Sonde Health is developing voice-based technologies capable of diagnosing medical conditions by analyzing vocal biomarkers—acoustic characteristics in the speakers voice.
The PureTech company announced today the licensing of an award winning audio analysis technology developed at the MIT Lincoln Laboratory. The technology has demonstrated best-in-class performance in screening for depression risk, and although mental health will be a primary focus area for Sonde moving forward, the company also plans to work with a range of other disease states, such as respiratory and cardiovascular conditions.
“Producing speech turns out to be one of the most complex biological functions that we do, and it requires an incredible coordination of large areas of your brain with your musculoskeletal system and your respiratory system,” explains Dr. Jim Harper, co-founder and COO of Sonde Health. “Research has shown that disease induced physiological changes in these systems can subtlety change the quality of the sound that you produce while speaking in ways that can be measured with modern computational methods.”
The idea that changes in vocal characteristics may indicate a particular disease state is nothing new. “As far back as the early 20th century, people described using characteristics of a patient’s speech to help assess whether or not they had depression,” says Harper. Now, companies like Sonde plan to use signal processing to extract acoustic features—such as pitch, rate of speech, and the linking of certain phonemes—to potentially diagnose medical conditions faster and with more accuracy.
The first generation of Sonde’s product will likely require active engagement with a piece of software on a mobile phone of similar device. The platform gleans information from changes in the acoustic features of the voice—not the content of the speech—so users will potentially be asked to provide speech samples in response to a predetermined prompt.
“As we move forward we are also using additional technologies that will allow us to collection and analyze voices passively without recording the voice,” says Harper.
Dr. Aimee Danielson, Director of the Women’s Mental Health Program at MedStar Georgetown University Hospital, sees potential in a diagnostic tool that objectively measures vocal biomarkers: “This would be particularly useful in conditions that are chronically underdiagnosed, like perinatal mood and anxiety disorders, including postpartum depression, and in other mental health and central nervous system disorders where there is a lack of objective and reliable screening and monitoring technologies,” she says.
Send this to a friend