Plain-English translation of NCT07500428 on ClinicalTrials.gov ↗ · Source last updated · Translation generated · How we translate trials
Read our Breast Neoplasms research guide →Researchers are creating a large database of breast ultrasound images to train and test artificial intelligence (AI) systems in reading these scans. The goal is to build a standardized way to measure how well AI performs at interpreting breast ultrasounds, especially in difficult cases where doctors disagree. This research focuses on improving AI's ability to help radiologists detect breast cancer more reliably.
Breast ultrasounds are highly dependent on the skill and experience of the radiologist reading them—two doctors can look at the same image and reach different conclusions. This variation can lead to missed diagnoses or unnecessary worry. By creating a standardized benchmark and testing AI systems, researchers hope to reduce this variability and give all patients more consistent, accurate care.
You likely qualify if…
You likely don't qualify if…
This study uses images that have already been collected—you would not need to have new ultrasounds performed. Researchers will review your existing breast ultrasound images from hospital records or public datasets, after all your personal information has been removed. Expert radiologists will carefully examine and label these images, and then AI systems will be trained and tested using them to see how accurately they can interpret breast ultrasounds.
AI-generated summary from trial data · Jul 9, 2026 · Not medical advice
China