EUVIMEDEUVIMED entdecken
Vergleichsstudie

SegRenal: AI-Driven segmentation of frozen sections in transplant kidney biopsies — A comparative analysis of deep learning models

Ibrahim Yilmaz, Heba M Alazab, Fatih Doganay, Bryan Dangott, Sam Albadri, Aziza Nassar, Fadi Salem, Zeynettin Akkus

Computer Methods and Programs in Biomedicine · 2026

DOI: 10.1016/j.cmpb.2025.109216 ↗

Kurz gefasst

Vergleichsstudie in der Zeitschrift Computer Methods and Programs in Biomedicine, 2026. Verfasst von Ibrahim Yilmaz und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Laboratory Medicine and Pathology, Mayo Clinic, Jacksonville, 32224, FL, USA. Hauptthemen laut MEDLINE: Deep Learning, Kidney Transplantation, Frozen Sections, Kidney. Die Arbeit nennt 51 Literaturangaben. Gefördert von: Department of Pathology and Laboratory Medicine, University of North Carolina School of Medicine, Mayo Clinic.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

BACKGROUND AND OBJECTIVE: Frozen section evaluation of donor kidney biopsies is vital for determining transplant suitability, yet remains challenging due to interobserver variability and freezing-related artifacts. While deep learning (DL) has been used for permanent sections, its application to frozen tissue is limited. We developed SegRenal, an artificial intelligence (AI)-based segmentation model for automated identification of glomeruli (non-sclerotic and sclerotic), arteries, and interstitial fibrosis and tubular atrophy (IFTA) in hematoxylin and eosin-stained frozen whole-slide images (WSIs). This study focuses on rigorous model adaptation, dataset development, cross-scanner performance evaluation, and integration into a clinical digital pathology workflow. METHODS: A total of 183 frozen WSIs were collected from two scanners (GT450 and Grundium) and manually annotated by expert renal pathologists. Three encoder-decoder architectures (UNet, ResNet-UNet, and DenseNet-UNet) were trained on patch-level data to compare binary and multiclass segmentation strategies. The best-performing configuration - pre-trained DenseNet with multiclass output - was further evaluated on 21 unseen WSIs scanned with both platforms. Preprocessing included downsampling, patch extraction, and annotation refinement. Performance was assessed using Dice score, precision, recall, and intraclass correlation coefficient (ICC). Bland-Altman analysis and scanner variability experiments were conducted. RESULTS: Pre-trained DenseNet multiclass segmentation yielded the best overall performance: Dice scores of 0.95 (glomeruli), 0.90 (sclerotic glomeruli), 0.80 (arteries), and 0.88 (IFTA). Recall reached 99.8% for glomeruli and 100% for arteries. Performance remained consistent across scanners. In several cases, the model detected structures initially missed by manual annotation, later confirmed by pathologists. CONCLUSIONS: SegRenal accurately segments key renal compartments in frozen biopsies and demonstrates robust cross-scanner performance. By automating tissue quantification, the model reduces variability and turnaround time, supporting fast and consistent intraoperative kidney transplant assessments.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningKidney TransplantationFrozen SectionsKidneyArtificial IntelligenceHumansBiopsyImage Processing, Computer-AssistedImage Interpretation, Computer-AssistedIntelligent Systems

Stichwörter der Autor:innen

ArteriesArtificial intelligence in pathologyDeep learningFrozen kidney biopsyGlomeruli segmentationInterstitial fibrosis and tubular atrophyKidney transplantPreimplantation assessmentSclerotic glomeruliWhole slide imaging

Autor:innen und Institutionen

  • Department of Laboratory Medicine and Pathology, Mayo Clinic, Jacksonville, 32224, FL, USA.
  • Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, 55905, MN, USA.
  • Department of Laboratory Medicine and Pathology, Mayo Clinic, Jacksonville, 32224, FL, USA. Electronic address: akkus.zeynettin@mayo.edu.

Förderung

  • Department of Pathology and Laboratory Medicine, University of North Carolina School of Medicine
  • Mayo Clinic

Zitieren

APA

Yilmaz, I., Alazab, H. M., Doganay, F., Dangott, B., Albadri, S., Nassar, A., Salem, F., & Akkus, Z. (2026). SegRenal: AI-Driven segmentation of frozen sections in transplant kidney biopsies — A comparative analysis of deep learning models. Computer Methods and Programs in Biomedicine. https://doi.org/10.1016/j.cmpb.2025.109216

Vancouver

Yilmaz I, Alazab HM, Doganay F, Dangott B, Albadri S, Nassar A, et al. SegRenal: AI-Driven segmentation of frozen sections in transplant kidney biopsies — A comparative analysis of deep learning models. Computer Methods and Programs in Biomedicine. 2026. doi:10.1016/j.cmpb.2025.109216

BibTeX
@article{Yilmaz2026,
  author = {Yilmaz, Ibrahim and Alazab, Heba M and Doganay, Fatih and Dangott, Bryan and Albadri, Sam and Nassar, Aziza and Salem, Fadi and Akkus, Zeynettin},
  title = {SegRenal: AI-Driven segmentation of frozen sections in transplant kidney biopsies — A comparative analysis of deep learning models},
  journal = {Computer Methods and Programs in Biomedicine},
  year = {2026},
  doi = {10.1016/j.cmpb.2025.109216}
}
Kostenlos starten →