[Development of the automated Patlak plot method and its verification in clinical examples]

Area of Science:

• Neuroimaging
• Radiochemistry
• Medical Physics

Context:

• The Patlak plot method is a standard clinical tool for estimating mean cerebral blood flow (mCBF).
• Current methods are susceptible to operator subjectivity and variability, impacting reproducibility.
• The conventional Patlak plot involves manual region of interest selection and graphic analysis.

Purpose:

• To develop and validate an automated Patlak plot method for calculating mCBF.
• To eliminate operator subjectivity and improve the reproducibility of mCBF measurements.
• To enhance the reliability and accuracy of cerebral blood flow quantification in clinical settings.

Summary:

• This study presents an automated algorithm for all three steps of the Patlak plot method: region of interest (ROI) selection in the cerebral hemisphere and aortic arch, and graphic analysis.
• Automation ensures that mCBF calculations are independent of operator experience and subjectivity.
• Verification demonstrated successful ROI automation, precise graphic analysis within 1-2% error, and a mean absolute percentage error of 3.1% compared to conventional methods.

Impact:

• The automated method significantly improves the objectivity and reproducibility of mCBF quantification.
• It addresses the inherent ambiguity in the conventional Patlak plot method, providing more reliable results.
• This advancement facilitates more consistent and accurate assessment of cerebral blood flow in clinical neuroimaging.