COMPARISON OF ULTRASOUND AND CT SCAN DETERMINED AAA DIAMETERS FOLLOWING ENDOVASCULAR AAA REPAIR

OBJECTIVE: To evaluate the correlation of ultrasound(US)-derived aortic aneurysm diameter measurements with centerline, three dimensional M2S CT measurements following endovascular aortic aneurysm repair(EVAR).
METHODS: A review of concurrent CT and US examinations from patients undergoing post-EVAR surveillance was performed. The aortic aneurysm diameter was defined as the minor axis on the reconstructed M2S centerline CT images. This was compared to ultrasound-derived minor and major axis measurements, as well as to the minor axis measurement on the conventional CT images. Linear regression analyses were performed to evaluate correlation, and Bland-Altman plots were performed to assess agreement between different imaging modalities and measurements.
RESULTS: The correlation coefficients from linear regression analyses were 0.915, 0.916, 0.937, and 0.940; between CT minor/US minor; CT minor/US major; CT centerline minor/US minor; CT centerline minor/US major; respectively. CT minor correlation with CT centerline minor was 0.937. Bland-Altman plots showed the mean difference of 1.70mm between US minor axis and CT centerline measurements, compared to 3.78mm between US major axis and CT centerline measurements. This suggested that US minor axis measurements were in better agreement with CT centerline measurements, than US major axis. Variability between major and minor US and CT centerline diameter measurements were high with a standard deviation of 3.93~4.84mm. However, high variability was also observed between CT and centerline CT measurements with a standard deviation of 2.86mm.

CONCLUSIONS: Minor axis aneurysm diameter measurement obtained by ultrasound for surveillance following EVAR correlates well with centerline M2S CT diameters and most accurately reflects the true aortic aneurysm diameter albeit with a significant standard deviation. CT minor axis measurements also correlate well with centerline M2S, but with some degree of variability.