Early experience utilizing artificial intelligence shows significant reduction in transfer times and length of stay in a hub and spoke model

Background Recently approved artificial intelligence (AI) software utilizes AI powered large vessel occlusion (LVO) detection technology which automatically identifies suspected LVO through CT angiogram (CTA) imaging and alerts on-call stroke teams. We performed this analysis to determine if utilization of AI software and workflow platform can reduce the transfer time (time interval between CTA at a primary stroke center (PSC) to door-in at a comprehensive stroke center (CSC)). Methods We compared the transfer time for all LVO transfer patients from a single spoke PSC to our CSC prior to and after incorporating AI Software (Viz.ai LVO). Using a prospectively collected stroke database at a CSC, demographics, mRS at discharge, mortality rate at discharge, length of stay (LOS) in hospital and neurological-ICU were examined. Results There were a total of 43 patients during the study period (median age 72.0 ± 12.54 yrs., 51.16% women). Analysis of 28 patients from the pre-AI software (median age 73.5 ± 12.28 yrs., 46.4% women), and 15 patients from the post-AI software (median age 70.0 ± 13.29 yrs., 60.00% women). Following implementation of AI software, median CTA time at PSC to door-in at CSC was significantly reduced by an average of 22.5 min. (132.5 min versus 110 min; p = 0.0470). Conclusions The incorporation of AI software was associated with an improvement in transfer times for LVO patients as well as a reduction in the overall hospital LOS and LOS in the neurological-ICU. More extensive studies are warranted to expand on the ability of AI technology to improve transfer times and outcomes for LVO patients.

Download Full-text

Abstract TMP62: Early Experience Utilizing Artificial Intelligence Shows Significant Reduction in Transfer Times and Length of Stay in a Hub and Spoke Model

Stroke ◽

10.1161/str.51.suppl_1.tmp62 ◽

2020 ◽

Vol 51 (Suppl_1) ◽

Author(s):

Ameer E Hassan ◽

Victor M Ringheanu ◽

Rani Rabah ◽

Laurie Preston ◽

Adnan I Qureshi

Keyword(s):

Artificial Intelligence ◽

Length Of Stay ◽

Time Interval ◽

Time Of Arrival ◽

Vessel Occlusion ◽

Hub And Spoke ◽

Ct Angiogram ◽

Detection Technology ◽

Stroke Center ◽

Transfer Times

Background: Viz.ai artificial intelligence (AI) software utilizes AI powered large vessel occlusion (LVO) detection technology which automatically identifies suspected LVO through CT angiogram (CTA) imaging and alerts on-call stroke teams. We performed this analysis to determine if utilization of AI software can reduce the time interval between CTA at a primary stroke center (PSC) and arrival time at a comprehensive stroke center (CSC). Methods: We compared time interval between CTA and time of arrival for all LVO transfer patients from a single spoke PSC to our CSC prior to (Feb. 2017 to Nov. 2018) and after (Nov. 2018 to May 2019) incorporating Viz.ai. Using a prospectively collected stroke database at a CSC, demographics, transfer time (CTA time to time of arrival at CSC), modified Rankin Scale at discharge (mRS dc), mortality rate at discharge, length of stay (LOS) in hospital and neurological ICU, and intracranial hemorrhage rates were examined. Results: There were a total of 43 patients during the study period (average age 70.77 ± 12.54 yrs., 51.16% women). Analysis of 28 patients from the pre-Viz.ai (average age 71.64 ± 12.28 yrs., 46.4% women), and 15 patients from the post-Viz.ai (average age 69.13 ± 13.29 yrs., 60.0% women); see Table 1 for comparison of baseline characteristics and outcomes. Following implementation of Viz.ai, CTA time at PSC to time of arrival at CSC was significantly reduced by an average of 66 min. (mean CTA to time of arrival, 171 min. vs 105 min; P= 0.0163); significant reductions were also noted in the overall LOS (9.7 days vs 7.2 days; P= 0.0324) and LOS in the neurological ICU (6.4 days vs 2.9 days; P= 0.0039). Conclusions: The incorporation of Viz.ai was associated with a significant improvement in transfer times for LVO patients as well as a significant reduction in the overall hospital LOS and LOS in the neurological ICU. More extensive studies are warranted to expand on the ability of AI technology to improve transfer times and outcomes for LVO patients.

