Interobserver variability in CT assessment of brain atrophy

Area of Science:

• Radiology
• Neurology
• Medical Imaging

Background:

• Brain atrophy estimation is crucial for diagnosing neurological conditions.
• Subjective assessment of brain atrophy on CT scans can be prone to variability.

Purpose of the Study:

• To evaluate interobserver variability in the subjective estimation of brain atrophy using CT scans.
• To compare subjective assessments with quantitative measurements.

Main Methods:

• Four neuroradiologists reviewed CT brain images from 150 patients without focal lesions.
• Quantitative measurements included frontal horn index and subarachnoid space area ratios.
• Interobserver agreement was assessed using kappa statistics.

Main Results:

• Fair agreement for atrophy presence (k=0.24) and degree (k=0.24).
• Moderate agreement for atrophy type (k=0.59).
• Significant correlation found between observer agreement and quantitative measurements.

Conclusions:

• Subjective estimation of brain atrophy by neuroradiologists is unreliable.
• Quantitative measurements are essential for accurate brain atrophy assessment, especially for clinical decision-making.