Synthetic Geometries Research Articles

Virtual Shape-Editing of Patient-Specific Vascular Models Using Regularized Kelvinlets.

Cardiovascular diseases, and the interventions performed to treat them, can lead to changes in the shape of patient vasculatures and their hemodynamics. Computational modeling and simulations of patient-specific vascular networks are increasingly used to quantify these hemodynamic changes, but they require modifying the shapes of the models. Existing methods to modify these shapes include editing 2D lumen contours prescribed along vessel centerlines and deforming meshes with geometry-based approaches. However, these methods can require extensive by-hand prescription of the desired shapes and often do not work robustly across a range of vascular anatomies. To overcome these limitations, we develop techniques to modify vascular models using physics-based principles that can automatically generate smooth deformations and readily apply them across different vascular anatomies. We adapt Regularized Kelvinlets, analytical solutions to linear elastostatics, to perform elastic shape-editing of vascular models. The Kelvinlets are packaged into three methods that allow us to artificially create aneurysms, stenoses, and tortuosity. Our methods are able to generate such geometric changes across a wide range of vascular anatomies. We demonstrate their capabilities by creating sets of aneurysms, stenoses, and tortuosities with varying shapes and sizes on multiple patient-specific models. Our Kelvinlet-based deformers allow us to edit the shape of vascular models, regardless of their anatomical locations, and parametrically vary the size of the geometric changes. These methods will enable researchers to more easily perform virtual-surgery-like deformations, computationally explore the impact of vascular shape on patient hemodynamics, and generate synthetic geometries for data-driven research.

Deep learning based assessment of hemodynamics in the coarctation of the aorta: comparison of bidirectional recurrent and convolutional neural networks.

The utilization of numerical methods, such as computational fluid dynamics (CFD), has been widely established for modeling patient-specific hemodynamics based on medical imaging data. Hemodynamics assessment plays a crucial role in treatment decisions for the coarctation of the aorta (CoA), a congenital heart disease, with the pressure drop (PD) being a crucial biomarker for CoA treatment decisions. However, implementing CFD methods in the clinical environment remains challenging due to their computational cost and the requirement for expert knowledge. This study proposes a deep learning approach to mitigate the computational need and produce fast results. Building upon a previous proof-of-concept study, we compared the effects of two different artificial neural network (ANN) architectures trained on data with different dimensionalities, both capable of predicting hemodynamic parameters in CoA patients: a one-dimensional bidirectional recurrent neural network (1D BRNN) and a three-dimensional convolutional neural network (3D CNN). The performance was evaluated by median point-wise root mean square error (RMSE) for pressures along the centerline in 18 test cases, which were not included in a training cohort. We found that the 3D CNN (median RMSE of 3.23mmHg) outperforms the 1D BRNN (median RMSE of 4.25mmHg). In contrast, the 1D BRNN is more precise in PD prediction, with a lower standard deviation of the error (±7.03mmHg) compared to the 3D CNN (±8.91mmHg). The differences between both ANNs are not statistically significant, suggesting that compressing the 3D aorta hemodynamics into a 1D centerline representation does not result in the loss of valuable information when training ANN models. Additionally, we evaluated the utility of the synthetic geometries of the aortas with CoA generated by using a statistical shape model (SSM), as well as the impact of aortic arch geometry (gothic arch shape) on the model's training. The results show that incorporating a synthetic cohort obtained through the SSM of the clinical cohort does not significantly increase the model's accuracy, indicating that the synthetic cohort generation might be oversimplified. Furthermore, our study reveals that selecting training cases based on aortic arch shape (gothic versus non-gothic) does not improve ANN performance for test cases sharing the same shape.

Generation of synthetic aortic valve stenosis geometries for in silico trials.

In silico trials are a promising way to increase the efficiency of the development, and the time to market of cardiovascular implantable devices. The development of transcatheter aortic valve implantation (TAVI) devices, could benefit from in silico trials to overcome frequently occurring complications such as paravalvular leakage and conduction problems. To be able to perform in silico TAVI trials virtual cohorts of TAVI patients are required. In a virtual cohort, individual patients are represented by computer models that usually require patient-specific aortic valve geometries. This study aimed to develop a virtual cohort generator that generates anatomically plausible, synthetic aortic valve stenosis geometries for in silico TAVI trials and allows for the selection of specific anatomical features that influence the occurrence of complications. To build the generator, a combination of non-parametrical statistical shape modeling and sampling from a copula distribution was used. The developed virtual cohort generator successfully generated synthetic aortic valve stenosis geometries that are comparable with a real cohort, and therefore, are considered as being anatomically plausible. Furthermore, we were able to select specific anatomical features with a sensitivity of around 90%. The virtual cohort generator has the potential to be used by TAVI manufacturers to test their devices. Future work will involve including calcifications to the synthetic geometries, and applying high-fidelity fluid-structure-interaction models to perform in silico trials.

