The University of Rochester in the United States has developed an artificial intelligence system that helps detect Parkinson’s disease from the voice. For the analysis, it is enough to say aloud the English phrase “The quick brown fox jumps over the lazy dog,” which contains all the letters of the alphabet.
The AI evaluates tone, rhythm and speech clarity. On a test set of 1,306 recordings, it achieved 85.7% accuracy, which is higher than that of non-specialist doctors (73.8%) and comparable to movement-disorder experts (79.6%).
The method is simple: all that is needed is a computer with a microphone and internet access. This is especially important for regions where neurologists are in short supply and a trip to a doctor takes a long time. The technology has limitations: it works only for English speech, about a quarter of cases remain undetected, and natural voice changes in older people sometimes interfere with the analysis. Even so, the system has shown no bias by sex, age or ethnicity and has already proven that it can be useful for mass screening.
Researchers are now adapting the technology to other languages. In a few years, an everyday “Good morning!” to a voice assistant may be not only a greeting, but also a quick test of nervous system health.
