Related Experiment Video
Updated: May 11, 2026
Automated Joint Space Detection Improves Bone Segmentation Accuracy
Published on: November 28, 2025
Evaluation of automated 2D-3D image overlay system utilizing subtraction of bone marrow image for EVAR: feasibility
1Department of Radiology, National Cerebral and Cardiovascular Center, Suita, Osaka, Japan.
Objective:
To evaluate the automated 2D-3D image overlay system ("3D Roadmap") for use during endovascular aneurysm repair (EVAR) in the hybrid operating theater.
Methods:
Datasets of preoperative CT images were modified to subtract dense bone marrow to improve the visualization of vasculature on the overlaid image, and allow for accurate navigation of the endovascular devices. The 3D-CT overlay image was registered on the 2D fluoroscopy image to mark the iliac crest and lumbar vertebrae on both images as landmarks.
Results:
Arteriography was performed only twice to confirm the precision of the position of renal artery and the final evaluation. Twenty patients underwent EVAR with Medtronic Endurant, Gore Excluder, or COOK Zenith using "3D Roadmap". The origin of the renal artery and iliac bifurcation were registered with complete accuracy in 10 patients (50%). The lower renal artery deviated toward the cranial side less than 3 mm in six patients. In all cases, EVAR was successful, and completed with the volume of contrast material limited to 43.8 ± 3.1 mL.
Conclusion:
"3D Roadmap" was confirmed to be valuable for visualization of vessel origin in a fused image and for reduction of contrast material during EVAR.