Representation of Fully Three‐Dimensional Interfacial Curvature in Pore‐Network Models

AbstractQuasi two‐dimensional approximations of interfacial curvature, present in current network models of multiphase flow in porous media, are extended to three dimensions. The new expressions for threshold capillary pressure are validated and calibrated using high‐resolution direct numerical simulations on synthetic geometries. The effects of pore‐space expansion and sagittal interface curvature on displacement are quantified, and are shown to be a key step in improving the physical accuracy of network models. Finally, the calibrated network model is used to obtain predictions for relative permeability and capillary pressure in a water‐wet Bentheimer sandstone. The predictions are compared to experimental measurements, revealing that the inclusion of three‐dimensional interfacial curvature leads to more accurate predictions.

Predicting ocean-induced ice-shelf melt rates using deep learning

Abstract. Through their role in buttressing upstream ice flow, Antarctic ice shelves play an important part in regulating future sea-level change. Reduction in ice-shelf buttressing caused by increased ocean-induced melt along their undersides is now understood to be one of the key drivers of ice loss from the Antarctic ice sheet. However, despite the importance of this forcing mechanism, most ice-sheet simulations currently rely on simple melt parameterisations of this ocean-driven process since a fully coupled ice–ocean modelling framework is prohibitively computationally expensive. Here, we provide an alternative approach that is able to capture the greatly improved physical description of this process provided by large-scale ocean-circulation models over currently employed melt parameterisations but with trivial computational expense. This new method brings together deep learning and physical modelling to develop a deep neural network framework, MELTNET, that can emulate ocean model predictions of sub-ice-shelf melt rates. We train MELTNET on synthetic geometries, using the NEMO ocean model as a ground truth in lieu of observations to provide melt rates both for training and for evaluation of the performance of the trained network. We show that MELTNET can accurately predict melt rates for a wide range of complex synthetic geometries, with a normalised root mean squared error of 0.11 m yr−1 compared to the ocean model. MELTNET calculates melt rates several orders of magnitude faster than the ocean model and outperforms more traditional parameterisations for > 96 % of geometries tested. Furthermore, we find MELTNET's melt rate estimates show sensitivity to established physical relationships such as changes in thermal forcing and ice-shelf slope. This study demonstrates the potential for a deep learning framework to calculate melt rates with almost no computational expense, which could in the future be used in conjunction with an ice sheet model to provide predictions for large-scale ice sheet models.

Contribution of Shape Features to Intradiscal Pressure and Facets Contact Pressure in L4/L5 FSUs: An In-Silico Study

Finite element models (FEMs) of the spine commonly use a limited number of simplified geometries. Nevertheless, the geometric features of the spine are important in determining its FEM outcomes. The link between a spinal segment’s shape and its biomechanical response has been studied, but the co-variances of the shape features have been omitted. We used a principal component (PCA)-based statistical shape modelling (SSM) approach to investigate the contribution of shape features to the intradiscal pressure (IDP) and the facets contact pressure (FCP) in a cohort of synthetic L4/L5 functional spinal units under axial compression. We quantified the uncertainty in the FEM results, and the contribution of individual shape modes to these results. This parameterisation approach is able to capture the variability in the correlated anatomical features in a real population and sample plausible synthetic geometries. The first shape mode (phi _1) explained 22.6% of the shape variation in the subject-specific cohort used to train the SSM, and had the largest correlation with, and contribution to IDP (17%) and FCP (11%). The largest geometric variation in (phi _1) was in the annulus-nucleus ratio.

