Multimodal GPT model for assisting thyroid nodule diagnosis and management
Jincao Yao, Yunpeng Wang, Zhikai Lei, Kai Wang, Na Feng, Fajin Dong, Jianhua Zhou, Xiaoxian Li, Xiang Hao, Jiafei Shen, Shanshan Zhao, Yuan Gao, Vicky Yang Wang, Di Ou, Xing Li, Yidan Lu, Li‐Yu Chen, Yang Chen, Liping Wang, Bojian Feng, Yahan Zhou, Chen Chen, Yuqi Yan, Zhengping Wang, Rongrong Ru, Yaqing Chen, Yanming Zhang, Ping Liang, Dong Xu
Abstract
Although using artificial intelligence (AI) to analyze ultrasound images is a promising approach to assessing thyroid nodule risks, traditional AI models lack transparency and interpretability. We developed a multimodal generative pre-trained transformer for thyroid nodules (ThyGPT), aiming to provide a transparent and interpretable AI copilot model for thyroid nodule risk assessment and management. Ultrasound data from 59,406 patients across nine hospitals were retrospectively collected to train and test the model. After training, ThyGPT was found to assist in reducing biopsy rates by more than 40% without increasing missed diagnoses. In addition, it detects errors in ultrasound reports 1,610 times faster than humans. With the assistance of ThyGPT, the area under the curve for radiologists in assessing thyroid nodule risks improved from 0.805 to 0.908 (p < 0.001). As an AI-generated content-enhanced computer-aided diagnosis (AIGC-CAD) model, ThyGPT has the potential to revolutionize how radiologists use such tools.