Automated Detection of Normal Pressure Hydrocephalus Using CT Imaging for Calculating the Ventricle-to-Subarachnoid...

Area of Science:

• Neurology
• Radiology
• Medical Imaging

Background:

• Normal pressure hydrocephalus (NPH) diagnosis is challenging due to overlapping symptoms with normal aging and dementia.
• Timely NPH identification is crucial for effective treatment, improving gait and quality of life.

Purpose of the Study:

• To develop and evaluate an automated marker for NPH detection using clinical CT images.
• To assess the utility of this marker in a large patient cohort.

Main Methods:

• Retrospective review of CT images from 306 NPH patients and controls.
• Automatic calculation of ventricle-to-subarachnoid volume ratio (VSR) using imaging software.
• Linear regression modeling to assess VSR as an NPH predictor, adjusted for age and sex.

Main Results:

• VSR was significantly higher in NPH patients compared to controls (P < .001).
• VSR showed high diagnostic accuracy for NPH with 94.1% sensitivity and 92.5% specificity.
• Area under the ROC curve was 0.99, indicating excellent performance.

Conclusions:

• Automated VSR assessment on head CT accurately identifies probable NPH with 93% accuracy.
• The findings support the generalizability of automated CT screening for NPH.
• The study highlights the clinical utility of ventricle-to-sulcal concordance as an imaging marker for NPH.