Related Experiment Video
Updated: May 4, 2026
09:33
Automated Slide Scanning and Segmentation in Fluorescently-labeled Tissues Using a Widefield High-content Analysis System
Published on: May 3, 2018
8.1K
A whole-slide foundation model for digital pathology from real-world data
Hanwen Xu1,2, Naoto Usuyama1, Jaspreet Bagga1
1Microsoft Research, Redmond, WA, USA.
Nature
|May 22, 2024
Summary
Prov-GigaPath, a new whole-slide pathology foundation model, overcomes computational challenges in digital pathology. It achieves state-of-the-art results on 25 of 26 tasks by analyzing entire gigapixel slides, not just samples.
Area of Science:
- Computational pathology
- Artificial intelligence in medicine
- Digital pathology image analysis
Background:
- Digital pathology gigapixel slides present significant computational challenges.
- Previous models often used tile subsampling, missing crucial whole-slide context.
- There is a need for models capable of analyzing entire digital pathology slides.
Purpose of the Study:
- To introduce Prov-GigaPath, a novel whole-slide pathology foundation model.
- To enable comprehensive analysis of gigapixel pathology slides.
- To advance the capabilities of artificial intelligence in digital pathology.
Main Methods:
- Pretraining Prov-GigaPath on 1.3 billion image tiles from 171,189 whole slides.
- Developing GigaPath, a vision transformer architecture adapted from LongNet for gigapixel slide analysis.
- Creating a benchmark with 9 cancer subtyping and 17 pathomics tasks using Providence and TCGA data.
Main Results:
- Prov-GigaPath achieved state-of-the-art performance on 25 out of 26 benchmark tasks.
- Significant improvements were observed over existing methods in 18 tasks.
- Demonstrated potential in vision-language pretraining by incorporating pathology reports.
Conclusions:
- Prov-GigaPath is an open-weight foundation model excelling in digital pathology tasks.
- The study highlights the importance of large-scale, real-world data and whole-slide analysis.
- Prov-GigaPath sets a new standard for AI in digital pathology.
Related Concept Videos
01:21Mechanistic Models: Overview of Compartment Models
587
Mechanistic models, a category encompassing both physiological and compartmental modeling, differ from empirical models' approaches to incorporating known factors about the systems being modeled. Empirical models describe data with minimal assumptions, while mechanistic models aim to provide a robust description of available data by specifying assumptions and integrating known factors about the system. Compartmental analysis is a key example of a mechanistic model in pharmacokinetics and...
587
01:06Model Approaches for Pharmacokinetic Data: Distributed Parameter Models
333
Pharmacokinetic models are mathematical constructs that represent and predict the time course of drug concentrations in the body, providing meaningful pharmacokinetic parameters. These models are categorized into compartment, physiological, and distributed parameter models.
The distributed parameter models are specifically designed to account for variations and differences in some drug classes. This model is particularly useful for assessing regional concentrations of anticancer or...
The distributed parameter models are specifically designed to account for variations and differences in some drug classes. This model is particularly useful for assessing regional concentrations of anticancer or...
333
01:27Pharmacodynamic Models: Overview
159
Pharmacodynamic (PD) responses describe the interaction between a drug and its biological target, culminating in a physiological effect. These responses can be classified into different types: continuous variables, such as blood glucose levels; categorical outcomes, like survival rates; and time-to-event metrics, such as disease progression. Understanding and modeling PD responses are critical for optimizing drug efficacy and safety.PD models describe the relationship between drug concentration...
159