Ellipsis Health made its way to the forefront of vocal biomarker technology through years of thorough research, testing, and development.
Although we only began commercializing our product in 2021, the team had been working tirelessly on research and validation — gathering data, refining algorithms, and strengthening models — since 2017. This meant we hit the market with an industry-leading amount of peer-reviewed studies and two approved patents to our name, as well as a clinically validated product based on:
A vast database of clinically labeled recordings from over 100,000 individuals representing different ages, genders, and socioeconomic statuses
An innovative blend of models to analyze both what someone says and how they say it
A screening and triaging product that payors, providers, and patients can easily integrate into their current methods
More than simply breaking barriers in speech technology, our commitment to science has meaningful impacts on individual patients, payors, and healthcare providers. We know that with every study, we can improve mental health screening for a greater number of people in a wider variety of ways.
In this article, we’re going to look at a series of use cases of our pioneering vocal biomarker technology through the lens of studies and partnerships that have contributed to our success. Covering landmark studies with leading institutions and integration into regional and national healthcare systems, our list of partners is a testament to the groundbreaking potential of our models and algorithms.
Ceras Health: Detecting depression in chronically ill patients
Ceras Health is a major player in finding innovative digital, AI, remote, and wearable health solutions, streamlining patient care for both patients and providers.
Ellipsis Health entered into a partnership with Ceras Health to focus specifically on the mental state of patients with chronic diseases such as hypertension, hyperlipidemia, diabetes, and rheumatoid arthritis.
Individuals with chronic physical illness may be susceptible to masking depression and anxiety due to stigma and other factors. In the USA, 27% of patients with chronic diseases also suffer from mental disorders[1] which, if not treated, increase the risk of premature death, frequent hospitalizations, and reduced quality of life[2].
By applying Ellipsis Health technology onto Ceras care management calls, we screened a total of 465 individuals and ultimately identified signs of depression in 110 (24%). With vocal biomarker technology in place, care managers can identify depression earlier without the need to deliver a cumbersome, lengthy mental health survey.
Through early identification, institutions can make significant cost savings through reduced readmissions and lengths of stay. At the same time, patients can expect more effective, targeted treatment, as well as a potential increase in life expectancy.[3]
In this partnership, Ellipsis Health offers Ceras Health a comprehensive dashboard that provides case managers and the care team with information critical to their jobs and that can improve the care of their patients.
Getting ahead of the game in perinatal depression
In another collaboration with a health system that began in 2024, we are screening for depression and anxiety that are common during pregnancy. In California alone, 100,000 or one in five mothers suffer perinatal depression, costing a massive $2.25 billion per year, according to the California Task Force on the Status of Maternal Mental Health Care.
We offer participants two ways of collecting mental health data:
In-call analysis during care calls and virtual encounters
Screening and monitoring during live interactions give care managers extensive data without adding time-consuming steps to their regular workflows. Standard calls are streamed, giving the care manager a real-time display of their patient’s mental health score. This data is transmitted to existing systems where clinical teams can track scores over time and set up alerts to catch high levels of depression and anxiety as they emerge.
Digital front doors or stand-alone apps
Using a custom-built speech journal, expectant mothers express themselves verbally throughout their pregnancy and after birth. Questions prompt users to speak for a short period about how they are feeling, before displaying a “mood word”, with a short explanation of their emotional state. This is followed by options to explore what is behind their mood and suggestions to improve it. Their providers simultaneously receive a clinical grade score of the severity of their depression and anxiety symptoms to aid in their further care.
More than a simple triaging tool, apps and digital front doors powered by Ellipsis Health work as a continuous mood journal during and after pregnancy. Health system partners receive an automatic flow of mental health data points as the tool invites users to send recordings to their healthcare team so they can reach out when they need help.
In a study of almost 2,000 pregnant women, three in five with postnatal depression present at nine to ten months did not report depression at two to six months. This statistic shows the importance of implementing screening at all points in a pregnancy journey - both before birth and after - to promote early intervention, support the mother and family, and reduce costly events like readmissions and emergency department visits.
Desert Oasis Healthcare: Engaging a senior population in app-based mental health monitoring
Desert Oasis Healthcare (DOHC), a California healthcare provider, collaborated with Ellipsis Health to study a cohort of 250 mainly senior individuals with a previous history of depression, as well as a control group without depression.
This study aimed to find ways of creating a screening app that goes beyond being user-friendly to make patients actively want to participate. Individuals with depression are generally less likely to engage with monitoring and treatment plans, making user engagement one of Ellipsis Health’s top priorities.
We asked participants to make a short voice recording every week for six weeks, answering open-ended questions. The resulting study[4], published in Frontiers in Psychology, showed general acceptance of the app, with many participants speaking for longer than the required time.
