Mitigating Catastrophic Forgetting in the Incremental Learning of Medical Images
Sara Yavari, Jacob Furst · 2025
Read Mitigating Catastrophic Forgetting in the Incremental Learning of Medical Images by Sara Yavari, Jacob Furst (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-20033.txt.
Description
Sara Yavari and Jacob Furst tackle catastrophic forgetting when medical image models learn new tasks in sequence. A model trained on a new hospital task often loses skill on earlier tasks. Storing all past data is often impossible across health centers, which makes the problem acute in medical imaging.
Their incremental method uses knowledge distillation with images generated from past tasks to guide new training. The test bed is prostate cancer detection in MRI scans from a multi center dataset. The approach improves performance and speeds convergence compared with plain sequential training. Checks on eye scans, a pathology image set, and a standard continual learning benchmark show the method stays robust across image types.
Catalog record
| CREATOR | Sara Yavari, Jacob Furst |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-20033