Regional ventilatory clearance by xenon scintigraphy: a critical evaluation of two estimation procedures

Estimates of ventilatory clearance are usually made by inspecting xenon washout images. Quantitative computer procedures have been described that produce regional clearance rates, yet their accuracy is not well established. We define a mathematical model for scintigraphic ventilation data based on 96 clinical studies, and with this model we test the accuracy of two procedures used to estimate ventilatory clearance. The least-squares curve-fitting technique for both washin and washout data has the same accuracy as a modified Stewart-Hamilton method (A/H) that uses washout data alone. Both procedures demonstrate relative errors of less than 5% and coefficients of variation of 10-20% when regions with equilibrium count rates of 3 cps and clearance times between 10 and 90 sec are examined. Because the A/H procedure is preferred for its simplicity and speed, we analyze two of its main sources of error: early washin/washout termination and background activity. To measure regional ventilation by the A/H procedure, we recommend: (a) washin and washout periods at least three times the largest clearance time of clinical interest; b) a regional equilibrium count rate of at least 3cps; and c) a 25- to 50-sec average of the equilibrium count rate.