Diagnostic Parkinson’s with audio analysis to intercept pre-tremor symptoms – Corriere.it

More than ten million people around the world live with Parkinson’s disease. Slowly developing neurodegenerative diseaseBut it is progressive, and includes functions such as motion control and balance. There is no cure, but if symptoms are caught early, the disease can be controlled. the early symptoms Non-motors can appearI am up to ten years before the diagnosis Of the disease. By the time the characteristic Parkinson’s symptoms such as tremor or stiffness appear, half of the dopamine-producing neurons have already died. Gianni PezzoliNeurologist, President of the Italian Parkinson’s Association and the Grisons L Parkinson’s disease.

difficulty speaking

These are the symptoms that characterize it
to Difficulty pronouncing words. The sound becomes more gradual WeakIt is often monotonous, with difficulty pronouncing words and sometimes opening the mouth. Usually the patient speaks in a way Less expressive, slower, more fragmented And this is not always easy to perceive with the naked ear. Hoarseness, stuttering, slurring of words, and loss of pauses between words become more pronounced as the disease progresses. Often the listener finds himself having to ask to raise his voice because he cannot hear.

Changes in language are gradual And a group of researchers from Kaunas University of Technology (KTU), Lithuania with colleagues from the University of Health Sciences worked meticulously to try Identifying early symptoms of Parkinson’s disease using speech dataChanges in speech often appear much earlier than motor function impairment, which is why speech impairment may be one of the first signs of illness. Rytis MakeleonasResearcher in the Department of Multimedia Engineering at KTU. Today, with the advancement of technologies, it has become possible to extract more and more information from the spoken language and the Lithuanian working group has been usedArtificial intelligence To analyze and evaluate speech signals and calculations are performed in seconds, facilitating diagnosis. We’re not creating an alternative route to necessary routine screenings — our method is designed to facilitate early diagnosis of disease and monitor the effectiveness of treatment, says Maskelinas.

in the future app

At the moment, the algorithm has been created (the study was published in applied Sciences) It has only been tested in patients diagnosed with Parkinson’s disease. Our system can distinguish people with Parkinson’s disease from healthy people using CMy voice is wide Explain Kipras PribuisProfessor of Otolaryngology and one of the study’s authors. In a soundproofed room, a microphone was used to record the speech of people with Parkinson’s disease and healthy people. The AI ​​algorithm learned to process the signal by evaluating the recordings. The researchers point out that the algorithm does not require particularly powerful hardware in the future It can also work with an app. Our findings have very high scientific potential, the researchers say, even if there is still a long way to go until we can apply the system in everyday practice. The next steps will be to expand the patient population to collect as much data as possible and it will also be necessary to check if the algorithm works well not only in laboratory settings but also in the doctor’s office or even in the patient’s home.