EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Auto-Segmentation via deep-learning approaches for the assessment of flap volume after reconstructive surgery or radiotherapy in head and neck cancer

Juliette Thariat, Zacharia Mesbah, Youssef Chahir, Arnaud Beddok, Alice Blache, Jean Bourhis, Abir Fatallah, Mathieu Hatt, Romain Modzelewski

Scientific Reports · 2025

DOI: 10.1038/s41598-025-08073-4 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Scientific Reports, 2025. Verfasst von Juliette Thariat und 8 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiotherapy, Centre François-Baclesse, Caen, France. jthariat@gmail.com. Hauptthemen laut MEDLINE: Head and Neck Neoplasms, Deep Learning, Plastic Surgery Procedures, Surgical Flaps. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 20 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: MIC / 2022 / Flaps-segmIA / 22CM048-00, Normastic MSc Grant.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reconstructive flap surgery aims to restore the substance and function losses associated with tumor resection. Automatic flap segmentation could allow quantification of flap volume and correlations with functional outcomes after surgery or post-operative RT (poRT). Flaps being ectopic tissues of various components (fat, skin, fascia, muscle, bone) of various volume, shape and texture, the anatomical modifications, inflammation and edema of the postoperative bed make the segmentation task challenging. We built a artificial intelligence-enabled automatic soft-tissue flap segmentation method from CT scans of Head and Neck Cancer (HNC) patients. Ground-truth flap segmentation masks were delineated by two experts on postoperative CT scans of 148 HNC patients undergoing poRT. All CTs and flaps (free or pedicled, soft tissue only or bone) were kept, including those with artefacts, to ensure generalizability. A deep-learning nnUNetv2 framework was built using Hounsfield Units (HU) windowing to mimic radiological assessment. A transformer-based 2D “Segment Anything Model” (MedSAM) was also built and fine-tuned to medical CTs. Models were compared with the Dice Similarity Coefficient (DSC) and Hausdorff Distance 95th percentile (HD95) metrics. Flaps were in the oral cavity ( N = 102), oropharynx ( N = 26) or larynx/hypopharynx ( N = 20). There were free flaps ( N = 137), pedicled flaps ( N = 11), of soft tissue flap-only ( N = 92), reconstructed bone ( N = 42), or bone resected without reconstruction ( N = 40). The nnUNet-windowing model outperformed the nnUNetv2 and MedSam models. It achieved mean DSCs of 0.69 and HD95 of 25.6 mm using 5-fold cross-validation. Segmentation performed better in the absence of artifacts, and rare situations such as pedicled flaps, laryngeal primaries and resected bone without bone reconstruction ( p < 0.01). Automatic flap segmentation demonstrates clinical performances that allow to quantify spontaneous and radiation-induced volume shrinkage of flaps. Free flaps achieved excellent performances; rare situations will be addressed by fine-tuning the network.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Head and Neck NeoplasmsDeep LearningPlastic Surgery ProceduresSurgical FlapsHumansMaleFemaleTomography, X-Ray ComputedMiddle AgedAgedAdult

Stichwörter der Autor:innen

FlapHead and neck neoplasmsNeural networks (Computer)SurgeryVolume

Autor:innen und Institutionen

  • Department of Radiotherapy, Centre François-Baclesse, Caen, France. jthariat@gmail.com.
  • Corpuscular Physics Laboratory, IN2P3, Ensicaen, CNRS UMR 6534, Caen, France. jthariat@gmail.com.
  • University of Normandy, Caen, France. jthariat@gmail.com.
  • GORTEC. 4bis rue Emile Zola, Tours, France. jthariat@gmail.com.
  • Laboratoire de physique Corpusculaire IN2P3/ENSICAEN/CNRS UMR 6534 - Normandie Université, Caen, France. jthariat@gmail.com.
  • LITIS - UR4108 - Quantif, University of Rouen, Rouen, France.

Förderung

  • MIC / 2022 / Flaps-segmIA / 22CM048-00, Normastic MSc Grant

Zitieren

APA

Thariat, J., Mesbah, Z., Chahir, Y., Beddok, A., Blache, A., Bourhis, J., Fatallah, A., Hatt, M., & Modzelewski, R. (2025). Auto-Segmentation via deep-learning approaches for the assessment of flap volume after reconstructive surgery or radiotherapy in head and neck cancer. Scientific Reports. https://doi.org/10.1038/s41598-025-08073-4

Vancouver

Thariat J, Mesbah Z, Chahir Y, Beddok A, Blache A, Bourhis J, et al. Auto-Segmentation via deep-learning approaches for the assessment of flap volume after reconstructive surgery or radiotherapy in head and neck cancer. Scientific Reports. 2025. doi:10.1038/s41598-025-08073-4

BibTeX
@article{Thariat2025,
  author = {Thariat, Juliette and Mesbah, Zacharia and Chahir, Youssef and Beddok, Arnaud and Blache, Alice and Bourhis, Jean and Fatallah, Abir and Hatt, Mathieu and Modzelewski, Romain},
  title = {Auto-Segmentation via deep-learning approaches for the assessment of flap volume after reconstructive surgery or radiotherapy in head and neck cancer},
  journal = {Scientific Reports},
  year = {2025},
  doi = {10.1038/s41598-025-08073-4}
}
Kostenlos starten →