Biology subjects

Makaroff, S. N.

Publications and source records attributed to Makaroff, S. N..

10 recordsLinked to original sources

bioRxiv2025

Construction of Accurate Turtle Cerebellum Models and their use in Simulating Neuromodulation using the BEM-FMM

Digital representations of Purkinje cells from turtle cerebellum, which have long been to study cellular transmission dynamics and neuroplasticity, are necessary to conduct simulations of neuromodulation methods, including Transcranial Magnetic Stimulation (TMS). Methods: Data collected using microCT with 0.6 um isometric resolution are processed, segmented, and arranged to enable high resolution simulation via the Boundary Element Fast Multipole Method. Results: Fully computer aided design compatible, manifold triangular surface meshes were generated from microCT data and used to simulate cellular electromagnetic response to TMS. Conclusion: Accurate Purkinje cells may be readily generated and manipulated to create representative models of turtle cerebellum. Significance: The techniques employed herein may be used to generate accurate surface mesh of Purkinje cells which can enable computational electromagnetics code verification and serve as the basis for skeletal models used for bidomain biophysical simulations in tools such as NEURON to study neurostimulation dynamics and neuromodulation.

neuroscience↗

bioRxiv2025

High-Definition MEG Source Estimation using the Reciprocal Boundary Element Fast Multipole Method

Magnetoencephalographic (MEG) source estimation relies on the computation of the gain (lead-field) matrix, which embodies the linear relationship between the amplitudes of the sources and the recorded signals. However, with a realistic forward model, the calculation of the gain matrix in a "direct" fashion is a computationally expensive task because the number of dipolar sources in standard MEG pipelines is often limited to [~]10,000. We propose a fast approach based on the reciprocal relationship between MEG and transcranial magnetic stimulation (TMS). This approach couples naturally with the charge-based boundary element fast multipole method (BEM-FMM), which allows us to efficiently generate gain matrices for high-resolution multi-layer non-nested meshes involving source spaces of up to a [~]1 million dipoles. We evaluate our approach by performing MEG source reconstruction against simulated data (at varying noise levels) obtained from the direct computation of MEG readings from 2000 different dipole positions over the cortical surface of 5 healthy subjects. Additionally, we test our methods with real MEG data from evoked somatosensory fields by right-hand median nerve stimulation in these same 5 subjects. We compare our experimental source reconstruction results against the standard MNE-Python source reconstruction pipeline.

neuroscience↗

bioRxiv2024

Statistical method accounts for microscopic electric field distortions around neurons when simulating activation thresholds

IntroductionNotwithstanding advances in computational models of neuromodulation, there are mismatches between simulated and experimental activation thresholds. Transcranial Magnetic Stimulation (TMS) of the primary motor cortex generates motor evoked potentials (MEPs). At the threshold of MEP generation, whole-head models predict macroscopic (at millimeter scale) electric fields (50-70 V/m) which are considerably below conventionally simulated cortical neuron thresholds (175-350 V/m). MethodsWe hypothesize that this apparent contradiction is in part a consequence of electrical field warping by brain microstructure. Classical neuronal models ignore the physical presence of neighboring neurons and microstructure and assume that the macroscopic field directly acts on the neurons. In previous work, we performed advanced numerical calculations considering realistic microscopic compartments (e.g., cells, blood vessels), resulting in locally inhomogeneous (micrometer scale) electric field and altered neuronal activation thresholds. Here we combine detailed neural threshold simulations under homogeneous field assumptions with microscopic field calculations, leveraging a novel statistical approach. ResultsWe show that, provided brain-region specific microstructure metrics, a single statistically derived scaling factor between microscopic and macroscopic electric fields can be applied in predicting neuronal thresholds. For the cortical sample considered, the statistical methods match TMS experimental thresholds. ConclusionsOur approach can be broadly applied to neuromodulation models, where fully coupled microstructure scale simulations may not be computationally tractable.

neuroscience↗

bioRxiv2024

Fast and Accurate EEG/MEG BEM-Based Forward Problem Solution for High-Resolution Head Models

A BEM (boundary element method) based approach is developed to accurately solve an EEG/MEG forward problem for a modern high-resolution head model in approximately 60 seconds using a common workstation. The method utilizes a charge-based BEM with fast multipole acceleration (BEM-FMM) and a "smart" mesh pre-refinement (called b-refinement) close to the singular source(s). No costly matrix-filling or direct solution steps typical for the standard BEM are required; the method generates on-skin voltages as well as MEG magnetic fields for high-resolution head models in approximately 60 seconds after initial model assembly. The method is verified both theoretically and experimentally.

neuroscience↗

bioRxiv2024

Accuracy of dipole source reconstruction in the 3-layerBEM model against the 5-layer BEM-FMM model

