Summary of Deep Evidential Learning For Radiotherapy Dose Prediction, by Hai Siong Tan et al.
Deep Evidential Learning for Radiotherapy Dose Prediction
by Hai Siong Tan, Kuancheng Wang, Rafe Mcbeth
First submitted to arxiv on: 26 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV); Medical Physics (physics.med-ph)
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 The proposed Deep Evidential Learning framework is applied to radiotherapy dose prediction, leveraging medical images from the Open Knowledge-Based Planning Challenge dataset. The model yields uncertainty estimates that correlate with prediction errors during network training, achieved by reformulating the loss function for stability. The results show that epistemic uncertainty is highly correlated with prediction errors, outperforming Monte-Carlo Dropout and Deep Ensemble methods in terms of association indices. Additionally, aleatoric uncertainty demonstrates a significant shift in its distribution when Gaussian noise is added to CT intensity, consistent with its interpretation as reflecting data noise. Overall, the framework exhibits statistical robustness in deep-learning models for radiotherapy dose prediction. |
Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper uses a new way to predict radiation doses based on medical images. They created a special learning system called Deep Evidential Learning that can estimate how certain it is about its predictions. The results show that this method is better than others at predicting radiation doses and understanding the uncertainty of those predictions. This could be important for making accurate decisions in cancer treatment. |
Keywords
» Artificial intelligence » Deep learning » Dropout » Loss function