Summary of Multi-scale Spatio-temporal Transformer-based Imbalanced Longitudinal Learning For Glaucoma Forecasting From Irregular Time Series Images, by Xikai Yang et al.
Multi-scale Spatio-temporal Transformer-based Imbalanced Longitudinal Learning for Glaucoma Forecasting from Irregular Time Series Images
by Xikai Yang, Jian Wu, Xi Wang, Yuchen Yuan, Ning Li Wang, Pheng-Ann Heng
First submitted to arxiv on: 21 Feb 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 This paper introduces a novel approach to glaucoma forecasting using a Multi-scale Spatio-temporal Transformer Network (MST-former). The MST-former is designed specifically for sequential image inputs, such as fundus images of the eye. It leverages multi-scale features and time attention to learn representative semantic information from historical fundus images and forecast the likelihood of glaucoma occurrence in the future. To address the challenges of irregular sampling nature and imbalanced class distribution, the paper introduces a temperature-controlled Balanced Softmax Cross-entropy loss. Experimental results on the Sequential fundus Images for Glaucoma Forecast (SIGF) dataset demonstrate the superiority of the proposed MST-former method, achieving an AUC of 98.6% for glaucoma forecasting. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research helps prevent blindness by predicting if someone will get a serious eye disease called glaucoma. The scientists created a special computer program that looks at old pictures of eyes to predict if someone might get glaucoma in the future. This is important because glaucoma can cause people to lose their sight if it’s not treated early enough. To make sure the program works well, the researchers had to solve some tricky problems, like dealing with images that were taken at different times and having a way to balance out the numbers when there are more “normal” pictures than “glaucoma” pictures. The new program is very good at predicting glaucoma and can even be used for other diseases, like Alzheimer’s. |
Keywords
» Artificial intelligence » Attention » Auc » Cross entropy » Likelihood » Softmax » Temperature » Transformer
