Arterial input function segmentation based on a contour geodesic model for tissue at risk identification in ischemic...

Area of Science:

• Medical Imaging
• Neuroscience
• Biomedical Engineering

Background:

• Perfusion parameters, including cerebral blood flow (CBF) and Tmax, are crucial for diagnosing and predicting outcomes in ischemic stroke.
• Accurate estimation of these parameters relies on the selection of the arterial input function (AIF), a critical step in clinical practice.

Purpose of the Study:

• To develop and validate a novel technique for arterial input function (AIF) selection in perfusion imaging.
• To improve the accuracy and reliability of perfusion parameter estimation for ischemic stroke assessment.

Main Methods:

• A variational segmentation model incorporating a distance function with geometric constraints was employed.
• Discrete total variation and energy minimization were used to identify arterial regions.
• Matrix analysis was applied to select the AIF with the highest peak within the segmented region.

Main Results:

• The proposed AIF selection method demonstrated superior arterial features, including higher peak position and faster attenuation, compared to existing methods.
• Perfusion parameters (mean CBF and Tmax) estimated using the novel AIF were higher than those from traditional methods.
• Ischemic regions were accurately delineated on perfusion maps derived from the proposed method.

Conclusions:

• The developed AIF segmentation framework outperforms current clinical standards for perfusion imaging.
• Perfusion parameters derived using the proposed AIF selection are more accurate and reliable.
• This method holds potential for integration into routine perfusion imaging calculations.