Summary of Finding Regions Of Interest in Whole Slide Images Using Multiple Instance Learning, by Martim Afonso et al.
Finding Regions of Interest in Whole Slide Images Using Multiple Instance Learning
by Martim Afonso, Praphulla M. S. Bhawsar, Monjoy Saha, Jonas S. Almeida, Arlindo L. Oliveira
First submitted to arxiv on: 1 Apr 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
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 A novel weakly supervised Multiple Instance Learning (MIL) approach is proposed for predicting cancer phenotype and identifying associated cellular morphologies in Whole Slide Images (WSI). The MIL method is applied to two prevalent cancer types, Invasive Breast Carcinoma (TCGA-BRCA) and Lung Squamous Cell Carcinoma (TCGA-LUSC), for tumor detection at low magnification levels and TP53 mutations at various levels. The results show that the proposed additive implementation of MIL matches the performance of a reference implementation with an AUC of 0.96, and is only slightly outperformed by Attention MIL with an AUC of 0.97. Notably, different AI architectures identify distinct sensitivities to morphological features (Regions of Interest, RoI) at different amplification levels, with TP53 mutation being most sensitive to features at higher applications where cellular morphology is resolved. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way to use artificial intelligence (AI) in medicine helps doctors better understand cancer cells. This method uses a type of AI called Multiple Instance Learning (MIL). MIL looks at many small parts of an image, like a microscope slide, to figure out what’s going on overall. In this case, the researchers used MIL to look at two types of breast cancer and lung cancer. They wanted to see if they could use AI to detect tumors and find specific genes that are mutated in these cancers. The results show that this method is pretty good at doing these things! It also helps doctors understand what’s going on at a cellular level, which is important for developing new treatments. |
Keywords
» Artificial intelligence » Attention » Auc » Supervised
