Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study
Abstract Background This study aimed to develop a BI-RADS network (DL-UM) via integrating ultrasound (US) and mammography (MG) images and explore its performance in improving breast lesion diagnosis and management when collaborating with radiologists, particularly in cases with discordant US and MG...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-05-01
|
| Series: | Breast Cancer Research |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s13058-025-02033-6 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850221789546807296 |
|---|---|
| author | Ziting Xu Shengzhou Zhong Yang Gao Jiekun Huo Weimin Xu Weijun Huang Xiaomei Huang Chifa Zhang Jianqiao Zhou Qing Dan Lian Li Zhouyue Jiang Ting Lang Shuying Xu Jiayin Lu Ge Wen Yu Zhang Yingjia Li |
| author_facet | Ziting Xu Shengzhou Zhong Yang Gao Jiekun Huo Weimin Xu Weijun Huang Xiaomei Huang Chifa Zhang Jianqiao Zhou Qing Dan Lian Li Zhouyue Jiang Ting Lang Shuying Xu Jiayin Lu Ge Wen Yu Zhang Yingjia Li |
| author_sort | Ziting Xu |
| collection | DOAJ |
| description | Abstract Background This study aimed to develop a BI-RADS network (DL-UM) via integrating ultrasound (US) and mammography (MG) images and explore its performance in improving breast lesion diagnosis and management when collaborating with radiologists, particularly in cases with discordant US and MG Breast Imaging Reporting and Data System (BI-RADS) classifications. Methods We retrospectively collected image data from 1283 women with breast lesions who underwent both US and MG within one month at two medical centres and categorised them into concordant and discordant BI-RADS classification subgroups. We developed a DL-UM network via integrating US and MG images, and DL networks using US (DL-U) or MG (DL-M) alone, respectively. The performance of DL-UM network for breast lesion diagnosis was evaluated using ROC curves and compared to DL-U and DL-M networks in the external testing dataset. The diagnostic performance of radiologists with different levels of experience under the assistance of DL-UM network was also evaluated. Results In the external testing dataset, DL-UM outperformed DL-M in sensitivity (0.962 vs. 0.833, P = 0.016) and DL-U in specificity (0.667 vs. 0.526, P = 0.030), respectively. In the discordant BI-RADS classification subgroup, DL-UM achieved an AUC of 0.910. The diagnostic performance of four radiologists improved when collaborating with the DL-UM network, with AUCs increased from 0.674–0.772 to 0.889–0.910, specificities from 52.1%–75.0% to 81.3%–87.5% and reducing unnecessary biopsies by 16.1%–24.6%, particularly for junior radiologists. Meanwhile, DL-UM outputs and heatmaps enhanced radiologists’ trust and improved interobserver agreement between US and MG, with weighted kappa increased from 0.048 to 0.713 (P < 0.05). Conclusions The DL-UM network, integrating complementary US and MG features, assisted radiologists in improving breast lesion diagnosis and management, potentially reducing unnecessary biopsies. |
| format | Article |
| id | doaj-art-276d65506573490e8dbe60d2c7e77131 |
| institution | OA Journals |
| issn | 1465-542X |
| language | English |
| publishDate | 2025-05-01 |
| publisher | BMC |
| record_format | Article |
| series | Breast Cancer Research |
| spelling | doaj-art-276d65506573490e8dbe60d2c7e771312025-08-20T02:06:36ZengBMCBreast Cancer Research1465-542X2025-05-0127111210.1186/s13058-025-02033-6Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective studyZiting Xu0Shengzhou Zhong1Yang Gao2Jiekun Huo3Weimin Xu4Weijun Huang5Xiaomei Huang6Chifa Zhang7Jianqiao Zhou8Qing Dan9Lian Li10Zhouyue Jiang11Ting Lang12Shuying Xu13Jiayin Lu14Ge Wen15Yu Zhang16Yingjia Li17Department of Ultrasound, Nanfang Hospital, Southern Medical UniversitySchool of Biomedical Engineering, Southern Medical UniversityDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityDepartment of Imaging, Zengcheng Branch of Nanfang Hospital, Southern Medical UniversityDepartment of Radiology, Nanfang Hospital, Southern Medical UniversityDepartment of Ultrasound, First People’s Hospital of FoshanDepartment of Medical Imaging, Nanfang Hospital, Southern Medical UniversityDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityDepartment of Ultrasound, Ruijin Hospital, Shanghai Jiaotong University School of MedicineShenzhen Key Laboratory for Drug Addiction and Medication Safety, Department of Ultrasound, Institute of Ultrasonic Medicine, Peking University Shenzhen Hospital, Shenzhen Peking University-The Hong Kong University of Science and Technology Medical CenterDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityDepartment of Medical