EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Label‐Free Morphometric Profiling of Renal Cell Carcinoma Using Digital Holographic Microscopy and Machine Learning

Eva R Schwark, Yazan A Almahdi, Ayah A Farhat, Besa Xhabija iD

Journal of Biophotonics · 2026-11 · 19 (11)

DOI: 10.1002/jbio.70379 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Biophotonics, online erschienen am 11.11.2026. Verfasst von Eva R Schwark und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Natural Sciences, University of Michigan-Dearborn, Dearborn, Michigan, USA. Hauptthemen laut MEDLINE: Carcinoma, Renal Cell, Holography, Machine Learning, Kidney Neoplasms. Die Arbeit nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: University of Michigan-Dearborn.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Renal cell carcinoma (RCC) has a substantial mortality rate among urologic cancers today, and our current knowledge of the disease includes few distinct biomarkers, making early diagnosis particularly difficult. In this study, we utilized digital holographic microscopy (DHM) and quantitative phase imaging (QPI) to investigate trends in cell morphology and optical properties across 38 morphological and optical features in renal carcinoma cell lines UO‐31 and SN12C, and renal epithelial cell line HK2. Dimensionality reduction techniques including PCA, t‐SNE, and Spectral Embedding Analysis enabled clearer differentiation between cancerous and normal cells. Our findings reveal distinct morphological profiles, with UO‐31 and SN12C characterized by greater elongation, coarser features, and lower optical thickness compared to HK2. A Random Forest ensemble classifier achieved a cross‐validated accuracy of 99.50% and AUC values exceeding 0.999 for all three cell lines, with Irregularity and Texture Clustershade emerging as the most influential discriminatory features.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Carcinoma, Renal CellHolographyMachine LearningKidney NeoplasmsMicroscopyHumansCell Line, TumorImage Processing, Computer-AssistedQuantitative Phase Imaging

Stichwörter der Autor:innen

cell line characterizationdigital holographic microscopydimensionality reductionlabel‐free imagingmachine learningmorphometric profilingphenotypic biomarkersquantitative phase imagingrandom forest classificationrenal cell carcinoma

Autor:innen und Institutionen

  • Department of Natural Sciences, University of Michigan-Dearborn, Dearborn, Michigan, USA.

ORCID-Profile: Besa Xhabija

Förderung

  • University of Michigan-Dearborn

Zitieren

APA

Schwark, E. R., Almahdi, Y. A., Farhat, A. A., & Xhabija, B. (2026). Label‐Free Morphometric Profiling of Renal Cell Carcinoma Using Digital Holographic Microscopy and Machine Learning. Journal of Biophotonics, 19(11). https://doi.org/10.1002/jbio.70379

Vancouver

Schwark ER, Almahdi YA, Farhat AA, Xhabija B. Label‐Free Morphometric Profiling of Renal Cell Carcinoma Using Digital Holographic Microscopy and Machine Learning. Journal of Biophotonics. 2026;19(11). doi:10.1002/jbio.70379

BibTeX
@article{Schwark2026,
  author = {Schwark, Eva R and Almahdi, Yazan A and Farhat, Ayah A and Xhabija, Besa},
  title = {Label‐Free Morphometric Profiling of Renal Cell Carcinoma Using Digital Holographic Microscopy and Machine Learning},
  journal = {Journal of Biophotonics},
  year = {2026},
  volume = {19},
  number = {11},
  doi = {10.1002/jbio.70379}
}
Kostenlos starten →