Comparison of four scatter correction methods for patient whole-body imaging during therapeutic trials with iodine-131.

BACKGROUND: Activity in regions of interest can be measured using serial whole-body scintigraphic images to estimate the dose received by a patient after therapeutic injections. As scatter and attenuation introduce biases in quantitative measurements, these phenomena need to be corrected to allow accurate determination of tracer concentration.
METHODS: The feasibility of iodine-131 whole-body imaging in list mode was studied over an extended spectrum (0-750 keV) in order to compare four scatter correction methods by the geometric mean approach (20%, Dual Energy Window, Triple Energy Window, and Spectral Factor Analysis methods). All data were corrected for attenuation using a Transmission Attenuation Correction prototype from Sopha Médical Vision international. The half-life of an iodine-131 standard source was calculated from scatter-corrected anterior views. Whole-body activities, using the Day 0, Hour 1, image as a reference (calibration from an administered dose) and an external calibration source (calibration from an imaged known-activity source), were calculated for three patients undergoing a radioimmunotherapy trial in order to assess the reliability of quantification by the geometric mean approach.
RESULTS: Patient studies confirmed the clinical feasibility of this type of acquisition. As expected, all methods allowed determination of an accurate half-life for the calibration source. A slight impact of scatter correction was observed in quantification with calibration from an administered dose. For quantification with calibration from an imaged known-activity source, whole-body activity was overestimated by +100% to +200% with the 20% window, depending on the size of the patient, whereas errors were about +50% with scatter correction. However, the influence of patient morphology was less marked when a scatter correction method was used.
CONCLUSIONS: When the geometric mean approach is used together with a sophisticated transmission acquisition device for quantification with calibration from an administered dose, the 20% energy window appears to be adequate. However, for quantification with calibration from an imaged known-activity source, accurate activity estimates cannot be obtained even when scatter correction is used to compensate for the influence of patient morphology. Copyright 2002 American Cancer Society.