Applying Computer Science to Anxiety, Depression, and Veteran Brain Health

Voice Vision Vitals for Depression, Anxiety, and Mood Disorders

Anxiety disorders and major depressive disorder are often overlooked, underdiagnosed, and untreated, even though these disorders are the primary cause of disability under the age of 40. About 18% and 7% of the population suffer from anxiety and major depressive disorders respectively, yet only 37% of people with anxiety disorder and 50% of people with major depressive disorder receive treatment. Untreated these can result in the inability to hold a job, financial problems, homelessness, divorce, broken relationships, inability to care for children, and even suicide.

Barriers to diagnosis and treatment include low self-perception of need, attitudinal barriers such as stigma or lack of trust in the healthcare system, structural barriers such as cost and low availability of doctors, and subtlety of early symptoms which are often confused with other disorders. These treatment barriers and lack of timely diagnoses are common to a spectrum of diseases, particularly in psychiatry, neurology, and cardiology.

New methods of disease screening that address these barriers to treatment are needed. This work is currently addressing the problem by developing automated screening techniques for anxiety disorder and major depressive disorder based on human speech and language. Multiple projects in this space (Voice Vitals: A New Approach for Anxiety and Depression Screening in the Era of COVID-19, Jump ARCHES 2021; Developing a Living Veteran Brain Bank, CSBS 2021; Voice Vitals, NCSA Faculty Fellows 2021, Amazon AWS Infrastructure 2021) are exploring curation of research datasets, development of machine models and tools that can predict presence and severity of depression and anxiety, development of secure computing infrastructure to support the work.

Veteran Brain Health

 Many health challenges that military Veterans face are physically “invisible” injuries sustained to the brain and psyche. About 20% of post-9/11 and 30% of Vietnam veterans suffer from major depression or post-traumatic stress disorder. Roughly 430,720 service members have been diagnosed with traumatic brain injury since 2000, 40% of whom are expected to develop post-traumatic epilepsy. Traumatic brain injury, an invisible injury often conflated with PTSD due to similar symptoms, is termed the “signature wound” for post-9/11 veterans.

The prevalence of such invisible injuries among service members and veterans, along with the difficulty diagnosing and treating such injuries, is problematic. Veterans with multiple traumatic brain injuries are twice as likely to consider suicide, and suicide remains the second leading cause of death for veterans. Understanding how trauma affects the brain, both physically and psychologically, and screening for these invisible injuries, is essential to advancing interventions and clinical treatments for veterans, service members, and their families.

ARI and the Center for Social & Behavioral Science are addressing the problem by building “A Living Veteran (ALiVe) Brain Bank” to study the impact of invisible injuries on Veterans and find ways to help. The current effort seeks to 1) curate content for ALiVe, focusing on interview data and standard mental and physical health screening data, and 2) conducting a pilot analysis of the data that will lead to the development of interactive “brain health mirror” tools that reveal invisible injuries. More specifically, this initial effort focuses on the exploration of relationships among speech, language, relevant survey data, and health conditions resulting from invisible injuries common to Veterans.

If you are interested in participating in either of these studies, please visit the Research Participation Opportunities page on the ARI website.

To learn more about current and recent projects at ARI that are focused on using algorithms and software to advance health, click here.