AI-Assisted Ultrasound Review Before Thyroid Surgery

Clinical Application of AI-assisted Ultrasound Technology in the Preoperative Evaluation of Thyroid Cancer

This prospective, single-center, single-arm study evaluates a locked artificial intelligence (AI) system as a surgeon-led second-read tool before thyroid surgery. Eligible participants scheduled for thyroid surgery undergo standard ultrasonography followed by a standardized AI-assisted repeat examination. The AI system evaluates thyroid nodules only; cervical lymph nodes are assessed by clinicians. The study assesses participant-level supplementary pathological examinations and treatment-decision changes and evaluates nodule-level diagnostic performance against final surgical histopathology.

Study Overview

• Status: Completed
• Conditions: Thyroid Cancer; Thyroid Nodule
• Intervention / Treatment: Diagnostic test: Locked AI-Assisted Thyroid Ultrasound Repeat Examination and Review

Detailed Description

This prospective, single-center, single-arm study evaluates a locked AI-assisted ultrasound system used during surgeon-led preoperative review in participants scheduled for thyroid surgery. Each participant undergoes standard preoperative ultrasonography followed by a standardized AI-assisted repeat examination. The locked system provides thyroid-nodule malignancy scores and classifications; it was not retrained or recalibrated during the study and did not assess cervical lymph nodes. Clinicians integrate standard ultrasound, cytology, clinical findings, and other information and retain responsibility for all supplementary examinations and final management decisions. The two primary outcomes are participant-level: (1) whether a participant undergoes an additional cytologic or pathologic examination and (2) whether a participant has at least one change in planned surgical management. Participants are counted once within each outcome, while event counts may be summarized separately. The embedded diagnostic analysis is nodule-level and uses unambiguously pathology-matched surgical histopathology as the reference standard. Only nodules with a definitive benign or malignant surgical histopathological diagnosis are included in binary diagnostic-performance calculations. Low-risk or borderline thyroid neoplasms, including follicular tumors of uncertain malignant potential (FT-UMP), well-differentiated tumors of uncertain malignant potential (WDT-UMP), and non-invasive follicular thyroid neoplasms with papillary-like nuclear features (NIFTP), and nodules without unambiguous lesion-level linkage are excluded from binary reference-standard analyses. These post-enrollment analysis exclusions do not alter the 515-participant actual enrollment or the participant-level workflow cohort. A participant may contribute another eligible nodule if that lesion has an unambiguous definitive diagnosis. Cervical lymph-node pathology is described only for nodes actually removed or sampled; no reference diagnosis is assigned to unsampled nodes.

• Study Type: Interventional
• Enrollment (Actual): 515
• Phase: Not Applicable

Contacts and Locations

Study Locations

• China
  • Fujian
    • Fuzhou, Fujian, China, 350001
      • Fujian Medical University Union Hospital

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion Criteria:

• Patients with preoperative pathological confirmation of thyroid malignant tumors undergoing surgical treatment.
• Patients with benign thyroid tumors, such as thyroid adenomas causing compressive symptoms, undergoing surgical treatment.
• Patients with complete and high-quality traditional two-dimensional color ultrasound images.
• Complete postoperative pathology reports.
• Willingness to participate in this clinical trial and signing of informed consent.

Exclusion Criteria:

• Patients with a history of neck surgery or radiotherapy.
• Patients with a history of malignant tumors in other parts of the body.
• Patients with thyroid dysfunction.
• Incomplete or poor-quality traditional two-dimensional color ultrasound images.
• Incomplete postoperative pathology reports.
• Refusal to participate in this clinical trial.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Diagnostic
• Allocation: N/A
• Interventional Model: Single Group Assignment
• Masking: None (Open Label)

Number of Arms

1

Arms and Interventions

Participant Group / Arm

Intervention / Treatment

Experimental: AI-Assisted Preoperative Review; Participants undergo standard preoperative thyroid ultrasonography followed by a standardized AI-assisted repeat examination and surgeon-led review incorporating the locked AI system's thyroid-nodule output. The AI system does not assess cervical lymph nodes. All supplementary examinations and final diagnostic and surgical decisions are made by clinicians.

Diagnostic test: Locked AI-Assisted Thyroid Ultrasound Repeat Examination and Review; After standard preoperative ultrasonography, the same participant underwent a standardized AI-assisted repeat examination. The locked system identified and classified thyroid nodules and provided malignancy-score information for surgeon-led review. It did not assess cervical lymph nodes and was not retrained or recalibrated during the study. Additional fine-needle aspiration, intraoperative pathologic examination, and surgical management were determined by clinicians using the complete clinical assessment.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Proportion of Participants Undergoing Supplementary Cytologic or Pathologic Examination	Number and proportion of enrolled participants who underwent at least one additional cytologic or pathologic examination, such as fine-needle aspiration or intraoperative frozen-section examination, after completion of conventional ultrasonography and AI-assisted review and before completion of surgery. Each participant is counted once regardless of the number of examinations. The decision remained clinician led; cervical lymph-node examinations were outside the AI system's task.	From completion of the AI-assisted repeat examination through completion of surgery
Proportion of Participants With a Change in Planned Surgical Management	Number and proportion of enrolled participants with at least one documented change in the surgeon-led planned surgical approach or extent after conventional ultrasonography, AI-assisted review, and any supplementary pathological examination, when performed. Changes may involve thyroid resection or cervical lymph-node management. Each participant is counted once regardless of the number of changes. The AI system provided thyroid-nodule information only and did not assess cervical lymph nodes.	From completion of the AI-assisted review through surgery

