Article
Radiology, Nuclear Medicine & Medical Imaging
Fang Xie, Yu-Kun Luo, Yu Lan, Xiao-Qi Tian, Ya-Qiong Zhu, Zhuang Jin, Ying Zhang, Ming-Bo Zhang, Qing Song, Yan Zhang
Summary: This study assessed the diagnostic efficacy of a computer-aided ultrasonic diagnosis system (CAD system) in differentiating benign and malignant thyroid nodules. The results showed that the CAD system, when used in combination with junior physicians, had higher accuracy and comparable performance to senior physicians, improving the diagnostic accuracy of junior physicians.
BMC MEDICAL IMAGING
(2022)
Article
Oncology
Huan Zheng, Zebin Xiao, Siwei Luo, Suqing Wu, Chuxin Huang, Tingting Hong, Yan He, Yanhui Guo, Guoqing Du
Summary: This study develops a computer aided diagnosis (CAD) system using deep learning to assist radiologists in diagnosing follicular thyroid carcinoma (FTC) on thyroid ultrasonography. The CAD system achieves better performance than radiologists and significantly improves their diagnosis of FTC. It provides a reliable reference for preoperative diagnosis of FTC and may assist in the development of a fast, accessible screening method for FTC.
FRONTIERS IN ONCOLOGY
(2022)
Article
Physiology
Keen Yang, Jing Chen, Huaiyu Wu, Hongtian Tian, Xiuqin Ye, Jinfeng Xu, Xunpeng Luo, Fajin Dong
Summary: This study compared the diagnostic results of S-thyroid, a computer-aided diagnosis (CAD) software, based on two mutually perpendicular planes. The results showed that the diagnosis of thyroid cancer in the CAD transverse plane was superior to that in the CAD longitudinal plane when using the TI-RADS classification. However, there was no difference in the diagnosis between the two planes when using risk. The combination of CAD transverse and longitudinal planes had the best diagnostic ability.
FRONTIERS IN PHYSIOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Yifei Chen, Dandan Li, Xin Zhang, Jing Jin, Yi Shen
Summary: A multi-view ensemble learning method is proposed to enhance the performance of models trained on small-dataset thyroid nodule ultrasound images. This method integrates diagnosis results from different sources and utilizes a voting mechanism to improve the accuracy of final results.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Endocrinology & Metabolism
Mengwen Xia, Fulong Song, Yongfeng Zhao, Yongzhi Xie, Yafei Wen, Ping Zhou
Summary: This study compared the performance of radiomics and computer-aided diagnosis (CAD) models based on ultrasonography (US) features in predicting malignancy in thyroid nodules, and evaluated their utility in thyroid nodule management. The results showed that both models had good diagnostic performance and could be used to optimize the recommendations of the American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS), reducing unnecessary biopsies.
FRONTIERS IN ENDOCRINOLOGY
(2023)
Article
Clinical Neurology
M. Han, E. J. Ha, J. H. Park
Summary: The study compared the diagnostic performance of a computer-aided diagnostic system and radiologist-based assessment for thyroid cancer detection. The computer-aided diagnostic system showed high sensitivity but lower specificity and accuracy compared to the radiologist.
AMERICAN JOURNAL OF NEURORADIOLOGY
(2021)
Article
Biology
Anca Emanuela Musetescu, Florin Liviu Gherghina, Lucian-Mihai Florescu, Liliana Streba, Paulina Lucia Ciurea, Alesandra Florescu, Ioana Andreea Gheonea
Summary: The study used CAD technology to detect lung nodules in patients with rheumatoid arthritis, and found that CAD can play a role in reducing the interpretation time of CT examinations, but there are also certain false positive and false negative rates.
Article
Endocrinology & Metabolism
Lin-Lin Zheng, Su-Ya Ma, Ling Zhou, Cong Yu, Hai-Shan Xu, Li-Long Xu, Shi-Yan Li
Summary: This study aimed to evaluate the ability of a computer-aided diagnosis system based on artificial intelligence (AI-CADS) to predict thyroid malignancy by analyzing different ultrasound sections of thyroid nodules. The study included patients with preoperative ultrasound data and postoperative pathological results, and assessed the diagnostic performance and consistency of AI-CADS in different sections. The results showed that the performance of AI-CADS was better in the transverse section, and certain ultrasonic features had higher diagnostic agreement.
