Skip to content
DraconiticResearch
Search Sources API
190,335 studies31 sourcesUpdated 59 min ago
Peer-reviewedCC BYnpj Digital Medicine

Translation of frozen sections into FFPE images for skin cancer resection margins using generative AI

  • Sang-Hoon Lee1
  • Yu-Sung Chu2
  • Yosep Chong3
  • Seong Min Hong3
  • Eung Ho Choi4
  • Seung-Phil Hong4
  • Taeyeong Kim1
  • Sejung Yang2
  • Minseob Eom1
4 affiliations
  1. 1Department of Pathology, Yonsei University Wonju College of Medicine
  2. 2Department of Precision Medicine, Yonsei University Wonju College of Medicine
  3. 3Department of Hospital Pathology, College of Medicine, The Catholic University of Korea
  4. 4Department of Dermatology, Yonsei University Wonju College of Medicine
Read full textPDFDOICite

Abstract

Frozen section (Frozen) analysis is essential for intraoperative margin assessment in skin cancer surgery, but assessment is limited by freezing artifacts that obscure cellular detail. Formalin-fixed paraffin-embedded (FFPE) histopathology, the gold standard, provides higher quality but requires >24 h for processing. Herein, we developed and validated generative AI models to translate Frozen images into AI-generated FFPE images (GenFFPE) using 2594 slides from 283 cases across five major skin cancer types. Four unpaired image-to-image models (CycleGAN, CUT, AIFFPE, and SANTA) were trained and compared using quantitative metrics and expert rankings; CUT demonstrated the best overall fidelity. External validation and a visual Turing test (accuracy 60.2%) confirmed image realism. Among 55 discrepant cases, GenFFPE-based reassessment increased diagnostic concordance by 53.3%. AI-based Frozen-to-FFPE translation is, thus, feasible and clinically meaningful, offering a potential tool to improve intraoperative diagnostic reliability and support decision-making for challenging tumor types such as extramammary Paget’s disease.

Related

  1. Multiple R1 margins are associated with shorter overall survival: national register-based study on pancreatic cancer
    BJS openPeer-reviewedCC BYFull textSurgery & anaesthesiaEurope PMC
  2. Non-Inferiority Margins in Randomized Controlled Trials in Abdominal Surgery- a Systematic Review
    C. Leonhardt, D. Birrer, M. F. Stauffer +5
    medRxivPreprintAll rights reservedFull textSurgery & anaesthesiamedRxiv
  3. Phyllodes Tumors of the Breast in 2026: A Contemporary Narrative Review of Diagnosis, Surgical Margins, Prognosis, and Emerging Therapies
    George Mpourazanis, Dimitrios Alefragkis, Angeliki Siafaka +11
    CureusPeer-reviewedCC BYFull textGeneral medicineEurope PMC
  4. Rethinking Surgical Margins in Colorectal Liver Metastases: Vessel-Skeletonized Parenchymal-Sparing Hepatectomy (VESPAH) and Treatment Continuity
    Yuzo Umeda, Tomokazu Fuji, Kosei Takagi +5
    CancersPeer-reviewedCC BY 4.0OncologyCrossref
  5. The impact of histological subtype on positive surgical margins in patients with renal cell carcinoma undergoing robot-assisted partial nephrectomy: a multicenter retrospective cohort study using propensity score matching (the MSUG 94 group)
    Suguru Oka, Koji Iinuma, Tomoki Taniguchi +11
    Scientific ReportsPeer-reviewedCC BYFull textGeneral medicineDOAJ

Cite

BibTeXRIS
BibTeX
@article{lee2026translation,
  title = {{Translation of frozen sections into FFPE images for skin cancer resection margins using generative AI}},
  author = {Sang-Hoon Lee and Yu-Sung Chu and Yosep Chong and Seong Min Hong and Eung Ho Choi and Seung-Phil Hong and Taeyeong Kim and Sejung Yang and Minseob Eom},
  journal = {npj Digital Medicine},
  year = {2026},
  month = jun,
  doi = {10.1038/s41746-026-02939-1},
  url = {https://doi.org/10.1038/s41746-026-02939-1},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Translation of frozen sections into FFPE images for skin cancer resection margins using generative AI
AU  - Sang-Hoon Lee
AU  - Yu-Sung Chu
AU  - Yosep Chong
AU  - Seong Min Hong
AU  - Eung Ho Choi
AU  - Seung-Phil Hong
AU  - Taeyeong Kim
AU  - Sejung Yang
AU  - Minseob Eom
PY  - 2026
DA  - 2026/06/01
JO  - npj Digital Medicine
DO  - 10.1038/s41746-026-02939-1
UR  - https://doi.org/10.1038/s41746-026-02939-1
AB  - Frozen section (Frozen) analysis is essential for intraoperative margin assessment in skin cancer surgery, but assessment is limited by freezing artifacts that obscure cellular detail. Formalin-fixed paraffin-embedded (FFPE) histopathology, the gold standard, provides higher quality but requires >24 h for processing. Herein, we developed and validated generative AI models to translate Frozen images into AI-generated FFPE images (GenFFPE) using 2594 slides from 283 cases across five major skin cancer types. Four unpaired image-to-image models (CycleGAN, CUT, AIFFPE, and SANTA) were trained and compared using quantitative metrics and expert rankings; CUT demonstrated the best overall fidelity. External validation and a visual Turing test (accuracy 60.2%) confirmed image realism. Among 55 discrepant cases, GenFFPE-based reassessment increased diagnostic concordance by 53.3%. AI-based Frozen-to-FFPE translation is, thus, feasible and clinically meaningful, offering a potential tool to improve intraoperative diagnostic reliability and support decision-making for challenging tumor types such as extramammary Paget’s disease.
SN  - 2398-6352
ER  -