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Distribution of C-TIRADS Categories Among Evaluated Thyroid Nodules	Number and percentage of evaluated thyroid nodules in each clinician-assigned C-TIRADS category during conventional ultrasonography. Nodule size, composition, echogenicity, margins, calcifications, and clinician-assessed cervical lymph-node findings were recorded as descriptive covariates and were not treated as separate outcome measures.	During standard preoperative ultrasonography, before the AI-assisted repeat examination
Proportion of Evaluated Thyroid Nodules Classified as AI-Positive	Number and percentage of evaluated thyroid nodules with a locked AI malignancy score >0.5. The AI system evaluated thyroid nodules only and did not assess cervical lymph nodes.	During the preoperative AI-assisted repeat examination, before surgery
Results of Supplementary Cytologic or Pathologic Examinations	Among participants who underwent at least one additional cytologic or pathologic examination, record the cytologic or histopathologic result for each sampled thyroid nodule or cervical lymph node. Results are summarized at the examination level as nonmalignant, indeterminate, malignant, or metastatic, as applicable.	From supplementary sampling to availability of the corresponding result, before or during surgery
Final Histopathological Diagnosis of Resected Thyroid Nodules	For each resected thyroid nodule linked unambiguously to the evaluated lesion, record the final diagnosis as benign, malignant, follicular tumor of uncertain malignant potential (FT-UMP), well-differentiated tumor of uncertain malignant potential (WDT-UMP), non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP), or another borderline diagnosis.	From surgery until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Nodule-Level Sensitivity of AI and C-TIRADS	Using definitive postoperative histopathology as the reference standard, sensitivity is TP/(TP + FN), expressed as a percentage, for the locked AI classification (positive if score >0.5) and conventional C-TIRADS (positive if category 4a or higher). Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage. These lesion-level exclusions do not alter the 515-participant workflow cohort.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Nodule-Level False-Negative Rate of AI and C-TIRADS	For each method, the false-negative rate is FN/(TP + FN), expressed as a percentage, among definitively malignant thyroid nodules. AI positivity is defined as score >0.5; conventional ultrasound positivity is C-TIRADS category 4a or higher. Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Nodule-Level Positive Predictive Value of AI and C-TIRADS	For each method, positive predictive value is TP/(TP + FP), expressed as a percentage, among thyroid nodules classified as positive. AI positivity is defined as score >0.5; conventional ultrasound positivity is C-TIRADS category 4a or higher. Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Nodule-Level Youden Index of AI and C-TIRADS	For each method, the Youden index is sensitivity + specificity - 1 using fixed binary thresholds. AI positivity is defined as score >0.5; conventional ultrasound positivity is C-TIRADS category 4a or higher. Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Distribution of Cervical Lymph-Node Metastasis Status Among Pathologically Examined Nodes	Among cervical lymph nodes actually removed or sampled and pathologically examined, record the number and percentage with and without metastasis. No pathologic reference diagnosis is assigned to unsampled lymph nodes.	From surgery or supplementary sampling until the corresponding pathology result became available; assessed through October 15, 2025, over the approximately 14-month study period.
Nodule-Level Specificity of AI and C-TIRADS	Using definitive postoperative histopathology as the reference standard, specificity is TN/(TN + FP), expressed as a percentage, for the locked AI classification (positive if score >0.5) and conventional C-TIRADS (positive if category 4a or higher). Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Nodule-Level Negative Predictive Value of AI and C-TIRADS	Using definitive postoperative histopathology as the reference standard, negative predictive value is TN/(TN + FN), expressed as a percentage, for the locked AI classification (positive if score >0.5) and conventional C-TIRADS (positive if category 4a or higher). Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Nodule-Level Accuracy of AI and C-TIRADS	Using definitive postoperative histopathology as the reference standard, accuracy is (TP + TN)/(TP + TN + FP + FN), expressed as a percentage, for the locked AI classification (positive if score >0.5) and conventional C-TIRADS (positive if category 4a or higher). Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.
Area Under the ROC Curve for the Continuous AI Malignancy Score	Using definitive postoperative histopathology as the reference standard, calculate the area under the receiver operating characteristic curve for the continuous locked AI malignancy score. Include only pathology-matched nodules with an unambiguous lesion-level link and a definitive benign or malignant diagnosis. Exclude FT-UMP, WDT-UMP, NIFTP, other borderline diagnoses, and nodules without unambiguous lesion-level linkage. No AUC is calculated for the single-threshold C-TIRADS classification.	From the preoperative examinations until the final postoperative histopathology report became available; assessed through October 15, 2025, over the approximately 14-month study period.

Collaborators and Investigators

• Sponsor: Fujian Medical University

Study record dates

Study Major Dates

• Study Start (Actual): September 1, 2024
• Primary Completion (Actual): September 28, 2025
• Study Completion (Actual): October 15, 2025

Study Registration Dates

• First Submitted: June 9, 2024
• First Submitted That Met QC Criteria: July 11, 2024
• First Posted (Actual): July 12, 2024

Study Record Updates

• Last Update Posted (Actual): August 7, 2026
• Last Update Submitted That Met QC Criteria: August 5, 2026
• Last Verified: August 1, 2026

More Information

Terms related to this study

• Keywords: Ultrasound; Artificial Intelligence; Thyroid Cancer; Preoperative Assessment
• Additional Relevant MeSH Terms: Endocrine System Diseases; Neoplasms by Site; Neoplasms; Endocrine Gland Neoplasms; Head and Neck Neoplasms; Thyroid Diseases; Thyroid Nodule; Thyroid Neoplasms
• Other Study ID Numbers: AI-base PCD

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: Individual participant data are not planned for external sharing because the dataset contains detailed clinical records and ultrasound images that may remain re-identifiable, and the available study records do not document participant consent specifically permitting external IPD sharing or institutional authorization for such sharing. Aggregate, non-identifiable results will be reported in publications.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No