• Workflow improvements from aut...

*Result*: Workflow improvements from automated large vessel occlusion detection algorithms are dependent on care team engagement.

Title:
Workflow improvements from automated large vessel occlusion detection algorithms are dependent on care team engagement.
Authors:
Ebirim EC; The University of Texas Medical Branch at Galveston, Galveston, Texas, USA., Le NM; Department of Neurology, UTHealth Houston McGovern Medical School, Houston, Texas, USA., Samaha JN; Department of Neurology, UTHealth Houston McGovern Medical School, Houston, Texas, USA., Azeem H; Department of Neurology, UTHealth Houston McGovern Medical School, Houston, Texas, USA., Iyyangar A; Department of Neurology, UTHealth Houston McGovern Medical School, Houston, Texas, USA., Ballekere AN; Department of Neurology, UTHealth Houston McGovern Medical School, Houston, Texas, USA., Dhanjani S; Rice University, Houston, Texas, USA., Giancardo L; McWilliams School of Biomedical Informatics at UTHealth Houston, Houston, Texas, USA., Lee E; Department of Neurology, UTHealth Houston McGovern Medical School, Houston, Texas, USA., Sheth SA; Department of Neurology, UTHealth Houston McGovern Medical School, Houston, Texas, USA Sunil.A.Sheth@uth.tmc.edu.
Source:
Journal of neurointerventional surgery [J Neurointerv Surg] 2026 Jan 13; Vol. 18 (2), pp. 385-389. Date of Electronic Publication: 2026 Jan 13.
Publication Type:
Journal Article; Randomized Controlled Trial; Multicenter Study
Language:
English
Journal Info:
Publisher: BMJ Publishing Group Country of Publication: England NLM ID: 101517079 Publication Model: Electronic Cited Medium: Internet ISSN: 1759-8486 (Electronic) Linking ISSN: 17598478 NLM ISO Abbreviation: J Neurointerv Surg Subsets: MEDLINE
Imprint Name(s):
Original Publication: London : BMJ Publishing Group, c2009-
MeSH Terms:
Ischemic Stroke*/surgery , Workflow* , Patient Care Team* , Machine Learning* , Algorithms*, Endovascular Procedures/methods ; Thrombectomy/methods ; Humans ; Female ; Aged ; Male ; Middle Aged ; Prospective Studies ; Time-to-Treatment ; Detection Algorithms
References:
Stroke. 2022 May;53(5):1651-1656. (PMID: 34865511)
AJNR Am J Neuroradiol. 2023 Jan;44(1):47-53. (PMID: 36574318)
J Neurointerv Surg. 2023 Jan;15(1):52-56. (PMID: 35086962)
Cureus. 2018 Mar 10;10(3):e2300. (PMID: 29755897)
Stroke. 2024 Jun;55(6):1609-1618. (PMID: 38787932)
Stroke Vasc Interv Neurol. 2022 Sep 07;2(5):e000224. (PMID: 41584017)
J Am Heart Assoc. 2014 Nov 11;3(6):e000963. (PMID: 25389281)
Cerebrovasc Dis Extra. 2023;13(1):41-46. (PMID: 36787716)
JAMA Neurol. 2023 Nov 1;80(11):1182-1190. (PMID: 37721738)
J Neurointerv Surg. 2016 May;8(5):447-52. (PMID: 26863106)
J Neurointerv Surg. 2020 Feb;12(2):156-164. (PMID: 31594798)
J Clin Med. 2023 May 30;12(11):. (PMID: 37297949)
AJNR Am J Neuroradiol. 2021 Jan;42(2):247-254. (PMID: 33384294)
Stroke. 2023 Jun;54(6):1505-1516. (PMID: 37216446)
Stroke. 2020 Oct;51(10):3133-3137. (PMID: 32842922)
Stroke. 2019 Nov;50(11):3093-3100. (PMID: 31547796)
Grant Information:
R01 NS138765 United States NS NINDS NIH HHS
Contributed Indexing:
Keywords: CT Angiography; Stroke; Technology; Thrombectomy
Entry Date(s):
Date Created: 20250127 Date Completed: 20260113 Latest Revision: 20260213
Update Code:
20260213
PubMed Central ID:
PMC12892295
DOI:
10.1136/jnis-2024-022896
PMID:
39870518
Database:
MEDLINE

*Further Information*

*Background: Automated machine learning (ML)-based large vessel occlusion (LVO) detection algorithms have been shown to improve in-hospital workflow metrics including door-to-groin time (DTG). The degree to which care team engagement and interaction are required for these benefits remains incompletely characterized.
Methods: This analysis was conducted as a pre-planned post-hoc analysis of a randomized prospective clinical trial. ML-based LVO detection software was implemented at four comprehensive stroke centers (CSCs) from January 1, 2021, to February 27, 2022. Patients were included if they underwent endovascular thrombectomy for LVO acute ischemic stroke. ML software utilization was quantified as the total number of active users and the ratio of the number of comments to the number of patients analyzed by the software by site per week. Primary outcome was the reduction in DTG relative to pre-ML implementation by hospital utilization level. Data are expressed as median (IQR).
Results: Among 101 patients who met the inclusion criteria, the median age was 71 years (IQR 59-79), with 48.5% being female. CSC 4 had the greatest number of total active users per week (32.5 (27.5-34.5)), and comment-to-patient ratio per week (5.8 (4.6-6.9)). Increased ML software utilization was associated with improvements in DTG reduction. For every 1 unit increase in the comment-to-patient ratio, DTG time decreased by 2.6 (95% CI -5.09 to -0.13) min, while accounting for site-level random effects. Number of users-to-patient was not associated with a reduction in DTG time (β=-0.22, 95% CI -1.78 to 1.33).
Conclusions: In this post-hoc analysis, user engagement with software, rather than total number of users, was associated with site-specific improvements in DTG time.
(© Author(s) (or their employer(s)) 2026. No commercial re-use. See rights and permissions. Published by BMJ Group.)*

*Competing interests: SAS reports funding from the National Institutes of Health as well as consultancy fees from Penumbra, Viz.AI and Imperative Care for unrelated topics.*