OBJECTIVE: To validate the harmonic-fundamental frequency ratio peak for user-independent differentiation of myocardial perfusion under clinically relevant levels of signal attenuation. METHODS: Radio frequency data were obtained by using apical long-axis scans in 11 open-chest pigs during continuous infusion of a contrast agent after left anterior descending artery occlusion. Silicone pads were interposed between the transducer and the heart to simulate levels of thoracic wall attenuation. Samples of image data from perfused and nonperfused regions were collected; values using harmonic-fundamental frequency ratio peak and conventional harmonic gray scale intensity techniques were calculated. RESULTS: At each attenuation level, the harmonic-fundamental frequency ratio peak value of perfused myocardium was higher than that of nonperfused myocardium (P < .0001). The variance of these values was smaller than that of the gray scale intensity values (P < .0001), with smaller overlap between harmonic-fundamental frequency ratio peak values differentiating perfused and nonperfused regions. In the receiver operating characteristic curves, this analysis had better diagnostic performance than gray scale analysis. In the optimal cutoff value, harmonic-fundamental frequency ratio peak analysis provided 87% sensitivity and 91% specificity; gray scale analysis had 80% sensitivity and 78% specificity. CONCLUSIONS: Harmonic-fundamental frequency ratio peak analysis differentiated perfused from nonperfused myocardium under clinically relevant attenuation conditions and provided higher sensitivity and specificity for perfusion determination in attenuated myocardium than did gray scale intensity analysis.
OBJECTIVE: To validate the harmonic-fundamental frequency ratio peak for user-independent differentiation of myocardial perfusion under clinically relevant levels of signal attenuation. METHODS: Radio frequency data were obtained by using apical long-axis scans in 11 open-chest pigs during continuous infusion of a contrast agent after left anterior descending artery occlusion. Silicone pads were interposed between the transducer and the heart to simulate levels of thoracic wall attenuation. Samples of image data from perfused and nonperfused regions were collected; values using harmonic-fundamental frequency ratio peak and conventional harmonic gray scale intensity techniques were calculated. RESULTS: At each attenuation level, the harmonic-fundamental frequency ratio peak value of perfused myocardium was higher than that of nonperfused myocardium (P < .0001). The variance of these values was smaller than that of the gray scale intensity values (P < .0001), with smaller overlap between harmonic-fundamental frequency ratio peak values differentiating perfused and nonperfused regions. In the receiver operating characteristic curves, this analysis had better diagnostic performance than gray scale analysis. In the optimal cutoff value, harmonic-fundamental frequency ratio peak analysis provided 87% sensitivity and 91% specificity; gray scale analysis had 80% sensitivity and 78% specificity. CONCLUSIONS: Harmonic-fundamental frequency ratio peak analysis differentiated perfused from nonperfused myocardium under clinically relevant attenuation conditions and provided higher sensitivity and specificity for perfusion determination in attenuated myocardium than did gray scale intensity analysis.