23rd February 2023
Mollie Hillis, Senior Account Executive

Artificial intelligence and diabetes; how advances in technology have liberated patients

Type 1 diabetes mellitus (T1DM) is a chronic autoimmune condition characterised by T cell mediated destruction of the β cells in the pancreas that affects the patient’s life in a multitude of ways1. Micro and macrovascular complications of hyperglycaemia include heart disease, blindness, neuropathy, loss of limb and impaired renal function1,2,3. The burden of the disease on the patient is high, with frequent monitoring of blood glucose levels and adjustment of insulin doses2. Over the last 10 years, the progression of the treatment of this condition has catapulted into the stratosphere, with advances in continuous glucose monitors (CGMs) and insulin pumps making the disease easier to manage and control than ever before2. The eventual goal would be a cure, but how far are we from that day?

In the early 2000s, the standard treatment of T1DM was a different story to today. Blood testing was done daily with finger pricks, the insulin’s available were less sophisticated and the thought of a CGM was not yet a reality2. Fast forward to 2023, and daily finger pricks are a thing of the past. Currently on the NHS, all people with T1DM have access to flash monitoring systems (scan a sensor on your arm and read your blood sugars on an LED device) and other CGM devices are becoming more readily available4. Alongside this, the structure of basal insulin has improved greatly, helping extend the drug’s half-life once within a patient and enabling glucose levels to remain steady and avoid hypoglycaemia5.

Alongside these advances, insulin pumps are now more readily available. These devices are attached to the patient via cannulation and secrete pre-adjusted volumes of insulin, and the patient is able to input variables to increase or decrease the speed of secretion – effectively removing the need for subcutaneous injection6. The first insulin pump was developed in the early 1960s; however this was a much larger device which was as large as an army backpack pack2,6. With vast technological developments, the devices are now approximately the size of small box of matches and work extremely efficiently2.

There are currently trials across NHS England for over 100,000 people for a closed-hybrid loop system (CLS) which in theory acts as an artificial pancreas2. The NHS has announced funding worth £2 million for a pilot roll out of this sci-fi development in which artificial intelligence (AI) communication between the insulin pump and the CGM allows for overall improved glucose control when compared to conventional insulin pump therapy2. There are several variations on this currently, all of which need the patient to continue to input variables to deliver insulin (such as when eating carbohydrates or exercising). The eventual holy grail would be a fully automated, closed system, where the user can live life without worrying about their glucose levels2,7.

Some may say we are living in the heyday of bio-tech developments within this field, with artificial intelligence pushing patients with T1DM ever closer to a ‘normal life’. The continuous innovation and development within diabetes research is bringing us a step closer to the so-called ‘artificial pancreas’. Although this has not yet been realised in clinical practice, the overarching success of hybrid closed loop systems will bring much needed relief from the never-ending burden of T1DM to patients.


  1. Akil et al, 2021 https://translational-medicine.biomedcentral.com/articles/10.1186/s12967-021-02778-6
  2. Templer, 2022 https://www.frontiersin.org/articles/10.3389/fendo.2022.919942/full
  3. Jonassen et al, 2008 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2494662/
  4. NHS England, 2022 https://www.england.nhs.uk/diabetes/digital-innovations-to-support-diabetes-outcomes/flash-glucose-monitoring/
  5. Woo et al, 2020 https://link.springer.com/article/10.1007/s13300-020-00915-w
  6. Yao et al, 2022 https://www.ncbi.nlm.nih.gov/books/NBK555961/
  7. Melmer et al, 2020 https://pubmed.ncbi.nlm.nih.gov/32996429/