EMbot’ – a smarter Mental Health chatbot for defense users

Requirement

There is a requirement to increase the number of serving and veteran personnel who engage with mental health support services through a readily deployable, persistently engaging, and emotionally smart application that leverages the benefits of Artificial Intelligence (AI). There is an additional requirement to provide the same capability to the clinician to aid triage, patient management, and as a new diagnostic tool.

Background:

Suicide rates increased by approximately 33% between 1999 and 2019 in the US, responsible for 47,500 deaths in 2019. This issue is more acute in the military where PTSD impacts significantly on service personnel and their families. Approximately 77% more ‘War on Terror’ veterans died from suicide than during deployment. Suicide and poor mental-wellbeing is a major concern to the Department of Defense (DOD), presenting significant challenges in both the medical and operational domains.

Image of serving US army personnel, a community challenged by adverse mental health effects exacerbated by often extreme operational environments

Despite an increase of awareness and some resource to combat suicide, the majority (~60%) of those who die from suicide have not engaged with mental health support services. Without ‘admission’ onto the patient’s pathway there is no subsequent diagnosis and treatment. This is a major healthcare issue where preventable negative mental health outcomes such as absence, addiction, self-harm, and ultimately suicide, are impacting operational readiness, recruitment & retention, and successful delivery of duty-of-care by the DOD.

Three key challenges prevent improved mental health outcomes:

  • Those who would benefit from professional mental health support are not aware that they should seek it.
  • Those who are aware they should seek help are unwilling to disclose mental health issues fearing loss of autonomy, over reaction, and stigma.
  • A lack of mental health consultants to provide effective care to a growing number of patients.

Solution:

EMbot’ is an AI-driven mental health monitoring and management solution operating as a stand-alone system or integrating into existing healthcare support systems. It helps more individuals and groups engage, reducing the personal, societal, organisational, and financial negative outcomes of harmful behaviour and suicide. DOD’s most important assets are people and data. Use of EMbot improves the virtuous cycle bettering use of both asset types to increase availability, performance, and care of serving and veteran personnel. It delivers improved situational awareness to commanders and support organisations, ensuring better understanding of operational readiness and opportunity for timely and effective interventions.

An easily accessible self-assessment/self-help chatbot app is linked to a clinical management app, reducing friction associated with adoption of new approaches by simplifying access to resources. Automated appointment, triaging, and diagnostic insight tools enabling clinicians to support more individuals more effectively.

Image of the EMbot UI providing the chatbot functionality to make human interaction with the AI assessment tool engaging and low-friction

Technology:

EMbot is a cloud-based downloadable app that collects conversational text using a smart chatbot, processes it using a Hybrid-AI Text Analytics (HTA) engine with ‘insight modules’ identifying behavioural indicators. It integrates ontological lenses tuneable to develop understanding of conditions, states, and triggers against specific scenarios – eg: alcohol/substance abuse, relationship breakdown, threat-to-life.

EMbot reports to the anonymous user and tracks their mental status based on patterns associated with severe impact. It provides them access to support resources, and should the user agree, can be monitored by clinical staff via a clinician’s patient management app, with automated appointment booking, triage, and insight tools to support diagnosis.

 

 

EMbot exceeds existing apps by understanding emotions, psychometrics and psycholingustics, improving long-term end-user adoption via a smarter chatbot engine.

 

The technology innovation is underpinned by cognitive models mapped into the solution. Easily adoptable, reducing recognised barrier to use for many self-help solutions, whilst enabling novel automation of specialist care.

The Akumen HTA engine is proven to extract intelligence from user conversations using established indicator libraries, across industries including Healthcare (NHS), Products/Services (Accenture), and PsyOps (UK Police).

Image of the EMbot clinician interface providing automated appointment, triaging, and diagnostic insight tools – designed to be easily integrated into existing healthcare systems

Development:

The EMbot development team is made up of 3 expert organisations:

Indigo Arc (US) develops, deploys, and supports integrated telemedicine and healthcare management software.

Akumen (UK) is an AI world-leader in forensic translation of human conversations into insights/intelligence.

BAR Associates (US/UK) supports product development, end-user networking, UK/US market entry, and commercial exploitation.

 

EMbot exists as a proof of concept system at TRL 4.

Further funding and end-user support will accelerate delivery and deployment to within 18 months.

For further information please contact:

Richard Thurston (US) and Andy Crouch (UK).