– Pratibha Gopalakrishna
Type-1 diabetes is an autoimmune disorder in which insulin-producing cells in the pancreas are attacked and destroyed by the body’s own immune cells. Therefore, people with Type-1 diabetes need to take insulin regularly to maintain their blood sugar levels to avoid hypoglycaemia (low blood sugar) or hyperglycaemia (high blood sugar).
To help such people, researchers from the Robert Bosch Centre for Cyber-Physical Systems at IISc, in collaboration with doctors at MS Ramaiah Medical College, have developed an artificial pancreas system that can help monitor and control blood sugar levels in real time.
According to the International Diabetes Federation’s Diabetes Atlas (2021), at least 74 million people in India have diabetes and this number is expected to grow to 125 million by 2045. Prolonged hyperglycaemia can cause conditions like retinopathy, neuropathy, and nephropathy, while severe hypoglycaemia can lead sufferers to lose consciousness, fall into a coma, and in extreme cases die.
The Artificial Pancreas (AP) setup developed by Radhakant Padhi, Professor in the Department of Aerospace Engineering and Robert Bosch Centre for Cyber-Physical Systems, and his team mimics the body’s own closed loop system that regulates insulin production. The system has three parts: a sensor, an insulin pump, and an Android app. The sensor is a small coin-like device with a tiny needle-like extension that can be stuck onto the skin similar to a patch or band-aid. It monitors glucose concentration in the subcutaneous tissue continuously. The sensor is connected to an insulin pump that can infuse insulin underneath the skin. The pump is a small rectangular device, similar to a cassette, that can be carried around in a pocket. The Android app determines how much insulin needs to be pumped into the body and can be downloaded on any mobile device to control the sensor and the pump.
The key component of the Android app is the Model Predictive Control (MPC) – a powerful algorithm that has been proven to be a good candidate for AP systems, according to the researchers. It predicts how much insulin is required based on the sensor’s data and sends the signal to the insulin pump. This predictive nature makes MPC a good algorithm for the AP system, since the blood glucose levels of Type-1 patients need to be continuously regulated.
In a pilot study, the researchers tested their AP system on 10 Type-1 diabetic patients at MS Ramaiah Medical College from January – March 2022 after obtaining clearance from the ethics committee and obtaining informed consent from the patients. Out of the 10 subjects, four were successful where the blood glucose was within normal ranges throughout the duration of the trial. Three out of the remaining six were partially successful but were abandoned halfway primarily due to hardware issues.
The researchers plan to make the AP system fully automated to work 24×7, as the current one operates over only one meal cycle in a day. This includes operations that will monitor insulin levels continuously, using AI-based techniques to improve the app algorithm, and developing another app through which caregivers can monitor the patient. “One of our main objectives is to customise it for Indian patients and make it cost-effective,” says Padhi. With this in mind, the researchers are collaborating with Amrita University to develop a cost-effective insulin pump, which is the vital component of the AP system.
Chandrasekhar A, Padhi R, Optimization Techniques for Online MPC in Android Smartphones for Artificial Pancreas: A Comparison Study, IFAC-PapersOnLine, 2022.
Chandrasekhar, Abishek, Divya Saini, and Radhakant Padhi. “An android phone based artificial pancreas system for type-1 diabetic patients.” 2021 IEEE 9th Region 10 Humanitarian Technology Conference (R10-HTC). IEEE, 2021.
Nath, Anirudh, Shivanagouda Biradar, Archana Balan, Rajeeb Dey, and Radhakant Padhi. Physiological models and control for type 1 diabetes mellitus: a brief review, IFAC-PapersOnLine, 2018.
Biradar S, Balan A, Padhi R, Dharamalingam M, Modified Bergman Minimal Model For Glucose-Insulin Dynamics And Estimation Of Model Parameters For Indian Population. Diabetes Technology & Therapeutics, 2019.
Ali, Sk Faruque, and Radhakant Padhi, Optimal blood glucose regulation of diabetic patients using single network adaptive critics, Optimal Control Applications and Methods, 2011.
Bhitre Nitin, and Radhakant Padhi, An adaptive insulin infusion approach for customized blood glucose regulation of type I diabetic patients, 2011 IEEE International Conference on Control Applications, 2011.
Indian Institute of Science
Phone Number: +91-80-22932756