Body contour determination and validation for bremsstrahlung SPECT imaging.

UNLABELLED

The purpose of this study was to validate a previously reported body contour measurement using Compton backscatter sources with bremsstrahlung SPECT imaging.

METHODS

Bremsstrahlung SPECT imaging was performed with 32P using a dual-headed camera system fitted with medium-energy, parallel-hole collimators. Two sources of 99mTc were placed directly on each collimator. Energy windows of 100 keV +/- 25% were used to image the 32P and also to record the Compton scatter from the 99mTc sources. Eleven patients enrolled in clinical Phase I therapeutic protocols were injected with 32P-chromic phosphate and SPECT images were acquired and reconstructed in the transaxial plane. The 32P distribution and the patient body contour were both visualized in these slices. The anteroposterior and lateral patient dimensions were measured by generating count profiles parallel to the anteroposterior and lateral body contour, respectively, at the midline in a transaxial slice. The distance in centimeters between the two centroids of each profile is representative of the anteroposterior and lateral dimensions and was determined for each patient. These anteroposterior and lateral dimensions were compared to the same distance measurements made in these patients by CT in an anatomically comparable transaxial slice. A cylindrical SPECT phantom was also studied to further validate the contour measurements.

RESULTS

The mean percent difference in the patient dimension measurements between SPECT and CT was -0.8% with a range of -8.5% to 9.9%. The percent difference between the known and SPECT measured dimensions in the cylindrical phantom was 0.5%.

CONCLUSION

The two external Compton scatter source method is accurate for determining the body contour.