Computer Vision-Based Recognition of Parathyroid Glands and Recurrent Laryngeal Nerves in Endoscopic Surgery

Official title

Prospective Multicenter Model Validation Study of the PTAIR Computer Vision Model for Recognition of Parathyroid Glands and Recurrent Laryngeal Nerves During Endoscopic Thyroid Surgery

Overview

This study aims to evaluate the performance of the PTAIR computer vision model for recognizing parathyroid glands and recurrent laryngeal nerves during endoscopic thyroid surgery. Multicenter intraoperative endoscopic surgery videos will be prospectively collected. Using the video recorder timeline as the unified time reference, the first recognition time of parathyroid glands and recurrent laryngeal nerves by PTAIR, junior physicians, and senior physicians will be recorded during surgery. The continuous recognition time of parathyroid glands and recurrent laryngeal nerves by PTAIR will also be recorded. This study focuses on the recognition performance of PTAIR in real-world multicenter endoscopic thyroid surgery settings. At this stage, PTAIR will not be used to guide intraoperative surgical decision-making, and the study will not evaluate the effect of PTAIR on clinical outcomes. The collected video data will be used to assess the performance of PTAIR under different centers, surgeons, equipment conditions, and surgical field conditions, and may provide data support for future model optimization.

Detailed description

Identification of the parathyroid glands and recurrent laryngeal nerves is an important component of safe thyroid surgery. Parathyroid glands are small and may resemble surrounding fat, lymph nodes, or soft tissue. The recurrent laryngeal nerve has a complex anatomical course and may be partially covered by fascia or insufficiently exposed in some cases, which may increase the difficulty of intraoperative recognition. Computer vision-based artificial intelligence models may help recognize these structures in endoscopic surgical images, but their performance in multicenter real-world surgical settings requires further validation.

Previous PTAIR studies mainly focused on parathyroid gland recognition in endoscopic thyroid surgery videos. Based on these prior studies, the present study will further evaluate the PTAIR model in a prospective multicenter setting, with a focus on recognition of parathyroid glands and recurrent laryngeal nerves. This stage of the study is designed as a model performance validation study and will not evaluate whether PTAIR changes intraoperative surgical decisions, parathyroid management, or postoperative clinical outcomes.

During endoscopic thyroid surgery, PTAIR, junior physicians, and senior physicians will identify parathyroid glands and recurrent laryngeal nerves based on the same intraoperative endoscopic visual field. All time-related outcomes will use the video recorder timeline as the unified reference. The study will record the first recognition time of the recurrent laryngeal nerve and parathyroid gland by PTAIR, junior physicians, and senior physicians, as well as the continuous recognition time of these structures by PTAIR. Expert-reviewed annotations will be used as the reference for subsequent model performance evaluation and confirmation of recognition results.

The study will also evaluate PTAIR performance under different exposure conditions, including cases with fascial coverage, incomplete exposure, or complex surgical field conditions. Multicenter video data will be used to assess model performance across different centers, surgeons, equipment settings, and operative field conditions, and may provide data for future model improvement and iteration.

Interventions

• Other PTAIR Computer Vision Model Assessment
  PTAIR is a computer vision model used to recognize parathyroid glands and recurrent laryngeal nerves in endoscopic thyroid surgery videos. In this observational study, PTAIR will be used to analyze recorder-based intraoperative videos and evaluate first recognition time, continuous recognition time, and recognition performance. PTAIR results will be compared with recognition by junior and senior physicians based on the same intraoperative endoscopic visual field and the same video recording time

Primary outcome measures

• First Recognition Time of the Recurrent Laryngeal Nerve [Time frame: Intraoperative period]
• First Recognition Time of the Parathyroid Gland [Time frame: Intraoperative period]
• Continuous Recognition Time of the Recurrent Laryngeal Nerve [Time frame: Intraoperative period]
• Continuous Recognition Time of the Parathyroid Gland [Time frame: Intraoperative period]

Secondary outcome measures (6)

• AP50 for Recurrent Laryngeal Nerve Recognition [Time frame: From completion of intraoperative video collection to completion of expert annotation and model analysis]
• AP50 for Parathyroid Gland Recognition [Time frame: From completion of intraoperative video collection to completion of expert annotation and model analysis]
• Comparison of First Recognition Time of the Recurrent Laryngeal Nerve Between PTAIR and Physicians [Time frame: Intraoperative period]
• Comparison of First Recognition Time of the Parathyroid Gland Between PTAIR and Physicians [Time frame: Intraoperative period]
• In Situ Recognition Rate of the Parathyroid Gland [Time frame: From completion of intraoperative video collection to completion of expert review]
• Detection Rate of Parathyroid Glands on the Surgical Specimen [Time frame: Intraoperative period, before removal of the surgical specimen from the body]

Eligibility criteria

Inclusion criteria

• Age 18 years or older.
• Patients with benign or malignant thyroid disease who are scheduled to undergo endoscopic thyroid surgery.
• Standardized recorder-based intraoperative video collection is planned during surgery.
• Patients agree to intraoperative video recording, de-identification, and research use.
• Basic perioperative clinical data are available.
• The participating center is able to follow the standardized procedures for video collection, data upload, data security, and annotation quality control.

Exclusion criteria

• Refusal to provide informed consent.
• Patients scheduled to undergo conventional open thyroid surgery.
• Standardized recorder-based intraoperative video collection cannot be performed.
• Intraoperative video quality is insufficient for model analysis or expert review.
• Key surgical phases are missing from the video recording.
• Expert annotation or review of the parathyroid glands or recurrent laryngeal nerves cannot be completed.
• Video de-identification fails.
• Severe missing perioperative clinical data.
• Other conditions judged by the investigators to be unsuitable for study inclusion.

Criteria are shown verbatim from the registry (in English). Final eligibility is always assessed by the study center.

Healthy volunteers: No

Study design

Observational model

Cohort

Study locations

China · 1 center

• Fujian Medical University Union Hospital — Fuzhou

Publications

• Wang B, Zheng J, Yu JF, Lin SY, Yan SY, Zhang LY, Wang SS, Cai SJ, Abdelhamid Ahmed AH, Lin LQ, Chen F, Randolph GW, Zhao WX. Development of Artificial Intelligence for Parathyroid Recognition During Endoscopic Thyroid Surgery. Laryngoscope. 2022 Dec;132(12):2516-2523. doi: 10.1002/lary.30173. Epub 2022 May 31. PMID 35638245
• Wang B, Yu JF, Lin SY, Li YJ, Huang WY, Yan SY, Wang SS, Zhang LY, Cai SJ, Wu SB, Li MY, Wang TY, Abdelhamid Ahmed AH, Randolph GW, Chen F, Zhao WX. Intraoperative AI-assisted early prediction of parathyroid and ischemia alert in endoscopic thyroid surgery. Head Neck. 2024 Aug;46(8):1975-1987. doi: 10.1002/hed.27629. Epub 2024 Feb 13. PMID 38348564

Identifiers

NCT: NCT07628543 · PTAIR 3.0

Primary sources (government registries)

• ClinicalTrials.gov (US NIH)

View this study on ClinicalTrials.gov ↗