|
Back to ePosters
Validation of Geometric Solid Models for Planning Endovascular Repair of Complex Aortic Aneurysms
Omid Jazaeri, MD1, Rajan Gupta, MD2, Jennifer L. Wagner, AAS, RT, BS3, Khurram Rasheed, MD4, Michael Richards, PhD4, Robin Shandas, PhD3, Ankur Chandra, MD4.
1University of Colorado School of Medicine, Department of Surgery, Aurora, CO, USA, 2University of Colorado School of Medicine, Department of Radiology, Aurora, CO, USA, 3University of Colorado School of Engineering, Department of Bioengineering, Aurora, CO, USA, 4University of Rochester School of Medicine, Department of Surgery, Rochester, NY, USA.
OBJECTIVES:
Human anatomy is often difficult to appreciate with 2-dimensional imaging. Preservation of spatial relationships is crucial to the success of endovascular aneurysmal repair. To gain an appreciation for complex aortic anatomy, optimize graft placement, and assist with surgical planning, 3-dimensional solid models should prove valuable tools for increasing efficacy and efficiency of vascular treatment plans as well as for promoting a personalized approach to treatment. However, there are currently no standards or validation techniques to evaluate 3D reconstruction algorithms and in silico modeling methods. Original algorithms were developed to generate solid models in a time and cost effective manner. These processes were tested in 5 individuals previously treated for complex aortic anatomy using fenestrated devices.
METHODS:
Computed tomography scans of 5 patients previously treated with fenestrated endografts were retrospectively modeled by both University of Colorado Department of Bioengineering and University of Rochester Center for Medical Technology and Innovation. Standard measurements for device planning, as directed per manufacturers IFU, were compared to 3-D measurements based on 2 different methods of generating 3-dimensional solid models. Differences in measurements and the resulting graft configurations were examined.
RESULTS:
Measurements obtained from computer models and 3-D physical models are congruent with conventional measurements for all 5 patients. Differences observed (0-5mm) are due to different methods of measurement (along surface versus centerline) and inter-observer error. These differences did not translate into different endograft configurations when compared to patients treated. Figure 1 shows data from a representative patient.
CONCLUSIONS:
Sizing measurements taken from solid models consistently correlate with standard sizing measurements supporting further investigation into use of 3-D models as important tools to improve accuracy in endograft customization for AAA repair. Understanding of measurement differences could be used to advance planning practices for sizing standard devices, investigational devices and physician modified endografts, as directed by an institutional IDE. Future work will include quantification of different error sources and comparison of measurement procedures using in silico and in vitro models.
Back to ePosters
|
