Automated Tumor Segmentation in Breast-Conserving Surgery Using Deep Learning on Breast Tomosynthesis DOI
Wen-Pei Wu,

Yu-Wen Chen,

Hwa‐Koon Wu

et al.

Deleted Journal, Journal Year: 2025, Volume and Issue: unknown

Published: March 3, 2025

Breast cancer is one of the leading causes cancer-related deaths among women worldwide, with approximately 2.3 million diagnoses and 685,000 in 2020. Early-stage breast often managed through breast-conserving surgery (BCS) combined radiation therapy, which aims to preserve breast's appearance while reducing recurrence risks. This study aimed enhance intraoperative tumor segmentation using digital tomosynthesis (DBT) during BCS. A deep learning model, specifically an improved U-Net architecture incorporating a convolutional block attention module (CBAM), was utilized delineate margins high precision. The system evaluated on 51 patient cases by comparing automated manually delineated contours pathological assessments. Results showed that proposed method achieved promising accuracy, Intersection over Union (IoU) Dice coefficients 0.866 0.928, respectively, demonstrating its potential improve margin assessment surgical outcomes.

Language: Английский

Automated Tumor Segmentation in Breast-Conserving Surgery Using Deep Learning on Breast Tomosynthesis DOI
Wen-Pei Wu,

Yu-Wen Chen,

Hwa‐Koon Wu

et al.

Deleted Journal, Journal Year: 2025, Volume and Issue: unknown

Published: March 3, 2025

Breast cancer is one of the leading causes cancer-related deaths among women worldwide, with approximately 2.3 million diagnoses and 685,000 in 2020. Early-stage breast often managed through breast-conserving surgery (BCS) combined radiation therapy, which aims to preserve breast's appearance while reducing recurrence risks. This study aimed enhance intraoperative tumor segmentation using digital tomosynthesis (DBT) during BCS. A deep learning model, specifically an improved U-Net architecture incorporating a convolutional block attention module (CBAM), was utilized delineate margins high precision. The system evaluated on 51 patient cases by comparing automated manually delineated contours pathological assessments. Results showed that proposed method achieved promising accuracy, Intersection over Union (IoU) Dice coefficients 0.866 0.928, respectively, demonstrating its potential improve margin assessment surgical outcomes.

Language: Английский

Citations

0