Summary of Improving Quality Control Of Whole Slide Images by Explicit Artifact Augmentation, By Artur Jurgas et al.
Improving Quality Control of Whole Slide Images by Explicit Artifact Augmentation
by Artur Jurgas, Marek Wodzinski, Marina D’Amato, Jeroen van der Laak, Manfredo Atzori, Henning Müller
First submitted to arxiv on: 17 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes a method for augmenting whole slide images with artifacts to address the problem of artifacts in histopathology image acquisition. The goal is to develop quality control algorithms that can automatically detect and correct these artifacts, reducing the need for human intervention and re-scanning. To achieve this, the authors create an external library of artifact types and use it to generate and blend artifacts into a given histopathology dataset. The resulting augmented datasets are then used to train artifact classification methods, which show significant improvement in AUROC scores (from 0.10 to 0.01) depending on the artifact type. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Artifacts in whole slide image acquisition can be a big problem! Right now, we need people to manually check and fix these issues, which takes up a lot of time and money. To make things better, this paper suggests creating fake artifacts (called “augmented datasets”) that can help train computers to automatically detect and correct these mistakes. The idea is to take an existing library of artifact types and add them to real histopathology images, so we can teach machines how to spot and fix issues. This could make it much easier and faster for doctors and researchers to get accurate results. |
Keywords
» Artificial intelligence » Classification