Mechanical evaluation of 3D printed biomimetic non-Euclidean saddle geometries mimicking the mantis shrimp

Engineering design has drawn inspiration from naturally occurring structures to advance manufacturing processes and products, termed biomimetics. For example, the mantis shrimp, order Stomatopoda, is capable of producing one of the fastest appendage strikes in the world with marginal musculoskeletal displacement. The extreme speed of the mantis shrimp’s raptorial appendage is due to the non-Euclidean hyperbolic paraboloid (i.e. saddle) shape within the dorsal region of the merus, which allows substantial energy storage through compression in the sagittal plane. Here, investigation of 3D printed synthetic geometries inspired by the mantis shrimp saddle geometry has revealed insights for elastic energy storage (i.e. spring-like) applications. Saddles composed of either a stiff or a flexible resin were investigated for spring response to explore the geometric effects. By modulating the saddle geometry and testing the spring response, it was found that, for the stiff resin, the spring constant was improved as the curvature of the contact and orthogonal faces were maximized and minimized, respectively. For the flexible resin, it was found that the spring constant increased by less than 250 N mm−1 as the saddle geometry changed, substantiating that the flexible component of mantis saddles does not contribute to energy storage capabilities. The geometries of two saddles from the mantis shrimp species O. scyllarus were estimated and exhibited similar trends to manufactured saddles, suggesting that modulating saddle geometry can be used for tailored energy storage moduli in spatially constrained engineering applications.

Dynamical evolution of fractal structures in star-forming regions

ABSTRACT The $\mathcal {Q}$-parameter is used extensively to quantify the spatial distributions of stars and gas in star-forming regions as well as older clusters and associations. It quantifies the amount of structure using the ratio of the average length of the minimum spanning tree, $\bar{m}$, to the average length within the complete graph, $\bar{s}$. The interpretation of the $\mathcal {Q}$-parameter often relies on comparing observed values of $\mathcal {Q}$, $\bar{m}$, and $\bar{s}$ to idealized synthetic geometries, where there is little or no match between the observed star-forming regions and the synthetic regions. We measure $\mathcal {Q}$, $\bar{m}$, and $\bar{s}$ over 10 Myr in N-body simulations, which are compared to IC 348, NGC 1333, and the ONC. For each star-forming region, we set up simulations that approximate their initial conditions for a combination of different virial ratios and fractal dimensions. We find that the dynamical evolution of idealized fractal geometries can account for the observed $\mathcal {Q}$, $\bar{m}$, and $\bar{s}$ values in nearby star-forming regions. In general, an initially fractal star-forming region will tend to evolve to become more smooth and centrally concentrated. However, we show that different initial conditions, as well as where the edge of the region is defined, can cause significant differences in the path that a star-forming region takes across the $\bar{m}{-}\bar{s}$ plot as it evolves. We caution that the observed $\mathcal {Q}$-parameter should not be directly compared to idealized geometries. Instead, it should be used to determine the degree to which a star-forming region is either spatially substructured or smooth and centrally concentrated.

PoreFlow-Net: A 3D convolutional neural network to predict fluid flow through porous media

We present the PoreFlow-Net, a 3D convolutional neural network architecture that provides fast and accurate fluid flow predictions for 3D digital rock images. We trained our network to extract spatial relationships between the porous medium morphology and the fluid velocity field. Our workflow computes simple geometrical information from 3D binary images to train a deep neural network (the PoreFlow-Net) optimized to generalize the problem of flow through porous materials. Our results show that the extracted information is sufficient to obtain accurate flow field predictions in less than a second, without performing expensive numerical simulations providing a speed-up of several orders of magnitude. We also demonstrate that our model, trained with simple synthetic geometries, is able to provide accurate results in real samples spanning granular rocks, carbonates, and slightly consolidated media from a variety of subsurface formations, which highlights the ability of the model to generalize the porous media flow problem. The workflow presented here shows the successful application of a disruptive technology (physics-based training of machine learning models) to the digital rock physics community.

DLITE Uses Cell-Cell Interface Movement to Better Infer Cell-Cell Tensions

Cell shapes and connectivities evolve over time as the colony changes shape or embryos develop. Shapes of intercellular interfaces are closely coupled with the forces resulting from actomyosin interactions, membrane tension, or cell-cell adhesions. Although it is possible to computationally infer cell-cell forces from a mechanical model of collective cell behavior, doing so for temporally evolving forces in a manner robust to digitization difficulties is challenging. Here, we introduce a method for dynamic local intercellular tension estimation (DLITE) that infers such evolution in temporal force with less sensitivity to digitization ambiguities or errors. This method builds upon previous work on single time points (cellular force-inference toolkit). We validate our method using synthetic geometries. DLITE’s inferred cell colony tension evolutions correlate better with ground truth for these synthetic geometries as compared to tension values inferred from methods that consider each time point in isolation. We introduce cell connectivity errors, angle estimate errors, connection mislocalization, and connection topological changes to synthetic data and show that DLITE has reduced sensitivity to these conditions. Finally, we apply DLITE to time series of human-induced pluripotent stem cell colonies with endogenously expressed GFP-tagged zonulae occludentes-1. We show that DLITE offers improved stability in the inference of cell-cell tensions and supports a correlation between the dynamics of cell-cell forces and colony rearrangement.

