Computer-interactive method for quantifying cerebrospinal fluid and tissue in brain CT scans: effects of aging

Area of Science:

• Medical Imaging
• Neuroscience
• Radiology

Background:

• Accurate quantification of cerebrospinal fluid (CSF) and brain tissue is crucial for diagnosing neurological conditions.
• Previous semiautomated systems for brain CT analysis had limitations in accuracy and artifact reduction.

Purpose of the Study:

• To present a modified semiautomated system for enhanced quantification of CSF and brain tissue on CT scans.
• To improve the accuracy of fluid-tissue differentiation and volumetric measurements, particularly in sulci.

Main Methods:

• A semiautomated system was modified to automatically strip the skull and apply a 2D high-pass filter to reduce spectral-shift artifact.
• Human scorers differentiate fluid and tissue by adjusting a one-bit display threshold on filtered images.
• Volumetric data were summed within or across sections for regions of interest like ventricles and sulci.

Main Results:

• The modified system demonstrated high correlations with experienced clinician rankings for both ventricular (r=0.88) and sulcal (r=0.83) size.
• Significant correlations were found between computed measures of ventricular size and age.
• Significant correlations were also found between sulcal widening and age.

Conclusions:

• The enhanced semiautomated system provides accurate and reliable quantification of CSF and brain tissue on CT scans.
• The improved accuracy in sulcal measurement is a major advantage of this modified technique.
• The system's ability to correlate with age-related changes in ventricular and sulcal size suggests its clinical utility.