Keywords:
Performed at one institution, Cross-sectional study, Retrospective, Image verification, Technology assessment, Technical aspects, Comparative studies, Echocardiography, CT, Cardiac, Cardiology
Authors:
A. inakami1, K. miyauchi1, Y. katsuyama1, R. ogawa1, M. HABUCHI1, T. washio1, S. Takahashi2, A. Matsumoto3; 1hyogo/JP, 2Suita/JP, 3nagoya/JP
DOI:
10.26044/ecr2020/C-02841
Results
Each three cases were excluded due to lack of echocardiography data for TAPSE and RVFAC.
Two cases were excluded due to measurement difficulties at CT.
Therefore, there are obtained 127 datasets for TAPSE and 127 datasets for RVFAC.
Calculation of Pearson’s correlation coefficient (r) revealed a moderate correlation between CT derived values and those with echocardiography (TAPSE, r=0.445, p<.001; RVFAC, r=0.469, p<.001) (Fig.6).
Fig. 6: Scatter plots. Correlation of CT values and Echocardiography values in TAPSE and RVFAC. The graphs show moderate correlation (TAPSE, r=0.445, p<.001; RVFAC, r=0.469, p<.001) for all displayed parameters.
Bland-Altman plots show good agreement for CT and echocardiography, although CT derived values were significantly smaller than those with echocardiography (TAPSE, Bias=-18.0% p<.001; RVFAC, Bias=-5.9% P<.001) (Fig.7).
Fig. 7: Bland-Altman plot. The graph shows good inter-technique agreement for CT and echocardiography,although MDCT derived values were significantly smaller than those with echocardiography (TAPSE, Bias=-18.0% p<.001; RVFAC, Bias=-5.9% P<.001).
The reproducibility of inter-observer measurement values was almost perfect for both TAPSE (ICC=0.919, α=0.941) and RVFAC (ICC=0.869, α=0.894).