Abstract Body: Introduction
Autoregulation-guided blood pressure (BP) management offers a promising alternative to targeting fixed BP thresholds after acute ischemic stroke. However, the time required to compute personalized BP goals has hindered its broad application at the bedside for post-thrombectomy patients. To address this gap, our study compared the ability of three computational algorithms to rapidly determine a personalized BP target without compromising reliability.

Methods
From our prospectively maintained stroke database, we identified 206 patients who underwent thrombectomy for an anterior circulation large-vessel occlusion and had continuous physiologic data recordings for 24 hours post-thrombectomy. The cerebral autoregulatory index was used to calculate and trend the mean arterial pressure (MAP) at which autoregulation was most preserved (MAPopt) using three distinct algorithms (Yale, modified COGiTATE phase II, and Beqiri et al. 2024). Personalized BP targets were compared for the initial hour following the computation of MAPopt with each algorithm and throughout the 24-hour post-thrombectomy period. We analyzed the time to compute the first personalized MAPopt target and the reliability of MAPopt values over 24 hours using Bland-Altman analysis and intraclass correlation coefficient (ICC) analysis. One-way repeated measures ANOVA was conducted to examine differences in MAPopt values.

Results
The first personalized BP target value could be calculated using the Yale algorithm after a median of 86 minutes, compared to 121 and 184 minutes for the Beqiri et al. and COGiTATE algorithms (p < .05). The Yale algorithm also significantly increased the number of patients for whom a personalized BP target could be calculated and provided more continuous MAPopt values (Table 1). The mean and standard deviations of MAPopt were similar across the algorithms over both the 1-hour and 24-hour periods (p > .05, Table 1). MAPopt values calculated by the Yale algorithm were consistent with those calculated by the other two algorithms, with a slight bias in the Bland-Altman analysis (0.05 mmHg for Yale vs. COGiTATE and 0.61 mmHg for Yale vs. Beqiri et al.) and an ICC of 0.89 (95% CI 0.86 - 0.91), p < .05.

Conclusions
Our results demonstrate that the Yale algorithm computes a more continuous MAPopt and achieves target values faster without compromising reliability, making it a viable option for early and reliable personalized BP management in post-thrombectomy patients.