u Abstract��� This paper provides a first description of a multiscale systems modeling approach applied to the congenital birth defect known as the tetralogy of Fallot. The multiscale approach adopted owes a lot to the effort of the world-wide physiome consortium and the work of research groups within the European Union on the Virtual Physiological Human. Both a spatial scale and time scale are used to establish the systems boundaries of the application. The tetralogy of Fallot includes up to four simultaneously occurring anatomic abnormalities that underpin the defect. The use of finite state machines and cellular automata pave the way to understand the processes in time and space that contribute to the defect. I. INTRODUCTION ULTISCALE systems have applications in many domains. The US Department of Energy commissioned a multiscale systems roadmap in 2004 that looked at likely developments for the following ten years or so from the perspective of mathematics. [1]. Validation of the ideas generated can now commence, as the first milestones were expected in 2010. These ideas include the need to work in interdisciplinary research teams the development of multiscale metrology to be able to demonstrate systems behaviors at different spatial scale in a concurrent fashion and development of novel performance metrics. This insight is timely, as progress has been made in all three areas, though with different degrees of success. It is now quite normal to work within a team comprising clinicians, engineers, and technologists system behavior at different spatial scales has been achieved in some bio-based applications and work remains on-going on the development of novel performance metrics. The multiscale approach adopted by us owes much to other methods, including those from systems engineering (e.g. integration technologies and information modeling) the Physiome project [2] (which has many models documented and archived at a web-site managed by researchers at the University of Auckland, New Zealand, and the University of Manuscript received April 1st, 2010. This work was supported in part by EPSRC Grant EP/E018521/1 ���Bridging the Gap��� award to R. Summers for an immersive secondment for J-M Schleich. R. Summers, T. Abdulla and R. Imms are with the Department of Electronic and Electrical Engineering, Loughborough University, Loughborough, UK. (R. Summers phone: +441509 635713 e-mail: R.Summers@lboro.ac.uk). G. Carrault and A. Hernandez are with the LTSI, University of Rennes 1, Rennes, F-35000, France L. Houyel is with the Marie-Lannelongue Hospital, Paris, F-92350, France. J-M. Schleich is with the Dept Cardiology, University Hospital of Rennes, Rennes, F-35065, France. Washington, USA [e.g. 3]) and the EU-funded Network of Excellence on the Virtual Physiological Human [4]. This paper is an initial attempt to integrate our knowledge and understanding of these approaches at a conceptual level. The common theme in the resulting models is that they describe physical phenomena at multiple scales simultaneously. The bulk of the work thus far in the public domain describes models in the multiscale spatial domain, yet multiple temporal scales also offer an opportunity to better understand physical processes. This paper sets out an initial approach to further understand a physiological system using both spatial and temporal domains. In fact the work described extends our existing knowledge of multiscale models of the heart [e.g. 5] into the developmental phases of heart anatomy and function. Whereas normal embryonic development of the heart is considered, this can be compared and contrasted to a congenital heart defect known as the tetralogy of Fallot. The clinical perspective is considered in the next section, which is followed by examples taken from multi-spatial scale models then multi-temporal scale models. Future work in the form of a roadmap is indicated before conclusions from the paper are drawn. II. CLINICAL PERSPECTIVE A. Normal Development The development of the embryonic heart commences in week 2 and is fully formed by week 8. This process is well documented [e.g. 6]. Week 2 of fetal life provides the first milestone of cardiac development as cells of the splanchnic mesoderm cluster to form the cardiogenic crescent at the cranial end of the embryo. The two primitive heart tubes join together at the median and ventral part of the embryo, thus forming the primitive heart tube. At this stage of development the first contractions occur, permitting actual blood circulation [7]. At the end of week 3 the heart tube folds into an S-shape, looping to the right (D-loop). This repositioning constitutes a crucial step towards the morphology of the heart because it brings the future heart chambers and their inflow and outflow tracts into their relative spatial positions. Throughout week 4 the ventricles grow considerably, in particular the right one by addition of myocardial cells by the second heart field. During that time, the embryo grows from 4 mm to around 50 mm in length. At the same time the apex of the ventricles balloon in sequence from the ventricular loop, leading to the development of the ventricular septum. Two processes in the development of the embryonic heart are important in the understanding of congenital heart diseases: looping and aortic wedging. Multiscale Systems Modeling of the Tetralogy of Fallot Ron Summers SMIEEE, Tariq Abdulla, Ryan Imms, Guy Carrault, Alfredo Hernandez, Lucile Houyel and Jean-Marc Schleich M 32nd Annual International Conference of the IEEE EMBS Buenos Aires, Argentina, August 31 - September 4, 2010 978-1-4244-4124-2/10/$25.00 ��2010 IEEE 752