AI Voice Tech in Healthcare — Adapted from the book “How AI Can Democratize Healthcare by Michael Ferro and Robin Farmanfarmaian
AI Voice Tech in Healthcare: Adapted from the book “How AI Can Democratize Healthcare by Michael Ferro and Robin Farmanfarmaian

AI Voice Technology in Healthcare: Five Startups On the Cutting-Edge


AI voice tech has been integrated into most American’s lives in some way, including smart speakers like Amazon’s Alexa (Echo) to smart phones assistants like Apple’s Siri. Beyond the basic smart speakers, AI voice tech has grown into new areas of healthcare, including vocal biomarkers.

What are Vocal Biomarkers?

Vocal biomarkers are metricizing a person’s voice to be used as a vital sign.

Vocal biomarkers translate and measure someone’s voice as data points, which can then be used to help lead to an overall diagnosis, the same way measuring vital signs like blood pressure, imaging, or blood lab testing can help lead to an overall diagnosis. A patient’s vocal biomarkers can include things like the speed of their speech, the way they string words together over a period of time, their tone, pitch, annunciation, and fluency. Once the voice aspects are metricized, it is easy to detect if any aspect of someone’s voice has changed day-to-day. Human beings can’t detect the small changes day-to day-that an AI software program can detect, making vocal biomarkers a great tool because they are able to quantifiably measure tiny changes and objectively compare it to earlier recorded data. Human beings must rely on subjective data when conversing.

Vocal biomarkers are a great tool because they objectively measure what was previously not easily or inexpensively quantifiable. Adapted from the book by Michael W Ferro and Robin Farmanfarmaian, “How AI Can Democratize Healthcare”
Ellipsis Health in the Medium article on AI Voice Technology by Michael W Ferro and Robin Farmanfarmaian

Ellipsis Health

Founder & CEO Mainul Mondal

Ellipsis Health is a vocal biomarker startup that has developed the only vital sign that has been clinically validated for mental health. Previously, there was no scalable, objective, and personalized measurement for mental health. Ellipsis Health solves that problem with AI-powered deep learning algorithms and the unique power of the human voice.

With a short voice sample, Ellipsis Health’s AI-based technology identifies, measures, and monitors the severity of depression and anxiety at scale. This creates a personalized, objective, and scalable clinical decision support tool (CDS).

“Assessing your behavioral wellbeing can be as easy as asking your smart speaker about the weather. Machine learning algorithms make it possible to use the words you say and how you say them to measure your level of depression and anxiety.” said Mainul Mondal, Founder & CEO of Ellipsis Health.

Ellipsis Health: How it Works from the voice tech Medium article by Robin Farmanfarmaian and Michael Ferro, jr
Ellipsis Health: How it Works from the Ellipsis Website
Hyfe AI in the Medium article on AI Voice Technology by Michael W Ferro and Robin Farmanfarmaian

Hyfe AI

Co-founders Joe Brew, Paul Riege, Iulian Circo
Chief Medical Officer Dr. Peter Small, Acoustic Epidemiologist

Coughs as Biomarkers

Hyfe AI detects and tracks coughs and cough frequency, remotely and in real-time using a smartphone or Hyfe AI’s autonomous wearable device. Their wearable tracker doesn’t need to connect to the internet, so it can be worn anywhere and anytime. One of their goals is to transform health systems by creating evidence-based acoustic tools for disease detection and tracking. Currently they have a free app that can be used as a wellness information tool. The app can record and monitor coughs, and users can elect to share the recordings with their care team. Now when a patient speaks with their physician, they can report the exact number of times they cough while awake and asleep. Cough frequency tracking provides objective data that can indicate whether a patient’s symptoms are worsening or improving, which could lead to a change in medication, environment, or additional medical tests.

Tracking the number of times a patient coughs is crucial for clinical trials as well. When testing a new medication for diseases and conditions like whooping cough (pertussis), coughing is a major symptom to track.

Hyfe AI image from the vocal biomarker article by Robin Farmanfarmaian and Michael W Ferro
From the Hyfe AI Website
Sonde Health Logo in the vocal biomarker article written by Michael W Ferro and Robin Farmanfarmaian

Sonde Health

Founder James Harper , CEO David Liu

Sonde is an AI voice technology and vocal biomarker startup that turns voice into a vital sign.

They have identified 4,000 different vocal features. Sonde can detect the smallest changes in someone’s voice, which could indicate that the person is about to start or has already started to experience some type of a symptom.

They already have products for depression, respiratory illness, sleepiness and hypertension detection and are working on many other diseases and disorders.

Sonde Health isn’t claiming to be a diagnostic. What the AI-enabled software does is measure the risk of disease by identifying symptoms through vocal analytics. Sonde Health also has applications for behavioral health, respiratory health, and one that measures the risk of congestive heart failure (CHF). One great feature is that Sonde Health can be easily integrated into other apps through a self-serve API.

From the Sonde Health Website

Hardware Components for Vocal Biomarker Collection

There are 2 main components to AI voice tech and smart speakers: the hardware component and the software component. In terms of the hardware, the difference in the speaker + microphone hardware is mostly aesthetic. Beyond the basic speaker + mic, the hardware components could include a smart tablet or screen and a video camera for applications like video calls, surfing the web, and displaying written information. The hardware could also be mobile, as in a mobile robot on wheels or with movable limbs.

Elli-Q in the AI Voice Technology article on Medium by Michael Ferro jr and Robin Farmanfarmaian

ElliQ by Intuition Robotics

Co-founder and CEO Dor Skuler

The ElliQ is a smart speaker, video camera, and tablet combination for aging-in-place that provides contextual health coaching through an empathetic digital companion. ElliQ provides entertainment and proactively engages the senior. It allows the senior to call people on video just by saying “ElliQ, please call my doctor”. ElliQ also gives vocal and written medication reminders and learns the senior’s schedule in order to determine if there might be a problem.

While companies like Google and Amazon’s AI voice tech is assistive, ElliQ is intended to be a companion.

Check out some articles here and here on ElliQ’s release in the spring of 2022.

Elli-Q Smart Speaker in the AI Voice — vocal biomarker article written by Michael Ferro and Robin Farmanfarmaian
iPal in the AI Voice Technology article on Medium by Michael Ferro jr and Robin Farmanfarmaian: iPal is made by Avatar

iPal by AvatarMind

Co-founder John Ostrem

Originally designed as a teachers’ aid for children, the iPal2 offers an open development platform allowing users to leverage brain games, music therapy and other content to engage and entertain older adults with various levels of cognitive ability. The iPal2 is a 40" tall humanoid robot with a 10" Android tablet on the chest, a hand-held remote control. In addition to Alexa, Zoom and other popular apps, AvatarMind includes proprietary apps to facilitate text-to-speech chats, changes in facial expressions and movement (using a smart phone or separate tablet). This easy-to-use robot can reduce stress on staff or volunteer caregivers when used for 1:1 visits or in group activities.