Learnable Total Variation with Lambda Mapping for Low-Dose CT Denoising

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Basak, Yusuf Talha, Unal, Mehmet Ozan, Ertas, Metin, Yildirim, Isa
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918316305547264
author Basak, Yusuf Talha
Unal, Mehmet Ozan
Ertas, Metin
Yildirim, Isa
author_facet Basak, Yusuf Talha
Unal, Mehmet Ozan
Ertas, Metin
Yildirim, Isa
contents While Total Variation (TV) excels in noise reduction and edge preservation, its reliance on a scalar regularization parameter limits adaptivity. In this study, we present a Learnable Total Variation (LTV) framework coupling an unrolled TV solver with a LambdaNet that predicts a per-pixel regularization map. The proposed framework is trained end-to-end to optimize reconstruction and regularization jointly, yielding spatially adaptive smoothing. Experiments on the DeepLesion dataset, using realistic LoDoPaB-CT simulation, show consistent gains over classical TV and FBP+U-Net, achieving up to +3.7 dB PSNR and 8% relative SSIM improvement. LTV provides an interpretable alternative to black-box CNNs for low-dose CT denoising.
format Preprint
id arxiv_https___arxiv_org_abs_2511_10500
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Learnable Total Variation with Lambda Mapping for Low-Dose CT Denoising
Basak, Yusuf Talha
Unal, Mehmet Ozan
Ertas, Metin
Yildirim, Isa
Computer Vision and Pattern Recognition
While Total Variation (TV) excels in noise reduction and edge preservation, its reliance on a scalar regularization parameter limits adaptivity. In this study, we present a Learnable Total Variation (LTV) framework coupling an unrolled TV solver with a LambdaNet that predicts a per-pixel regularization map. The proposed framework is trained end-to-end to optimize reconstruction and regularization jointly, yielding spatially adaptive smoothing. Experiments on the DeepLesion dataset, using realistic LoDoPaB-CT simulation, show consistent gains over classical TV and FBP+U-Net, achieving up to +3.7 dB PSNR and 8% relative SSIM improvement. LTV provides an interpretable alternative to black-box CNNs for low-dose CT denoising.
title Learnable Total Variation with Lambda Mapping for Low-Dose CT Denoising
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.10500