The new method developed by scientists; It has the potential to eliminate distance, cost and time barriers.
Diabetes The disease occurs when the level of glucose, that is, sugar, in the blood rises above normal and as a result, glucose is found in the urine, which should not normally contain glucose. It is one of the most common diseases in our country and even around the world.
Even though you may think we don’t come across it very often, data is available all over the world. more than 400 million He states that he has diabetes. And although it may seem like a simple disease, 1 person dies every 6 seconds in the world due to diabetes-related reasons. That’s why it’s critical to diagnose it. The most recent step taken by scientists is to improve the work of both doctors and patients. will make it easier looks like.
Voice recognition and artificial intelligence come together to diagnose diabetes.
In the study published in Mayo Clinic Proceedings, scientists examined the time between 6 and 10 seconds. audio recordings used with simple health information. As a result, he developed an artificial intelligence model that can distinguish whether individuals have type 2 diabetes. The model in question is for men 86% accuracy rate While this rate increases to 89% for women.
Klick Labs researchers who conducted the study included both type 2 diabetic and healthy participants. 267 people included in the study. The researchers asked each of these people to record a certain sentence 6 times a day for 2 weeks. Researchers, who obtained more than 18,000 records at the end of a 2-week period, found that those without diabetes and those without diabetes between the sounds He analyzed 14 acoustic differences.
RELATED NEWS
A New Method Has Been Announced That Could Clearly Explain How Life Occurred
Surprisingly, these differences, which cannot be distinguished by the human ear, also differed between men and women with diabetes. As a result, as we mentioned before, this artificial intelligence model; It showed an accuracy rate of 89% for women and 86% for men. From here on, the team wants to use the diagnosis method through voice recognition in different areas. Voice diagnostics of pre-diabetes and hypertension are some of these areas.