness over time. It showed a satisfactory performance even when the blood glucose was challenged by exercise or alcohol.
BACKGROUND: A learning-type artificial pancreas has been proposed to exploit the repetitive nature in the blood glucose dynamics. We clinically evaluated the efficacy of the learning-type artificial pancreas. METHODS: We conducted a pilot clinical study in 10 participants of mean age 36.1 years (standard deviation [SD] 12.7; range 16-58) with type 1 diabetes. Each trial was conducted for eight consecutive mornings. The first two mornings were open-loop to obtain the individualized parameters. Then, the following six mornings were closed-loop, during which a learning-type model predictive control algorithm was employed to calculate the insulin infusion rate. To evaluate the algorithm's robustness, each participant took exercise or consumed alcohol on the fourth or sixth closed-loop day and the order was determined randomly. The primary outcome was the percentage of time spent in the target glucose range of 3.9-8.0 mmol/L between 0900 and 1200 h. RESULTS: The percentage of time with glucose spent in target range was significantly improved from 51.6% on day 1 to 71.6% on day 3 (mean difference between groups 17.9%, confidence interval [95% CI] 3.6-32.1; P = 0.020). There were no hypoglycemic episodes developed on day 3 compared with two episodes on day 1. There was no difference in the percentage of time with glucose spent in target range between exercise day versus day 5 and alcohol day versus day 5. CONCLUSIONS: The learning-type artificial pancreas system achieved good glycemic regulation and provided increased effectiveness over time. It showed a satisfactory performance even when the blood glucose was challenged by exercise or alcohol.
Entities:
Keywords:
Alcohol; Artificial pancreas; Closed-loop; Hypoglycemia; Learning-type model predictive control; Type 1 diabetes
Authors: Eyal Dassau; Jordan E Pinsker; Yogish C Kudva; Sue A Brown; Ravi Gondhalekar; Chiara Dalla Man; Steve Patek; Michele Schiavon; Vikash Dadlani; Isuru Dasanayake; Mei Mei Church; Rickey E Carter; Wendy C Bevier; Lauren M Huyett; Jonathan Hughes; Stacey Anderson; Dayu Lv; Elaine Schertz; Emma Emory; Shelly K McCrady-Spitzer; Tyler Jean; Paige K Bradley; Ling Hinshaw; Alejandro J Laguna Sanz; Ananda Basu; Boris Kovatchev; Claudio Cobelli; Francis J Doyle Journal: Diabetes Care Date: 2017-10-13 Impact factor: 19.112