Download Full-text

Abstract P248: CSC Implementation of Artificial Intelligence Software Significantly Improves Door-In to Groin Puncture Time Interval and Recanalization Rates

Stroke ◽

10.1161/str.52.suppl_1.p248 ◽

2021 ◽

Vol 52 (Suppl_1) ◽

Author(s):

Ameer E Hassan ◽

Victor M Ringheanu ◽

Laurie Preston ◽

WONDWOSSEN TEKLE

Keyword(s):

Artificial Intelligence ◽

Treatment Time ◽

Care Center ◽

Mass Effect ◽

Comprehensive Care ◽

Time Interval ◽

Vessel Occlusion ◽

Ct Angiogram ◽

Detection Technology ◽

The Mean

Introduction: Viz.ai artificial intelligence (AI) software utilizes AI powered large vessel occlusion (LVO) detection technology which automatically identifies suspected LVO through CT angiogram (CTA) imaging and alerts on-call stroke teams. We performed this analysis to determine if utilization of this AI software can reduce the door-in to groin puncture time interval within the comprehensive care center (CSC) for patients arriving at the CSC for endovascular treatment. Methods: We compared the time interval between door-in to groin puncture for all LVO transfer patients who arrived at our comprehensive care center for approximately two years prior to and after the implementation of the AI software in November of 2018. Using a prospectively collected database at a CSC, demographics, door-in to groin time, modified Rankin Scale at discharge (mRS dc), mortality rate at discharge, length of stay (LOS) in hospital, mass effect, and hemorrhage rates were examined. Results: There were a total of 188 patients during the study period (average age 69.26 ± 14.55, 42.0% women). We analyzed 86 patients from the pre-AI (average age 68.53 ± 13.13, 40.7% women) and 102 patients from the post-AI (average age 69.87 ± 15.75, 43.1% women); see Table 1 for comparison of baseline characteristics and outcomes. Following the implementation of the AI software, the mean door-in to groin puncture time interval within the CSC significantly improved by 86.7 minutes (206.6 vs 119.9 minutes; p < 0.0001); significant improvements were also noted in the rate of good recanalization (mTICI 2B-3) for patients in the post-AI population (p=0.0364). Conclusion: The incorporation of the AI software was associated with a significant improvement in treatment time within the CSC as well as significantly higher rates of good recanalization for patients treated. More extensive studies are warranted to expand on the ability of AI technology to improve transfer times and outcomes for LVO patients.

Download Full-text

Abstract P266: The Implementation of Artificial Intelligence Significantly Reduces Door-In Door-Out Times in Primary Care Center Prior to Transfer

Stroke ◽

10.1161/str.52.suppl_1.p266 ◽

2021 ◽

Vol 52 (Suppl_1) ◽

Author(s):

Ameer E Hassan ◽

Victor M Ringheanu ◽

Laurie Preston ◽

WONDWOSSEN TEKLE

Keyword(s):

