Data Diet: Can Trimming PET/CT Datasets Enhance Lesion Segmentation?
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912129731264512 |
|---|---|
| author | Jaus, Alexander Reiß, Simon Kleesiek, Jens Stiefelhagen, Rainer |
| author_facet | Jaus, Alexander Reiß, Simon Kleesiek, Jens Stiefelhagen, Rainer |
| contents | In this work, we describe our approach to compete in the autoPET3 datacentric track. While conventional wisdom suggests that larger datasets lead to better model performance, recent studies indicate that excluding certain training samples can enhance model accuracy. We find that in the autoPETIII dataset, a model that is trained on the entire dataset exhibits undesirable characteristics by producing a large number of false positives particularly for PSMA-PETs. We counteract this by removing the easiest samples from the training dataset as measured by the model loss before retraining from scratch. Using the proposed approach we manage to drive down the false negative volume and improve upon the baseline model in both false negative volume and dice score on the preliminary test set. Code and pre-trained models are available at github.com/alexanderjaus/autopet3_datadiet. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_13548 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Data Diet: Can Trimming PET/CT Datasets Enhance Lesion Segmentation? Jaus, Alexander Reiß, Simon Kleesiek, Jens Stiefelhagen, Rainer Image and Video Processing Computer Vision and Pattern Recognition In this work, we describe our approach to compete in the autoPET3 datacentric track. While conventional wisdom suggests that larger datasets lead to better model performance, recent studies indicate that excluding certain training samples can enhance model accuracy. We find that in the autoPETIII dataset, a model that is trained on the entire dataset exhibits undesirable characteristics by producing a large number of false positives particularly for PSMA-PETs. We counteract this by removing the easiest samples from the training dataset as measured by the model loss before retraining from scratch. Using the proposed approach we manage to drive down the false negative volume and improve upon the baseline model in both false negative volume and dice score on the preliminary test set. Code and pre-trained models are available at github.com/alexanderjaus/autopet3_datadiet. |
| title | Data Diet: Can Trimming PET/CT Datasets Enhance Lesion Segmentation? |
| topic | Image and Video Processing Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2409.13548 |