To build on the progress made through the DOHC study, we are continuing to find new ways to boost engagement with our platforms through elements such as automated notifications, seamless onboarding processes, and shorter 1-minute voice samples.
University of Michigan: Uncovering psychological distress in adolescents and young adults diagnosed with cancer
Toward the other end of the age range, we are collaborating with the University of Michigan on a first-of-its-kind study[5] to better engage adolescent and young adult (AYA) patients diagnosed with cancer in their mental health monitoring.
Despite the importance of tracking psychological distress during an AYA’s cancer journey, many centers do not have sufficient resources to tackle mental health concerns. This is exacerbated by pen-and-paper screening and monitoring techniques based on survey responses that are often unreliable. In addition, AYAs are the most likely demographic to disengage from traditional methods[6], making the delivery method just as important as the accuracy of our vocal biomarker models. The study with the University of Michigan therefore analyzes practical elements such as the user interface and the voice collection feasibility.
The AYA population, largely digital natives, are especially receptive to an app-based tool for psychological monitoring, which provides compelling insights into applications beyond this pilot study. Creating the right circumstances for AYAs to participate in their own mental health monitoring makes patients feel more engaged in the process and helps providers maintain high standards of monitoring without significant resource reallocation.
Augmedix: Making clinician-patient relationships more human and less administrative
Augmedix’s [1] mission is to use industry-leading technology to enable healthcare systems, physician practices, and hospitals to save time, increase productivity, and improve patient satisfaction.
In a strategic partnership that began in 2023, Augmedix is incorporating Ellipsis Health’s mental health screenings during clinical encounters to accelerate the identification and triaging of issues such as depression and anxiety. With an average 11-year delay between initial symptoms of mental illness and treatment, early identification can help healthcare providers manage resources and close the gap.
This collaboration draws on our work in developing application programming interfaces (APIs) that allow partners like Augmedix to seamlessly integrate our vocal biomarker technology into their existing systems. In addition to ambient medical documentation, Augmedix providers receive their patient’s anxiety and depression scores, equipping physicians with a more complete picture of the individual in front of them.
The result is more informed medical decisions that enable patients to receive referrals or interventions earlier than they would without the efficiencies that Ellipsis Health brings. It also marks a shift away from PHQ and GAD surveys which are time-consuming and require verbatim, skilled delivery. But far from replacing mental health professionals, our founder and CEO, Mainul Mondal, describes our partnership with Augmedix as “like having a mental health copilot supporting care during every visit.”
Scaling up into the future
Despite the giant leaps we’ve already taken in the vocal biomarker space, we believe we’re only beginning to tap into the potential of this technology.
With an ethos founded in a commitment to science, we plan to continue breaking new ground in vocal biomarker technology, changing the game for payors and providers, and achieving better outcomes for patients.
If you are interested in joining us on this mission, reach out to our team to discuss a potential partnership.
[1] Daré, L. O., Bruand, P. E., Gérard, D., Marin, B., Lameyre, V., Boumédiène, F., & Preux, P. M. (2019). Co-morbidities of mental disorders and chronic physical diseases in developing and emerging countries: a meta-analysis. BMC public health, 19(1), 304. https://doi.org/10.1186/s12889-019-6623-6
[2] National Alliance on Mental Illness California. (n.d.). What is mental illness? Retrieved from https://namica.org/what-is-mental-illness/#:~:text=Without%20treatment%2C%20the%20consequences%20of,and%20poor%20qualit y%20of%20life.
[3] Voinov, B., Richie, W.D., & Bailey, R.K. (2013). Depression and chronic diseases: it is time for a synergistic mental health and primary
care approach. The primary care companion for CNS disorders, 15(2), PCC.12r01468. https://doi.org/10.4088/PCC.12r01468
[4] Lin, D., Nazreen, T., Rutowski, T., Lu, Y., Harati, A., Shriberg, E., Chlebek, P. and Aratow, M. (2022) Feasibility of a Machine Learning-Based Smartphone Application in Detecting Depression and Anxiety in a Generally Senior Population. Front. Psychol. 13:811517. doi: 10.3389/fpsyg.2022.811517
[5] Zhang, A., Kamat, A., Acquati, C., Aratow, M., Kim, J. S., DuVall, A. S., & Walling, E. (2022). Evaluating the Feasibility and Acceptability of an Artificial-Intelligence-Enabled and Speech-Based Distress Screening Mobile App for Adolescents and Young Adults Diagnosed with Cancer: A Study Protocol. Cancers, 14(4). https://doi.org/10.3390/cancers14040914
[6] Sansom-Daly, U.M.; Wakefield, C.E. Distress and Adjustment among Adolescents and Young Adults with Cancer: An Empirical
and Conceptual Review. Transl. Pediatrics 2013, 2, 167. [CrossRef]