Artificial Intelligence can detect diabetes by listening to the voice

Isbel Lázaro.

Inteligencia Artificial

Inspenet, December 23, 2023.

A recent study published in the specialized journal Mayo Clinic Proceedings: Digital Health reveals that it is now possible to use Artificial Intelligence to analyze a brief sequence of speech and accurately identify the presence of type 2 diabetes in a person. This technology aims to facilitate the identification of individuals who have not yet been diagnosed with diabetes.

Approximately 240 million adults around the world live with diabetes without being aware of it and according to the International Diabetes Association, about 90% of diabetes cases correspond to type 2. People with this type of diabetes face an increased risk of heart and vascular diseases, such as heart attacks, strokes, and circulatory problems in the legs and feet.

Figure AI
robot blando de Toyota
modulo lunar Odysseus interrumpe su actividad
Renault nuevo auto electrico R5
Trinidad y Tobago
Figure AI raised more than $650 million to commercialize humanoids
Meet Punyo: Toyota’s new soft robot
Forced end! The Odysseus lunar module interrupts its activity
Renault presented its mythical new electric car R5
Trinidad and Tobago hires two companies to tackle oil spill
previous arrowprevious arrow
next arrownext arrow

How does this Artificial Intelligence work?

During voice frequency analysis, Artificial Intelligence examines changes in the voice that are imperceptible to the human ear. Recordings of telephone conversations are used to carry out this analysis, evaluating aspects such as intonation, rhythm, pauses and tones.

Likewise, recently developed Artificial Intelligence has the ability to diagnose type 2 diabetes from a voice recording that lasts between six and ten seconds. By combining this voice data with basic health information such as age, gender, height and weight, AI can identify the presence of the disease. Diagnostic accuracy is surprisingly high, as voice changes vary between men and women. The diagnostic accuracy rate was 89% in women and 86% in men.

To develop this Artificial Intelligence, a team of researchers led by Jaycee Kaufman from the Ontario Technological University recorded the voices of 267 individuals, some without diabetes and others with a previous diagnosis of type 2 diabetes. Over a two-week period, these participants recited a short phrase on their smartphones six times a day.

From more than 18,000 voice samples collected, they identified 14 acoustic characteristics that present significant differences between people with and without type 2 diabetes . Jaycee Kaufman, lead author of the study and researcher at Klick Labs, highlighted that current detection methods can be expensive and require time and travel, while voice technology has the potential to overcome these barriers.

In the future, Klick Labs intends to explore whether other conditions, such as high blood pressure and prediabetes, can also be detected through voice analysis. However, they emphasize that, even if the results are accurate, it is always necessary for medical professionals to confirm the diagnoses, which also applies to the use of artificial intelligence to diagnose mental illnesses.

Don’t miss any of our posts and follow us on social media!






Source: =Artificial %20inteligencia% %20puede% 20detect, but %20presenta% %20grandes% %C3% A9n 20risks.&text=The %20diagn% C3 %B3sticos% 20m %C3% A9dic %20mediante% 20an %C3% A1lysis %20de% 20voice %20son% 20every 20time %20vez% %C3% A1s %20accurate

Share this news on your social networks

Rate this post
1 star2 stars3 stars4 stars5 stars (No rating yet)