Imaging, Nanfang Hospital, Southern Medical UniversitySchool of Biomedical Engineering, Southern Medical UniversityDepartment of Ultrasound, Nanfang Hospital, Southern Medical UniversityAbstract Background This study aimed to develop a BI-RADS network (DL-UM) via integrating ultrasound (US) and mammography (MG) images and explore its performance in improving breast lesion diagnosis and management when collaborating with radiologists, particularly in cases with discordant US and MG Breast Imaging Reporting and Data System (BI-RADS) classifications. Methods We retrospectively collected image data from 1283 women with breast lesions who underwent both US and MG within one month at two medical centres and categorised them into concordant and discordant BI-RADS classification subgroups. We developed a DL-UM network via integrating US and MG images, and DL networks using US (DL-U) or MG (DL-M) alone, respectively. The performance of DL-UM network for breast lesion diagnosis was evaluated using ROC curves and compared to DL-U and DL-M networks in the external testing dataset. The diagnostic performance of radiologists with different levels of experience under the assistance of DL-UM network was also evaluated. Results In the external testing dataset, DL-UM outperformed DL-M in sensitivity (0.962 vs. 0.833, P = 0.016) and DL-U in specificity (0.667 vs. 0.526, P = 0.030), respectively. In the discordant BI-RADS classification subgroup, DL-UM achieved an AUC of 0.910. The diagnostic performance of four radiologists improved when collaborating with the DL-UM network, with AUCs increased from 0.674–0.772 to 0.889–0.910, specificities from 52.1%–75.0% to 81.3%–87.5% and reducing unnecessary biopsies by 16.1%–24.6%, particularly for junior radiologists. Meanwhile, DL-UM outputs and heatmaps enhanced radiologists’ trust and improved interobserver agreement between US and MG, with weighted kappa increased from 0.048 to 0.713 (P < 0.05). Conclusions The DL-UM network, integrating complementary US and MG features, assisted radiologists in improving breast lesion diagnosis and management, potentially reducing unnecessary biopsies.https://doi.org/10.1186/s13058-025-02033-6Neural networksUltrasonographyDigital mammographyClinical decision-makingBreast tumours |
| spellingShingle | Ziting Xu Shengzhou Zhong Yang Gao Jiekun Huo Weimin Xu Weijun Huang Xiaomei Huang Chifa Zhang Jianqiao Zhou Qing Dan Lian Li Zhouyue Jiang Ting Lang Shuying Xu Jiayin Lu Ge Wen Yu Zhang Yingjia Li Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study Breast Cancer Research Neural networks Ultrasonography Digital mammography Clinical decision-making Breast tumours |
| title | Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study |
| title_full | Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study |
| title_fullStr | Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study |
| title_full_unstemmed | Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study |
| title_short | Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study |
| title_sort | optimizing breast lesions diagnosis and decision making with a deep learning fusion model integrating ultrasound and mammography a dual center retrospective study |
| topic | Neural networks Ultrasonography Digital mammography Clinical decision-making Breast tumours |
| url | https://doi.org/10.1186/s13058-025-02033-6 |
| work_keys_str_mv | AT zitingxu optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT shengzhouzhong optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT yanggao optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT jiekunhuo optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT weiminxu optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT weijunhuang optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT xiaomeihuang optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT chifazhang optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT jianqiaozhou optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT qingdan optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT lianli optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT zhouyuejiang optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT tinglang optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT shuyingxu optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT jiayinlu optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT gewen optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT yuzhang optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy AT yingjiali optimizingbreastlesionsdiagnosisanddecisionmakingwithadeeplearningfusionmodelintegratingultrasoundandmammographyadualcenterretrospectivestudy |