A Neural-Operator Surrogate for Platelet Deformation Across Capillary Numbers
Marco Laudato · 2025
Read A Neural-Operator Surrogate for Platelet Deformation Across Capillary Numbers by Marco Laudato (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-20341.txt.
Introduction
Marco Laudato, in a solo paper from twenty twenty-five, trained a fast stand in model for platelet shape change. Platelets are the small blood cells that help form clots. True models of clots need cell level truth at organ level cost, a gap that is hard to bridge. Laudato trained a neural operator on runs from a physics code. The runs spanned ten stiffness levels and a wide range of flow rates.
The network reads shear stress, bond stiffness, time, and start points of the membrane. It gives back the full three dimensional shape change. Errors stayed near one percent for the middle case, and speed rose by four to five orders. Readers who model clots get a near exact tool at a far lower cost.
Catalog record
| CREATOR | Marco Laudato |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-20341
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A Neural-Operator Surrogate for Platelet Deformation Across Capillary Numbers Marco Laudato Abstract Reliable multiscale models of thrombosis require platelet-scale fidelity at organ-scale cost, a gap that scientific machine learning has the potential to narrow. We train a DeepONet surrogate on platelet dynamics generated with LAMMPS for platelets spanning ten elastic moduli and capillary numbers (0.07 - 0.77). The network takes in input the wall shear stress, bond stiffness, time, and initial particle coordinates and returns the full three-dimensional deformation of the membrane. Mean-squared-error minimization with Adam and adaptive learning-rate decay yields a median displacement error below 1%, a 90th percentile below 3%, and a worst case below 4% over the entire calibrated range while accelerating computation by four to five orders of magnitude. Leave-extremes-out retraining shows graceful extrapolation: the held-out stiffest and most compliant platelets retain sub-3% median error and an 8% maximum. Error peaks coincide with transient membrane self-contact, suggesting improvements via graph neural trunks and physics-informed torque regularization. These results classify the surrogate as high-fidelity and position it for seamless coupling with continuum CFD, enabling platelet-resolved hemodynamic simulations in patient-specific geometries and opening new avenues for predictive thrombosis modeling. 1 Introduction Scientific machine learning is challenging the long-standing computational barrier that separated molecular-scale platelet physics from organ-scale blood-flow models. Physics-informed [ 1 ] and operator-learning architectures (e.g., DeepONet [ 2 ] ) can learn the nonlinear map between membrane-level mechanics and flow conditions directly from high-fidelity particle- or molecular-dynamics data and then inject that knowledge into a continuum computational fluid dynamics (CFD) solver at virtually zero computational cost. Recent platelet-specific implementations [ 3 , 4 ] show that a single DeepONet evaluation reproduces full platelet dynamics at molecular level with sub-percent error while accelerating the micro-solver by four to five orders of magnitude, potentially enabling simulations with clinically realistic platelet counts and geometries that were previously infeasible [ 5 ] . The application of such surrogates in a multi-scale model loop can be impactful for thrombosis modeling [ 6 ] , where clot initiation depends on nanometer-scale activation [ 7 ] yet is modulated by vessel-scale haemodynamics [ 8 ] . Bridging these two complementary descriptions of thrombosis is crucial [ 9 ] . Continuum-mechanics models excel at capturing haemodynamics in both idealized [ 10 , 11 , 12 ] and complex, patient-specific geometries [ 13 , 14 , 15 , 16 ] , making them indispensable for clinically relevant scenarios, while molecular-dynamics simulations deliver high-fidelity in the mechanochemical behaviour of individual blood components, including receptor binding, membrane deformation, and chemical signalling [ 17 , 18 , 19 ] . By coupling operator-learned surrogates like DeepONet into a multiscale loop, it is possible to unify the geometric flexibility of continuum approaches with the atomistic accuracy of molecular models, thereby obtaining a holistic (and computationally tractable) framework for thrombus initiation and progression. The fidelity of such multiscale framework is bounded by the surrogate’s ability to emulate the underlying molecular dynamics; rigorous benchmarking across the governing dimensionless groups is therefore mandatory. In this study we