Vocal Biomarkers


Brain / Nervous System

Vocal Biomarkers
Depression, Suicidality, Parkinson's Disease

Brain / Nervous System

Vocal Biomarkers
INDICATION(S):Depression, Suicidality, Parkinson's Disease
NAME:Sonde (93.1%*)
STAGE:Phase 1
Vocal Biomarkers

Our Sonde affiliate has the potential to fundamentally change the way mental and physical health is monitored, managed, and diagnosed, beginning with conditions that affect the neurological, muscular, and respiratory systems required for speech production. We are advancing a proprietary voice-based technology, developed internally and licensed from the Massachusetts Institute of Technology (MIT) Lincoln Laboratory, which has been tested in over three thousand individuals to date. This development work has demonstrated the potential to use commonly owned devices like smartphones to effectively screen and monitor a range of diseases using information obtained from an individual’s voice.

  • Patient Need & Market Potential
    • High-tech devices that continuously stream sensor data are ubiquitous, but there remains a major gap in converting this information into broadly actionable health insights. The lag between onset of disease and accurate diagnosis and beginning of treatment can be measured in years for many high-burden health conditions including depression, Alzheimer’s disease, multiple sclerosis, and Parkinson’s disease, to name just a few. Near-continuous health information, powered by our technology, has the potential to improve diagnosis, monitoring, and treatment of high-cost conditions, broadly improving outcomes and care efficiency.
    • For example, despite the high burden of depression and the recommendation that the 270 million U.S. adults and adolescents receive depression screening once a year, fewer than 10 percent of patients are screened and only about half of those diagnosed receive care as recommended. There is an escalating demand for solutions to improve screening rates and facilitate more rapid intervention that could potentially save lives.
    • Development of effective therapies for central nervous system diseases and disorders is hampered by the high cost and inherent variability of these diseases and the reference diagnostic measures used to characterize them. Objective digital tools that can augment and perhaps one day replace the current clinical endpoints with novel measures that can be measured with more meaningful accuracy and less burden can improve patient enrollment and drug development for a range of important conditions.
  • Our Approach to Solving the Problem
    • Our proprietary technology is being developed to enable a range of consumer devices such as smartphones and smart speakers to provide effective disease screening and management solutions based on an analysis of seconds of voice capture. By tailoring the information produced from these objective voice measures to correlate with existing screening and diagnostic measures that integrate seamlessly with patient care flows and individuals’ daily lives, we are creating services to address a range of health care needs from depression to respiratory and aging related conditions.
  • Intellectual Property
    • We have broad intellectual property coverage worldwide, currently owning or having exclusive rights to three issued patents and twelve patent applications in four families of patent filings. We have filed several patent applications covering a number of facets of its technology in addition to the IP that was licensed from MIT.
    • The first U.S. patent, broadly covering methods of assessing mental or physical conditions of a subject using speech audio waveforms extracted from phonetic structures, or phones, was issued in September 2017.
    • A second U.S. patent, covering phonological biomarkers that can be used in the screen and monitor of physical and psychological disorders, such as major depressive disorder, was issued in April 2018.
    • We were recently granted an Australian patent covering the use of vocal biomarker technology to detect physical or psychological conditions in the human body. The patent provides coverage until 2034.
  • Team
    • Key advisors include Dr. Maurizio Fava, Director of the Division of Clinical Research of the Massachusetts General Hospital (MGH) Research institute and Executive Vice Chair of the MGH Department Psychiatry; Dr. Harry Leider, Chief Medical Officer and Group Vice President of Walgreens; Dr. Ian Gotlib, Professor and Chair of the Department of Psychology at Stanford University and Director of the Stanford Mood and Anxiety Disorders Laboratory; Dr. Helen Christensen, Director and Chief Scientist of the Black Dog Institute and Professor of Mental Health at the University of New South Wales; Dr. Aimee Danielson, Founder and Director of the Women’s Mental Health Program at MedStar Georgetown University Hospital; Dr. Julien Epps, Associate Professor in Signal Processing with the School of Electrical Engineering and Telecommunications at UNSW Australia; and Dr. Robert Horvitz, Board Advisor & Scientific Advisory Board Chair at PureTech Health, Nobel Laureate, HHMI, MIT.

    • Sonde is led by Dr. Jim Harper, Dr. Eric Elenko, and Mr. Yogendra Jain.
  • Milestones Achieved
    • A scalable cross-platform mobile research app and administrator interface were made available to academic collaborators and study participants. This has allowed us to build a mobile depression and speech research corpus which has gathered data from over 3,000 volunteers to date.
    • Our technology has demonstrated best-in-class accuracy for recognizing depression in individuals and estimating the severity of their symptoms from brief samples of speech.
  • Collaborations
    • We are collaborating with the University of New South Wales (UNSW) and Black Dog Institute to create the first mobile device-based automatic assessment of depression from acoustic speech. UNSW was awarded a Linkage Project, funded by the Australian Government through the Australian Research Council (ARC) for international collaboration and partnership in research and innovation. The Linkage Project aims to support long-term strategic research alliances between organisations, in order to apply knowledge to highly technological and high-risk problems.
    • Pilot studies using our core technology have also demonstrated the potential to detect and objectively measure symptoms in a range of important conditions including depression, mild traumatic brain injury (mTBI), concussion, cognitive impairment and Parkinson’s disease.
    • To increase efforts to accelerate understanding and use of vocal biomarker technology for mental and physical health, we have entered into collaborative partnerships with leading institutions, including UMass Memorial Medical Center, Yale University, Partners MGH and other multiple ex-U.S. hospitals, clinics and academic medicine centers.

Detection and quantification of subtle characteristic changes in the voice can accurately reveal important aspects of an individual’s changing health. Our platform is designed to rapidly measure a number of health-related features from just a few seconds of speech – without requiring analysis of the content of speech so user privacy can be protected. With today’s estimated four billion voice queries a day expected to double by 2020, we are advancing a leading platform that could potentially uncover important health content currently being overlooked from these voice interactions.

Applying cutting edge machine learning to voice and health