The impact of second harmonic imaging on stress echocardiography reading

Noncontrast harmonic significantly improves overall image quality in echocardiography. The aim of this study was to assess whether harmonic imaging, without contrast, impacts on interobserver variability and diagnostic accuracy of "beginners" in stress echocardiography.

Images at rest and peak stress were obtained in digitized format in 15 consecutive patients (10 males, 5 females, mean age 66 +/- 9 years) and analyzed by 5 inexperienced observers (stress echo beginners) who were blinded to the imaging modality (standard versus harmonic imaging). Each observer graded the image quality as 1 = uninterpretable up to 5 = excellent, in a total of 240 segments.

The mean image quality per segment was 2.9 for conventional technology and increased up to 3.6 for harmonic imaging (p < 0.001). The interobserver agreement (> or = 4 out of 5 readers) rose from 46 to 60%. The percentage of uninterpretable segments was 7.5% at rest and 8.9% at peak stress of conventional imaging, and decreased to 4.6 and 6.5%, respectively (p < 0.05) by second harmonic technology. The unanimous reading of two additional independent expert observers was arbitrarily assumed to be the "gold standard" to verify the accuracy of reading of the 5 beginners. The 5 beginners showed poor diagnostic accuracy with fundamental imaging (73%), and they did not improve with second harmonic imaging (70%, NS vs fundamental).

Noncontrast second harmonic imaging improves image quality over conventional imaging. The improved quality of the image deflated interobserver variability but did not determine per se an improvement in the diagnostic accuracy of nonexperienced readers.