Disappearance of a hyperfunctioning thyroid nodule.

(CS) and without (C) are shown. The numbers indicate those studies in which there was agreement in diagnosis. The table also shows the relationship of these three techniques to the final diagnosis (FD) made for the patient in each case. The final diagnosis made use of other information such as additional perfusion views, chest radiographs, and patient examination. Film (F) relates best to final diagnosis (FD) because of its superior resolution over computer images. Diagnosis from computerized background-subtracted images (CS) was more closely related to the film diagnosis than the background-included images (C), but only modestly so. We believe this is because of the law governing Compton scatter (2). (The minimum energy of a singly scattered 140-keV photon still falls above the pulseheight window normally used for Xe-133.) This phenomenon is illustrated in Fig. 1. The image produced by the photons scattered from Tc-99m into the Xe-133 window (b) has little of the definition seen in the perfusion image (a). As a result, the first washin image with background subtraction (d) is similar to the first washin image without background subtraction (c), even though 49% of the counts in the first washin image are due to Tc-99m scattered photons. We submit that on film the trained eye can make adjustments for this kind of background. Although the numbers presented are small, they support our opinion that Xe-133 ventilation studies performed after perfusion studies give a better indication of disease if recorded on film, rather than computer, even if the computer images are background-subtracted . F. S. PRATO R. MACKENZIE St. Joseph's Hospital London, Ontario, Canada