|
| EPSRC Reference: |
EP/F011628/1 |
| Title: |
Chaste - Developing software for realistic heart simulations |
| Principal Investigator: |
Dr J M Pitt-Francis |
| Other Investigators: |
|
| Researcher Co-investigator: |
|
| Project Partner: |
|
| Department: |
Computing Laboratory |
| Organisation: |
University of Oxford |
| Scheme: |
Standard Research |
| Starts: |
01 October 2007 |
Ends: |
31 March 2009 |
Value (£): |
268,208
|
| EPSRC Research Topic Classifications: |
| Developmental Biology and Physiology |
High End Computing |
|
| EPSRC Industrial Sector Classifications: |
| Healthcare |
Information Technologies |
|
| Related Grants: |
|
| Panel History: |
| Panel Date | Panel Name | Outcome |
|
16 Apr 2007
|
HPC Software Development (Science)
|
Announced
|
|
|
Summary |
Our aim is to develop the next generation of modelling software to allow in silico modelling of the whole heart to move on to a new level of sophistication and applicability. We have chosen to focus on the heart simulation work since this is the most mature area of systems-level physiological modelling, and provides extremely challenging computational problems. The proposed research will utilise recent developments, both within our group and elsewhere, in numerical analysis, software engineering, and algorithm design to develop a new approach which will yield fully tested and robust codes that are several orders of magnitude faster than existing codes. Combined with the next generation of HPC hardware, this will open up completely novel avenues of physiological research. We will test these codes in collaboration with leading theoretical modelling and experimental research groups, by simulating a range of important research questions which cannot be addressed with the current generation of modelling software. The development and optimisation of the underlying numerical algorithms and their parallel implementation will be undertaken in collaboration with Fujitsu's European Laboratory, ensuring that the resulting code base is of production standard and can be used with confidence by the heart modelling community.
|
| Final Report Summary |
Our aim is to develop the next generation of modelling software to allow in silico modelling of the whole heart to move on to a new level of sophistication and applicability. We have chosen to focus on the heart simulation work since this is the most mature area of systems-level physiological modelling, and provides extremely challenging computational problems. The proposed research will utilise recent developments, both within our group and elsewhere, in numerical analysis, software engineering, and algorithm design to develop a new approach which will yield fully tested and robust codes that are several orders of magnitude faster than existing codes. Combined with the next generation of HPC hardware, this will open up completely novel avenues of physiological research. We will test these codes in collaboration with leading theoretical modelling and experimental research groups, by simulating a range of important research questions which cannot be addressed with the current generation of modelling software. The development and optimisation of the underlying numerical algorithms and their parallel implementation will be undertaken in collaboration with Fujitsu's European Laboratory, ensuring that the resulting code base is of production standard and can be used with confidence by the heart modelling community.
At the close of this project:
The Chaste software is now released under the LGPL Open Source software licence. Release 1.0 of Chaste in late March 2009 coincided with submission of the source code to the Computer Physics Communications Program Library, submission of a companion paper to the Computer Physics Communications journal and an international workshop for users and developers. Chaste is used as a training tool in our research group, has been adopted by other researchers in the community and is the software of choice for the European Framework 7 Virtual Physiological Human "preDiCT" project.
The submission to CPC was accepted by the software library curators and by a review process. It attracted favorable anonymous peer review comments: "This manuscript describes a truly unique piece of software that has the potential to be incredibly valuable for researchers in a wide range of fields. The manuscript is well-written and both the physiological and numerical underpinnings of this software are sound. The software package is complicated but the authors provide documentation and support which allows a novice to begin to run simulations in a very short time. The great value of the Chaste software is that it is both well-engineered and open source. There are no equivalent publicly-available packages that can match the functionality and extensibility of the Chaste code."
The workshop (23-25 March 2009) was funded by this EPSRC grant. We attracted around 50 participants from UK, Europe, USA; from academia, our collaborators from Fujitsu Laboratories Europe, from pharmaceutical companies and from the US Food and Drug Administration. Participants were invited to comment on Chaste in discussions throughout the Workshop and later via a web-based questionnaire. There were many positive comments on both the software and the workshop.
Comment from an overseas academic attending the workshop:
"I personally thank to the organizers of the workshop and developers of chaste. I greatly thank the developers for releasing this software under LGPL since that will tremendously help our research community and rapid progress in the cardiac electrophysiology field."
Comment from an researcher in the industrial section attending the workshop:
"This is an incredibly valuable research tool and the workshop is one way to ensure that others use it and continue its development. The experiment in releasing it as open source is great, and I hope to be an active part of it."
|
| Further Information: |
|
| Organisation Website: |
http://www.ox.ac.uk |
|
|