Hepatic perfusion index in cirrhotic livers--investigation of imaging and analytical procedures.

Dynamic hepatic scintigraphy was performed in a group of cirrhotic patients to evaluate the optimum imaging and analytical procedures necessary for the measurement of the hepatic perfusion index (HPI). Patients were studied in the posterior (n = 19) and the anterior (n = 14), positions, with either 0.2 or 0.5 ml of 99Tcm sulphur colloid administered as a rapid bolus injection. In each subject, three ROIs (small, medium and large) were drawn over the liver, and time-activity perfusion curves were generated. Analytical techniques were developed to allow flexibility in selecting the arterial and portal venous phases of the liver perfusion curve. The quality of the bolus, expressed as the full width at half-maximum of the left ventricular time--activity curve, was independent of the bolus volumes and patient positioning. The dispersion in the data and the inter-observer variation were less in the anterior view using medium and large ROIs, compared with the anterior small ROI and all the posterior ROI sizes. A time delay between liver and kidney arterial phases, if ignored, produced statistically significant effects on the values of the HPI. We conclude that HPI investigations are best performed in the anterior projection. Data analysis using a large liver ROI is preferred, and flexible data-processing techniques are recommended, particularly in the presence of a liver and kidney arterial time delay.