Artificial Intelligence ◽

Primary Care ◽

Care Center ◽

Time Interval ◽

Primary Care Center ◽

Vessel Occlusion ◽

Discharge Rates ◽

Ct Angiogram ◽

Detection Technology ◽

The Mean

Introduction: Viz.ai artificial intelligence (AI) software utilizes AI powered large vessel occlusion (LVO) detection technology which automatically identifies suspected LVO through CT angiogram (CTA) imaging and alerts on-call stroke teams. We performed this analysis to determine if utilization of AI software can reduce the door-in door-out (DIDO) time interval within the primary care center (PSC) prior to transfer to the comprehensive care center (CSC). Methods: We compared the time interval between door-in and door-out for all LVO transfer patients from a single spoke PSC to our CSC prior to (Feb. 2017 to Nov. 2018) and after (Nov. 2018 to June 2020) incorporating Viz.ai. Using a prospectively collected stroke database at a CSC, demographics, DIDO time at PSC, modified Rankin Scale at discharge (mRS dc), mortality rate at discharge, length of stay (LOS) in hospital and neurological ICU, and intracranial hemorrhage rates were examined. Results: There were a total of 63 patients during the study period (average age 69.99 ± 13.72, 55.56% women). We analyzed 28 patients from the pre-AI (average age 71.64 ± 12.28, 46.4%), and 35 patients from the post-AI (average age 68.67 ± 14.88, 62.9% women); see Table 1 for comparison of baseline characteristics and outcomes. Following the implementation of the AI software, the mean DIDO time interval within the PSC significantly improved by 102.3 minutes (226.7 versus 124.4 minutes; p=0.0374); significant reductions were not noted in mRS at discharge, rates of hemorrhage, or mortality. Conclusion: The incorporation of the AI software was associated with a significant improvement in DIDO times within the PSC and may lead to significant improvements in functional outcome and mortality in transfer patients. More extensive studies are warranted to expand on the ability of AI technology to improve transfer times and outcomes for LVO patients.

Download Full-text

Real-World Experience with Artificial Intelligence-Based Triage in Transferred Large Vessel Occlusion Stroke Patients

Cerebrovascular Diseases ◽

10.1159/000515320 ◽

2021 ◽

pp. 1-6

Author(s):

Jacob R. Morey ◽

Xiangnan Zhang ◽

Kurt A. Yaeger ◽

Emily Fiano ◽

Naoum Fares Marayati ◽

...

Keyword(s):

Artificial Intelligence ◽

Clinical Outcomes ◽

Warning System ◽

Large Vessel ◽

Time Interval ◽

Large Vessel Occlusion ◽

Time Intervals ◽

Vessel Occlusion ◽

Left Behind ◽

Stroke Center

Background and Purpose: Randomized controlled trials have demonstrated the importance of time to endovascular therapy (EVT) in clinical outcomes in large vessel occlusion (LVO) acute ischemic stroke. Delays to treatment are particularly prevalent when patients require a transfer from hospitals without EVT capability onsite. A computer-aided triage system, Viz LVO, has the potential to streamline workflows. This platform includes an image viewer, a communication system, and an artificial intelligence (AI) algorithm that automatically identifies suspected LVO strokes on CTA imaging and rapidly triggers alerts. We hypothesize that the Viz application will decrease time-to-treatment, leading to improved clinical outcomes. Methods: A retrospective analysis of a prospectively maintained database was assessed for patients who presented to a stroke center currently utilizing Viz LVO and underwent EVT following transfer for LVO stroke between July 2018 and March 2020. Time intervals and clinical outcomes were compared for 55 patients divided into pre- and post-Viz cohorts. Results: The median initial door-to-neuroendovascular team (NT) notification time interval was significantly faster (25.0 min [IQR = 12.0] vs. 40.0 min [IQR = 61.0]; p = 0.01) with less variation (p < 0.05) following Viz LVO implementation. The median initial door-to-skin puncture time interval was 25 min shorter in the post-Viz cohort, although this was not statistically significant (p = 0.15). Conclusions: Preliminary results have shown that Viz LVO implementation is associated with earlier, more consistent NT notification times. This application can serve as an early warning system and a failsafe to ensure that no LVO is left behind.

Download Full-text

Impact of Viz LVO on Time-to-Treatment and Clinical Outcomes in Large Vessel Occlusion Stroke Patients Presenting to Primary Stroke Centers

10.1101/2020.07.02.20143834 ◽

2020 ◽

Author(s):

Jacob R. Morey ◽

Emily Fiano ◽

Kurt A. Yaeger ◽

Xiangnan Zhang ◽

Johanna T. Fifi

Keyword(s):

