Power of PTAIR 2.0 in Prediction, Identification, and Ischemia Assessment of the Parathyroid Gland Under Endoscopic Thyroid Surgery

Abstract

Objective: This video demonstrates the use of PTAIR 2.0, which is an artificial intelligence model for the prediction, identification, and ischemia assessment of parathyroid glands in endoscopic thyroid surgery. Methods: PTAIR 2.0 is an artificial intelligence model for parathyroid gland recognition during surgery. In total, 1641 patients who underwent endoscopic thyroid surgery at our hospital from August 2012 to December 2021 were selected. YOLOX was used to train this version. The hindsight strategy was used for the labeling of parathyroid gland prediction data.1–3 Results: PTAIR achieved the AP50 values for the prediction, identification, and ischemia assessment of the parathyroid glands of 94.1%, 98.9%, and 92.1%, respectively. Conclusions: PTAIR, which was established through computer vision, provides a novel methodology for the prediction, identification, and ischemia assessment of parathyroid glands in endoscopic thyroid surgery. PTAIR can be used to detect parathyroid glands and provide decision-making in the autotransplantation of parathyroid glands in endoscopic thyroid surgery. Acknowledgments: On behalf of the PTAIR Research Group, Yi-Jian Li from Fuzhou University, Jun-Yi Huang and Chi-Peng Zhou from Fujian Medical University Union Hospital, and Jing Zheng from Zhangzhou Affiliated Hospital of Fujian Medical University have also contributed to this video. Si-Ying Lin, Meng-Yao Li, and Jia-Fan Yu contributed equally to this study and are considered co-first authors. PTAIR is a registered trademark of Fujian Medical University Union Hospital. All other authors have no conflicts of interest to disclose. Funding Information: This research was supported by the China Scholarship Council; the Joint Funds for the Innovation of Science and Technology, Fujian Province (Grant Nos. 2018Y9093, 2018Y0915, and 2018Y9018); sponsored by Fujian Provincial Health Technology Project (Grant Nos. 2019-CX-16 and 2018-CX-17). Runtime of video: 3 mins 23 secs An abstract based on this video was posted for presentation at the 2022 American Thyroid Association Annual Meeting.