A Multi-Institution Analysis Demonstrates that Intelligent Maps Improve Intraoperative Safety and Efficiency During
Dean J Arnaoutakis1, Rohini J Patel2, Mohammed Hamouda2
1Division of Vascular Surgery, University of South Florida.
Introduction:
Augmented intelligence (AI) tools have become popular additions to medicine. AI can transform operating rooms with advanced visualization and aid surgeons with decision-making. The AI tool we are investigating in this study allows patient specific maps to be created with a cloud computing software based on preoperative CT scans which are superimposed with fluoroscopic images in the operating room. The image fusion arguably permits easier identification and cannulation of visceral vessels. A unique feature of this software is that the map can be manually moved and adjusted in the operating room as the aorta is manipulated. The goal of our study is to evaluate the impact of AI on the intraoperative outcomes of fenestrated/branched endovascular aortic repairs (F/BEVARs).
Methods:
This is a multi-institution review of patients who underwent a F/BEVAR at tertiary care centers from August 2015-December 2022. Patients were stratified based on the use of AI during F/BEVAR. Primary outcomes included operative time, radiation exposure, fluoroscopy time, and contrast use. Secondary outcomes included 30-day complications. Multivariate linear regression models were used to evaluate the association between AI and outcomes after adjustment for potential confounders.
Results:
There were 344 patients included in this analysis; 248 (72.1%) patients underwent procedures using AI, and 96 (27.9%) had procedures without AI. There were no significant differences in baseline characteristics between the two groups. There was a significant difference in aneurysm diameter at time of repair, 65.3mm in the AI group and 64.1mm in the non-AI group, p=0.047. On univariate analysis there was not a significant difference in operative time, fluoroscopy time or radiation exposure in the non-AI group compared to the AI group. After adjusting for potential confounders, we found that the AI group had a significant reduction in contrast use - 62.4cc (p<0.0001), radiation exposure - 916.4 mGy (p=0.002), and fluoroscopy time - 16.6 minutes (p=0.001). Additionally, no significant difference was found in 30-day events between the groups.
Conclusion:
Incorporating AI technology into the operating room during F/BEVAR can result in a significant reduction in contrast use, radiation exposure and fluoroscopy time. Therefore, AI can improve safety in the operating room for both patient and providers and should be considered as an adjunct to F/BEVAR cases. Additional studies are needed to confirm our findings and further extrapolate this benefit.