A Multi-Agent Simulation of Kidney Function for Medical Education

Introduction

Renal physiology is exceptionally difficult to learn. The countercurrent exchange system of the kidney was ranked number one in basic science difficulty and number two in clinical difficulty in a 1990 survey of medical educators (Dawson-Saunders, Feltovich, Coulson, & Steward, 1990). Project TOUCH was a multiyear collaboration between the University of Hawaii and the University of New Mexico in which virtual reality (VR) applications were developed to advance medical education. One application was a 3D VR fly-through model of the kidney (Alverson & Saiki, 2006). Each kidney is composed of millions of nephrons. Our simulation shows the details of one nephron, while adjacent nephrons are shown without detail for context. The simulation prototype has subsequently been enhanced with a new user interface and gaming motifs. The new system underwent heuristic and usability tests that are described in this chapter.

The kidney's function is to eliminate water soluble wastes from the body (Banasik, 2000). We chose a multi-agent system implementation because it results in more realistic and detailed model. Agents represent the motion of molecules and fluids within a nephron’s tubules. The agents move dynamically and intelligently within their environment to simulate different kidney function, such as water reabsorption and waste secretion.

The rest of this paper is organized as follows. First, multiagent systems and the renal system are described. Second, we describe the implementation and characteristics of our simulation system. Then the visualization of difficult concepts to aid understanding is described. Finally, the usability experiments conducted and the results obtained are described followed by future trends for agents in medical simulations and finally, the conclusion.