Abstract
Purpose The purpose of this study is to evaluate the time reduction from hospital door to endovascular treatment (EVT) when using a non‐contrast CT (NCCT) basis artificial intelligence (AI) solution to classify and notify patients with anterior large vessel occlusion (LVO). It has the advantage of improving the prognosis by reducing the time required for treatment if patients of suspected LVO can be quickly screened and notified to clinicians in step of the NCCT test. Therefore, it was compared the time from the door of emergency room (ER) to reperfusion before and after to use of AI solution for evaluating clinical effectiveness aspects. Materials and Methods Patients over 19 years of age who visited a thrombectomy capable stroke center (TSC) with acute stroke symptoms and underwent EVT were enrolled this study. For the Post‐AI group, cases were collected prospectively after applying the AI solution [2022.05.01~2023.12.31], and cases with similar age, gender, and NIHSS with Post‐AI group were retrospectively collected as the Pre‐AI group [2020.05.01~2022.04.31]. The main comparison point was time difference from ER door to endovascular treatment between Pre‐AI and Post‐AI groups. Additionally, time from ER door to CT scan, CT scan to Stroke team treatment (STT), and STT to EVT were also compared between groups. Differences in time between groups were analyzed using an unpaired t‐test with Welch's correction. Results A total of 25 EVT cases were enrolled as the Post‐AI group, and 70 cases were retrospectively selected after 1:3 propensity score matching with age, gender and NIHSS in EVT cases. As a result of primary endpoint, time from ER door to EVT was significantly different between Pre‐AI (174.7±75.0 min.) and Post‐AI (147.7±31.6 min) [p=0.0155]. In detail results of time comparison, the time from CT scan to Stroke team treatment (20.2±7.9 min. vs. 35.4±41.3 min., p=0.0043) and the time from CT scan to EVT (127.7±29.2 min. vs. 153.9±71.1 min., p=0.0127) were significantly reduced in Post‐AI group. Conclusions and Limitations It has been confirmed that quickly screening LVO patients and notifying clinicians by an AI solution at the very front of the clinical process can significantly reduce the time from ER admission to EVT, and it will influence to prognosis outcome. Therefore, it is highly recommended to use for effective workflow in clinical environment. This study demonstrated how AI improved the hyperacute endovascular treatment workflow, by showing the impact on reducing ER door‐to‐reperfusion time. Although the study results were limited in generalizability as it was evaluated in a single TSC, we expect that it will be particularly valuable in remote regions where clinical experts may be limited.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call