Aim: Deep inferior epigastric perforator (DIEP) flap breast reconstruction is considered the gold standard for autologous reconstruction. Preoperative perforator mapping using computed tomography angiography (CTA) remains labor-intensive, time-consuming, and subject to interobserver variability. Automated computer-aided detection (CAD) systems may help standardize and accelerate this process. This study aimed to develop and evaluate a proof-of-concept automated CAD pipeline for CTA-based perforator mapping in DIEP flap planning. Methods: A retrospective dataset of 504 CTA scans acquired for DIEP flap planning was analyzed. Fifty-five scans were manually annotated for perforator segmentation, and 100 scans were annotated for umbilicus landmark detection. A dual maximum intensity projection (MIP) depth-aware annotation workflow was introduced to standardize vessel labeling. The automated pipeline combined anatomical region of interest (ROI) localization with deep-learning-based vessel segmentation using a 3D Swin UNETR (Swin Transformer-based) model. Performance was evaluated using the Dice similarity coefficient (Dice), centerline Dice, recall, and the 95th percentile Hausdorff distance (HD95). Results: Depth-aware annotation reduced labeling time by approximately 60\%-70\%. ROI localization was successful in all scans (28 ± 5 s), and umbilicus localization achieved an error of approximately 2 mm. The Swin UNETR model achieved a median Dice score of 0.58, outperforming Attention U-Net. Continuity-aware training improved Dice to 0.60 and recall to 0.58, while multiclass segmentation improved performance in adipose tissue. Conclusion: This study demonstrates the feasibility of an automated CAD pipeline integrating standardized annotation, anatomical ROI localization, and deep-learning-based vessel segmentation for DIEP flap planning. This represents an important step toward faster, more reproducible, and clinically scalable CTA-based perforator mapping.
Kapila, AK, Marcos, DL, Ceranka, J, Brussaard, C, Boonen, PT, Ledegen, L, Vandemeulebroucke, J & Hamdi, M 2026, 'Automated CTA-based perforator mapping for DIEP flap planning in breast cancer reconstruction', Artificial Intelligence Surgery, vol. 6, no. 2, pp. 255-267. https://doi.org/10.20517/ais.2026.01
Kapila, A. K., Marcos, D. L., Ceranka, J., Brussaard, C., Boonen, P. T., Ledegen, L., Vandemeulebroucke, J., & Hamdi, M. (2026). Automated CTA-based perforator mapping for DIEP flap planning in breast cancer reconstruction. Artificial Intelligence Surgery, 6(2), 255-267. https://doi.org/10.20517/ais.2026.01
@article{524a2404d6254f2b9163c78c66d9297f,
title = "Automated CTA-based perforator mapping for DIEP flap planning in breast cancer reconstruction",
abstract = "Aim: Deep inferior epigastric perforator (DIEP) flap breast reconstruction is considered the gold standard for autologous reconstruction. Preoperative perforator mapping using computed tomography angiography (CTA) remains labor-intensive, time-consuming, and subject to interobserver variability. Automated computer-aided detection (CAD) systems may help standardize and accelerate this process. This study aimed to develop and evaluate a proof-of-concept automated CAD pipeline for CTA-based perforator mapping in DIEP flap planning. Methods: A retrospective dataset of 504 CTA scans acquired for DIEP flap planning was analyzed. Fifty-five scans were manually annotated for perforator segmentation, and 100 scans were annotated for umbilicus landmark detection. A dual maximum intensity projection (MIP) depth-aware annotation workflow was introduced to standardize vessel labeling. The automated pipeline combined anatomical region of interest (ROI) localization with deep-learning-based vessel segmentation using a 3D Swin UNETR (Swin Transformer-based) model. Performance was evaluated using the Dice similarity coefficient (Dice), centerline Dice, recall, and the 95th percentile Hausdorff distance (HD95). Results: Depth-aware annotation reduced labeling time by approximately 60\%-70\%. ROI localization was successful in all scans (28 ± 5 s), and umbilicus localization achieved an error of approximately 2 mm. The Swin UNETR model achieved a median Dice score of 0.58, outperforming Attention U-Net. Continuity-aware training improved Dice to 0.60 and recall to 0.58, while multiclass segmentation improved performance in adipose tissue. Conclusion: This study demonstrates the feasibility of an automated CAD pipeline integrating standardized annotation, anatomical ROI localization, and deep-learning-based vessel segmentation for DIEP flap planning. This represents an important step toward faster, more reproducible, and clinically scalable CTA-based perforator mapping.",
keywords = "artificial intelligence, Breast cancer reconstruction, computed tomographic angiography, computer-aided detection, deep learning, DIEP flap surgery, perforator segmentation, computer, computer aided detection system, electronics, Meta AI, Pythonidae, Python 3.11, PyTorch, algorithm, anatomy, architecture, article, breast cancer, breast cancer reconstruction, clinical practice, cross validation, decision making, deep inferior epigastric perforator flap, digital imaging and communications in medicine, entropy, esthetic surgery, ethics, experience, human, information science, learning, morphology, performance, planning, radiologist, reproducibility, statistical analysis, training, vascular anatomy, warm up, workflow",
author = "Kapila, \{Ayush K.\} and Marcos, \{Diego Lamtenzan\} and Jakub Ceranka and Carola Brussaard and Boonen, \{Pieter Thomas\} and Laure Ledegen and Jef Vandemeulebroucke and Moustapha Hamdi",
note = "Publisher Copyright: {\textcopyright} The Authors 2026.",
year = "2026",
month = jun,
day = "1",
doi = "10.20517/ais.2026.01",
language = "English",
volume = "6",
pages = "255--267",
journal = "Artificial Intelligence Surgery",
issn = "2771-0408",
publisher = "OAE Publishing",
number = "2",
}