Benchmarking Hierarchical Image Pyramid Transformer for the classification of colon biopsies and polyps in histopathology images

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Contreras, Nohemi Sofia Leon, D'Amato, Marina, Ciompi, Francesco, Grisi, Clement, Aswolinskiy, Witali, Vatrano, Simona, Fraggetta, Filippo, Nagtegaal, Iris
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929357344210944
author Contreras, Nohemi Sofia Leon
D'Amato, Marina
Ciompi, Francesco
Grisi, Clement
Aswolinskiy, Witali
Vatrano, Simona
Fraggetta, Filippo
Nagtegaal, Iris
author_facet Contreras, Nohemi Sofia Leon
D'Amato, Marina
Ciompi, Francesco
Grisi, Clement
Aswolinskiy, Witali
Vatrano, Simona
Fraggetta, Filippo
Nagtegaal, Iris
contents Training neural networks with high-quality pixel-level annotation in histopathology whole-slide images (WSI) is an expensive process due to gigapixel resolution of WSIs. However, recent advances in self-supervised learning have shown that highly descriptive image representations can be learned without the need for annotations. We investigate the application of the recent Hierarchical Image Pyramid Transformer (HIPT) model for the specific task of classification of colorectal biopsies and polyps. After evaluating the effectiveness of TCGA-learned features in the original HIPT model, we incorporate colon biopsy image information into HIPT's pretraining using two distinct strategies: (1) fine-tuning HIPT from the existing TCGA weights and (2) pretraining HIPT from random weight initialization. We compare the performance of these pretraining regimes on two colorectal biopsy classification tasks: binary and multiclass classification.
format Preprint
id arxiv_https___arxiv_org_abs_2405_15127
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Benchmarking Hierarchical Image Pyramid Transformer for the classification of colon biopsies and polyps in histopathology images
Contreras, Nohemi Sofia Leon
D'Amato, Marina
Ciompi, Francesco
Grisi, Clement
Aswolinskiy, Witali
Vatrano, Simona
Fraggetta, Filippo
Nagtegaal, Iris
Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
Training neural networks with high-quality pixel-level annotation in histopathology whole-slide images (WSI) is an expensive process due to gigapixel resolution of WSIs. However, recent advances in self-supervised learning have shown that highly descriptive image representations can be learned without the need for annotations. We investigate the application of the recent Hierarchical Image Pyramid Transformer (HIPT) model for the specific task of classification of colorectal biopsies and polyps. After evaluating the effectiveness of TCGA-learned features in the original HIPT model, we incorporate colon biopsy image information into HIPT's pretraining using two distinct strategies: (1) fine-tuning HIPT from the existing TCGA weights and (2) pretraining HIPT from random weight initialization. We compare the performance of these pretraining regimes on two colorectal biopsy classification tasks: binary and multiclass classification.
title Benchmarking Hierarchical Image Pyramid Transformer for the classification of colon biopsies and polyps in histopathology images
topic Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2405.15127