Experimental models of brain injury.

General categories of experimental brain injury models are reviewed regarding their clinical significance, and two new models are presented that use different methodology to produce injury. This report describes and characterizes the pathophysiologic changes produced by a novel fluid percussion (FP) method and a controlled cortical impact (CI) technique, both developed at the General Motors Research Laboratories (GMRL). The new models are compared to prior experimental brain injury techniques in relation to ongoing physical and analytical modeling used in automotive safety research by GMRL. Experimental results from our laboratory indicate that although the FP technique, currently the most widely used method for producing brain injury, is useful for producing graded injury responses systemically and centrally, it is not well-suited for detailed biomechanical analyses. This conclusion is based on high-speed cineradiographic studies where the physiologic saline in the FP cannula was substituted with a radiopaque contrast medium (Conray 1:1 dilution/saline). High speed x-ray movies (1000 fps) were taken of the fluid percussion pulse (1.5-3.4 atm/20 msec) in sagittal, dorsal, and frontal planes of orientation. When viewed together, the cineradiography revealed a complex, dynamic interaction between the injected fluid and the skull/cranial contents. Rapid lateral and anterior/posterior epidural fluid flow suggest that the pathology and dysfunction following FP brain injury reflects diffuse mechanical loading of the brain. Because fluid is used to transfer mechanical energy to brain tissue, and because fluid flow characteristics (i.e., direction, velocity, and displacement) are dependent on the brain geometry and species used, accurate analytical and biomechanical analyses of the resultant injury would be difficult at best. In contrast, the cortical impact model of experimental brain injury uses a known impact interface and a measurable, controllable impact velocity and cortical compression. These controlled variables enable the amount of deformation and the change in deformation over time to be accurately determined. In addition, the CI model produces graded, reproducible cortical contusion, prolonged functional coma, and extensive axonal injury, unlike the FP technique. The quantifiable nature of the single mechanical input used to produce the injury allows correlations to be made between the amount of deformation and the resultant pathology and functional changes.(ABSTRACT TRUNCATED AT 400 WORDS)