Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
August 22, 2026Precision Clinical MedicineOpen Access

Artificial intelligence-based multimodal integration of ultrasound and digital breast tomosynthesis for breast-level risk classification

View Full Paper
Ask AI
Bookmark
Share

Authors

YTYujie TanMacau University of Science and TechnologyZHZhenjun HuangSun Yat-sen UniversityJLJunwei LiGuangxi Medical University

Discussion

Loading...

Member takes

Implication

Validation study demonstrates high diagnostic specificity using multimodal ultrasound and digital breast tomosynthesis deep learning models, suggesting improved accuracy in breast cancer triage.

Key Points

  • Develop and evaluate a parallel-branch deep learning framework combining ultrasound, digital mammography, and digital breast tomosynthesis for breast-level cancer risk classification.
  • Trained a deep learning framework on 2,187 breasts and evaluated six single- and dual-modality model configurations.
  • Evaluated diagnostic performance using an internal validation cohort (N=632 breasts) and an independent pathology-confirmed cohort (N=500 breasts).
  • In the internal validation cohort, the US–DBT model achieved an AUC of 0.944 (95% CI, 0.926–0.963), sensitivity of 0.860 (95% CI, 0.805–0.904), and specificity of 0.904 (95% CI, 0.871–0.930).
  • In the pathology-confirmed cohort, US–DBT achieved an AUC of 0.934 (95% CI, 0.913–0.955), PPV of 0.958 (95% CI, 0.931–0.977), sensitivity of 0.850 (95% CI, 0.807–0.887), and significantly higher specificity (0.955; 95% CI, 0.927–0.975; adjusted P < 0.001) compared to single and dual modalities.

Cite This Study

Tan et al. (2026) studied this question.

synapsesocial.com/papers/6a895f62ca7ade938187e0behttps://doi.org/10.1093/pcmedi/pbag023
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Radiomics analysis based on DBT and US for differentiating benign from malignant breast lesions: a multicenter study2026
  2. 2Integrating Artificial Intelligence-Powered Ultrasound With Mammography to Improve Differentiation of Benign and Malignant Non-Mass Breast Lesions: A Multimodal Diagnostic Approach2026
  3. 3Deep learning combining mammography and ultrasound images to predict the malignancy of BI-RADS US 4A lesions in women with dense breasts: a diagnostic study2024 · 21 citations
  4. 4A Multi-Modal AI System for Screening Mammography: Integrating 2D and 3D Imaging to Improve Breast Cancer Detection in a Prospective Clinical Study2025 · 4 citations
  5. 5Constructing the Optimal Classification Model for Benign and Malignant Breast Tumors Based on Multifeature Analysis from Multimodal Images2024 · 11 citations