EUVIMEDEUVIMED entdecken
Crossref

Deep Learning‐Based Analysis of Microscopic Hyperspectral Images of HER2 ‐Stained Breast Cancer Sections for Molecular Status Prediction

Lin Yan, Yan Wan, Haonan Liu, Wenlou Liu, Yang Zhao, Hongmei Wang iD, Juanjuan Tang, Xuanyu Chen iD, Zhengxiang Han iD, Xiaobing Qin iD

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

DOI: 10.1002/jbio.70384 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Biophotonics, online erschienen am 06.11.2026. Verfasst von Lin Yan und 9 weiteren Autor:innen. Beteiligte Einrichtung: Department of Oncology, The First College of Clinical Medicine, Xuzhou Medical University, Xuzhou, China. Hauptthemen laut MEDLINE: Erb-b2 Receptor Tyrosine Kinases, Breast Neoplasms, Deep Learning, Hyperspectral Imaging. Die Arbeit nennt 20 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

HER2 assessment in breast cancer relies on immunohistochemistry (IHC) with reflex in situ hybridization for equivocal cases, while conventional bright‐field imaging provides limited quantitative optical information. We investigated microscopic hyperspectral imaging (HSI) of HER2‐stained sections for slide‐level molecular status prediction. HSI data from 43 cases (21 HER2‐positive, 22 HER2‐negative; 400–1000 nm, 128 bands) were acquired between January 2024 and August 2025. After calibration, optical‐density conversion, tissue masking, and Savitzky–Golay smoothing, we developed SpectralMix‐MIL, a weakly supervised multiple‐instance learning model combining learnable spectral projection ( K = 32) with attention pooling. On a patient‐level 7:3 train/test split, SpectralMix‐MIL achieved an AUROC of 0.923 and AUPRC of 0.910. Repeated stratified fivefold cross‐validation yielded mean AUROC 0.904 ± 0.041 and AUPRC 0.892 ± 0.045, while leave‐one‐out cross‐validation yielded AUROC 0.897. These findings support microscopic HSI for quantitative breast cancer molecular phenotyping, pending larger external validation.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Erb-b2 Receptor Tyrosine KinasesBreast NeoplasmsDeep LearningHyperspectral ImagingImage Processing, Computer-AssistedMicroscopyHumansFemaleStaining and Labeling

Stichwörter der Autor:innen

HER2breast cancerhyperspectral imagingpathologyspectral analysis

Substanzen

Erb-b2 Receptor Tyrosine KinasesERBB2 protein, human

Autor:innen und Institutionen

  • Department of Oncology, The First College of Clinical Medicine, Xuzhou Medical University, Xuzhou, China.
  • Department of Oncology, Xuzhou Medical University, Xuzhou, China.
  • Department of Oncology, The Affiliated Hospital of Xuzhou Medical University, Xuzhou, China.

ORCID-Profile: Hongmei Wang, Xuanyu Chen, Zhengxiang Han, Xiaobing Qin

Zitieren

APA

Yan, L., Wan, Y., Liu, H., Liu, W., Zhao, Y., Wang, H., Tang, J., Chen, X., Han, Z., & Qin, X. (2026). Deep Learning‐Based Analysis of Microscopic Hyperspectral Images of HER2 ‐Stained Breast Cancer Sections for Molecular Status Prediction. Journal of Biophotonics, 19(11). https://doi.org/10.1002/jbio.70384

Vancouver

Yan L, Wan Y, Liu H, Liu W, Zhao Y, Wang H, et al. Deep Learning‐Based Analysis of Microscopic Hyperspectral Images of HER2 ‐Stained Breast Cancer Sections for Molecular Status Prediction. Journal of Biophotonics. 2026;19(11). doi:10.1002/jbio.70384

BibTeX
@article{Yan2026,
  author = {Yan, Lin and Wan, Yan and Liu, Haonan and Liu, Wenlou and Zhao, Yang and Wang, Hongmei and Tang, Juanjuan and Chen, Xuanyu and Han, Zhengxiang and Qin, Xiaobing},
  title = {Deep Learning‐Based Analysis of Microscopic Hyperspectral Images of HER2 ‐Stained Breast Cancer Sections for Molecular Status Prediction},
  journal = {Journal of Biophotonics},
  year = {2026},
  volume = {19},
  number = {11},
  doi = {10.1002/jbio.70384}
}
Kostenlos starten →