Gaussian Splatting-Based Registration for PET Image Correction: A Proof-of-Concept Study
Bibliographic record
Abstract
Attenuation, scatter, and random coincidence corrections are essential for quantitative accuracy in positron emission tomography (PET). Conventional correction methods often rely on external anatomical imaging, which can introduce additional cost, radiation dose, and potential misalignment. In scenarios where transmission scans are unavailable, such as portable or organ-specific PET systems, alternative correction strategies are needed. We propose a novel approach based on Gaussian Splatting, a recent differentiable rendering technique from computer vision, adapted here for full PET image correction directly in the image domain. Our method represents both corrected and uncorrected PET images as sets of 2D Gaussians, each defined by position, scale, orientation, and colour. A template representation is first trained using paired corrected and uncorrected PET images, then fine-tuned on a new patient using only their uncorrected scan. This process enables direct correction of attenuation, scatter, and random coincidences without requiring additional data. We evaluate our method on six template-patient combinations using head and neck PET images. While diffusion models achieve higher accuracy, our method yields competitive results (SSIM: 0.965 ± 0.014, PSNR: 25.57 ± 1.97) with significantly lower data and computational requirements. This proof of concept highlights the potential of Gaussian Splatting as a data-efficient, flexible, and interpretable solution for PET correction, particularly in transmission-less imaging settings.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".