Clinical benefits of precise delay correction for CBF measurement in O-15 water PET study.

Objectives A new method of delay time estimation was proposed for measuring precise cerebral blood flow (CBF) and arterial blood volume (V0) using O-15 water PET.

Methods Nineteen patients with unilateral arterial stenoocclusive lesions were studied for hemodynamic status before treatment. The delay time of each pixel was calculated using least squares fitting with an arterial blood input curve adjusted to the internal carotid artery (ICA) count at the skull base. Pixel-by-pixel delay estimation provided a delay map image which could be used for CBF and V0 image calculation precisely using a 2-compartment model, and the values from this method were compared with those from the slice-by-slice correction method. The differences in hemodynamic parameters from O-15 gas PET scans were also compared between ipsilateral and contralateral sides.

Results The affected side showed longer delay times compared with the contralateral cerebral hemisphere. Although the mean cortical CBF values were not different between the two methods, the slice-by-slice delay correction overestimated in the hypoperfused area. The scatter plot of V0 pixel values showed significant difference between the two correction methods, where the slice-by-slice delay correction significantly overestimated in the whole brain. Hemodynamic impairment expressed by increase in oxygen extraction fraction (OEF) was correlated with the delay time of tracer arrival in the ipsilateral hemisphere (r = 0.37, p < 0.05).

Conclusions The precise estimation of hemodynamic parameters such as CBF and V0 by using the new method should be ideal and beneficial for clinical use. Pixel-by-pixel delay estimation also provides a new image of delay time which would be useful in the clinical assessment of CVD.