An artificial intelligence system would be able to measure in real time the glucose level of children with type 1 diabetes. Thus, it could more precisely deliver the dose of insulin they need throughout the day.
- Treatments for type 1 diabetes in children are less effective than in adolescents or adults.
- Researchers have therefore developed an artificial intelligence system capable of estimating their needs more precisely.
In France, the incidence of type 1 diabetes was estimated at 19.1 per 100,000 in children aged 6 months to 14 years in 2015, according to Public Health France. It is the most common type of diabetes in this age group. This one is characterized by a high level of sugar – or glucose – in the blood, hyperglycemia. It is due to a deficiency of insulin, a hormone produced by the pancreas and which regulates blood sugar levels. This illness is caused by the dysfunction of certain cells of the immune system, the T lymphocytes. The problem is that they identify the cells of the pancreas as cells foreign to the patient’s body and eliminate them. Thus, the pancreas can no longer regulate the blood sugar level, which is therefore too high.
Conventional treatments not suitable for children under 7 years old
Currently, the treatment offered to children aims to control blood sugar, that is to say that their blood sugar level remains at the correct level. For this, they must receive an insulin supply via a treatment which can be either conventional insulin therapy – 2 to 3 subcutaneous injections of regular insulin and 1 to 2 injections of slow or intermediate insulin per day – or functional insulin therapy, i.e. say 4 or 5 subcutaneous injections a day or installing an insulin pump that delivers them alone. The doctor, in agreement with the child and the family, chooses the best treatment according to his pathology. Nevertheless, in those under seven years old, these devices do not seem sufficiently precise because their activity rhythm and diet vary too much throughout the day. In fact, their insulin levels change much more than teenagers or adults. In the endthey have a greater risk of being in hyperglycemia or hypoglycemia, a level of sugar that is too low in the blood.
A new artificial intelligence system…
But a new treatment may soon be available for these young children with type 1 diabetes. According to a study published in the journal New England Journal of Medicine, an artificial intelligence system called CamAPS FX would be able to predict the insulin needs of these young patients in real time, thus mimicking the role that the pancreas would have if they were not sick. To do this, CamAPS FX alone measures children’s blood sugar levels and determines the exact dose of insulin to deliver to them.
…able to better adjust the insulin dose to their real needs
If they are equipped with this device, children no longer have to think about what they eat because the intelligence system regulates their blood sugar levels on their own. For example, if the young patient takes a very sweet snack, CamAPS FX will automatically detect a hyperglycemia and controlling it by sending insulin through the body. Another advantage: as the system is connected to the body of the carrier, the analysis of the necessary dose will be more precise and therefore more adapted to the real needs of the child, compared to existing treatments.
So far, clinical trials conducted on diabetic children under the age of seven have obtained very good results: the young patients are less often in hyperglycemia than with their usual treatment.
.
