TY - JOUR T1 - Preliminary assessment of image quality, comparing 2 acquisition speeds in whole body bone scan in a last generation gamma camera and new processing software JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 1197 LP - 1197 VL - 60 IS - supplement 1 AU - Camilo Garcia AU - Marianne Tondeur Y1 - 2019/05/01 UR - http://jnm.snmjournals.org/content/60/supplement_1/1197.abstract N2 - 1197Aim: To prospectively evaluate the quality of images comparing standard acquisition speed with increased acquisition speed in whole body bone scan (WBS) using a 870DR® SPECT/CT Camera (GE healthcare, Milwaukee, USA) with Low Energy High Resolution and Sensitivity (LEHRS) collimators and Clarity 2D® (C2D) processing software.Introduction: Major advances in hardware such as, gamma cameras and collimators have been made in the last decades. In parallel, calculation capacities and software development have continued to advance with encouraging results. New generation of Digital Ready (DR) Cameras using LEHRS collimators and new processing image software C2D are reputed to enable decreasing scan times and/or administered activity without compromising image quality. C2D processing is applied to planar images with purpose to diminish statistical noise without degradation of image resolution and contrast. It combines a linear non-anisotropic edge preserving iterative filter with contrast enhancement procedure. The filtering procedure iteratively updates each pixel using and. signal strength and local variation in the pixel neighboring area. The contrast enhancement is done by deconvolution with empirical point spread function. At the last stage -the processed image is linearly blended with original image to preserve noise texture. Methods: Two senior nuclear medicine specialists independently and blinded concerning the acquisition characteristics, levels of C2D applied and the score of the other observer, compared quality images of 10 WBS using 2 speed acquisitions, classical 15cm/min and faster acquisition at 20cm/min in same patients. All examinations were scored with a 5 levels scale (1: poor/not clinically acceptable - 5: excellent quality). After this first analysis, the quality of the same WBS acquisitions were compared using 5 levels of C2D (no Clarity®, 20, 40, 60, 80 and 100% blending). All WBS were scored with the same scale and finally, a “best choice” concerning quality was made by each observer. Results: All images were scored as good (score=3) to excellent quality (score =5). Using a Wilcoxon signed rank test, a statistical analysis for two related samples, we did not find any significant difference between the scores, p=0.73 (median value of the differences 0.5 in favor of the 20 cm/min speed). When applying a McNemar test on binary evaluations (score of 3 or 4 versus 5), we find in 10 occurrences that quality was assessed as better with the 20 cm/min speed compared to 7 occurrences with the reverse situation. The comparison is nonsignificant (p=0.47). No obvious preferences for the level of C2D was found, but, 40% blending had the highest frequency (35% of the best choice selections). However, we found significant interobserver differences (p<0.0001), observer ‘G’ preferred lower levels of blending (median of 40% blending) while observer ‘T’ preferred high values (median of 80% blending). There was no detectable difference concerning the level of C2D and the speed of the acquisition (p=0.31). The observed median is 50 for images acquired at a speed of 15 cm/min while it is 40 for the speed of 20 cm/min. If we stratify the analysis by speed of image acquisition and considering together, the observer has a significant impact on the “best choice” parameter (p<0.001) while this is not the case for speed of acquisition (p=0.07). Conclusions: There is no indication that the 20 cm/min provides inferior image quality in the 870DR camera using LEHRS collimators, as compared to standard time acquisition. No obvious preferences for the level of blending of C2D were found, but 40% blending was selected as ‘best choice’ most frequently. We did not find any significant dependence regarding image quality scores on the level of C2D and speed of the acquisition. However, there is a strong impact of the observer on the preferred value for choosing a blending level for the C2D software. ER -