Abstract Background and purpose The "average pixel intensity" (API) method, which is applied to the continuous Doppler spectral signal in mitral regurgitation (MR), has been proposed as a new method for automated MR grading. The objective of this study is to compare the API method with the effective regurgitant orifice area (EROA) for the detection of moderate-severe or severe MR (grades III or IV/IV) using open-source software (ImageJ). Methods Prospective single-center study (n=267) including patients with different degrees of MI, with a subgroup with cardiac magnetic resonance (CRM, n=55). Of the total sample, API was measured in 148 patients, and EROA in 159. Significant MR by echocardiography was defined according to a multiparameter criterion (American Society of Echocardiography - ASE algorithm), and by CRM according to regurgitant volume (≥50 ml) or regurgitant fraction (≥40%). The association between quantitative parameters (API and EROA) and the diagnosis of moderate-severe MI according to each imaging technique (echocardiography vs CMR) was evaluated using ROC curves. Results The prevalence of significant MR was 34.1% by echocardiography, and 21.8% by CMR. In 24% of cases the MR mechanism was functional. The median [Q1-Q3] API was 112 [141-166] units, and EROA 0.17 [0.10-0.30] cm2. The correlation between API and EROA was low (r = 0.34; p < 0.01). For the diagnosis of moderate-severe MR, EROA performed better than API taking the ASE algorithm as a reference (AUC 0.89 vs 0.75; Figure 1A), but both parameters showed a similar performance when compared to the CMR criterion ( AUC 0.80 vs 0.75; Figure 1B). Conclusions The low correlation between the API method and EROA raises their potential complementarity in the diagnosis of MR. Regarding the ASE multiparametric criterion by echocardiography, the EROA is superior to the API (being a key element in this algorithm). However, both methods present a similar association with the diagnosis of moderate-severe MR by CMR, and the assessment of both parameters could improve the agreement between echocardiography and CMR.
Read full abstract