Quantitative CT densitometry for predicting intracerebral hemorrhage growth

Area of Science:

• Neurology
• Radiology
• Medical Imaging

Background:

• Intracerebral hemorrhage (ICH) growth is a key predictor of patient outcomes, including disability and mortality.
• Accurate prediction of hematoma expansion is crucial for effective clinical management.

Purpose of the Study:

• To investigate if noncontrast quantitative CT densitometry (qCTD) of hyperacute ICH can reflect active bleeding.
• To determine if qCTD parameters improve predictive models for hematoma growth.

Main Methods:

• Analysis of 81 baseline CT scans from the placebo arm of a phase IIb trial, obtained within 3 hours of ICH onset.
• Generation of Hounsfield unit histograms to derive qCTD parameters: mean, SD, coefficient of variation, skewness, and kurtosis.
• Statistical modeling incorporating qCTD parameters, baseline volume (BV), and time-to-scan (TTS) to predict hematoma growth.

Main Results:

• The coefficient of variation of ICH attenuation was the strongest individual predictor of hematoma growth (adjusted R(2) = 0.107, P = .002).
• A combined model using coefficient of variation, BV, and TTS significantly improved growth prediction (adjusted R(2) = 0.202, P = .009).
• qCTD enhanced the accuracy of 24-hour growth predictions by up to 47%.

Conclusions:

• Heterogeneous ICH attenuation on early CT scans (within 3 hours) predicts hematoma expansion.
• qCTD may serve as a marker for active bleeding processes in the early stages of ICH.
• Further research is needed to evaluate the integration of qCTD into routine clinical imaging protocols for ICH.