A Closest Point Proposal for MCMC-based Probabilistic Surface Registration

Madsen, Dennis; Morel-Forster, Andreas; Kahr, Patrick; Rahbani, Dana; Vetter, Thomas; Lüthi, Marcel

doi:10.1007/978-3-030-58520-4_17

Cited by 10 publications

(9 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A property of iteratively re-estimating the correspondence points is that the optimization can be non-monotonically decreasing, which can help avoid getting stuck in local optima, as highlighted in [2]. Besides being able to generalize existing deterministic registration algorithms, GiNGR also facilitates probabilistic registration as introduced in [29,33].…”

Section: Methodsmentioning

confidence: 99%

“…introduced in [29] and generalizes it to be able to use different correspondence and uncertainty estimation functions.…”

Section: Gingr Gingrmentioning

confidence: 99%

“…In Fig. 5, we compare ICP and CPD registrations over 5000 different reference and target femur pairs from [29] using both the deterministic and the probabilistic settings of GiNGR. For this experiment, the femurs are first aligned based on 6 landmarks, before the registration begins.…”

Section: Icp and Cpd Comparison In Gingrmentioning

confidence: 99%

“…Many of the most popular point registration methods are summarized in [30,44]. However, they do not include surface registration methods such as [1,2,7,12,14,23,29,45]. With so many different methods, it can be difficult to know the difference between all of them and therefore also which one to choose for a specific problem.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

GiNGR: Generalized Iterative Non-Rigid Point Cloud and Surface Registration Using Gaussian Process Regression

Madsen¹,

Jonathan²,

Morel-Forster³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

In this paper, we unify popular non-rigid registration methods for point sets and surfaces under our general framework, GiNGR. GiNGR builds upon Gaussian Process Morphable Models (GPMM) and hence separates modeling the deformation prior from model adaptation for registration. In addition, it provides explainable hyperparameters, multi-resolution registration, trivial inclusion of expert annotation, and the ability to use and combine analytical and statistical deformation priors. But more importantly, the reformulation allows for a direct comparison of registration methods. Instead of using a general solver in the optimization step, we show how Gaussian process regression (GPR) iteratively can warp a reference onto a target, leading to smooth deformations following the prior for any dense, sparse, or partial estimated correspondences in a principled way. We show how the popular CPD and ICP algorithms can be directly explained with GiNGR. Furthermore, we show how existing algorithms in the GiNGR framework can perform probabilistic registration to obtain a distribution of different registrations instead of a single best registration. This can be used to analyze the uncertainty e.g. when registering partial observations. GiNGR is publicly available and fully modular to allow for domain-specific prior construction.

show abstract

Section: Methodsmentioning

confidence: 99%

“…introduced in [29] and generalizes it to be able to use different correspondence and uncertainty estimation functions.…”

Section: Gingr Gingrmentioning

confidence: 99%

Section: Icp and Cpd Comparison In Gingrmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

GiNGR: Generalized Iterative Non-Rigid Point Cloud and Surface Registration Using Gaussian Process Regression

Madsen¹,

Jonathan²,

Morel-Forster³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…To demonstrate the utility of the DMFC-GPM framework in making predictions from observations of data, we adopt the Markov Chain Monte Carlo (MCMC) and Metropolis Hastings algorithms for the model fitting as in [32], [33], [34]. Markov Chain Monte Carlo methods are computational tools to perform approximate inference with intractable probabilistic models.…”

Section: Fitting the Model To Observationsmentioning

confidence: 99%

Dynamic Multi-object Gaussian Process Models

Fouefack¹,

Borotikar²,

Douglas³

et al. 2020

Medical Image Computing and Computer Assisted Intervention – MICCAI 2020

View full text Add to dashboard Cite

In model-based medical image analysis, three features of interest are the shape of structures of interest, their relative pose, and image intensity profiles representative of some physical property. Often, these are modelled separately through statistical models by decomposing the object's features into a set of basis functions through principal geodesic analysis or principal component analysis. However, analysing multiple objects in an image using multiple single object models may lead to large errors and uncertainties, especially around organ boundaries. A question that comes to mind is what kind of advantages can be gained from combining the three features of interest in the same statistical space for analysing images. This study presents a statistical modelling method for automatic learning of shape, pose and intensity features in medical images which we call the Dynamic multi feature-class Gaussian process models (DMFC-GPM). A DMFC-GPM is a Gaussian process (GP)-based model with a shared latent space that encodes linear and non-linear variation. Our method is defined in a continuous domain with a principled way to represent shape, pose and intensity feature classes in a linear space, based on deformation fields. A deformation field-based metric is adapted in the method for modelling shape and intensity feature variation as well as for comparing rigid transformations (pose). Moreover, DMFC-GPMs inherit properties intrinsic to GPs including marginalisation and regression. Furthermore, they allow for adding additional pose feature variability on top of those obtained from the image acquisition process; what we term as permutation modelling. For image analysis tasks using DMFC-GPMs, we adapt Metropolis-Hastings algorithms making the prediction of features fully probabilistic. We validate the method using controlled synthetic data and we perform experiments on bone structures from CT images of the shoulder to illustrate the efficacy of the model for pose and shape feature prediction. The model performance results suggest that this new modelling paradigm is robust, accurate, accessible, and has potential applications in a multitude of scenarios including the management of musculoskeletal disorders, clinical decision making and image processing.

show abstract