A robust virtual-camera 3D shape reconstruction of deforming bubbles/droplets with additional physical constraints

Bubbles and droplets in strong turbulence exhibit non-affine deformation, especially near a breakup event when the strong necking and interfacial instability begin to play roles. The virtual-camera (VC) method was designed and implemented to mitigate the limited-angle reconstruction problem for complex bubble/droplet geometries. The VC method incorporates the additional physical constraint of minimum surface curvature into reconstruction. This framework helps to reduce the reconstruction uncertainty for experiments studying multiphase flows that use a limited number of cameras, non-ideal camera positions, or the combination of both. The method was tested with three synthetic geometries– sphere, ellipsoid, and dumbbell–that represent the undeformed, gently-deformed, and severely-deformed bubbles. In addition, a DNS dataset of bubbles with a 2% void fraction in turbulent channel flow was also utilized to generate a synthetic dataset for tests, and the volume overestimation was found to decrease from over 20% to about 10% after implementing the VC method. Finally, the method was applied to the experimental dataset of bubble-turbulence interaction, and the result shows over 20% improvement in volume estimation. An indirect uncertainty quantification that relies on the bubble aspect ratio and orientation has also been developed to estimate the bubble reconstruction uncertainty in each frame. This allows a more selective postprocessing step to obtain the statistics of deformation and breakup based on the quality of reconstruction.

Towards pore network modelling of spontaneous imbibition: contact angle dependent invasion patterns and the occurrence of dynamic capillary barriers

Imbibition is an important process encountered in many porous media applications. At the pore scale, pore network models (PNM) are computationally efficient and can model drainage accurately. However, using PNM to model imbibition still remains a challenge due to the complexities encountered in understanding pore-scale flow phenomena related to pore body filling (PBF) and snap-off along with the relative competition between these events. In this work, we use direct numerical simulations (DNS) to revisit the basic principles of PBF in a two-dimensional synthetic pore geometry. We notice that PBF during spontaneous imbibition is dependent on several parameters such as shape of the transition zone, contact angle and the fluid properties like density. The interactions between these parameters are investigated in a quantitative manner. We demonstrate the existence of a critical contact angle \U0001d703c and a barrier contact angle \U0001d703b. \U0001d703c depends on the shape of the pore geometry, whereas \U0001d703b depends on the pore geometry, contact angle and fluid properties. For a system comprising of light fluids, \U0001d703b is only slightly larger than \U0001d703c; whereas for a system occupied by dense fluids, \U0001d703b is notably larger than \U0001d703c. The contact angle of the wetting phase \U0001d703 in relation to \U0001d703c and \U0001d703b decides if the wetting phase can imbibe a pore body. Imbibition always occurs if \U0001d703 < \U0001d703c. For \U0001d703 > \U0001d703c, we observe capillary barrier zones in which capillary forces accompany viscous forces to resist spontaneous imbibition. For this case, we observe smooth transition of the meniscus curvature while the meniscus enters and exits capillary barrier zones. For \U0001d703c ≤ \U0001d703 ≤ \U0001d703b, inertia assists the wetting phase to overcome resisting forces and imbibe the pore space. For \U0001d703 > \U0001d703b, the resisting forces dominate over inertia so that the wetting phase cannot imbibe the pore space. For the synthetic pore geometries investigated, we provide analytical and semi-analytical expressions to determine \U0001d703c and the position of capillary barrier zones respectively. The barrier contact angle \U0001d703b is computed numerically for several inertial systems and for various shapes of the synthetic pore geometry. The results of this quantitative analysis can be utilised to improve the existing pore filling rules and predictive capabilities of PNM used for two-phase flows.

