Feasibility of tomographic freehand three-dimensional ultrasound for surveillance of abdominal aortic aneurysms after endovascular repair

Abstract

Objectives

This study aimed to investigate the feasibility of a commercially available tomographic freehand three-dimensional ultrasound (3D-US) system for surveillance of abdominal aortic aneurysms (AAAs) after endovascular repair (EVAR).

Methods

In 10 patients who underwent EVAR, a 3D-US scan was obtained post-operatively in addition to regular computed tomography angiography (CTA). Five independent observers evaluated 12 EVAR surveillance parameters for the 10 scans, resulting in a total of 600 individual observations.

Results

The diameters of the neck, sac and iliac arteries were evaluable in most of the observations (neck: 80%, AAA sac: 98%, right CIA: 90%, left CIA: 68%). The diameter measurements of the 3D-US strongly correlated with the diameters measured on CTA (ρ = 0.90, p < .001). The general offset of 3D-US compared to CTA was −4.4 mm and the proportional bias was 3%. For all observers, the diameter measurements were significantly strongly correlated (O1:ρ = 0.85, O2:ρ = 0.88, O3:ρ = 0.93, O4:ρ = 0.96, O5:ρ = 0.86, p < .001 for all). The inter-observer reliability was excellent with an overall concordance correlation coefficient of 0.98. The start of the endograft, the lowest renal artery, the distance between these landmarks, and the proximal and distal sealing zones were evaluable in fewer than half of the observations (42%, 12%, 10%, 32%, 32%, 26%).

Conclusion

3D-US after EVAR is feasible. Diameters measured in 3D-US correlate strongly with CTA-based diameters and have a good interrater variability. However, the sealing zones are difficult to assess on 3D-US. This technique could be a useful addition to duplex ultrasound to facilitate offline 3D analysis, increase measurement reproducibility, enable volume measurements, and minimise the use of harmful CTA for surveillance after EVAR.