Extracranial internal carotid artery calcium volume measurement using computer tomography

Area of Science:

• Medical Imaging
• Cardiovascular Radiology
• Stroke Prevention

Background:

• Carotid artery calcium volume on contrast-enhanced CT is a key stroke risk indicator.
• Automated systems are needed for fast and accurate calcium volume computation.
• Contrast agents can obscure calcium, necessitating correction methods.

Purpose of the Study:

• To present an automated and semi-automated system for carotid artery calcium volume calculation.
• To address challenges posed by contrast agent interference in CT scans.
• To evaluate the accuracy and reliability of the developed systems against manual measurements.

Main Methods:

• Segmentation of calcium regions using Hounsfield Unit-based thresholding.
• Correction of calcium regions using neighboring slices to account for contrast agent variations.
• Development of both fully automated and semi-automated (fallback) systems.
• Validation against manual calcium volume tracings by a neuroradiologist.

Main Results:

• The automated system achieved correlation coefficients (CC) of 0.89-0.90 with manual readings.
• The semi-automated system achieved higher CC of 0.96 with manual readings.
• No significant difference was found between automated, semi-automated, and manual methods (t-test).
• Excellent intra-observer reliability (CC 0.98) was demonstrated.

Conclusions:

• The semi-automated method for carotid artery calcium volume is acceptable and aligns well with manual assessments.
• Both automated and semi-automated systems provide reliable calcium volume estimations.
• Further validation of the semi-automated protocol is recommended.