Noninvasive measurement of arterial blood plasma concentration of iodinated contrast agents from CT scans of human brain.

OBJECTIVE: Our goal was to assess the accuracy of estimating the time course of the arterial plasma concentration of meglumine iothalamate from cranial CT images of different vascular structures in the brain. MATERIALS AND METHODS: Dynamic CT studies of transcapillary transport in various brain lesions were analyzed. Vascular structures in the brain were identified and classified in three categories: arteries, veins, and venous sinuses. Systemic venous blood samples were taken prior to the infusion of meglumine iothalamate and 10 min after completion of the infusion and used as a calibration for the volume averaging fraction of the image of each vascular structure. A time course of plasma meglumine iothalamate concentration for each of the vascular categories in the CT images was obtained and compared with a variety of methods. RESULTS: Significant differences were found for measurement of plasma meglumine iothalamate concentration from different vascular categories. There was also a disparity between the volume averaging fraction that we calculated and what would be expected due to the measured systemic hematocrit for all vascular structures. CONCLUSION: The use of images of veins and venous sinuses consistently underestimated the arterial concentration around the peak values. Correcting the imaged venous sinus values with the measured systemic hematocrit was even less reliable. The most accurate method of determining arterial plasma concentration of meglumine iothalamate from CT images of brain was to correct the identified arterial vessels for volume averaging.