Deep Learning-assisted Diagnosis of Breast Lesions on Ultrasound Images: Multivendor, Multicenter Dataset (study data)
2025-12-05https://doi.org/10.1148/atlas.1764971966407
63
Overview
Schema Version
https://atlas.rsna.org/schemas/2025-11/dataset.json
Name
Deep Learning-assisted Diagnosis of Breast Lesions on Ultrasound Images: Multivendor, Multicenter Dataset (study data)
Link
https://dx.doi.org/10.1148/ryai.220185
Indexing
Keywords: Ultrasound, Breast, Diagnosis, Breast Cancer, Deep Learning, B-mode, Color Doppler, Multicenter, Multivendor
Content: BR, US, IN
RadLex: RID10326, RID29920, RID10904, RID34616, RID10909
SNOMED: 254837009, 427785007, 254848002
Author(s)
Huiling Xiang
Xi Wang
Min Xu
Yuhua Zhang
Shue Zeng
Chunyan Li
Lixian Liu
Tingting Deng
Guoxue Tang
Cuiju Yan
Jinjing Ou
Qingguang Lin
Jiehua He
Peng Sun
Anhua Li
Hao Chen
Pheng-Ann Heng
Xi Lin
Organization(s)
Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
Zhejiang Laboratory, Hangzhou, China
Department of Radiation Oncology, Stanford University School of Medicine, Stanford, Palo Alto, California, USA
Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong, China
The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, China
The Third People's Hospital of Zhengzhou, Cancer Hospital of Henan University, Zhengzhou, China
Hubei Cancer Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China
The Hong Kong University of Science and Technology, Hong Kong, China
License
Text: © 2023 by the Radiological Society of North America, Inc.
Contact
Corresponding author email: nc.gro.ccusys@ixnil
Funding
National Natural Science Foundation of China (82171955); Natural Science Foundation of Guangdong Province (2021A1515012476).
Ethical review
Ethical approval was obtained from all participating hospitals; the need for written informed consent was waived as the study presented no more than minimal risk.
Comments
Retrospective, multicenter, multivendor breast ultrasound study including B-mode and color Doppler images with pathology-confirmed labels (benign vs malignant).
Date
Published: 2023-07-12
References
[1] Xiang H, Wang X, Xu M, Zhang Y, Zeng S, Li C, Liu L, Deng T, Tang G, Yan C, Ou J, Lin Q, He J, Sun P, Li A, Chen H, Heng P-A, Lin X. "Deep Learning-assisted Diagnosis of Breast Lesions on US Images: A Multivendor, Multicenter Study". Radiology: Artificial Intelligence. 2023 Sep;5(5):e220185. 2023-07-12. doi:10.1148/ryai.220185. PMID: 37795135. PMCID: PMC10546363.
Dataset
Motivation
Develop and validate a deep learning model for breast tumor diagnosis on multicenter, multivendor ultrasound data and assess its effect on readers of differing experience.
Sampling
Retrospective inclusion of diagnostic US examinations with biopsy- or surgery-proven lesions from March 2011 to August 2018 across four hospitals.
Partitioning scheme
Primary site data randomly split 7:1:2 into training, validation, and internal test sets. Three external hospital datasets used as independent test sets.
Missing information
Image file formats, resolutions, and full per-site demographic breakdowns are not specified.
Relationships between instances
Multiple (multiview) images per lesion; lesion-level predictions obtained by averaging image-level probabilities.
External data
Data generated or analyzed during the study are available from the corresponding author by request.
Confidentiality
Retrospective clinical imaging with pathology-confirmed labels; static images used.
Sensitive data
Clinical images of human subjects; pathology-confirmed diagnoses.