ASNC 2014 Posters
Versatility of Emory Toolbox 4.0 shown at ASNC in scientific posters
(September 20, 2014 – Atlanta, GA) -- David Cooke, Emory University, presented two scientific posters at ASNC. The first poster looked at the clinical benefits of MPI gated studies with 8 frames per cardiac cycle for calculating diastolic function. In conjunction with colleagues at Emory University, Cooke studied 121 patients with end-stage renal disease and normal LVEF by echo (>50%). Each patient had same-day 99mTc low-dose rest / high-dose stress protocol, and standard SPECT. Emory Toolbox 4.0 was used to analyze the resting gated short-axis slices. The time-volume curves were used for calculating diastolic function parameters. The Fourier analysis using 3 harmonics was used to approximate each time-volume curve. In a previous study, J Chen et al. demonstrated that 3-harmonic Fourier analysis of a time-volume curve from an 8-frame gated MPI study is able to detect phase offsets of as little as 5.6⁰. That study showed the value of replacing the 8 discrete time points with a continuous harmonic function (J Nucl Cardiol 15(3):383-391, 2008).
"Many labs are still acquiring gated studies at 8 frames/cardiac cycle because of the increased counts in each gated bin compared to studies acquired with 16 or more frames," says Cooke. "We explored whether 3-harmonic Fourier diastolic analysis of 8-frame gated MPI studies can be used to help overcome this limited sampling deficiency." The authors concluded that for adults with end-stage renal disease parameters of diastolic function can be successfully extracted from 8-frame gated 99mTc MPI studies.
ASNC abstract 322-70.
As second scientific poster at ASNC demonstrated the accuracy of a new left ventricular. automatic base-detection algorithm in Emory Toolbox 4.0 for PET myocardial perfusion studies. This comparison study by David Cooke, and colleagues at Emory University and Syntermed, Inc. analyzed 100 clinical 82Rb stress/rest patients. All the studies were read visually by the same expert reader. Each study was then re-processed using Emory Cardiac Toolbox 4.0 and the original base-detection algorithm, and a second time using the new base-detection algorithm. "In this study we report on how the accuracy of clinical decision support is affected by proper base selection. Within PET MPI we often see atria and other background features which sometimes caused the original base-detection algorithm to locate the base too far out," says Cooke. "The new algorithm now employs multiple thresholds that are confined to specific myocardial walls as well as the valve-plane and is correctly finding the base further in, significantly impacting the quantitation and interpretation of the study."
The authors concluded the new automatic base-detection algorithm coupled with Syntermed IDS improves accuracy and provides diagnostic results consistent with those from an expert processing.
ASNC abstract 322-71.
D. Cooke*1, R. D. Folks1, F. Esteves1, R. Eppes2, E. V. Garcia1; 1Emory University, Atlanta, GA, 2Syntermed, Inc., Atlanta, GA
Additional references relevant to ASNC Abstract 322-71
1. Garcia EV, Klein J, Esteves FP, Cooke CD, Manatunga D, Del’Aune C, Eppes R, Folks R. LVX: A Novel Decision Support System for Cardiac Image Interpretation and Reporting. J Nucl Cardiol Suppl 19(4):838, 2012.
2. Esteves FP, Sultana R, Cooke CD, Folks RD, Garcia EV. Diagnostic Accuracy Comparison of LVX Versus ECTb4 on Rest/Stress Rb-82 Myocardial Perfusion 3D PET. J Nucl Cardiol Suppl 20(4):689, 2013.