U-TTT: Towards Generalizable PET Image Denoising via Test-Time Training

Hugging Face Daily Papers Papers
pet-image-denoising test-time-training deep-learning medical-imaging distribution-shift self-supervision dual-domain-adaptation

Summary

This paper introduces U-TTT, a U-shaped deep learning model with test-time training layers and dual-domain adaptation for robust PET image denoising under distribution shifts, achieving state-of-the-art performance across different dose levels and scanner types.

Existing deep learning models for Positron Emission Tomography (PET) image denoising often suffer from severe performance degradation under distribution shifts, fundamentally restricting their robust clinical deployment. This lack of generalization stems from the conventional paradigm of fixed-parameter models that cannot adapt to variations in test data (e.g., dose levels or scanner types) after training. To overcome this limitation and achieve robust generalization, we introduce U-TTT, a novel U-shaped model that integrates Test-Time Training (TTT) layers to dynamically adjust model parameters during inference through self-supervision, thereby adapting to the specific characteristics of each test instance. Furthermore, to comprehensively capture the complex degradations of 3D PET data, U-TTT features a dual-domain adaptation mechanism comprising a Spatial Test-Time Training (S-TTT) layer and a Frequency Test-Time Training (F-TTT) layer. The S-TTT layer captures and corrects spatial structural degradations, while the F-TTT layer suppresses global noise spectra and restores delicate high-frequency details. Extensive experiments demonstrate that U-TTT achieves state-of-the-art PET denoising performance and exhibits superior generalization under challenging distribution shifts, including both unseen dose levels and unseen scanners. Our code will be available at https://github.com/Yaziwel/U-TTT.
Original Article
View Cached Full Text

Cached at: 06/10/26, 09:44 AM

Paper page - U-TTT: Towards Generalizable PET Image Denoising via Test-Time Training

Source: https://huggingface.co/papers/2606.11032

Abstract

A novel U-shaped deep learning model with test-time training layers and dual-domain adaptation mechanisms achieves robust PET image denoising under distribution shifts.

Existing deep learning models forPositron Emission Tomography(PET) imagedenoisingoften suffer from severe performance degradation underdistribution shifts, fundamentally restricting their robust clinical deployment. This lack of generalization stems from the conventional paradigm offixed-parameter modelsthat cannot adapt to variations in test data (e.g., dose levels or scanner types) after training. To overcome this limitation and achieve robust generalization, we introduce U-TTT, a novelU-shaped modelthat integratesTest-Time Training(TTT) layers to dynamically adjust model parameters during inference throughself-supervision, thereby adapting to the specific characteristics of each test instance. Furthermore, to comprehensively capture the complex degradations of 3D PET data, U-TTT features adual-domain adaptationmechanism comprising aSpatial Test-Time Training(S-TTT) layer and aFrequency Test-Time Training(F-TTT) layer. The S-TTT layer captures and correctsspatial structural degradations, while the F-TTT layer suppressesglobal noise spectraand restores delicatehigh-frequency details. Extensive experiments demonstrate that U-TTT achieves state-of-the-art PETdenoisingperformance and exhibits superior generalization under challengingdistribution shifts, including both unseen dose levels and unseen scanners. Our code will be available at https://github.com/Yaziwel/U-TTT.

View arXiv pageView PDFGitHub2Add to collection

Get this paper in your agent:

hf papers read 2606\.11032

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.11032 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.11032 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.11032 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Linearizing Vision Transformer with Test-Time Training

Hugging Face Daily Papers

This paper proposes a method to convert pretrained Softmax attention models into linear-complexity Test-Time Training (TTT) architectures, achieving comparable text-to-image quality to fine-tuned Softmax models while significantly accelerating inference. The approach is validated by linearizing Stable Diffusion 3.5, resulting in SD3.5-T^5 with 1.32x speedup at 1K resolution.

Back to Repair: A Minimal Denoising Network\ for Time Series Anomaly Detection

Hugging Face Daily Papers

This paper introduces JuRe (Just Repair), a minimal denoising network for time series anomaly detection that matches or exceeds complex neural baselines on the TSB-AD and UCR benchmarks, demonstrating that a proper manifold-projection training objective is more important than architectural complexity.