ObjectiveTo compare cortical dipole fitting spatial accuracy between the widely used yet highly simplified 3-layer and modern more realistic 5-layer BEM-FMM models with and without adaptive mesh refinement (AMR) methods. MethodsWe generate simulated noiseless 256-channel EEG data from 5-layer (7-compartment) meshes of 15 subjects from the Connectome Young Adult dataset. For each subject, we test four dipole positions, three sets of conductivity values, and two types of head segmentation. We use the boundary element method (BEM) with fast multipole method (FMM) acceleration, with or without (AMR), for forward modeling. Dipole fitting is carried out with the FieldTrip MATLAB toolbox. ResultsThe average position error (across all tested dipoles, subjects, and models) is [~]4 mm, with a standard deviation of [~]2 mm. The orientation error is [~]20{degrees} on average, with a standard deviation of [~]15{degrees}. Without AMR, the numerical inaccuracies produce a larger disagreement between the 3- and 5-layer models, with an average position error of [~]8 mm (6 mm standard deviation), and an orientation error of 28{degrees} (28{degrees} standard deviation). ConclusionsThe low-resolution 3-layer models provide excellent accuracy in dipole localization. On the other hand, dipole orientation is retrieved less accurately. Therefore, certain applications may require more realistic models for practical source reconstruction. AMR is a critical component for improving the accuracy of forward EEG computations using a high-resolution 5-layer volume conduction model. SignificanceImproving EEG source reconstruction accuracy is important for several clinical applications, including epilepsy and other seizure-inducing conditions.

bioinformatics↗

bioRxiv2024

Electromagnetic Modeling within a Microscopically Realistic Brain - Implications for Brain Stimulation

Modeling brain stimulation at the microscopic scale may reveal new paradigms for various stimulation modalities. We present the largest map to date of extracellular electric field distributions within a layer L2/L3 mouse primary visual cortex brain sample. This was enabled by the automated analysis of serial section electron microscopy images with improved handling of image defects, covering a volume of 250 x 140 x 90 m3. The map was obtained by applying a uniform brain stimulation electric field at three different polarizations and accurately computing microscopic field perturbations using the boundary element fast multipole method. We used the map to identify the effect of microscopic field perturbations on the activation thresholds of individual neurons. Previous relevant studies modeled a macroscopically homogeneous cortical volume. Our result shows that the microscopic field perturbations - an electric field spatial noise with a mean value of zero - only modestly influence the macroscopically predicted stimulation field strengths necessary for neuronal activation. The thresholds do not change by more than 10% on average. Under the stated limitations and assumptions of our method, this result justifies the conventional theory of "invisible" neurons embedded in a macroscopic brain model for transcranial magnetic and transcranial electrical stimulation. However, our result is solely sample-specific and largely neglects the effect of the microcapillary network. Furthermore, we only considered the uniform impressed field and a single- pulse stimulation time course. Significance statementThis study is arguably the first attempt to model brain stimulation at the microscopic scale, enabled by automated analysis of modern scanning electron microscopy images of the brain. It concentrates on modeling microscopic perturbations of the extracellular electric field caused by the physical cell structure and is applicable to any type of brain stimulation. Data availability statementPost-processed cell CAD models (383, stl format), microcapillary CAD models (34, stl format), post-processed neuron morphologies (267, swc format), extracellular electric field and potential distributions at different polarizations (267x3, MATLAB format), *.ses projects files for biophysical modeling with Neuron software (267x2, Neuron format), and computed neuron activating thresholds at different conditions (267x8, Excel tables, without the sample polarization correction from Section 2.8) are made available online through BossDB, a volumetric open-source database for 3D and 4D neuroscience data.

neuroscience↗

bioRxiv2023

Software Package for Transcranial Magnetic Stimulation Coil and Coil Array Analysis and Design

ObjectiveThis study aims to describe a MATLAB software package for transcranial magnetic stimulation (TMS) coil analysis and design. ApproachElectric and magnetic fields of the coils as well as their self- and mutual (for coil arrays) inductances are computed, with or without a magnetic core. Solid and stranded (Litz wire) conductors are also taken into consideration. The starting point is the centerline of a coil conductor(s), which is a 3D curve defined by the user. Then, a wire mesh and a computer aided design (CAD) mesh for the volume conductor of a given cross-section (circular, elliptical, or rectangular) are automatically generated. Self- and mutual inductances of the coil(s) are computed. Given the conductor current and its time derivative, electric and magnetic fields of the coil(s) are determined anywhere in space. Computations are performed with the fast multipole method (FMM), which is the most efficient way to evaluate the fields of many elementary current elements (current dipoles) comprising the current carrying conductor at a large number of observation points. This is the major underlying mathematical operation behind both inductance and field calculations. Main ResultsThe wire-based approach enables precise replication of even the most complex physical conductor geometries, while the FMM acceleration quickly evaluates large quantities of elementary current filaments. Agreement to within 0.74% was obtained between the inductances computed by the FMM method and ANSYS Maxwell 3D for the same coil model. Although not provided in this study, it is possible to evaluate non-linear magnetic cores in addition to the linear core exemplified. An experimental comparison was carried out against a physical MagVenture C-B60 coil; the measured and simulated inductances differed by only 1.25%, and nearly perfect correlation was found between the measured and computed E-field values at each observation point. SignificanceThe developed software package is applicable to any quasistatic inductor design, not necessarily to the TMS coils only.

neuroscience↗

bioRxiv2023

An Adaptive H-Refinement Method for the Boundary Element Fast Multipole Method for Quasi-static Electromagnetic Modeling

