Conference proceeding
Spectral U-Net: Enhancing Medical Image Segmentation via Spectral Decomposition
Proceedings (International Symposium on Biomedical Imaging), pp.1-5
04/14/2025
DOI: 10.1109/ISBI60581.2025.10981199
Abstract
This paper introduces Spectral U-Net, a novel deep learning network based on spectral decomposition, by exploiting Dual Tree Complex Wavelet Transform (DTCWT) for down-sampling and inverse Dual Tree Complex Wavelet Transform (iDTCWT) for up-sampling. We devise the corresponding Wave-Block and iWave-Block, integrated into the U-Net architecture, aiming at mitigating information loss during down-sampling and enhancing detail reconstruction during up-sampling. In the encoder, we first decompose the feature map into high and low-frequency components using DTCWT, enabling down-sampling while mitigating information loss. In the decoder, we utilize iDTCWT to reconstruct higher-resolution feature maps from down-sampled features. Evaluations on the Retina Fluid, Brain Tumor, and Liver Tumor segmentation datasets with the nn U-Net framework demonstrate the superiority of the proposed Spectral U-Net.
Details
- Title: Subtitle
- Spectral U-Net: Enhancing Medical Image Segmentation via Spectral Decomposition
- Creators
- Yaopeng Peng - University of Notre DameDanny Z. Chen - University of Notre DameMilan Sonka - University of Iowa
- Resource Type
- Conference proceeding
- Publication Details
- Proceedings (International Symposium on Biomedical Imaging), pp.1-5
- DOI
- 10.1109/ISBI60581.2025.10981199
- ISSN
- 1945-7928
- eISSN
- 1945-8452
- Publisher
- IEEE
- Number of pages
- 5
- Grant note
- R01-EB004640 / NIH (10.13039/100000002)
- Language
- English
- Date published
- 04/14/2025
- Academic Unit
- Roy J. Carver Department of Biomedical Engineering; Electrical and Computer Engineering; Radiation Oncology; Fraternal Order of Eagles Diabetes Research Center; Injury Prevention Research Center; Ophthalmology and Visual Sciences
- Record Identifier
- 9984825532302771
Metrics
39 Record Views