AI-COA®: using AI to measure depression and anxiety severity from video interviews

Symptom Evaluation and Quantification for Unified Outcomes and Individualized Assessment

This will test whether an AI tool can reliably measure depression and anxiety severity from video interviews in adults who have just started a new treatment.

Quick facts

• Phase: NA
• Study type: Interventional
• Enrollment: 120 (estimated)
• Ages: 18 Years to 65 Years
• Sex: All
• Sponsor: Deliberate Solutions Inc. (industry)
• Locations: 1 site (Houston, Texas)
• Trial ID: NCT06923995 on ClinicalTrials.gov

What this trial studies

This interventional project uses an AI-based Clinical Outcome Assessment (AI-COA®) that analyzes recorded video interviews to generate symptom severity scores for depression and anxiety. Adults aged 18–65 with a baseline HAM-D17 > 10 who have begun a new treatment within the prior 2–3 weeks and who can use a webcam and microphone are enrolled and complete remote interviews and online surveys. The AI outputs are compared with standard clinician-based ratings and symptom measures to determine consistency and objectivity across participants. The AI-COA® approach has been accepted into the FDA's Innovative Science and Technology Approaches for New Drugs (ISTAND) program as a novel measurement method.

Who should consider this trial

Good fit: Ideal candidates are English-speaking adults 18–65 living in the United States with HAM-D17 > 10 who have started a new depression or anxiety treatment within 2–3 weeks and have a computer with a working webcam, microphone, and stable internet.

Not a fit: Patients who cannot use a webcam/microphone, have cognitive or sensory impairments that prevent meaningful video interviews, are non-English speakers, or are not U.S. residents are unlikely to benefit from this enrollment.

Why it matters

Potential benefit: If successful, this could provide more consistent, objective, and scalable symptom measurements to improve monitoring and comparability across depression and anxiety studies and care.

How similar studies have performed: Early research on AI analysis of speech, facial expression, and behavior for mood disorders shows promising signals, but large-scale clinical validation of these tools is still limited.

Eligibility criteria

Inclusion Criteria:

* English fluency
* 18 to 65 years of age.
* HAM-D 17 \> 10
* Starting, or has started, a new treatment for depression or anxiety within 2-3 weeks of enrollment
* Access to a laptop or other computer with a well functioning microphone and webcam, and a stable Internet connection
* Stated willingness to comply with all study procedures and availability for the duration of the study
* Resides in the United States at the time of consent and during completion of study

Exclusion Criteria:

* Any cognitive impairment that limits ability to provide informed consent or authorization
* Vulnerable or protected populations (e.g. prisoners)
* Impairment that would prevent participants from completing an online survey and/or engaging in clinician assessment interviews (e.g., visual impairment, motor impairment, hearing impairment)
* Acute intoxication at the time of the assessments
* Concurrent medication/treatment:

  * Receiving any fast-acting treatment for depression or anxiety (e.g. ketamine, psychedelics, deep brain stimulation, etc.) in between initial baseline assessment (A1) and Restest (A2) , or in between Follow up assessment (B1) and Retest (B2)
  * Anxiolytics: Use of benzodiazepines or other anxiety-reducing medications that could affect speech or motor activity, within the past 4 weeks.
  * Antipsychotics and Mood Stabilizers: Medications that can alter cognitive and motor functions, within the past 6 weeks.
  * Stimulants: Use of medications like methylphenidate or amphetamines that affect energy levels and behavior, within the past 2 weeks
  * Epilepsy medication: seizure activity or medication side effects that may alter behavior, within the past 4 weeks.
* Any history or evidence of any of the following conditions:

  * Neurodevelopmental, Neurocognitive, Neurodegenerative or movement disorders including, but not limited to:
  * Tourette's syndrome
  * Multiple Sclerosis
  * Amyotrophic Lateral Sclerosis
  * Parkinson's Disease
  * Stroke
  * Traumatic Brain Injury
  * Facial paralysis.
* Conditions with vocal cord impact:

  * Vocal cord injury or cerebrovascular accident or head trauma with residual dysarthria in the past year
  * Disorders that may impact vocal cords such as acute or chronic laryngitis, vocal cord paresis or paralysis, or spasmodic dysphonia
  * Past or active heavy smokers (an average of \>20 cigarettes per day)
* Schizophrenia Spectrum and Other Psychotic Disorders:

  * Individuals with a current or past diagnosis of Schizophrenia, Schizoaffective Disorder, or other psychotic disorders, including Delusional Disorder and Brief Psychotic Disorder.
  * Current hypomanic episode as defined by DSM-5 criteria. This includes those with Bipolar I Disorder, Bipolar II Disorder, or Cyclothymic Disorder who are not in a stable mood state at the time of assessment.
  * Chronic Pain Conditions: Such as fibromyalgia, which may affect facial expressions and vocal tone.
  * Prosthetic Facial Devices: Could affect facial recognition algorithms. Cosmetic Procedures: Such as Botox injections or facelifts that can impact facial expressions.
  * Dental Work: Major procedures that might affect speech.
* Participants who have previously participated in another research project by Deliberate.

Where this trial is running

Houston, Texas

• Baylor College of Medicine — Houston, Texas, United States (RECRUITING)

Study contacts

• Principal investigator: Marc Aafjes — Deliberate Solutions Inc.
• Study coordinator: Eric Storch, PhD

Conditions: Depression Disorders, Anxiety Disorders