Comparison of local sine wave modeling with harmonic phase analysis for the assessment of myocardial strain

PURPOSE: To compare local sine-wave modeling (SinMod) with harmonic phase analysis (HARP), for assessment of left ventricular (LV) circumferential strain (εcc) from tagged cardiovascular magnetic resonance images. MATERIALS AND METHODS: Mid-ventricular spatial modulation of magnetization was performed in 60 participants (15 each with hypertrophic, dilated or ischemic cardiomyopathy and 15 healthy controls) at 1.5 Tesla. Global and segmental peak transmural εcc were measured using HARP and SinMod. Repeated measurements were performed on 25% of examinations to assess observer variability. Effect of contrast was assessed in 10 additional patients. RESULTS: SinMod showed a high level of agreement with HARP for global εcc (mean difference -0.02, 95% limits of agreement -6.46 to 6.43%). Agreement was much lower for segmental εcc. Both methods showed excellent observer agreement for global εcc (intraclass correlation coefficient >0.75). Observer agreement for segmental εcc was also excellent with SinMod, but was significantly lower with HARP. Analysis time was significantly shorter using SinMod. Pre- and postcontrast εcc measurements were not significantly different using either technique, although postcontrast measurements showed greater variability with HARP. CONCLUSION: SinMod and HARP-based measurements of global εcc have a high level of agreement, but segmental agreement is substantially lower. SinMod has generally lower observer variability, is faster and is less affected by contrast, but requires further validation.