Citation

BibTex format

@inproceedings{Herrero:2013,
author = {Herrero, P and Georgiou, P and Oliver, N and Reddy, M and El, Sharkawy M and Pesl, P and Johnston, D and Toumazou, C},
title = {In-Silico Comparison of a Bio-Inspired Glucose Controller vs. a PID controller with Insulin Feedback},
url = {https://dl.dropbox.com/u/7629368/BIAP-vs-PID-ATTD2013.pdf},
year = {2013}
}

RIS format (EndNote, RefMan)

TY  - CPAPER
AB - A recently developed bio-inspired glucose controller (BIAP) based on a model of β-cell physiology is compared against a recently published PID controller with insulin feedback (PID-IF) using the UVa/Padova T1DM metabolic simulator.Both controllers were tuned using a 24-hour scenario containing a 40g(10am) carbohydrates meal and initial blood glucose of 160mg/dl. A single tuning parameter was adjusted to stabilize glucose around 100mg/dl, avoiding dropping below 80mg/dl. To compare controllers, a 24-hour scenario containing 3 meals (30g(6am), 40g(2pm), 20g(10pm)) was employed. In both controllers, a partial bolus corresponding to 50% of the required insulin to cover a meal was delivered at the time of ingestion. Standard metrics provided by the simulator were employed for comparison purposes.See table for results of BIAP and PID-IF in n=10 adult, n=10 adolescent and n=10 children in-silico subjects. Mean blood glucose with PID-IF was lower than with BIAP in the adult cohort, percentage of time below target with BIAP was lower in the children cohort, percentage of time within target with BIAP was higher in the adolescents cohort and risk index was lower with BIAP in the adult and adolescent groups.The tuning parameter in BIAP correlated very well with the insulin sensitivity factor of the test subjects (R2=0.9), while the correlation observed within PID-IF was lower (R2=0.6).The BIAP glucose controller performed comparably to the PID-IF controller in in-silico setting with superior metrics in an adolescent cohort. The BIAP controller can be easily tuned by using a simple correlation with the insulin sensitivity factor.
AU - Herrero,P
AU - Georgiou,P
AU - Oliver,N
AU - Reddy,M
AU - El,Sharkawy M
AU - Pesl,P
AU - Johnston,D
AU - Toumazou,C
PY - 2013///
TI - In-Silico Comparison of a Bio-Inspired Glucose Controller vs. a PID controller with Insulin Feedback
UR - https://dl.dropbox.com/u/7629368/BIAP-vs-PID-ATTD2013.pdf
ER -

Contact us

Centre for Bio-Inspired Technology
Imperial College London
Bessemer Building
South Kensington
SW7 2AZ, UK

Tel: +44 (0)207 594 0701
Fax: +44 (0)207 594 0704

E-mail: bioinspired@imperial.ac.uk