Funded by the Stuart Scott Memorial Cancer Research Fund
Voice is an incredible tool that we use every day to express our feelings and even our health. You can often tell if someone is stressed, happy, or not feeling well just by the way they sound. In fact, over 50 different diseases can cause noticeable changes in a person’s voice. One of the most serious conditions that affect the voice is laryngeal cancer, or cancer of the voice box. This type of cancer can cause major changes in a person’s voice, even in the early stages. Unfortunately, if it is not caught early, it can lead to the loss of voice, difficulty swallowing, and too many cases, death. The earlier we detect laryngeal cancer, the more treatment can preserve the voice and improve survival rates. The problem is that while many people with laryngeal cancer experience voice changes, most people who have a change in their voice do not have cancer. This creates a challenge for primary care doctors, who need to identify the rare instances where voice change suggests something more serious. To help with this, we are developing a machine learning tool that can listen to voice recordings and help doctors figure out when a patient may be at high risk for laryngeal cancer. This could help detect cancer earlier and save lives by making it easier for doctors to know when to refer patients to specialists for further testing.
