HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Rollo, F. D. Articles by Schulz, A. G. Search for Related Content PubMed PubMed Citation Articles by Rollo, F. D. Articles by Schulz, A. G.

Effect of Pulse-Height Selection on Lesion Detection Performance

The Johns Hopkins Medical Institutions, Baltimore, Maryland
The Johns Hopkins University Applied Physics Laboratory, Silver Spring, Maryland

Correspondence: For reprints contact: A. G. Schulz, The Johns Hopkins University Applied Physics Laboratory, 8621 Georgia Ave., Silver Spring, Md. 20910.

ABSTRACT

As indicated earlier, various workers have considered the effect of photon energy-window selection in scanning systems. However, systematic measurements of the effects of energy-window selection on fluctuation noise, resolution, and contrast in scan images have been lacking, and no measurements have been made of the resultant effects on observer performance in detecting lesions. In this study experimental results were used to generate simulated scans for observer tests to show how energy-window selection affects an observer's ability to detect the presence of a lesion in a scan.

The results of the observer performance studies with ¹⁹⁷Hg show peaks in the lesion-detection test scores at window baseline settings of 65 and 45 keV. The optimum value of the window baseline shifts from 65 to 45 keV as the test lesion was changed from 1.6 to 1.9 cm. The lower energy performance peak is postulated to result from the inclusion of iodine escape peak photons at the lower energy window setting. Additional frontscatter material significantly reduced the overall detection rate. With additional frontscatter material the probability of correctly reading the scan has very little dependence on the window baseline setting.

The results for ^99mTc showed a marked variation in detection scoring as a function of energy-window baseline setting. The observer performance peaked at a baseline setting of 130 keV and dropped by a factor of two for baseline settings only 10 keV on either side of this value. The detection performance decreases as the window baseline is lowered and then rises at a value near the iodine escape peak in a manner similar to the ¹⁹⁷Hg case. Unfortunately, the data were not taken below 100 keV. This behavior in both cases is confirmed by the narrow energy interval line-spread functions which show that the line-spread functions associated with the iodine escape peak are sharper than those appearing at intermediate energies near the photopeak.

Detectability of lesions is considerably more dependent upon window setting when using ^99mTc compared with ¹⁹⁷Hg. Also, the general level of detectability is much higher for ^99mTc when optimum window settings are used for both radionuclides. This observation is consistent with the superior signal contrasts measured with this nuclide compared with ¹⁹⁷Hg.

For the restricted parameters of scattering, geometry, and collimator tested, the energy-window baseline for ^99mTc is critical and should be set at 130 keV for optimum performance in detecting small focal lesions.

Detection performance with ¹⁹⁷Hg does not vary greatly as the energy window is changed. The selection of a best operating point for the window baseline is not obvious. It appears that for relatively small lesions behind a thick layer of scattering medium the usual setting in the range of 60–65 keV is best. It is possible that for certain lesion and collimator combinations the 45-keV baseline setting would give better performance.

Since the "organ" used in these studies was comparable to the dimensions of the response function tails with the thicknesses of frontscattering material employed, it is expected that for larger organs the scattering effects will be slightly more pronounced. In these cases the level of detection would be slightly decreased, but the general results found above with energy-window variation would not be altered.

The rise in detectability as the energy window baseline is lowered to include the iodine escape peak indicates that a dual-energy window might provide improved performance for nuclides with primary photons in the energy range below 150 keV. The line-spread functions for narrow energy intervals showed that fewer of the photons at the iodine escape peak energy are scattered than is the case for photons intermediate in energy between this value and the photopeak energy. The observer detection scores were consistent with these results. Therefore a dual window which accepts the photopeak and the iodine escape peak and rejects intermediate energies may give considerable performance improvement for sodium iodide detectors.

FOOTNOTES

^* Present address: Div. Nuclear Medicine, Upstate Medical Center, Syracuse, N.Y.