Diffusion tortuosity in complex porous media from pore-scale numerical simulations

Diffusive transport is an important mechanism of mass transfer in gas or liquid phases confined in porous media. The continuous presence of solids in porous media causes diffusion paths to deviate from straight lines, leading to hindered diffusion characterized by the tortuosity of the media. In this study, the correlation between porosity and tortuosity in complex porous media is studied numerically using a random walk particle tracking method. Porous media studied include synthetic geometries based on three-dimensional Voronoi tessellations taking the conformations of networks of tubes, fissures, and fibers. Simulation results show that homogeneous networks of fissures are the least tortuous, homogeneous networks of tubes are intermediate, and fibrous geometries are the most tortuous. The effect of heterogeneities on the tortuosity is also investigated using both synthetic geometries and geometries of real rocks. It is found that vuggy porosity increases the tortuosity. Specifically, the larger the vug size relative to the inter-vug connections, the higher the tortuosity. Digitalized rock samples have higher tortuosities than the synthetic geometries because of the additional impact of imperfect pore connectivity.

Generalized network modeling of capillary-dominated two-phase flow.

We present a generalized network model for simulating capillary-dominated two-phase flow through porous media at the pore scale. Three-dimensional images of the pore space are discretized using a generalized network-described in a companion paper [A. Q. Raeini, B. Bijeljic, and M. J. Blunt, Phys. Rev. E 96, 013312 (2017)2470-004510.1103/PhysRevE.96.013312]-which comprises pores that are divided into smaller elements called half-throats and subsequently into corners. Half-throats define the connectivity of the network at the coarsest level, connecting each pore to half-throats of its neighboring pores from their narrower ends, while corners define the connectivity of pore crevices. The corners are discretized at different levels for accurate calculation of entry pressures, fluid volumes, and flow conductivities that are obtained using direct simulation of flow on the underlying image. This paper discusses the two-phase flow model that is used to compute the averaged flow properties of the generalized network, including relative permeability and capillary pressure. We validate the model using direct finite-volume two-phase flow simulations on synthetic geometries, and then present a comparison of the model predictions with a conventional pore-network model and experimental measurements of relative permeability in the literature.

INFLUENCE OF THE LOCAL MEAN CURVATURE ON THE ABDOMINAL AORTIC ANEURYSM STRESS DISTRIBUTION

An abdominal aortic aneurysm (AAA) is a permanent focal dilatation of the abdominal aorta of at least 1.5 times its normal diameter. Although the criterion of maximum diameter is still used in clinical practice to decide when to proceed with surgical intervention, numerical studies have demonstrated the importance of other geometric factors. In this work, the influence of the local mean curvature on AAA stress distribution has been analyzed in synthetic AAA geometries via finite element analysis. The results show a significant correlation between this geometric parameter and stress, suggesting that local mean curvature should also be considered along with the diameter criterion when making decisions about surgery.

Vertical seismic profile migration velocity analysis via residual moveout in receiver domain common image gathers

The success of imaging vertical seismic profile (VSP) data critically depends upon obtaining and using an accurate subsurface velocity model. First-arrival-time tomography can be used to derive velocities above the receivers. Reflection time tomography can be used to derive velocities below the receivers, but it requires reflection time picking and association with appropriate horizons. We described an alternative, faster way to perform velocity analysis by assessing the residual moveout after migrating walk-away VSP data. We have derived residual moveout equations for migrated common receiver gathers assuming a constant velocity model with a single horizontal reflector. The observed residual moveout in common image gathers in the receiver domain will be aligned and flat if the migration velocity is correct. If the residual moveout is not flat, by digitizing the extreme point on the unstacked, migrated traces in each common receiver gather, we could invert for the velocity of the target layer using our new residual moveout equations. We then extended this method to handle depth and laterally varying layer geometries using an iterative root-mean-square velocity scheme and a layer-stripping strategy. We combined the power of a prestack VSP depth migration using complicated velocity structures with straightforward velocity updating equations. For a range of synthetic cases and geometries, the estimated velocities were within 3% of the actual value.

Geometric and oceanographic controls on melting beneath Pine Island Glacier

Observations beneath the floating section of Pine Island Glacier have revealed the presence of a subglacial ridge which rises up to 300 m above the surrounding bathymetry. This topographic feature probably served as a steady grounding line position until sometime before the 1970s, when an ongoing phase of rapid grounding line retreat was initiated. As a result, a large ocean cavity has formed behind the ridge, strongly controlling the ocean circulation beneath the ice shelf and modulating the ocean water properties that cause ice melting in the vicinity of the grounding line. In order to understand how melt rates have changed during the various phases of cavity formation, we use a high-resolution ocean model to simulate the cavity circulation for a series of synthetic geometries. We show that the height of the ridge and the gap between the ridge and ice shelf strongly control the inflow of warm bottom waters into the cavity, and hence the melt rates. Model results suggest a rapid geometrically controlled increase of meltwater production at the onset of ice thinning, but a weak sensitivity to geometry once the gap between the ridge and ice shelf has passed a threshold value of about 200 m. This provides evidence for a new, coupled, ice-ocean feedback acting to enhance the initial retreat of an ice stream from a bedrock high. The present gap is over 200 m, and our results suggest that observed variability in melt rates is now controlled by other factors, such as the depth of the thermocline.

