Combining ultrasound and cytology whole-slide images to predict cancer risk in indeterminate (Bethesda III) thyroid nodules

Multimodal Assessment of Malignancy in Atypia of Undetermined Significance in Thyroid Nodules Using Ultrasound and Cytology Whole-Slide Images

Observational Sun Yat-Sen Memorial Hospital of Sun Yat-Sen University · NCT07488325

This project will test whether a deep learning model that merges ultrasound pictures and cytology whole-slide images can better predict which Bethesda III thyroid nodules are cancerous.

Quick facts

Study typeObservational
Enrollment396 (estimated)
Ages18 Years to 80 Years
SexAll
SponsorSun Yat-Sen Memorial Hospital of Sun Yat-Sen University Academic / other
Locations1 site (Guangzhou, Guangzhou)
Trial IDNCT07488325 on ClinicalTrials.gov

What this trial studies

This single-center retrospective diagnostic accuracy project builds separate deep learning models using ultrasound images, cytology whole-slide images, and a fusion model that combines both modalities to predict malignancy in Bethesda III thyroid nodules. Consecutive patients with a Bethesda III cytology result who underwent fine-needle aspiration and subsequent thyroid surgery at Sun Yat-sen Memorial Hospital between January 2016 and December 2024 were included if preoperative ultrasound images were available. Postoperative histopathology or BRAFV600E mutation status served as the reference standard, and model performance was measured using receiver operating characteristic curve analysis. Decision curve analysis was used to estimate clinical utility and the study compares diagnostic accuracy of the ultrasound-only, cytology-only, and combined fusion models.

Who should consider this trial

Good fit: Patients with a Bethesda III (atypia of undetermined significance) cytology result who have at least one preoperative ultrasound image and a definitive postoperative histopathology result or BRAFV600E mutation status are the intended population.

Not a fit: People without available or usable preoperative ultrasound images, without cytology whole-slide images, or who never undergo surgery or mutation testing would not be eligible and are unlikely to benefit from this model.

Why it matters

Potential benefit: If successful, this approach could improve preoperative cancer risk estimates and help reduce unnecessary diagnostic surgery for people with indeterminate Bethesda III thyroid nodules.

How similar studies have performed: Prior artificial intelligence work has shown promise using ultrasound or cytology images separately for thyroid nodule classification, but combining ultrasound and whole-slide cytology images is relatively novel and not yet widely validated.

Eligibility criteria

Show full inclusion / exclusion criteria
Inclusion Criteria:

* (a) Patients who underwent fine-needle aspiration of thyroid nodules and received a cytological diagnosis of Bethesda Ⅲ.
* (b) Patients who subsequently underwent thyroid surgery with a definitive histopathological diagnosis of benign or malignant lesions,.
* (c) Availability of at least one ultrasound image acquired prior to surgery.

Exclusion Criteria:

* (a) Patients without confirmation by surgical pathology,.
* (b) Patients without any available or usable ultrasound images.

Where this trial is running

Guangzhou, Guangzhou

Study contacts

How to participate

  1. Review the eligibility criteria above with your treating physician.
  2. Visit the official trial page on ClinicalTrials.gov for the most current contact information and recruitment status.
  3. Contact the listed study coordinator or principal investigator to request pre-screening. Pre-screening is free and never obligates you to enroll.
Conditions Thyroid NoduleAtypia of Undetermined SignificanceMultimodal Deep LearningUltrasoundCytology Whole-Slide Image
Last reviewed 2026-06-10 by the Find a Trial editorial team. Information on this page is for educational purposes and is not medical advice. Always consult qualified healthcare professionals about clinical trial participation.