Clinical manifestations (functional deficit) of spinal cord diseases such as multiple sclerosis or traumatic cord injury are thought to be closely related to the degree of axonal disruption and/or demyelination- the ability of spinal cord axons to conduct nerve impulses is dependent on the fidelity of the myelin sheaths. Conventional magnetic resonance (MR) images do not provide information regarding axonal integrity. However, MR-based measurement of apparent diffusion coefficients (ADC) can exploit the intrinsic anisotropy of white matter-the myelin sheaths restrict water diffusion perpendicular to the uninjured axons, but do not affect diffusion parallel to the axons. Thus ADC measured longitudinally (lADC) and transverse (tADC) to the axon fiber axis may be used to detect and characterize white matter damage by quantitating the loss of normal white matter diffusion anisotropy. However, ADC is also sensitive to changes in the relative intracellular/extracellular water content; therefore, the combined effects of axonal swelling, extracellular edema, and demyelination in spinal cord disease complicate the interpretation of MRI-based measurements of ADC. The primary objective is to utilize a computer model to determine the dependence of MRI-based measurements of tADC and 1ADC in spinal cord white matter on axonal morphology (distribution of axons, cellular volume fraction) and axonal integrity (permeability). A computer simulation of diffusion among axons has been developed that uses microscopic images of white matter as input; thus no assumptions are made concerning the axonal arrangement, cellular volume, or myelin sheath thicknesses. Preliminary results exhibit excellent agreement between the computer calculations and ADC measured in spinal cord. Computer simulations indicate that the competing affects of axonal swelling and demyelination can be deconvolved, without making explicit assumptions for the values of the intrinsic intracellular and extracellular diffusion coefficients. The model will be validated by comparing computer predictions to measured ADC in animal models of multiple sclerosis (EAE in rats and mice) and traumatic cord injury (weight-drop injury in rats). The remainder of the research effort is directed at determining the sensitivity of ADC to pathology in an animal model of early multiple sclerosis. A reliable model for white matter diffusion, which can relate measured ADC values to underlying pathologic changes at the cellular level, would one day allow the radiologist to infer axonal integrity, and thus enhance prognostic capability of MRI in spinal cord disease.