Pore-to-core simulations of flow with large velocities using continuum models and imaging data

We consider computational modeling of flow with small and large velocities at porescale and at corescale, and we address various challenges in simulation, upscaling, and modeling. While our focus is on voxel-based data sets from real porous media imaging, our methodology is verified first on synthetic geometries, and we analyze various scaling and convergence properties. We show that the choice of a voxel-based grid and representative elementary volume size can lead up to 10–20 % difference in calculated conductivities. On the other hand, the conductivities decrease significantly with flow rates, starting in a regime usually associated with the onset of inertia effects. This is accompanied by deteriorating porescale solver performance, and we continue our experiments up until about 50 % reduction in conductivities, i.e., to Reynolds number just under 1. To account for this decrease, we propose a practical power-based fully anisotropic non-Darcy model at corescale for which we calculate the parameters by upscaling.

Reactive transport in porous media: Pore-network model approach compared to pore-scale model

Accurate determination of three macroscopic parameters governing reactive transport in porous media, namely, the apparent solute velocity, the dispersion, and the apparent reaction rate, is of key importance for predicting solute migration through reservoir aquifers. Two methods are proposed to calculate these parameters as functions of the Péclet and the Péclet-Dahmköhler numbers. In the first method called the pore-scale model (PSM), the porous medium is discretized by the level set method; the Stokes and convection-diffusion equations with reaction at the wall are solved by a finite-difference scheme. In the second method, called the pore-network model (PNM), the void space of the porous medium is represented by an idealized geometry of pore bodies joined by pore throats; the flow field is computed by solving Kirchhoff's laws and transport calculations are performed in the asymptotic regime where the solute concentration undergoes an exponential evolution with time. Two synthetic geometries of porous media are addressed by using both numerical codes. The first geometry is constructed in order to validate the hypotheses implemented in PNM. PSM is also used for a better understanding of the various reaction patterns observed in the asymptotic regime. Despite the PNM approximations, a very good agreement between the models is obtained, which shows that PNM is an accurate description of reactive transport. PNM, which can address much larger pore volumes than PSM, is used to evaluate the influence of the concentration distribution on macroscopic properties of a large irregular network reconstructed from microtomography images. The role of the dimensionless numbers and of the location and size of the largest pore bodies is highlighted.

Point set augmentation through fitting for enhanced ICP registration of point clouds in multisensor coordinate metrology

In multisensor coordinate metrology scenarios involving the fusion of homogenous data, specifically 3D point clouds like those originated by CMMs and structured light scanners, the problem of registration, i.e. the proper localization of the clouds in the same coordinate system, is of central importance. For fine registration, known variants of the Iterative Closest Point (ICP) algorithm are commonly adopted; however, no attempt seems to be done to tweak such algorithms to better suit the distinctive multisensor nature of the data. This work investigates an original approach that targets issues which are specific to multisensor coordinate metrology scenarios, such as coexistence of point sets with different densities, different spatial arrangements (e.g. sparse CMM points vs. gridded sets from light scanners), and different noise levels associated to the point sets depending on the metrological performances of the sensors involved. The proposed approach is based on combining known ICP variants with novel point set augmentation techniques, where new points are added to existing sets with the purpose of improving registration performance and robustness to measurement error. In particular, augmentation techniques based on advanced fitting solutions promote a paradigm shift for registration, which is not seen as a geometric problem consisting in moving point sets as close as possible to each other, but as a problem where it is not the original points, but the underlying geometries that must be brought together. In this work, promising combinations of ICP and point augmentation techniques are investigated through the application to virtual scenarios involving synthetic geometries and simulated measurements. Guidelines for approaching registration problems in industrial scenarios involving multisensor data fusion are also provided. Highlights? Point set registration in multisensor coordinate metrology is investigated. ? Iterative Closest Point (ICP) combined with point set augmentation is proposed. ? Point set augmentation improves ICP registration performance. ? Augmentation based on fitting makes ICP more robust to measurement error.