New York ◽

Clinical Outcomes ◽

Large Vessel ◽

Similar Proportion ◽

Time Interval ◽

Large Vessel Occlusion ◽

Vessel Occlusion ◽

Time To Treatment ◽

Comprehensive Stroke Center ◽

Stroke Center

AbstractIntroductionRandomized controlled trials have demonstrated the importance of time-to-treatment on clinical outcomes in large vessel occlusion (LVO) stroke. Delays in interventional radiology (INR) consultation are associated with a significant delay in overall time to endovascular treatment (EVT). Delays in EVT are particularly prevalent in Primary Stroke Centers (PSC), hospitals without thrombectomy capability onsite, where the patient requires transfer to a Thrombectomy Capable or Comprehensive Stroke Center for EVT. A novel computer aided triage system, Viz LVO, assists in early notification of the PSC stroke team and affiliated INR team. This platform includes an image viewer, communication system, and an artificial intelligence algorithm that automatically identifies suspected LVO strokes on CTA imaging and rapidly triggers alerts.HypothesisViz LVO will decrease time-to-treatment and improve clinical outcomes.MethodsA prospectively maintained database was assessed for all patients who presented to a PSC currently utilizing Viz LVO in the Mount Sinai Health System in New York and underwent EVT following transfer for LVO stroke between October 1, 2018 and March 15, 2020. There were 42 patients who fit the inclusion criteria and divided into pre- and post-Viz ContaCT implementation by comparing the periods of October 1, 2018, to March 15, 2019, “Pre-Viz”, and October 1, 2019, to March 15, 2020, “Post-Viz.” Time intervals and clinical outcomes were collected and compared.ResultsThe Pre- and Post-Viz cohorts were similar in terms of gender, age, proportion receiving IV-tPA, and proportion with revascularization of TICI > 2B. The presenting NIHSS and pre-stroke mRS scores were not statistically different.The median initial door-to-INR notification was significantly faster in the post-Viz cohort (21.5 minutes vs 36 minutes; p=0.02). The median initial door-to-puncture time interval was 20 minutes shorter in the Post-Viz cohort, but this was not statistically significant (p=0.20).The 5-day NIHSS and discharge mRS were both significantly lower in the Post-Viz cohort (p=0.02 and p=0.03, respectively). The median 90-day mRS scores were also significantly lower post-Viz implementation, although a similar proportion received a good outcome (mRS score ≤ 2) (p=0.02 and p=0.42, respectively).ConclusionsEVT is a time-sensitive intervention that is only available at select stroke centers. Significant delays in time-to-treatment are present when patients require transfer from PSCs to a EVT capable stroke center. In a large health care system, we have shown that Viz LVO implementation is associated with improved time to INR notification and clinical outcomes. Viz LVO has the potential for wide-spread improvement in clinical outcomes with implementation across large hub and stroke systems across the country.

Download Full-text

Abstract TP6: Factors Influencing Infarct Progression During Ischemic Stroke Patients Transfer: Collaterals Lost is Brain Lost

Stroke ◽

10.1161/str.48.suppl_1.tp6 ◽

2017 ◽

Vol 48 (suppl_1) ◽

Author(s):

Gregoire Boulouis ◽

Arne Lauer ◽

Ahmer Khawdja Siddiqui ◽

Andreas Charidimou ◽

Robert Regenhardt ◽

...

Keyword(s):

