PT  - JOURNAL ARTICLE
AU  - Milad El Haddad
AU  - Tine De Backer
AU  - Marc De Buyzere
AU  - Daniel Devos
AU  - Abigail Swillens
AU  - Patrick Segers
AU  - Frank Timmermans
TI  - Grading of mitral regurgitation based on intensity analysis of the continuous wave Doppler signal
AID  - 10.1136/heartjnl-2016-309510
DP  - 2017 Feb 01
TA  - Heart
PG  - 190--197
VI  - 103
IP  - 3
4099  - http://heart.bmj.com/content/103/3/190.short
4100  - http://heart.bmj.com/content/103/3/190.full
SO  - Heart2017 Feb 01; 103
AB  - Objectives Echocardiographic methods are used to quantify mitral regurgitation (MR) severity; however, their applicability, accuracy and reproducibility have been debated. We aimed to develop and validate a novel custom-made transthoracic echocardiographic method for grading MR severity based on average pixel intensity (API) analysis of the continuous wave (CW) Doppler envelope.Methods MR was assessed in 290 patients using API, colour Doppler imaging, vena contracta width (VCW) and proximal iso-velocity surface area (PISA) method. For the validation of the API method, a pulsatile in vitro cardiac phantom was used.Results Indices of MR severity, such as left ventricular and atrial dimension, pulmonary arterial pressure, significantly cosegregate with API severity (p≤0.002). The API method showed a linear correlation with colour Doppler (r=0.79), VCW (r=0.68), PISA-effective regurgitant orifice area (r=0.72) and PISA-regurgitant volume (r=0.67); p<0.001 for all. The API was significantly more applicable than VCW (95% vs 75% of all patients; p<0.001) and PISA-based methods (65%; p<0.001). Additionally, the API showed a stronger intraobserver and interobserver agreement compared with other methods. Finally, in the in vitro validation, API values showed a strong linear correlation with increasing regurgitant volumes (r=0.81; p<0.001).Conclusions We showed the clinical feasibility and in vitro validation of a novel digital quantitative echocardiographic method to grade MR severity. This method is more applicable and has less interobserver and intraobserver variability compared with current quantitative methods.