Decoding Islet Algorithms in Biosensors for open and closed-Loop glycemia control – DIABLO

Chronic diseases are the main cause of death with a major burden on health budgets and economic losses. In TYPE 1 DIABETES MELLITUS destruction of pancreatic beta-cells leads to absolute insulin deficiency and concerns 5-10% of the current 371 million diabetes patients, expected to rise to 592 million by 2035 worldwide. T1DM starts generally in children/young adults with dire consequences and requires long-term solutions.

CONTINUOUS GLUCOSE MONITORING (CGM) linked to insulin delivery presents a major recent advance but suboptimal therapeutic insulin levels and life threatening hypoglycemia are still serious concerns. Current technology uses subcutaneous glucose measuring via electrochemical electrodes and algorithms to predict insulin dosage. This ARTIFICIAL PANCREAS is limited by its glucose sensor which does not capture other nutrients such as lipids or the general body status relayed by hormonal signals, conveying altered needs during stress, physical activity, cycle etc. Therefore NUMEROUS ALGORITHMS have been developed which still imperfectly forecast and need considerable user intervention precluding closed loop function.

IN CONTRAST, 0.5 billion years of evolution have shaped islets as “in-born” sensors/actuators in glucose homeostasis. ISLET ENDOGENOUS ALGORITHMS encode electrical activity for biphasic oscillatory insulin secretion and insulin safety mechanisms but underlying mechanisms are not fully known. We have already developed, patented and published a HIGH-RESOLUTION MICROFLUIDIC MULTI-ELECTRODE ARRAY recording and an ON-LINE REAL TIME ANALYSIS SYSTEM for islet electrical activity as biosensor.

DIABLO IS A MULTIDISCIPLINARY PROJECT of experts in CGMS diabetes therapy, islet electrophysiology, electronic devices/systems for real time interaction with biology, control automation for operational autonomy of complex safety-critical systems. We aim ultimately at a bio-microelectronic hybrid sensor and robust controller for autonomous, continuous hormone replacement therapy responsive to different every-day life situations through three main objectives:

OBJECTIVE 1: Assess the DIABLO human in silico biosensor model in physiological conditions (glucose, hormones). Using high-resolution islet data (generated by us) and parametric/non parametric identification techniques, we will design a mathematical model of islet sensing able to reproduce the observed data, predict their future behavior and explore underlying mechanisms.

OBJECTIVE 2: Enhance the reference whole human body single-input T1D simulator (T1DMS) by integrating the biosensor model from Obj. 1 and an innovative multi-parametric controller. Ensure the controller can be embedded into a hardware device.

OBJECTIVE 3: Design the DIABLO biosensor device and assess it in in vivo and extra corpore experiments (microdialysis). Validate the controller via a CLINICLA TRIAL with diverse daily scenarios, followed by a comparison to classical CGMS and evaluation of the benefits of our 1) multi-parameter sensor and 2) robust multi-input controller.

DIABLO will impact RESEARCH, HEALTHCARE TECHNOLOGY and DIABETES THERAPY by (i) DECODING of endogenous islets algorithms suitable also for other CGMS devices, (ii) ; 2) High-level synthesis techniques for EMBEDDED INTEGRATED CIRCUITS able to optimally process biosignals in real time, (iii) a NEW BIO-ELECTRONIC SENSOR to investigate and maintain glucose homeostasis for application in man, (iv) artificial pancreas algorithms by A NEW CLOSED-LOOP CONTROL with stability and robustness to sudden changes in a multivariate environment; (v) as high-resolution technique and modeling in development of human ORGANS-ON-CHIP, to provide unique insights into drug actions and genetic alterations.