FRONTIERS IN ENDOCRINOLOGY
(2023)
Article
Oncology
Liu Gong, Ping Zhou, Jia-Le Li, Wen-Gang Liu
Summary: This study aimed to assess the efficacy of a computer-aided diagnosis (CAD) system in distinguishing between benign and malignant thyroid nodules in the context of Hashimoto's thyroiditis (HT) and reducing unnecessary biopsies. The CAD system showed high accuracy and improved diagnostic performance compared to radiologists, leading to a reduction in unnecessary biopsies of benign lesions.
FRONTIERS IN ONCOLOGY
(2023)
Article
Acoustics
Minxia Hu, Suting Zong, Ning Xu, Jinzhen Li, Chunxia Xia, Fengxia Yu, Qiang Zhu, Hanxue Zhao
Summary: The CAD system has comparable sensitivity to an experienced radiologist in diagnosing thyroid malignancies concurrent with HT, but lower specificity. For less experienced radiologists, the CAD system can help improve diagnostic performance by increasing sensitivity and accuracy in assessing thyroid nodules with diffusely altered parenchyma.
JOURNAL OF ULTRASOUND IN MEDICINE
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Jiyoung Yoon, Eunjung Lee, Sang-Wook Kang, Kyunghwa Han, Vivian Youngjean Park, Jin Young Kwak
Summary: This study evaluated the role of radiomics score using US images in predicting malignancy in AUS/FLUS and FN/SFN nodules. Results showed that nodules with higher radiomics score were more likely to be malignant. The predictive model combining radiomics score with clinical variables had significantly higher accuracy compared to the model with clinical variables alone.
EUROPEAN RADIOLOGY
(2021)
Review
Engineering, Biomedical
Yasaman Sharifi, Mohamad Amin Bakhshali, Toktam Dehghani, Morteza DanaiAshgzari, Mahdi Sargolzaei, Saeid Eslami
Summary: The study systematically reviewed the technical characteristics of deep learning applications on ultrasound images of thyroid nodules, emphasizing the necessity of addressing existing barriers such as data limitations, generating public and valid datasets, and determining standard evaluation metrics. Recommendations were made to utilize complementary information with multi-modal images to enhance the diagnostic accuracy of deep learning models.
BIOCYBERNETICS AND BIOMEDICAL ENGINEERING
(2021)
Review
Pathology
Tapoi Dana Antonia, Lambrescu Ioana Maria, Gheorghisan-Galateanu Ancuta-Augustina
Summary: Thyroid cancer is a common endocrine malignancy with a rising incidence. The gold standard for preoperative diagnosis is fine needle aspiration (FNA) biopsy, but it often gives indeterminate results. Other diagnostic tools such as ultrasonography, elastography, immunohistochemical analysis, genetic testing, and core needle biopsy have been developed to improve preoperative accuracy and aid in selecting appropriate cases for surgery.
PATHOLOGY RESEARCH AND PRACTICE
(2023)
Article
Oncology
Xiaowen Liang, Yingmin Huang, Yongyi Cai, Jianyi Liao, Zhiyi Chen
Summary: The study investigated the efficiency of the AI-Sonic CAD system for the detection and diagnosis of thyroid nodules, with findings showing that the system had better performance in recommending FNA compared to experts and novices. Additionally, the CAD system based on deep learning showed improved diagnostic accuracy and feasibility, helpful in avoiding unnecessary FNA.
FRONTIERS IN ONCOLOGY
(2021)
Article
Hematology
Dan Wang, Chong-Ke Zhao, Han-Xiang Wang, Feng Lu, Xiao-Long Li, Le-Hang Guo, Li-Ping Sun, Hui-Jun Fu, Yi-Feng Zhang, Hui-Xiong Xu
Summary: This study evaluated the role of computer-aided diagnosis (CAD) technique in predicting malignancy for cytologically indeterminate thyroid nodules (TNs). The results showed that combining CAD with radiologists' diagnosis can improve diagnostic performance and specificity.
CLINICAL HEMORHEOLOGY AND MICROCIRCULATION
(2022)