# Gaussian Primitive Optimized Deformable Retinal Image Registration

arXiv: https://arxiv.org/abs/2508.16852 · HTML: https://mlsanity.com/paper/2508.16852/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 32 references that are in the map.
- First posted on arXiv 23 Aug 2025.

## Builds on (most cited first, 23 of 23)

- [SuperPoint: Self-Supervised Interest Point Detection and Description](https://mlsanity.com/paper/1712.07629/) (2017)
- [LightGlue: Local Feature Matching at Light Speed](https://mlsanity.com/paper/2306.13643/) (2023)
- [RoMa: Robust Dense Feature Matching](https://mlsanity.com/paper/2305.15404/) (2023)
- [A Deep Learning Framework for Unsupervised Affine and Deformable Image Registration](https://mlsanity.com/paper/1809.06130/) (2018)
- [XFeat: Accelerated Features for Lightweight Image Matching](https://mlsanity.com/paper/2404.19174/) (2024)
- [Unsupervised 3D End-to-End Medical Image Registration with Volume Tweening Network](https://mlsanity.com/paper/1902.05020/) (2019)
- [Closing the Gap between Deep and Conventional Image Registration using Probabilistic Dense Displacement Networks](https://mlsanity.com/paper/1907.10931/) (2019)
- [Affine Medical Image Registration with Coarse-to-Fine Vision Transformer](https://mlsanity.com/paper/2203.15216/) (2022)
- [GLAMpoints: Greedily Learned Accurate Match points](https://mlsanity.com/paper/1908.06812/) (2019)
- [Semi-Supervised Keypoint Detector and Descriptor for Retinal Image Matching](https://mlsanity.com/paper/2207.07932/) (2022)
- [Learning a Model-Driven Variational Network for Deformable Image Registration](https://mlsanity.com/paper/2105.12227/) (2021)
- [Voxelmorph++ Going beyond the cranial vault with keypoint supervision and multi-channel instance optimisation](https://mlsanity.com/paper/2203.00046/) (2022)
- [Spatially Covariant Image Registration with Text Prompts](https://mlsanity.com/paper/2311.15607/) (2023)
- [MemWarp: Discontinuity-Preserving Cardiac Registration with Memorized Anatomical Filters](https://mlsanity.com/paper/2407.08093/) (2024)
- [Embedding Gradient-based Optimization in Image Registration Networks](https://mlsanity.com/paper/2112.03915/) (2021)
- [Spatially Covariant Lesion Segmentation](https://mlsanity.com/paper/2301.07895/) (2023)
- [Recursive Deformable Image Registration Network with Mutual Attention](https://mlsanity.com/paper/2206.01863/) (2022)
- [RetinaRegNet: A Zero-Shot Approach for Retinal Image Registration](https://mlsanity.com/paper/2404.16017/) (2024)
- [Slicer Networks](https://mlsanity.com/paper/2401.09833/) (2024)
- [TaGAT: Topology-Aware Graph Attention Network For Multi-modal Retinal Image Fusion](https://mlsanity.com/paper/2407.14188/) (2024)
- [Optimal Transport-based Graph Matching for 3D retinal OCT image registration](https://mlsanity.com/paper/2203.00069/) (2022)
- [Unsupervised Deformable Image Registration with Structural Nonparametric Smoothing](https://mlsanity.com/paper/2506.10813/) (2025)
- [Superimposition of eye fundus images for longitudinal analysis from large public health databases](https://mlsanity.com/paper/1607.01971/) (2016)

## Built on by (most cited first, 1 of 1)

- [Disentangling Progress in Medical Image Registration: Beyond Trend-Driven Architectures towards Domain-Specific Strategies](https://mlsanity.com/paper/2512.01913/) (2025)

## Abstract (arXiv)

Deformable retinal image registration is notoriously difficult due to large homogeneous regions and sparse but critical vascular features, which cause limited gradient signals in standard learning-based frameworks. In this paper, we introduce Gaussian Primitive Optimization (GPO), a novel iterative framework that performs structured message passing to overcome these challenges. After an initial coarse alignment, we extract keypoints at salient anatomical structures (e.g., major vessels) to serve as a minimal set of descriptor-based control nodes (DCN). Each node is modelled as a Gaussian primitive with trainable position, displacement, and radius, thus adapting its spatial influence to local deformation scales. A K-Nearest Neighbors (KNN) Gaussian interpolation then blends and propagates displacement signals from these information-rich nodes to construct a globally coherent displacement field; focusing interpolation on the top (K) neighbors reduces computational overhead while preserving local detail. By strategically anchoring nodes in high-gradient regions, GPO ensures robust gradient flow, mitigating vanishing gradient signal in textureless areas. The framework is optimized end-to-end via a multi-term loss that enforces both keypoint consistency and intensity alignment. Experiments on the FIRE dataset show that GPO reduces the target registration error from 6.2,px to ~2.4,px and increases the AUC at 25,px from 0.770 to 0.938, substantially outperforming existing methods. The source code can be accessed via https://github.com/xintian-99/GPOreg.