Leveraging Pathology Knowledge to Enhance Visual-Language Representation Learning for Computational Pathology
Introducing structured pathology knowledge can significantly enhance visual-language representation learning for computational pathology tasks.
Feature Re-Embedding: Improving Computational Pathology Performance through Online Instance Feature Fine-Tuning
The core message of this paper is that incorporating an online instance feature re-embedding module, such as the proposed Re-embedded Regional Transformer (R2T), can significantly improve the performance of multiple instance learning (MIL) models in computational pathology tasks by enabling supervised fine-tuning of the instance features.
Generating Annotated Colorectal Cancer Tissue Images from Gland Layout
A framework that can generate realistic colorectal tissue images along with corresponding glandular masks, controlled by the input gland layout.
ReStainGAN: Generierung von in-silico IHC-Bildern für die Nukleussegmentierung
Generierung von in-silico IHC-Bildern zur Nukleussegmentierung durch ReStainGAN.
ReStainGAN: Leveraging IHC to IF Stain Domain for in-silico Data Generation
In-silico 데이터 생성을 위한 IHC에서 IF 염색 도메인 활용
MambaMIL: Computational Pathology Long Sequence Modeling with Sequence Reordering
MambaMIL incorporates Sequence Reordering Mamba to enhance long sequence modeling in computational pathology.
Generalizable Whole Slide Image Classification with Fine-Grained Visual-Semantic Interaction
Die Arbeit präsentiert das FiVE-Framework für die Klassifizierung von Whole Slide Images, das robuste Generalisierbarkeit und starke Übertragbarkeit zeigt.
PathoTune: Adapting Visual Foundation Model to Pathological Specialists
Efficiently adapting generalist foundation models to specialized pathological tasks through multi-modal prompt tuning is crucial for superior performance in computational pathology.
WEEP: Spatial Interpretation of Weakly Supervised CNN Models in Computational Pathology
Proposing the Wsi rEgion sElection aPproach (WEEP) for spatial interpretation of weakly supervised CNN models in computational pathology.
Duplex to Monoplex IHC Image Translation with Auxiliary CycleGAN Guidance
Duplex to monoplex IHC image translation is improved through the use of auxiliary CycleGAN guidance, addressing the ambiguous mapping challenge.
