Comparison of blind deconvolution- and Patlak analysis-based methods for determining vascular permeability

Area of Science:

• Physiology
• Biophysics
• Medical Imaging

Background:

• Vascular permeability is crucial for tissue homeostasis and disease.
• Accurate measurement of vascular permeability requires complex modeling.
• Current methods often rely on the arterial input function, limiting clinical application.

Purpose of the Study:

• To develop and validate a computational algorithm for determining vascular permeability constants.
• To assess the algorithm's performance without needing the arterial input function.
• To compare the algorithm's accuracy against established methods like Patlak analysis.

Main Methods:

• Development of a "blind" deconvolution algorithm for analyzing time-lapse imaging data.
• Modeling of bidirectional solute transport between capillaries and surrounding tissue.
• Comparison with algorithms assuming unidirectional transport and standard Patlak analysis.

Main Results:

• The blind deconvolution algorithm accurately determines vascular permeability constants from imaging data.
• It demonstrates superior accuracy compared to Patlak analysis, especially with solute delay and dispersion.
• Bidirectional transport modeling showed advantages over unidirectional models in specific scenarios.

Conclusions:

• Blind deconvolution is a robust method for vascular permeability mapping.
• This approach offers improved accuracy over traditional methods under clinically relevant conditions.
• The algorithm has the potential to enhance the detection of compromised vascular barriers in tissues.