Overview

Distributed Simulation of Integrated Human Function

Principal Investigator:
James E. Coolahan, Ph.D.

Organization:
Johns Hopkins University Applied Physics Laboratory

NASA Taskbook Entry

---

Technical Summary

Specific Aims:

1. To develop, at JHU, an experimentally-based computational model of the human ventricular myocyte using cells isolated from tissue biopsies performed in patients; and to develop a finite-element model of the geometry and fiber structure of the human heart using diffusion-tensor imaging data to be collected at JHU, fit by a finite-element model to be developed at the University of California, San Diego (UCSD).
2. To develop a distributed simulation of human cardiac function, incorporating a simulation of the human cardiac ventricular cell resident at JHU based on the model discussed above and a simulation of coupled cardiac mechanical and electrical function resident at UCSD, with distributed simulation control based at JHU/APL;
3. To select other appropriate cardiovascular system models that can be represented over time using simulations, and integrate them into a distributed simulation of cardiovascular function; and
4. To select bone and muscle models that can be represented over time using simulations and integrate them into a multi-function distributed simulation representative of the full integrated human function simulations that will be needed for long-duration space flight.

Main Findings in Reporting Year and Contributions to Answering Critical Questions:
In the Winslow lab at JHU, we have been successful in obtaining from the National Disease Registry Institute (NDRI) two normal human hearts from autopsies of human subjects for which the cause of death was un-related to cardiovascular disease..(NDRI provides extensive clinical information on patient health and cause of death, which allows us to screen for hearts that are representative of the normal state.) One of these two hearts has now been imaged at a spatial resolution of 350x350x800um. The diffusion tensor magnetic resonance imaging (DTMRI) data for this heart are being analyzed, and a finite-element model of heart geometry and fiber structure is being developed. We anticipate that reconstruction data on this heart will be available shortly on the website www.ccbm.jhu.edu. Imaging of the second normal heart is planned. This data will greatly enhance the ability to model dysrhythmias as an aid in determining the associated mechanisms and susceptibilities.

At JHU/APL, we have implemented a computationally-efficient Hybrid Cellular Automata (HCA) model of cardiac electrical activation in 3D cardiac muscle with arbitrary local fiber orientations and local conductivity tensors (that incorporate fiber sheet structure). We demonstrated a model for normal cardiac muscle and also models of muscle with altered electrical properties in a human-size schematic model with 20 million HCA elements on a 250 μm grid, with a left ventricle volume of 105 ml, and a right ventricle volume of 60 ml. We were able to simulate one second of physiologic time in one hour on a 3-GHz personal computer (and in 5 minutes for a 1.1-million-element model). We created a visualization of the 3D time-varying data using a PC-based volumetric rendering technique, produced a parallelized version of the code for a massively parallel processor implementation and initiated a collaboration with the NASA Ames Research Center High-Performance Computing (HPC) center. This medium-fidelity model will permit a very time-efficient means to model dysrhythmias, as an aid in determining the associated mechanisms and susceptibilities.

In collaboration with NSBRI researchers at the University of Washington, Case Western Reserve University (CWRU), and the Massachusetts Institute of Technology (MIT), we have developed an integrated physiological simulation of the cycle ergometer exercise protocol used by U.S. astronauts since the Skylab program. We have used a building-block approach where models of key physiological functions were known to exist within the NSBRI community, we employed those models with minimum modifications; where there were gaps, we developed as simple a functional representation as possible. The integrated simulation includes a cardiovascular system simulation (based on MITs Research Cardiovascular Simulator, RCVSIM), a simulation of local blood flows, a whole-body metabolism simulation (from CWRU) with embedded skeletal muscle energetics, and a respiration simulation. To connect these various components, we have used the High Level Architecture (HLA) standard for enabling simulation interoperability - originally developed in the U.S. Department of Defense and now an IEEE standard. Each integrated simulation component (federate in HLA parlance) executes on a different personal computer, all connected using a local area network. As a part of the exercise simulation effort, we collaborated with John A. Rummel, Ph.D., Deputy Director SA/Space and Life Sciences, NASA/JSC, in recovery of the NASA/Skylab Whole-Body Algorithm (WBA) code. Dr. Rummel provided a 20-year-old magnetic tape of archived WBA FORTRAN code, which we recovered to stable disk storage and returned to Dr. Rummel. We performed a preliminary assessment of the feasibility of future incorporation of a modified version of the WBA respiratory model in our exercise federation.

As a test case, we executed the federated simulation using cycle ergometer data previously collected in a laboratory setting at CWRU. We have also compared the values of heart rate generated by the simulation with those recorded during the exercise tests at CWRU on each of two subjects. Preliminary results show differences between the simulated and measured heart rates of between 5 and 15% during the 15-minute active exercise period. The simulation provides a first step in better understanding the multi-system physiologic responses during the astronaut exercise protocol that has been in use for over 25 years, which can be built upon in developing the improved countermeasures and in understanding the multi-system impact of cardiovascular alterations.

Unique Claims of Study:
The two heart reconstruction data sets from DTMRI will be the first examples of full reconstruction of the fiber organization of the human heart ever achieved.

The 3D HCA heart model provides a means of simulating cardiac electrical activation that executes sufficiently quickly to permit studies spanning minutes of physiologic time duration for a variety of conditions representing the possible altered electrical physiology associated with prolonged exposure to microgravity. We specifically investigated abnormalities that manifest as Q-T interval prolongation on the electrocardiogram and used this model to explore mechanisms connecting QT-interval prolongation to T-Wave alternans, which in a clinical setting can be a potent predictor of risk of ventricular arrhythmia, if detected at low heart rates during exercise stress testing.

The HLA-based federated simulation of human exercise is the first known computer-based simulation of the cycle ergometer exercise protocol used by U.S. astronauts. It also represents the first demonstration of DoD-developed interoperable simulation technology applied to a significant biomedical research application.

 

---

Earth Applications

Both the DTMRI-imaged reconstructed human heart model and the 3D HCA model of cardiac electrical activation have application to studies of dysrhythmias in Earth-based applications.

The multi-system physiological simulation of human exercise can be adapted to other earth-based exercise protocols involving a work rate stimulus.

---

This project's funding ended in 2004