Biomedical subjects

Andrew Pullan

Publications and source records attributed to Andrew Pullan.

3 recordsLinked to original sources

PubMed2003

A deformable finite element derived finite difference method for cardiac activation problems.

We present a finite element (FE) derived finite difference (FD) technique for solving cardiac activation problems over deforming geometries using a bidomain framework. The geometry of the solution domain is defined by a FE mesh and over these FEs a high resolution FD mesh is generated. The difference points are located at regular intervals in the normalized material space within each of the FEs. The bidomain equations are then transformed to the embedded FD mesh which provides a solution space that is both regular and orthogonal. The solution points move in physical space with any deformation of the solution domain, but the equations are set up in such a way that the solution is invariant as it is constructed in material space. The derivation of this new solution technique is presented along with a series of examples that demonstrate the accuracy of this bidomain framework.

Action Potentials↗

PubMedDate not supplied

Modelling gastrointestinal bioelectric activity.

The development of an anatomically realistic biophysically based model of the human gastrointestinal (GI) tract is presented. A major objective of this work is to develop a modelling framework that can be used to integrate the physiological, anatomical and medical knowledge of the GI system. The anatomical model was developed by fitting derivative continuous meshes to digitised data taken from images of the visible man. Structural information, including fibre distributions of the smooth muscle layers and the arrangement of the networks of interstitial cells of Cajal, were incorporated using published information. A continuum modelling framework was used to simulate electrical activity from the single cell to the whole organ and body. Also computed was the external magnetic field generated from the GI electrical activity. The set of governing equations were solved using a combination of numerical techniques. Activity at the (continuum) cell level was solved using a high-resolution trilinear finite element procedure that had been defined from the previously fitted C1 continuous anatomical mesh. Multiple dipolar sources were created from the excitation waves which were embedded within a coupled C1 continuous torso model to produce both the cutaneous electrical field and the external magnetic field. Initial simulations were performed using a simplified geometry to test the implementation of the numerical solution procedure. The numerical procedures were shown to rapidly converge with mesh refinement. In the process of this testing, errors in a long standing analytic solution were identified and are corrected in Appendix B. Results of single cell activity were compared to published results illustrating that the key features of the slow wave activity were successfully replicated. Simulations using a two-dimensional slice through the gastric wall produced slow wave activity that agreed with the known frequency and propagation characteristics. Three-dimensional simulations were also performed using the full stomach mesh and results illustrated the slow wave propagation throughout the stomach musculature.

Animals↗

PubMedDate not supplied

Torso coupling techniques for the forward problem of electrocardiography.

The calculation of body surface potentials from a known cardiac source is traditionally formulated in terms of a two step process. The first step involves the generation of some form of equivalent cardiac source (typically dipole based) at a resolution significantly lower than that of a continuum cell. The second then places that source into a volume conductor within which the potential fields are calculated. This approach does not properly capture the feedback between the torso and the extracellular potential field. Presented here are the details of two new methods which calculate continuous potential fields throughout the torso that are the direct result of cardiac cellular electrical activity. These new methods are termed the Boundary Iteration Method and the Direct Assembly Method. While these two methods return essentially identical answers, there is a definite tradeoff between computational speed and memory overhead with the direct assembly method proving faster but requiring significantly more memory. Examples are given that demonstrate the convergence and accuracy of these methods in an idealized coupled torso system. These techniques are then applied to an anatomically based model of a slice through a human male torso.

Abdomen↗