An integrated AI model to improve diagnostic accuracy of ultrasound and output known risk features in suspicious thyroid nodules
European Radiology · 2022-03 · 32 (3) · 2120-2129
DOI: 10.1007/s00330-021-08298-7 ↗Kurz gefasst
Fachartikel in der Zeitschrift European Radiology, online erschienen am 18.10.2021. Verfasst von Juan Wang und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Ultrasound, the Second Affiliated Hospital, Medical School of Xi'an Jiaotong University, Xi'an, 710004, China. Hauptthemen laut MEDLINE: Thyroid Neoplasms, Thyroid Nodule. Die Arbeit wurde bisher 40-mal zitiert (Crossref) und nennt 20 Literaturangaben. Gefördert von: Innovative Research Group Project of the National Natural Science Foundation of China.
Abstract (Originalsprache)
OBJECTIVES: From the viewpoint of ultrasound (US) physicians, an ideal thyroid US computer-assisted diagnostic (CAD) system for thyroid cancer should perform well in suspicious thyroid nodules with atypical risk features and be able to output explainable results. This study aims to develop an explainable US CAD model for suspicious thyroid nodules. METHODS: A total of 2992 solid or almost-solid thyroid nodules were analyzed retrospectively. All nodules had pathological results (1070 malignancies and 1992 benignities) confirmed by ultrasound-guided fine-needle aspiration cytology and histopathology after thyroidectomy. A deep learning model (ResNet50) and a multiple risk features learning ensemble model (XGBoost) were used to train the US images of 2794 thyroid nodules. Then, an integrated AI model was generated by combining both models. The diagnostic accuracies of the three AI models (ResNet50, XGBoost, and the integrated model) were predicted in a testing set including 198 thyroid nodules and compared to the diagnostic efficacy of five ultrasonographers. RESULTS: The accuracy of the integrated model was 76.77%, while the mean accuracy of the ultrasonographers was 68.38%. Of the risk features, microcalcifications showed the highest contribution to the diagnosis of malignant nodules. CONCLUSIONS: The integrated AI model in our study can improve the diagnostic accuracy of suspicious thyroid nodules and output the known risk features simultaneously, thus aiding in training young ultrasonographers by linking the explainable results to their clinical experience and advancing the acceptance of AI diagnosis for thyroid cancer in clinical practice. KEY POINTS: • We developed an artificial intelligence (AI) diagnosis model based on both deep learning and multiple risk feature ensemble learning methods. • The AI diagnosis model showed higher diagnostic accuracy for suspicious thyroid nodules than ultrasonographers. • The AI diagnosis model showed partial explainability by outputting the known risk features, thus aiding young ultrasonic doctors in increasing the diagnostic level for thyroid cancer.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Ultrasound, the Second Affiliated Hospital, Medical School of Xi'an Jiaotong University, Xi'an, 710004, China.
- School of Automation Science and Engineering, Faculty of Electronic and Information Engineering, Xi'an Jiaotong University, Xi'an, 710049, Shaanxi, China.
- The Institute of Medical Science, The University of Tokyo, Shirokanedai 4-6-1, Minato-ku, Tokyo, 108-8639, Japan.
- Precision Medicine Institute, Western China Science and Technology Innovation Harbor, Xi'an Jiaotong University, Xi'an, 712000, China.
- Department of Computer Science, Columbia University, 500 West 120 Street, New York, NY, 10027, USA.
- Institute of Artificial Intelligence and Robotics, College of Artificial Intelligence, Xi'an Jiaotong University, Xi'an, 710049, China. dushaoyi@gmail.com.
Förderung
- Innovative Research Group Project of the National Natural Science Foundation of China
Zitieren
APA
Wang, J., Jiang, J., Zhang, D., Zhang, Y. Z., Guo, L., Jiang, Y., Du, S., & Zhou, Q. (2022). An integrated AI model to improve diagnostic accuracy of ultrasound and output known risk features in suspicious thyroid nodules. European Radiology, 32(3), 2120-2129. https://doi.org/10.1007/s00330-021-08298-7
Vancouver
Wang J, Jiang J, Zhang D, Zhang YZ, Guo L, Jiang Y, et al. An integrated AI model to improve diagnostic accuracy of ultrasound and output known risk features in suspicious thyroid nodules. European Radiology. 2022;32(3):2120-2129. doi:10.1007/s00330-021-08298-7
BibTeX
@article{Wang2022,
author = {Wang, Juan and Jiang, Jue and Zhang, Dong and Zhang, Yao-Zhong and Guo, Long and Jiang, Yusheng and Du, Shaoyi and Zhou, Qi},
title = {An integrated AI model to improve diagnostic accuracy of ultrasound and output known risk features in suspicious thyroid nodules},
journal = {European Radiology},
year = {2022},
volume = {32},
number = {3},
pages = {2120--2129},
doi = {10.1007/s00330-021-08298-7}
}
EUVIMED