Fu, Yunguan;
Li, Yiwen;
Saeed, Shaheer U;
Clarkson, Matthew J;
Hu, Yipeng;
(2024)
Importance of Aligning Training Strategy with Evaluation for Diffusion Models in 3D Multiclass Segmentation.
In: Mukhopadhyay, A and Oksuz, I and Engelhardt, S and Zhu, D and Yuan, Y, (eds.)
Deep Generative Models (DGM4MICCAI 2023).
(pp. pp. 86-95).
Springer
Preview |
Text
arXiv 2303.06040v3.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Recently, denoising diffusion probabilistic models (DDPM) have been applied to image segmentation by generating segmentation masks conditioned on images, while the applications were mainly limited to 2D networks without exploiting potential benefits from the 3D formulation. In this work, we studied the DDPM-based segmentation model for 3D multiclass segmentation on two large multiclass data sets (prostate MR and abdominal CT). We observed that the difference between training and test methods led to inferior performance for existing DDPM methods. To mitigate the inconsistency, we proposed a recycling method which generated corrupted masks based on the model’s prediction at a previous time step instead of using ground truth. The proposed method achieved statistically significantly improved performance compared to existing DDPMs, independent of a number of other techniques for reducing train-test discrepancy, including performing mask prediction, using Dice loss, and reducing the number of diffusion time steps during training. The performance of diffusion models was also competitive and visually similar to non-diffusion-based U-net, within the same compute budget. The JAX-based diffusion framework has been released at https://github.com/mathpluscode/ImgX-DiffSeg.
| Type: | Proceedings paper |
|---|---|
| Title: | Importance of Aligning Training Strategy with Evaluation for Diffusion Models in 3D Multiclass Segmentation |
| Event: | 3rd Workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention (DGM4MICCAI) at the 26th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) |
| Location: | CANADA, Vancouver |
| Dates: | 8 Oct 2023 |
| ISBN-13: | 978-3-031-53766-0 |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1007/978-3-031-53767-7_9 |
| Publisher version: | https://doi.org/10.1007/978-3-031-53767-7_9 |
| Language: | English |
| Additional information: | This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. |
| Keywords: | Abdominal CT, Computer Science, Computer Science, Artificial Intelligence, Computer Science, Theory & Methods, Diffusion Model, Image Segmentation, Prostate MR, Science & Technology, Technology |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Med Phys and Biomedical Eng |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10214039 |
Archive Staff Only
 |
View Item |
