BioAtt: Anatomical Prior Driven Low-Dose CT Denoising

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kim, Namhun, Cho, UiHyun
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909561890275328
author Kim, Namhun
Cho, UiHyun
author_facet Kim, Namhun
Cho, UiHyun
contents Deep-learning-based denoising methods have significantly improved Low-Dose CT (LDCT) image quality. However, existing models often over-smooth important anatomical details due to their purely data-driven attention mechanisms. To address this challenge, we propose a novel LDCT denoising framework, BioAtt. The key innovation lies in attending anatomical prior distributions extracted from the pretrained vision-language model BiomedCLIP. These priors guide the denoising model to focus on anatomically relevant regions to suppress noise while preserving clinically relevant structures. We highlight three main contributions: BioAtt outperforms baseline and attention-based models in SSIM, PSNR, and RMSE across multiple anatomical regions. The framework introduces a new architectural paradigm by embedding anatomic priors directly into spatial attention. Finally, BioAtt attention maps provide visual confirmation that the improvements stem from anatomical guidance rather than increased model complexity.
format Preprint
id arxiv_https___arxiv_org_abs_2504_01662
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BioAtt: Anatomical Prior Driven Low-Dose CT Denoising
Kim, Namhun
Cho, UiHyun
Image and Video Processing
Computer Vision and Pattern Recognition
Deep-learning-based denoising methods have significantly improved Low-Dose CT (LDCT) image quality. However, existing models often over-smooth important anatomical details due to their purely data-driven attention mechanisms. To address this challenge, we propose a novel LDCT denoising framework, BioAtt. The key innovation lies in attending anatomical prior distributions extracted from the pretrained vision-language model BiomedCLIP. These priors guide the denoising model to focus on anatomically relevant regions to suppress noise while preserving clinically relevant structures. We highlight three main contributions: BioAtt outperforms baseline and attention-based models in SSIM, PSNR, and RMSE across multiple anatomical regions. The framework introduces a new architectural paradigm by embedding anatomic priors directly into spatial attention. Finally, BioAtt attention maps provide visual confirmation that the improvements stem from anatomical guidance rather than increased model complexity.
title BioAtt: Anatomical Prior Driven Low-Dose CT Denoising
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.01662