Ischemic Stroke ◽

Vascular Occlusion ◽

Time Interval ◽

Stroke Severity ◽

Collateral Flow ◽

Stroke Patients ◽

Vessel Occlusion ◽

Main Determinants ◽

Initial Score ◽

Stroke Center

Introduction: When transferred from a referring hospital (RH) to a thrombectomy capable stroke center (TCSC), patients with initially favorable imaging profile (ASPECT score ≥6) often demonstrate infarct progression significant enough to make them ineligible at arrival. We sought to determine the clinical and imaging factors associated with this phenomenon in transferred ischemic stroke patients. Methods: We identified adult stroke patients transferred from one of 30 RH between 2010 and 2016 for which (1) a RH computed tomography (CT) and (2) a CT Angiography (CTA) at arrival were available for review. ASPECT scores were evaluated by 2 raters. The adequacy of leptomeningeal collateral flow was rated as none/poor, decreased, adequate or augmented per the Maas et al (Stroke 2009), modified scale. ASPECTS decay was defined as an ASPECT initial score ≥6 worsening between RH and TCSC CTs to a score <6. Results: A total of 330 patients were included in the analysis (mean age 70.2 ± 14.2, 43.3% females). Univariable subgroup analyses showed that patients with ASPECTs decay were more likely to be females (55% vs 40%, p=0.02), not on anticoagulants (4% vs 15%, p=0.01), and with higher initial NIHSS (Median [IQR] 19 [15.3-22] vs 11 [6-17], p<0.001), hyperdense vessel sign on initial CT (71% vs 26%, p<0.001) and poor collaterals on CTA (72% vs 19%, p<0.001). In multivariable models, higher NIHSS, lower baseline ASPECTs, CTA evidence of a proximal occlusion, and none/poor collaterals were strong predictors of ASPECTs decay, with collateral status demonstrating the highest odds ratio (aOR 10.3, 95%CI: [4.1-29], p<0.001). Similar results were found after stratification by vessel occlusion level. Conclusion: In ischemic stroke patients transferred for thrombectomy, poor collateral flow, stroke severity and proximal vascular occlusion, but not time interval, are the main determinants of ASPECTs decay.

Download Full-text

Automated emergent large vessel occlusion detection by artificial intelligence improves stroke workflow in a hub and spoke stroke system of care

Journal of NeuroInterventional Surgery ◽

10.1136/neurintsurg-2021-017714 ◽

2021 ◽

pp. neurintsurg-2021-017714

Author(s):

Lucas Elijovich ◽

David Dornbos III ◽

Christopher Nickele ◽

Andrei Alexandrov ◽

Violiza Inoa-Acosta ◽

...

Keyword(s):

Artificial Intelligence ◽

Usual Care ◽

Linear Regression Analysis ◽

System Of Care ◽

Large Vessel ◽

Arterial Puncture ◽

Large Vessel Occlusion ◽

Vessel Occlusion ◽

Hub And Spoke ◽

Time Savings

BackgroundEmergent large vessel occlusion (ELVO) acute ischemic stroke is a time-sensitive disease.ObjectiveTo describe our experience with artificial intelligence (AI) for automated ELVO detection and its impact on stroke workflow.MethodsWe conducted a retrospective chart review of code stroke cases in which VizAI was used for automated ELVO detection. Patients with ELVO identified by VizAI were compared with patients with ELVO identified by usual care. Details of treatment, CT angiography (CTA) interpretation by blinded neuroradiologists, and stroke workflow metrics were collected. Univariate statistical comparisons and linear regression analysis were performed to quantify time savings for stroke metrics.ResultsSix hundred and eighty consecutive code strokes were evaluated by AI; 104 patients were diagnosed with ELVO during the study period. Forty-five patients with ELVO were identified by AI and 59 by usual care. Sixty-nine mechanical thrombectomies were performed.Median time from CTA to team notification was shorter for AI ELVOs (7 vs 26 min; p<0.001). Door to arterial puncture was faster for transfer patients with ELVO detected by AI versus usual care transfer patients (141 vs 185 min; p=0.027). AI yielded a time savings of 22 min for team notification and a 23 min reduction in door to arterial puncture for transfer patients.ConclusionsAI automated alerts can be incorporated into a comprehensive stroke center hub and spoke system of care. The use of AI to detect ELVO improves clinically meaningful stroke workflow metrics, resulting in faster treatment times for mechanical thrombectomy.

Download Full-text

Abstract WP61: Artificial Intelligence for Automated Detection of Large Vessel Occlusion Strokes

Stroke ◽

10.1161/str.51.suppl_1.wp61 ◽

2020 ◽

Vol 51 (Suppl_1) ◽

Author(s):

Catarina Perry da Câmara ◽

Gabriel M Rodrigues ◽

Clara Barreira ◽

Mehdi Bouslama ◽

Leonardo Pisani ◽

...

