PET-Adapter: Test-Time Domain Adaptation for Full and Limited-Angle PET Image Reconstruction
Rüveyda Yilmaz, Yuli Wu, Johannes Stegmaier, Volkmar Schulz
TLDR
PET-Adapter is a test-time domain adaptation framework that improves PET image reconstruction from phantom-trained models to diverse clinical data.
Key contributions
- Enables test-time adaptation of phantom-trained PET models to diverse clinical datasets without retraining.
- Introduces layer-wise low-rank anatomical conditioning for robust domain adaptation.
- Uses OSEM-based warm-starting to reduce diffusion steps from 50 to 2, boosting computational efficiency.
- Achieves superior 3D reconstruction in full and limited-angle PET, enhancing clinical feasibility.
Why it matters
This paper addresses a critical challenge in deep learning for medical imaging: generalization to unseen clinical data. By enabling test-time adaptation without retraining or ground truth, PET-Adapter makes advanced PET reconstruction models more practical and efficient for diverse clinical applications, significantly reducing computational burden.
Original Abstract
Positron Emission Tomography (PET) image reconstruction is inherently challenged by Poisson noise and physical degradation factors, which are further exacerbated in limited-angle acquisitions. While deep learning methods demonstrate promising performance, their generalization to unseen clinical data distributions remains limited without extensive retraining. We propose PET-Adapter, a test-time domain adaptation framework for generative PET reconstruction models pretrained solely on phantom data. Our method enables adaptation to clinical datasets with varying anatomies, tracers, and scanner configurations without requiring paired ground truth. PET-Adapter introduces layer-wise low-rank anatomical conditioning during adaptation and Ordered Subset Expectation Maximization-based warm-starting that initializes the generation from physics-informed reconstructions, reducing diffusion steps from 50 to 2 without compromising quality. Experiments across multiple clinical datasets demonstrate superior 3D reconstruction performance in both full-angle and limited-angle settings, highlighting the clinical feasibility and computational efficiency of the proposed approach.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.