The use of reinforcement learning algorithms to meet the challenges of an artificial pancreas.

Blood glucose control, for example, in diabetes mellitus or severe illness, requires strict adherence to a protocol of food, insulin administration and exercise personalized to each patient. An artificial pancreas for automated treatment could boost quality of glucose control and patients' independence. The components required for an artificial pancreas are: i) continuous glucose monitoring (CGM), ii) smart controllers and iii) insulin pumps delivering the optimal amount of insulin. In recent years, medical devices for CGM and insulin administration have undergone rapid progression and are now commercially available. Yet, clinically available devices still require regular patients' or caregivers' attention as they operate in open-loop control with frequent user intervention. Dosage-calculating algorithms are currently being studied in intensive care patients [1] , for short overnight control to supplement conventional insulin delivery [2] , and for short periods where patients rest and follow a prescribed food regime [3] . Fully automated algorithms that can respond to the varying activity levels seen in outpatients, with unpredictable and unreported food intake, and which provide the necessary personalized control for individuals is currently beyond the state-of-the-art. Here, we review and discuss reinforcement learning algorithms, controlling insulin in a closed-loop to provide individual insulin dosing regimens that are reactive to the immediate needs of the patient.

Local tolerance of intraarticular administration of lornoxicam into the rabbit knee joint.

The local tolerability of lornoxicam (Xefo) after single and repeated intraarticular administration was assessed in the rabbit and compared to established standard therapies (hyaluronic acid--Synvisc and the glucocorticoid triamcinolone--Triam), and the results are discussed in the context of the literature. Two local tolerance studies were performed using five male rabbits per group. Lornoxicam and competitor products were administered into the right knee joint in a volume of 500 µL. The contralateral left knee joint of the same animal was used as the control and was injected with water for injection. Three out of five animals were killed 72 h after the last administration, whereas the remaining two animals were subjected to a 2- or 6-week recovery period in the first and the second study, respectively. Findings revealed adaptive changes related to the mechanical irritation of the injection and to adaptive responses of the synoviocytes, but no signs of toxicity to bone or chondrotoxicity. Toxicokinetic analysis showed a fast and almost complete absorption of lornoxicam from the joints into the systemic circulation. As a conclusion, repeated intraarticular administration of lornoxicam was well tolerated in rabbits.