focus on the capillary number, Ca^{*} , that can be interpreted as the ratio of viscous to elastic forces [ 20 ] acting on the platelet. This number governs platelet and red blood cell deformation and margination in shear flows, as summarized in Table 1 and extensively characterized in prior work on blood-cell suspensions. Application Ca^{*} [-] Reference Red-blood-cell organization in straight micro-capillaries 0.07 [ 21 ] Platelet mechanotransduction 0.2 (derived) [ 22 ] Red-blood-cell diffusivity and margination 0.3 [ 23 ] Platelet - Red-blood-cell interaction 0.7 [ 24 ] Table 1 : Examples of dimensionless capillary numbers in micro- and nano-scale blood-flow studies, related to the range considered in this work. As Ca increases, platelets undergo tank-treading motions and flattening that enhance their lateral drift toward the vessel wall (margination), whereas at low Ca they retain a more rigid, discoidal shape with limited near-wall transport and adhesive surface area [ 25 ] . These deformation-driven changes also modulate platelet–red cell collision rates and shear-gradient diffusivity, both of which sensitively scale with Ca^{*} and determine how frequently platelets encounter and adhere to the endothelium [ 26 ] . In thrombosis, where clot initiation is based on the interplay between shear-driven transport and receptor–ligand binding kinetics, capturing the full spectrum of Ca^{*} -dependent behaviors is essential. Systematic studies of Ca^{*} therefore provide the mechanistic underpinning for training and validating neural-operator surrogates like DeepONet across the physiologically relevant deformation regimes encountered in vivo. In this study, we quantify the accuracy of a DeepONet-based surrogate in capturing the time‐dependent deformation of a platelet suspended in Couette flow. The platelet is represented by a network of roughly 20 000 particles linked by harmonic springs, with variations in the capillary number achieved by adjusting the spring constant. At each time step, the neural operator receives the current flow conditions and membrane parameters and returns the platelet’s instantaneous particle configuration. To assess the surrogate’s robustness, we systematically evaluate model performance across a range of capillary numbers, conducting both interpolation tests within the training regime and extrapolation tests beyond it. This framework allows us to determine the surrogate’s predictive limits and to establish guidelines for its deployment in multiscale thrombosis simulations. The platelet’s particle-based model and the governing haemodynamics are presented in Sec. 2 . In Sec. 3 , we detail the neural operator architecture applied in this work. Finally, Sec. 4 discusses the surrogate’s error across varying capillary numbers. 2 Particle–based platelet model The learning dataset is generated with LAMMPS [ 27 ] by coupling dissipative particle dynamics (DPD) [ 28 ] for the surrounding blood flow to a spring-network representation of the platelet, following the workflow established in our previous studies. In our initial work [ 3 ] , we sampled 101 discrete shear stresses (50–250 Pa) but recorded only the end‐state platelet shapes after one Jeffery orbit, whereas a subsequent study [ 4 ] resolved the full time evolution at a single shear stress (50 Pa). In the present work, we capture complete deformation trajectories across 10 different capillary numbers, obtained by varying the harmonic‐bond stiffness. In total, 101 distinct time instants are sampled for the 10 values of the harmonic-bond elastic constant K (Table 2 ). The resulting dimensional capillary number Ca^{*}=\frac{\mu\,\dot{\gamma}\,a}{G_{s}} (1) covers the physiologically relevant range 0.07<Ca^{*}<0.7 , where \mu is the blood viscosity, \dot{\gamma} the imposed shear rate, a the platelet’s characteristic radius, and G_{s} the in-plane surface shear modulus linked to K . Shear stress \sigma [Pa] Bond constant K [N m -1 ] Resulting Ca^{*} 50 K=\{0.0003-0.0030\} 0.07–0.70 Table 2 : Parameter sweep defining the capillary-number dataset. The bond constant K is varied from 0.0003 to 0.0030 N m -1 in steps of 0.0003 N m -1 . 2.1 Simulation domain and boundary conditions The computational box measures 16\times 16\times 8~$\mathrm{\SIUnitSymbolMicro m}$ (Fig. 1 ). Opposing translational velocities U are prescribed at the top and bottom walls to realize a Couette flow; the required velocity for a target shear stress \sigma follows the linear relation U\;=\;\frac{\sigma\,L}{2\mu}, (2) with L=16~$\mathrm{\SIUnitSymbolMicro m}$ the wall spacing. Periodic boundary conditions are enforced in the remaining directions. The time step \Delta t=2.4\times 10^{-10}\,$\mathrm{s}$ ensures numerical stability across all