New Delhi– Artificial Intelligence (AI) could help identify early-stage laryngeal cancer — cancer of the voice box — by analyzing the sound of a patient’s voice, according to new research by US scientists.
Laryngeal cancer is a significant public health concern, with an estimated 1.1 million cases worldwide in 2021 and about 100,000 related deaths. Major risk factors include smoking, excessive alcohol consumption, and human papillomavirus (HPV) infection. Prognosis varies widely — from 35 percent to 78 percent five-year survival — depending on tumor stage and location.
Researchers at Oregon Health & Science University demonstrated that AI can detect abnormalities in the vocal folds, such as benign lesions (nodules or polyps) or early cancer, by analyzing vocal biomarkers. Their proof-of-principle findings, published in Frontiers in Digital Health, suggest AI could become a non-invasive screening tool for detecting early warning signs of laryngeal cancer.
“Here we show that with this dataset we could use vocal biomarkers to distinguish voices from patients with vocal fold lesions from those without such lesions,” said Dr. Phillip Jenkins, a postdoctoral fellow in clinical informatics at Oregon.
The team analyzed 12,523 voice recordings from 306 participants across North America, including individuals with laryngeal cancer, benign vocal fold lesions, and other voice disorders such as spasmodic dysphonia and unilateral vocal fold paralysis.
They examined acoustic features such as mean fundamental frequency (pitch), jitter (pitch variation), shimmer (amplitude variation), and harmonic-to-noise ratio (HNR) — a measure of the balance between harmonic and noise components in speech.
The study found significant differences in HNR and pitch between men with no voice disorder, men with benign lesions, and men with laryngeal cancer. No similarly informative features were identified in women, though researchers believe larger datasets may reveal such differences.
The findings indicate that variation in HNR could be useful for tracking the progression of vocal fold lesions and detecting laryngeal cancer early — at least in men — offering potential for AI-assisted voice screening in the future. (Source: IANS)