ObjectiveIn our recent work pertinent to modeling of brain stimulation and neurophysiological recordings, substantial modeling errors in the computed electric field and potential have sometimes been observed for standard multi-compartment head models. The goal of this study is to quantify those errors and, further, eliminate them through an adaptive mesh refinement (AMR) algorithm. The study concentrates on transcranial magnetic stimulation (TMS), transcranial electrical stimulation (TES), and electroencephalography (EEG) forward problems. ApproachWe propose, describe, and systematically investigate an AMR method using the Boundary Element Method with Fast Multipole Acceleration (BEM-FMM) as the base numerical solver. The goal is to efficiently allocate additional unknowns to critical areas of the model, where they will best improve solution accuracy. The implemented AMR methods accuracy improvement is measured on head models constructed from 16 Human Connectome Project subjects under problem classes of TES, TMS, and EEG. Errors are computed between three solutions: an initial non-adaptive solution, a solution found after applying AMR with a conservative refinement rate, and a "silver-standard" solution found by subsequent 4:1 global refinement of the adaptively-refined model. Main ResultsExcellent agreement is shown between the adaptively-refined and silver-standard solutions for standard head models. AMR is found to be vital for accurate modeling of TES and EEG forward problems for standard models: an increase of less than 25% (on average) in number of mesh elements for these problems, efficiently allocated by AMR, exposes electric field/potential errors exceeding 60% (on average) in the solution for the unrefined models. SignificanceThis error has especially important implications for TES dosing prediction - where the stimula t ion strength plays a central role - and for EEG lead fields. Though the specific form of the AMR method described here is implemented for the BEM-FMM, we expect that AMR is applicable and even required for accurate electromagnetic simulations by other numerical modeling packages as well.

neuroscience↗

bioRxiv2023

An In-Silico Testbed for Fast and Accurate MR Labeling of Orthopaedic Implants

AO_SCPLOWBSTRACTC_SCPLOWOne limitation on the ability to monitor health in older adults using Magnetic Resonance (MR) imaging is the presence of implants, where the prevalence of implantable devices (orthopedic, cardiac, neuromodulation) increases in the population, as does the pervasiveness of conditions requiring MRI studies for diagnosis (musculoskeletal diseases, infections, or cancer). The present study describes a novel multiphysics implant modeling testbed using the following approaches with two examples: - an in-silico human model based on the widely available Visible Human Project (VHP) cryo-section dataset; - a finite element method (FEM) modeling software workbench from Ansys (Electronics Desktop/Mechanical) to model MR radio frequency (RF) coils and the temperature rise modeling in heterogeneous media. The in-silico VHP Female model (250 parts with an additional 40 components specifically characterizing embedded implants and resultant surrounding tissues) corresponds to a 60-year-old female with a body mass index (BMI) of 36. The testbed includes the FEM-compatible in-silico human model, an implant embedding procedure, a generic parameterizable MRI RF birdcage two-port coil model, a workflow for computing heat sources on the implant surface and in adjacent tissues, and a thermal FEM solver directly linked to the MR coil simulator to determine implant heating based on an MR imaging study protocol. The primary target is MR labeling of large orthopaedic implants. The testbed has very recently been approved by the US Food and Drug Administration (FDA) as a medical device development tool (MDDT) for 1.5 T orthopaedic implant examinations.

biophysics↗

bioRxiv2022

High-Resolution EEG Source Reconstruction with Boundary Element Fast Multipole Method Using Reciprocity Principle and TES Forward Model Matrix

BackgroundAccurate high-resolution EEG source reconstruction (localization) is important for several tasks, including rigorous and rapid mental health screening. ObjectiveThe present study has developed, validated, and applied a new source localization algorithm utilizing a charge-based boundary element fast multipole method (BEM-FMM) coupled with the Helmholtz reciprocity principle and the transcranial electrical stimulation (TES) forward solution. MethodsThe unknown cortical dipole density is reconstructed over the entire cortical surface by expanding into global basis functions in the form of cortical fields of active TES electrode pairs. These pairs are constructed from the reading electrodes. An analog of the minimum norm estimation (MNE) equation is obtained after substituting this expansion into the reciprocity principle written in terms of measured electrode voltages. Delaunay (geometrically balanced) triangulation of the electrode cap is introduced first. Basis functions for all electrode pairs connected by the edges of a triangular mesh are precomputed and stored in memory. A smaller set of independent basis functions is then selected and employed at every time instant. This set is based on the highest voltage differences measured. ResultsThe method is validated against the classic, yet challenging problem of median nerve stimulation and the tangential cortical sources located at the posterior wall of the central sulcus for an N20/P20 peak (2 scanned subjects). The method is further applied to perform source reconstruction of synthesized tangential cortical sources located at the posterior wall of the central sulcus (12 different subjects). In the second case, an average source reconstruction error of 7 mm is reported for the best possible noiseless scenario. ConclusionsOnce static preprocessing with TES electrodes has been done (the basis functions have been computed), our method requires fractions of a second to complete the accurate high-resolution source localization.

neuroscience↗