‘Virtual Artery’ takes us closer to improved prediction of treatment side-effects
Researchers have developed what they refer to as a ‘Virtual Artery’, a multiscale model incorporating data from physics, chemistry and biology. The team has managed to replicate single cell activity in the arterial wall (endothelial cells, smooth muscle cells) and in the flowing blood (red blood cells, platelets, white blood cells). They were investigating tunica media (the middle layer of the artery made up of muscle and tissue), which in the model was composed of spherical cells forming a hexagonal close-packed lattice, designed to mimic the mechanics of quiescent smooth muscle cells (SMCs). Treatment for stenosis, when a body channel abnormally constricts, often necessitates using balloon angioplasty and stenting. This in turn can result in so-called ‘in-stent restenosis’, SMC growth in a coronary artery. The research team set out to develop a means to statistically predict the likelihood of someone developing this dangerous post-treatment side effect, which can often result in further remedial surgery. A ‘Virtual Artery’ modelling individual cells and their interactions The research, partly funded by the EU, was recently outlined in a paper published in ‘Royal Society Publishing’. It explains how, after previous applications of the model in two-dimensional simulations, the three-dimensional version has now been validated against data gleaned from tests performed on dissected strips of tunica media. The team’s modelling system treated individual SMCs as entities which interact with each other, while also catering for the fact that individual cells undergo their own cycles. In this way they were able to represent more fully the mechanical and biochemical environment of cells. The modelled stretching tests were compared to those conducted in vitro, where dissected strips of tunica media were stretched in longitudinal and circumferential directions. High performance computing is the game-changer Using this modelling approach, the team were also able to study the transport of platelets in aneurisms, by simulating the mechanical properties of single red blood cells and platelets and coupling them with flow of blood plasma. They are currently working to further integrate biology and chemistry inputs, which will enable them to capture the processes linked to thrombosis. The applicability of the methodology to various biomedical processes reflects the aims of the EU-funded projects, COMPBIOMED and COMPAT, whose work contributed to these published results. The COMPBIOMED (A Centre of Excellence in Computational Biomedicine) project was set up specifically to tap the increasing power of high performance computers, to extend modelling accuracy towards generating a better understanding of cardiovascular, molecularly-based and neuro-musculoskeletal medicine. The ultimate aim is to develop an automated workflow whereby data taken from an individual patient can be input and processed to generate predicted health outcomes, offering the prospect of a more personalised medicine. One of the central challenges in building these predictive models is the complexity involved in replicating the variables inherent in multiscale, physical processes, which span time and location. The COMPAT (Computing Patterns for High Performance Multiscale Computing) project contributed knowledge gleaned from its reusable High Performance Multiscale Computing algorithms, scalable to so-called exascale (at least a billion billion calculations per second) systems. The project aims to develop software which will transform computer simulations into a predictive science. For more information, please visit: COMPBIOMED project website COMPAT project website
Countries
Netherlands, United Kingdom