Article Text
PDF
Abstract
Introduction Determining the ideal approach for mechanical thrombectomy is critical to the success of the procedure. Currently, an operator must decide between femoral or radial approach by evaluating preoperative vascular imaging. This choice is usually determined by the operator’s subjective experience. However, a more objective approach is possible by leveraging quantitate evaluation of the arterial anatomy. Many features that an operator qualitatively considers, such as the arch type or carotid tortuosity, can be described numerically. In this study, we evaluated a group of objective anatomical metrics to determine which correlated best with the operator’s perception of navigational difficulty.
Materials and Methods We deidentified a set of 47 anterior circulation thrombectomy cases with associated CTA images and operative notes of perceived navigational difficulty and time metrics (i.e., time to first pass). Difficulty was assessed for navigating the aortic arch, the neck, and the intracranial region, and respective scores ranged from 1 (easy) to 5 (impossible). With equal weighting, we combined each score to produce a single value for each case. We used Kendall’s tau, an analog to Pearson correlation for qualitative data, to measure correlations. Using 3DSlicer and MATLAB, we segmented major blood vessels, measured their centerlines and diameters, before generating a centerline for each possible path (i.e.: descending aortic arch to left internal carotid artery). For a given path, we measured its tortuosity index. We also computed local bending energy, which highlighted the 2 cm centerline window that exhibited the sharpest and greatest changes in direction. Finally, we again used Kendall’s tau to relate these metrics to the perceived difficulty of their paths.
Results When averaging across difficulty scores, time to first pass yielded a moderate correlation ( with perceived difficulty. Tortuosity index () and local bending energy () exhibited a much stronger relationship to the difficulty scores. The diagram on the right exhibits two paths with equal tortuosity; however energy is greater in the path with more turns.
Conclusion Surgical notes exhibited moderate correlation with duration. However, the bending energy of a centerline exhibited the strongest relationship to perceived navigational difficulty, indicating it may highlight features obfuscated in traditional metrics. Future studies will consider other metrics such catheter or wire properties to understand how the chosen devices impacted intervention difficulty.
Disclosures J. Rifkin: None. T. Ullah: None. N. McCann: None. A. Nilforoush: None. M. Panzer: None. R. Kellogg: None.