Caption LungAI for Shortness of Breath

This study is testing a software called Caption LungAI. This software helps healthcare professionals who are not ultrasound experts to take high-quality lung ultrasound images. The study wants to see if Caption LungAI can help detect B-lines (a sign of fluid in the lungs) in people experiencing shortness of breath. You might be able to join if you are over 18, have shortness of breath, and your doctors suspect you have B-lines. The study is looking for 220 participants, but its current recruitment status is unclear.

Study design
This interventional study plans to enroll 220 participants. It compares lung ultrasounds performed by an expert with those done by a non-expert using Caption LungAI.
What's involved
You would undergo two lung ultrasound exams. One exam will be done by an expert, and the other by a healthcare provider using Caption LungAI.
Compensation
Not stated in the trial record.
Follow-up
Your diagnostic image quality and B-line detection will be measured for up to 24 weeks after completing the study.

AI-generated from the public study record. Only the study team can confirm whether you're eligible — confirm details with them before making decisions.

NCT05992324

A Validation Study to Evaluate the Performance of Caption Health Lung Guidance and Interpretation

Recruiting
NAAges 18+InterventionalDiagnostic
Caption Health, Inc.
~220 participants
Updated 2024-11-26 on ClinicalTrials.gov
What's tested:Caption LungAI

At a glance

Recruiting sites
1 of 4 listed sites are recruiting right now
RecruitingSuspended, closed, or not yet open
What they're measuring
Diagnostic Image Quality (Trained Healthcare Professional)
Measured over Up to 24 weeks from completion of the study.
+3 more outcomes measured
Shortness of Breath

NCT05992324

Where you'd take part

This study runs at 4 sites. They're the same protocol — you choose where, and that choice sets who your contact draft is addressed to.

  • Yale New Haven Hospital

    New Haven, Connecticutstudy coordinator listed

    Recruiting

  • Northwestern University

    Chicago, Illinoisno site contact published

    Completed

  • Rush University Medical Center

    Chicago, Illinoisno site contact published

    Completed

  • The Moses H. Cone Memorial Hospital

    Greensboro, North Carolinano site contact published

    Completed

Sites open and close at different times, so the status above is per site — it can differ from the study's overall status.

Opens a ready-to-send draft in your own email app — review before sending.

Want this trial checked against your situation?

Add a private profile and we'll compare every criterion below against your situation — and tell you which ones are met, uncertain, or excluding.

Check eligibility for this trial ~2 min · HIPAA-protected · delete anytime
Eligibility criteria

Inclusion

Patients over the age of 18
Patients presenting to the hospital or outpatient setting with shortness of breath and suspected B-lines.

Exclusion

Patients in extremis/in whom a research lung ultrasound would not normally be performed due to other priorities
  • Diagnostic Image Quality (Trained Healthcare Professional)Up to 24 weeks from completion of the study.

    Evaluate the product's ability to assist the user to capture a LUS image with diagnostic quality. Percentage of zones with diagnostic image quality \> 0.80

  • B-Lines DetectionUp to 24 weeks from completion of the study.

    Evaluate the product's ability to retrospectively detect present B-lines on images acquired by a local expert with no Caption Lung AI assistance. AUROC \> 0.80; Sensitivity (Se) \> 0.80; Specificity (Sp) \> 0.75

  • B-Line SignificanceUp to 24 weeks from completion of the study.

    Evaluate the product's ability to retrospectively detect B-lines of significant severity on images acquired by a local expert with no Caption Lung AI assistance. AUROC \> 0.80; Sensitivity (Se) \> 0.80; Specificity (Sp) \> 0.75

  • Remote Reader PerformanceUp to 24 weeks from completion of the study.

    Evaluate the product's impact on a remote expert reader's ability to interpret LUS images. Difference Between Aided \& Unaided Groups: AUROC \> 0.00; Sensitivity (Se) \> 0.00; Specificity (Sp) \> 0.00