Keyword(s):

Artificial Intelligence ◽

Detrimental Effect ◽

Poor Quality ◽

Automated Detection ◽

Large Vessel ◽

Large Vessel Occlusion ◽

Accurate Identification ◽

Vessel Occlusion ◽

The Mean ◽

Stroke Center

Introduction: Identification of Large Vessel Occlusion (LVO) in acute ischemic stroke (AIS) patients is critical for proper decision-making. Limited availability of trained experts and delays in LVO recognition can have a detrimental effect on outcomes. We sought to evaluate an artificial intelligence-based algorithm for LVO detection in AIS. Methods: A retrospective analysis of a prospectively-collected database of AIS patients admitted to a large volume stroke center between 2014-2018 was performed. Experienced vascular neurologists graded CTA for presence and site of LVO. Concurrently, studies were analyzed by the Viz-LVO Algorithm® version 1.4 (GA) - a convolutional neural network programmed to detect occlusions from the internal carotid artery terminus (ICA-T) to the sylvian fissure, which would include all MCA M1-segment and most M2-segment lesions. CTA readings were categorized as LVOs (ICA-T, MCA-M1, MCA-M2) versus non-LVOs/more distal occlusions. Comparisons between human and AI-based readings were done by accuracy analysis and calculating Cohen’s kappa. Results: A total of 610 CTAs were analyzed. The AI algorithm rejected 3.4% of the CTAs due to poor quality. Viz-LVO identified LVOs with an overall sensitivity of 81.3%, specificity of 87.8%, and accuracy of 83.2% (AUC 0.845 (95%CI:0.81-0.88, p<0.001). Table 1 shows the results per occlusion site. Accuracy was higher for ICA-T and M1 occlusions as compared to M2 occlusions. The mean run time of the algorithm was 2.78±0.5minutes. Conclusion: Our study demonstrates that automated AI reading allows for fast and accurate identification of LVO strokes. Future efforts should be made to improve the detection of the more distal occlusions.

Download Full-text

Abstract P149: Regional Transfer Protocols Can Improve Transfer Times and Reduce Length of Stay for Thrombectomy Candidates Seen on Telestroke

Stroke ◽

10.1161/str.52.suppl_1.p149 ◽

2021 ◽

Vol 52 (Suppl_1) ◽

Author(s):

Michele M Joseph ◽

Amanda L Jagolino-Cole ◽

Alyssa D Trevino ◽

Liang Zhu ◽

Alicia M Zha ◽

...

Keyword(s):

Length Of Stay ◽

Emergency Rooms ◽

Stroke Patients ◽

Arrival Times ◽

Time Periods ◽

Comprehensive Stroke Center ◽

Stroke Center ◽

One Year ◽

The Impact ◽

Transfer Times

Introduction: Our telestroke (TS) network instituted a regional transfer protocol (RTP) that allows for stroke patients in need of higher level of care to be pre-accepted and transferred to the nearest appropriate comprehensive stroke center (CSC). We studied the impact of the RTP on resource utilization and time metrics in patients transferred for evaluation of intra-arterial thrombectomy (IAT). Before the RTP, all potential IAT patients were transferred to one central CSC. After the RTP was initiated, the network had the capability to transfer to two additional CSCs within the same health system that are strategically located in the Houston area. Methods: We identified patients evaluated via TS in spoke emergency rooms that were subsequently transferred for IAT evaluation from 1/1/2016 to 12/31/2017 - one year prior and one year after the RTP. Baseline demographic characteristics, transfer and IAT metrics, and outcomes were compared for the two time periods. Results: Of 220 patients, 102 patients were transferred pre-RTP, and 120 were transferred to the three CSCs post-RTP. There were no significant differences in baseline characteristics, except fewer patients received tPA post-RTP (Table 1). In total, 30 patients (29%) pre-RTP and 42 patients (35%) post-RTP underwent IAT (p=0.38). Post-RTP, there was a trend toward faster travel times (median 40 vs 32 minutes, p=.07) and transfer initiation times to hub arrival times (median 109 vs 100.5 minutes, p=0.09). Door to groin puncture times were not statistically different between the two time periods. Post-RTP patients had a significantly shorter length of stay (median 6 vs 5 days, p=0.03). Conclusions: Regional transfer protocols can potentially help reduce transfer times and length of stay for stroke patients at CSCs that were initially seen by TS at community hospitals; however, larger sample size is needed to study its impact on other IAT-related metrics and clinical outcomes.

Download Full-text