The Sound of Your Voice May Reveal That You Have Diabetes
Scientists have discovered that artificial intelligence (AI) can diagnose type 2 diabetes through voice analysis. Researchers at Luxembourg Institute of Health have proposed an innovative method using AI technology that analyzes voice pitch, intensity, and tone. Traditionally, type 2 diabetes diagnosis relies on blood tests and evaluating the patient's overall condition, but this new approach could serve as an alternative screening strategy.
The study involved 607 participants, both diabetic and non-diabetic, who provided 25-second voice samples by reading a specific text. The AI model accurately identified diabetes in 66% of women and 71% of men. Dr. Abir Elbeji and Dr. Guy Fagherazzi, leading the research project, suggest that although this AI voice diagnostic might not replace blood tests entirely, it holds promise as a screening tool to identify at-risk individuals and undiagnosed cases. Given that approximately 240 million adults worldwide remain unaware of their condition due to mild or absent symptoms, early detection through this technology could significantly reduce complications associated with type 2 diabetes.
Further research is needed to enhance the model's effectiveness, but the potential accessibility and convenience of voice diagnostics present substantial benefits. Fagherazzi points out that people with diabetes exhibit unique vocal traits probably stemming from chronic high blood sugar levels, dehydration, and neuropathies, all contributing to distinguishable voice differences. The application could also be particularly precise among older women and individuals with hypertension.
While current diabetes detection methods like blood tests remain standard, the AI method showed a 93% agreement rate with the American Diabetes Association questionnaire-based risk score. However, experts caution that more validation in real-world populations is required before adopting voice diagnostics broadly.
This cutting-edge research reflects a broader tendency among scientists to explore voice as a diagnostic tool for various diseases, including Parkinson's, heart disease, and depression. Voice diagnostics are still in their nascent stage, but continuous advancements promise a future where a simple voice recording could screen for significant health conditions reliably.
Earlier, SSP wrote that Air France brought Starlink to its entire fleet.