Skip to content

Feasibility Study on the Application of ACR-TIRADS Classification System for Thyroid Nodules Based on Multimodal Vision-Language Models

Feasibility Study on the Application of ACR-TIRADS Classification System for Thyroid Nodules Based on Multimodal Vision-Language Models

Status
Active, not recruiting
Phases
Unknown
Study type
Observational
Source
ChiCTR
Registry ID
ChiCTR2600115944
Enrollment
Unknown
Registered
2026-01-04
Start date
2026-02-01
Completion date
Unknown
Last updated
2026-01-05

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Thyroid cancer

Interventions

Gold Standard:Expert Standard: The ACR-TIRADS classification of thyroid nodules after evaluation by experts.
Index test:AI-Based ACR-TIRADS Classification

Sponsors

Shanghai General Hospital
Lead Sponsor

Eligibility

Sex/Gender
All
Age
18 Years to 80 Years

Inclusion criteria

Inclusion criteria: 1.Aged greater than or equal to 18 years old, with thyroid nodules detected for the first time via thyroid ultrasound examination; 2.The maximum diameter of the nodule > 5 mm, and the ultrasound images clearly display the core ultrasound features of ACR-TIRADS; 3.The patient has signed the informed consent form.

Exclusion criteria

Exclusion criteria: 1.Severe masking of nodule features by diffuse thyroid lesions; 2.(A history of previous thyroid surgery or radioactive iodine therapy; 3.Severe artifacts in the images (determined by 2 physicians as unevaluable).

Design outcomes

Primary

MeasureTime frame
Weighted kappa coefficient;Sensitivity and specificity;

Secondary

MeasureTime frame
Area under the curve;Consistency of diagnosis between models / Consistency of diagnosis between models and physicians;Accuracy;Positive predictive value, negative predictive value;

Countries

China

Contacts

Public ContactLiu Long

Shanghai General Hospital

13817309541@163.com+86 13817309541

Outcome results

None listed

Source: ChiCTR (via WHO ICTRP) · Data processed: Feb 4, 2026