In vitro and clinical validation of different correction algorithms for the two-dimensional proximal isovelocity surface area method in a low-velocity flow field for quantifying tricuspid regurgitation
Yu Liu,Li Liu,Beiqi Chen,Yuanfeng Wu,Rui Zhao,Wuxu Zuo,Quan Li,Fangmin Meng,Dehong Kong,Cuizhen Pan,Lili Dong,Xianhong Shu
DOI: https://doi.org/10.21037/qims-22-1311
2024-01-03
Quantitative Imaging in Medicine and Surgery
Abstract:Background: The 2-dimensional proximal isovelocity surface area (2D PISA) method underestimates tricuspid regurgitation (TR) severity. Previously proposed correction algorithms should be further scrutinized.Methods: Two correction algorithms were tested. One approach involves dividing the 2D PISA effective regurgitant orifice area by a constant of 0.7 (EROA0.7). Another approach involves multiplying the unadjusted EROA by Vorifice/(Vorifice - Valiasing), where Vorifice denotes the TR jet velocity, and Valiasing represents the color aliasing velocity (EROAVo-Va). In vitro validation was performed on a commercially available multifunctional valve tester with different size orifices and peak pressure gradients. A true EROA was derived through the regurgitant volume (RVol) calculated from the tester. For clinical validation, RVol was calculated as the difference between the overall stroke volume and the forward stroke volume of the right ventricle. Volumetric EROA was derived by dividing the RVol by the TR velocity-time integral (VTI). The vena contracta area (VCA) was obtained through direct planimetry with 3D echocardiography. The mean of volumetric EROA and VCA served as the reference in clinical validation.Results: Excellent correlation between the calculated EROAs and the true EROA was observed in vitro (r=0.98, r=0.97, and r=0.98 for uncorrected EROA, EROAVo-Va, and EROA0.7, respectively; all P values <0.0001). EROAVo-Va underestimated the true EROA and averaged 33% (P=0.3163), while EROA0.7 overestimated the true EROA and averaged 8% (P=0.0032). Clinically, these methods consistently exhibited a notable underestimation that varied with the reference EROA. This systematic underestimation was mitigated by both algorithms when either the VCA (biases of 19.6, 15.1, and 11.8 mm2 for uncorrected EROA, EROAVo-Va, and EROA0.7, respectively) or the volumetric EROA (biases of 10.1, 5.6, and 2.3 mm2 for uncorrected EROA, EROAVo-Va, and EROA0.7, respectively) served as the reference. Their ability to distinguish severe TR was similar, with area under the curve values of 0.905, 0.903, and 0.893 for uncorrected EROA, EROAVo-Va, and EROA0.7, respectively. No statistically significant differences were observed for diagnostic accuracy (all P values >0.05).Conclusions: Using a correction factor of 0.7 in quantifying TR provides similar accuracy when compared to other techniques. This represents a valuable clinical tool for quickly correcting the underestimation of the 2D PISA method in TR. This simple method may increase the frequency of applying the correction and earlier recognition of patients with severe TR.
radiology, nuclear medicine & medical imaging