Abstract
CT-cisternography performed on 35 neurological cases with spinal administration of meglumine iocarmate (Dimer-X) or metrizamide (Amipaque) was analysed digitally with a series of our original programs developed using the system computer of an EMI-1010 scanner. RI-cisternography with In-111 was performed for comparison in four of these cases, within 2 weeks after CT-cisternography.
Since the attenuation coefficients (CT-No.) correlated well with the concentration of iodine dye in CSF, temporal changes of the average CT-No. in geometrically defined areas (ROI) within lateral ventricles and various cisterns present useful information for the differential diagnosis of pathological CSF kinetics. This density method, however, may be available only for cases with marked dilatation of the CSF space which allows placing of large ROIs of sufficient size to compensate for a wide variation of CT-No. due to statistic errors of radiation and excluding the partial volume effect of bone or brain tissue.
The principle of our weight method is to analyze the weight of dye contained in the CSF of each slice. The CT-weight, and integrals of (CT-No.)×(pixel No.) and of CSF space of each slice were assumed to be parallel to the weight of the dye. This was calculated for a certain range of CT-No., excluding data involved in the partial volume effect of bone or air and was plotted on a logarithmic scale.
A close similarity was observed between the temporal changes of total CT-weight of dye and total counts per minute in RI-cisternography in the same case.
Total CT-weights of the upper and lower halves of slices show characteristic patterns of hydrocephalus, ventricular reflux of dye, impairment of absorption, etc, especially when combined with the temporal changes in CT-weight of the normal CSF range.
Digital analysis of CT-cisternography gives valuable information for objective and quantitative evaluation of the pathological process of CSF kinetics.
