Cerebral Artery Vasospasm Detection Using Transcranial Doppler Signal Analysis

Area of Science:

• Neurology
• Biomedical Engineering
• Medical Imaging

Background:

• Aneurysmal subarachnoid hemorrhage (aSAH) can lead to silent cerebral artery vasospasm (sCAS), causing severe outcomes like cerebral ischemia and death.
• Transcranial Doppler (TCD) ultrasound is a noninvasive tool for early sCAS detection, even before clinical symptoms appear.
• Current TCD monitoring is limited by the need for expert sonographers and physicians, hindering widespread, continuous use.

Purpose of the Study:

• To develop an automated algorithm for detecting cerebral artery vasospasm using TCD ultrasound data.
• To overcome the existing barriers in TCD monitoring for sCAS detection in aSAH patients.

Main Methods:

• Extracted various signal features including energy, entropy, spectral characteristics, and Mel frequency cepstral coefficients.
• Applied principal component analysis (PCA) for dimensionality reduction of the extracted features.
• Trained a decision-tree classifier using the selected features for automated sCAS detection.

Main Results:

• The developed algorithm demonstrated high accuracy in detecting cerebral artery vasospasm.
• Achieved an overall sensitivity of 87.5% and a specificity of 89.74% for sCAS detection.
• The feature extraction and classification approach proved effective for automated TCD analysis.

Conclusions:

• The automated algorithm shows significant potential for development into a continuous cerebral artery vasospasm monitoring system.
• This technology could improve patient outcomes by enabling earlier and more frequent sCAS detection.
• Automation in TCD analysis may reduce reliance on specialized personnel, increasing accessibility to monitoring.