Using brain scans and machine learning to track emotional ups and downs in bipolar disorder

Decoding Emotional Dynamics Driving Mood Instability in Bipolar Disorder

This project will test whether fMRI plus machine learning can read moment-by-moment emotions and whether boosting positive emotions can make mood and brain activity more stable in adults with bipolar I or II (currently depressed or mixed) compared with healthy volunteers.

Quick facts

• Phase: Not applicable
• Study type: Interventional
• Enrollment: 72 (estimated)
• Ages: 18 Years to 65 Years
• Sex: All
• Sponsor: Laureate Institute for Brain Research, Inc. (Academic / other)
• Locations: 1 site (Tulsa, Oklahoma)
• Trial ID: NCT07221864 on ClinicalTrials.gov

What this trial studies

This is a two-visit neuroimaging protocol at the Laureate Institute for Brain Research enrolling 36 adults with bipolar I or II in a depressed or mixed state and 36 healthy control participants. Participants complete questionnaires and clinical interviews, then undergo MRI including resting-state and task-based fMRI, diffusion imaging, structural scans, and physiological monitoring while performing an affect regulation task. Investigators will apply machine-learning methods to decode emotional states from brain activity, compare dynamic metrics (e.g., metastability, fractal scaling) between groups, and test whether experimentally amplifying positive emotion shifts neural and emotional dynamics toward greater stability. Behavioral and self-report emotional measures will be aligned with neural signals to link subjective experience to brain-state trajectories.

Who should consider this trial

Good fit: Ideal participants are adults aged 18–65 who either have bipolar I or II and are currently in a depressive or mixed state with moderate symptoms, or are healthy volunteers with no psychiatric history, are English-fluent, and can undergo MRI.

Not a fit: People who are manic, have severe unstable medical conditions, are unable to undergo MRI (e.g., certain metal implants, severe claustrophobia), or cannot comply with study procedures are unlikely to benefit from participation.

Why it matters

Potential benefit: If successful, the work could identify brain-based markers of mood instability and suggest emotion-focused approaches to reduce mood swings in bipolar disorder.

How similar studies have performed: Previous fMRI and machine-learning work has shown that emotional states can be decoded in healthy people and some clinical groups, but applying these dynamic metrics specifically to mood instability in bipolar disorder and testing positive-emotion amplification as a stabilizing intervention is relatively novel.

Eligibility criteria

Inclusion Criteria

1. Age 18 to 65 years
2. Male or female
3. BMI between 18.5 and 38.0 kg/m2 at Screening
4. Capable of understanding and complying with study requirements
5. Fluent in English
6. Able to provide informed consent

   BD Group:
7. Meet the DSM-5 diagnostic criteria for BD-I or BD-II who are currently depressed or mixed state defined by the Mini-International Neuropsychiatric Interview (MINI)
8. Moderate or greater depressive symptom severity (MADRS ≥ 15 or PHQ-9 ≥ 10)

   HC Group:
9. No current or past psychiatric disorder (verified by MINI)

Exclusion Criteria

1. No telephone or easy access to a telephone
2. Significant medical problems as identified by the medical screening questionnaire: e.g. a history of unstable liver or renal insufficiency; glaucoma; significant and unstable cardiac, vascular, pulmonary, gastrointestinal, endocrine, neurologic, hematologic, rheumatologic, or metabolic disturbance; or any other condition that, in the opinion of the investigator, would make participation not be in the best interest (e.g., compromise the well-being) of the participant or that could prevent, limit, or confound the protocol-specified assessments
3. A positive test for drugs of abuse, including alcohol (breath test), cocaine, opiates, amphetamines, methamphetamines, phencyclidine, benzodiazepines, barbiturates, methadone, and oxycodone
4. Drug or alcohol intoxication (based on positive UTOX or breathalyzer test at screening or study session) or reported alcohol/drug withdrawal, last cannabis use must be \>48 hours prior to study session.
5. Current DSM-5 diagnosis of a psychosis spectrum disorder or moderate to severe substance use disorder
6. Moderate to severe traumatic brain injury or other neurocognitive disorder with evidence of neurological deficits, neurological disorders, or severe or unstable medical conditions that might be compromised by participation in the study (to be determined by primary care provider)
7. Current significant suicidal ideation or suicide attempt within the past 3 months.
8. Change in the dose or prescription of a medication within the 6 weeks before enrolling in the study that could affect brain functioning, e.g., anxiolytics, antipsychotics, antidepressants, or mood stabilizers
9. Taking drugs that affect the fMRI hemodynamic response (e.g., methylphenidate, acetazolamide, excessive caffeine intake \> 1000 mg/day)
10. MRI contraindications as documented on the MR Environment Screening
11. Unwillingness or inability to complete any of the major aspects of the study protocol, including magnetic resonance imaging (i.e., due to claustrophobia), or behavioral assessment. However, failing to complete some individual aspects of these assessment sessions will be acceptable (i.e., being unwilling to answer individual items on some questionnaires or being unwilling to complete a behavioral task)
12. Non-correctable vision or hearing problems

Where this trial is running

Tulsa, Oklahoma

• Laureate Institute for Brain Research — Tulsa, Oklahoma, United States (Recruiting)

Study contacts

• Study coordinator: Masaya Misaki Study Primary Investigator, Ph.D.

Conditions Bipolar Disorder I or II; Healthy; neuroimaging; fMRI; Bipolar Disorder I; Bipolar Disorder II; machine learning; emotion regulation