Cover Image

Teacher-student collaborated multiple instance learning for pan-cancer PDL1 expression prediction from histopathology slides

Programmed cell death ligand 1 (PDL1), as an important biomarker, is quantified by immunohistochemistry (IHC) with few established histopathological patterns. Deep learning aids in histopathological assessment, yet heterogeneity and lacking spatially resolved annotations challenge precise analysis....

Full description

Saved in:
Bibliographic Details
Main Authors: Jin, Darui (Author) , Liang, Shangying (Author) , Shmatko, Artem (Author) , Arnold, Alexander (Author) , Horst, David (Author) , Grünewald, Thomas G. P. (Author) , Gerstung, Moritz (Author) , Bai, Xiangzhi (Author)
Format: Article (Journal)
Language:English
Published: 09 April 2024
In: Nature Communications
Year: 2024, Volume: 15, Pages: 1-14
ISSN:2041-1723
DOI:10.1038/s41467-024-46764-0
Online Access:Verlag, kostenfrei, Volltext: https://doi.org/10.1038/s41467-024-46764-0
Verlag, kostenfrei, Volltext: https://www.nature.com/articles/s41467-024-46764-0
Get full text
Author Notes:Darui Jin, Shangying Liang, Artem Shmatko, Alexander Arnold, David Horst, Thomas G. P. Grünewald, Moritz Gerstung & Xiangzhi Bai
Description
Summary:Programmed cell death ligand 1 (PDL1), as an important biomarker, is quantified by immunohistochemistry (IHC) with few established histopathological patterns. Deep learning aids in histopathological assessment, yet heterogeneity and lacking spatially resolved annotations challenge precise analysis. Here, we present a weakly supervised learning approach using bulk RNA sequencing for PDL1 expression prediction from hematoxylin and eosin (H&E) slides. Our method extends the multiple instance learning paradigm with the teacher-student framework, which assigns dynamic pseudo-labels for intra-slide heterogeneity and retrieves unlabeled instances using temporal ensemble model distillation. The approach, evaluated on 12,299 slides across 20 solid tumor types, achieves a weighted average area under the curve of 0.83 on fresh-frozen and 0.74 on formalin-fixed specimens for 9 tumors with PDL1 as an established biomarker. Our method predicts PDL1 expression patterns, validated by IHC on 20 slides, offering insights into histologies relevant to PDL1. This demonstrates the potential of deep learning in identifying diverse histological patterns for molecular changes from H&E images.
Item Description:Gesehen am 28.10.2024
Physical Description:Online Resource
ISSN:2041-1723
DOI:10.1038/s41467-024-46764-0

Similar Items