AI-powered voice analysis may revolutionise early diagnosis of Parkinson's disease

Artificial Intelligence (AI)-based algorithms that can detect subtle changes in a person's voice and detect Parkinson's disease may be emerging as a potential new diagnostic tool for the neurodegenerative disorder, according to a study.

The disease currently affects more than 8.5 million people. The incidence of Parkinson's has doubled in the past 25 years and now contributes to at least 330,000 deaths each year worldwide.

However, traditional diagnostic methods are often complex and slow, delaying early detection, according to researchers from Iraq and Australia who reviewed the advancements in AI techniques to detect Parkinson's disease.

The paper, presented to the Fifth Scientific Conference for Electrical Engineering Techniques Research in Baghdad, showed that all the evidence shows that AI-powered voice analysis could revolutionise early Parkinson's diagnosis and remote monitoring of the neurodegenerative disorder.

Speech impairments are often the first indicators of the fastest-growing neurological disease in the world.

“Vocal changes are early indicators of Parkinson's disease, including small variations in pitch, articulation, and rhythm, due to diminished control over vocal muscles,” Ali Al-Naji, a medical instrumentation engineer from Middle Technical University (MTU) in Baghdad.

“By analysing these acoustic features, AI models can detect subtle, disease-related vocal patterns long before visible symptoms appear,” Al-Naji said.

Simple voice recordings from Parkinson's patients and healthy controls are used to train the AI tools primarily using machine learning and deep learning algorithms.

Relevant features, such as pitch, speech distortions, and changes in vowels are extracted by these algorithms and then categorised in the voice recordings with remarkable accuracy -- as high as 99 per cent in one research study.

Researchers noted that while Parkinson's has no cure, early diagnosis, and intervention can improve quality of life and slow the progression of symptoms.

Besides early detection, “AI may also help monitor patients from a distance, reducing the need for in-person visits,” Al-Naji said, while acknowledging the need for further studies on larger, more diverse populations.