A fully-automated computer-assisted method of CT brain scan analysis for the measurement of cerebrospinal fluid spaces and brain absorption density

Publisher Website
Google Scholar
Abstract
Computer-assisted methods of CT brain scan analysis offer considerable advantages over visual inspection, particularly in research; and several semi-automated methods are currently available. A new computer-assisted program is presented which provides fully automated processing of CT brain scans, depending on “anatomical knowledge” of where cerebrospinal fluid (CSF)-containing spaces are likely to lie. After identifying these regions of interest quantitative estimates are then provided of CSF content in each slice in cisterns, ventricles, Sylvian fissure and interhemispheric fissure. Separate measures are also provided of mean brain density in each slice. These estimates can be summated to provide total ventricular and total brain volumes. The program shows a high correlation with measures derived from mechanical planimetry and visual grading procedures, also when tested against a phantom brain of known ventricular volume. The advantages and limitations of the present program are discussed.

This publication has 23 references indexed in Scilit: