First AI study of its kind now underway

December 8, 2021


The multi-year collaboration will evaluate AI-based measurement of cognitive behavioral therapy to improve care.

December 8, 2021 – Seattle, WA. Lyssn, a Seattle-based behavioral healthcare technology company, has kicked off a research partnership called AFFECT: AI-Based Fidelity Feedback to Enhance CBT Treatment. Funded by the National Institute of Mental Health (NIMH) through a Small Business Technology Transfer program (STTR), AFFECT will build off the existing Lyssn AI platform to develop an enhanced Cognitive Behavioral Therapy (CBT) tool.

“We’re excited to be able to offer the Lyssn technology to this study and a large publicly funded behavioral health system,” said David Atkins, Lyssn CEO and a principal investigator for the study. “The new AI-based tool will be made specifically for the needs of providers within the community and designed to support clinicians in offering high quality services.”

The scalable fidelity tool will estimate the effectiveness of CBT in recorded therapeutic sessions and support behavioral health providers in offering high-quality training, and improved outcomes, supervision, case management, and quality assurance.

“Mental health systems around the world have invested billions of dollars to increase access to evidence-based mental health practices, but our ability to measure the quality of that care has been very limited. For that reason, the NIMH identified a pressing public health need: to be able to measure at scale what is happening within therapy sessions,” said Dr. Torrey Creed, PhD, Senior Implementation Advisor for Lyssn.

“Before Lyssn’s AI platform, this was an aspirational goal because there just weren’t tools to evaluate the quality of psychotherapy at scale. But now with the development of Lyssn, we have the ability to quickly and accurately assess how skillfully these practices are being delivered in therapy sessions. Providing therapists with this feedback creates an opportunity for them to refine and sharpen their skills, ultimately providing the best possible care for clients,” Atkins said.

The four-year study will be conducted in two phases. Phase one, which is underway and expected to last a year, will work from the existing Lyssn platform to build an enhanced CBT measurement tool: LyssnCBT.

LyssnCBT will be designed, refined, and evaluated for implementation with input from user-centered design focus groups and by working in collaboration with the Philadelphia Department of Behavioral Health and Intellectual disAbility Services, community mental health therapists, supervisors, administrators, and others.

“As a part of refining and testing the Lyssn platform, we will also look at ways to ensure the AI is sensitive to diversity and cultural differences,” said Dr. Creed, a principal investigator for the research and director of the Penn Collaborative for CBT and Implementation Science, which is leading the clinical trial. “One of Philadelphia’s great strengths is our culturally diverse community. It’s so important that Lyssn’s AI algorithms continue to be built in a way that reflects and honors the diversity of providers and clients across the spectrum. Partnering with our public mental health system is a step toward that kind of inclusion.”

Once phase one is completed, the study will move on to phase two. During this stage, LyssnCBT will be evaluated in a randomized trial in five community mental health clinics within Philadelphia. Beginning with one clinic, and eventually rolling out to all five, the platform will be evaluated for its acceptability and feasibility for community providers, its effectiveness in improving clinicians CBT skills and clinical outcomes, and treatment retention.

“What we learn in this study could change the way quality in mental health is assessed,” said Dr. Creed. “Having a tool that provides automated, reliable and scalable CBT fidelity feedback could help us better provide high-quality training, effective supervision, and ensure quality care.”



University of Pennsylvania Financial Disclosure
Dr. Creed is a Penn faculty member and a paid consultant of Lyssn with an equity stake in the company.

About Lyssn
Lyssn offers a HIPAA, FERPA and GDPR-compliant AI platform to help behavioral health care providers and payors improve services and client outcomes. Secure recording, session assessment, fidelity monitoring and training of evidence-based clinical practices has never been easier. Lyssn was created by trained clinicians, academic researchers, and machine learning experts who believe in leveraging the power of science and technology to change behavioral health care for the better. For more information about Lyssn, visit And follow Lyssn on LinkedIn, Facebook, Instagram, and Twitter at