The Study Aims to Improve the Accuracy of Detecting Spina Bifida During Early Ultrasound Scans. to Achieve This, an AI Model Has Been Developed to Provide Feedback About the Presence of Spina Bifida. a RCT Has Been Designed to Compare the Effectiveness of AI Feedback with No AI Feedback.

NCT06566014 — Not applicable, Completed, 38 participants, Copenhagen Academy for Medical Education and Simulation
ClinicalTrials Daily  ·  Areas  ·  Conditions  ·  Drugs  ·  Sponsors  ·  

Evaluation of XAI-assisted Spina Bifida Diagnosis

RegistryNCT06566014
StatusCompleted
PhaseNot applicable
Study typeInterventional; randomized, parallel, masking: single, diagnostic
Enrollment38 (Actual)
SponsorCopenhagen Academy for Medical Education and Simulation (Other)
ConditionsSpina Bifida
InterventionsEvaluation of XAI-assisted spina bifida diagnosis (Other)
Datesstart 2024-07-01, primary completion 2024-10-30, completion 2024-10-30, first posted 2024-08-22, last updated 2024-12-04
Locations1 site in Denmark

Related Papers Related Patents Related Grants Related Experts
All trials for Spina Bifida All trials of Evaluation of XAI-assisted spina bifida diagnosis

-Summary

The study aims to improve the accuracy of detecting spina bifida during early ultrasound scans. To achieve this, an AI model has been developed to provide feedback about the presence of spina bifida. A RCT has been designed to compare the effectiveness of AI feedback with no AI feedback.

Expand all sections  ·  Collapse all

+Detailed description

The study aims to improve the accuracy of detecting spina bifida during early ultrasound scans. To achieve this, an AI model has been developed to provide feedback about the presence of spina bifida. A randomized controlled trial (RCT) has been designed to compare the effectiveness of AI feedback with no AI feedback. Forty fetal medicine specialists are randomly assigned to receive either the AI feedback or no feedback, and the two groups are compared to determine if the feedback improved the accuracy of detecting spina bifida.

+Arms and interventions

ArmTypeInterventionsDescription
AI feedbackExperimentalOther: Evaluation of XAI-assisted spina bifida diagnosisThe participant will receive AI feedback upon completing the task of analyzing 20 images. The AI feedback will include a prediction (Normal/Spina Bifida) along with a confidence score ranging from 0.0 to 1.0, where 0.0 indicates the lowest confidence and 1.0 indicates the highest confidence.
No AI feedbackPlacebo comparatorOther: Evaluation of XAI-assisted spina bifida diagnosisThe participants will complete the task of analyzing 20 images without any AI feedback.
  • Evaluation of XAI-assisted spina bifida diagnosis (Other): AI feedback

+Outcome measures

Primary outcomes

  • Evaluation of XAI-assisted spina bifida diagnosis [One month after the survey is distributed.]: The study aims to evaluate whether AI feedback improve the accuracy of diagnosing spina bifida by comparing the number of correct and incorrect responses in a task involving 20 images.

+Eligibility

Sex: All, accepts healthy volunteers

Inclusion Criteria:

* Obstetricians

Exclusion Criteria:

* Fetal medicine specialists

+Locations (1)

FacilityCityCountryStatus
Copenhagen University Hospital, RigshospitaletCopenhagenDenmark