Search

Background Delivery of acute stroke endovascular intervention can be challenging because it requires complex coordination of patient and staff across many different locations. In this proof-of-concept paper we (a) examine whether WiFi fingerprinting is a feasible machine learning (ML)-based real-time location system (RTLS) technology that can provide accurate real-time location information within a hospital setting, and (b) hypothesize its potential application in streamlining acute stroke endovascular intervention. Methods We conducted our study in a comprehensive stroke care unit in Melbourne, Australia that offers a 24hour mechanical thrombectomy service. ML algorithms including K-nearest neighbors, decision tree, random forest, support vector machine and ensemble models were trained and tested on a public WiFi dataset and the study hospital WiFi dataset. The hospital dataset was collected using the WiFi explorer software (version 3.0.2) on a MacBook Pro (AirPort Extreme, Broadcom BCM43xx1.0). Data analysis was implemented in the Python programming environment using the scikit-learn package. The primary statistical measure for algorithm performance was the accuracy of location prediction. Results ML-based WiFi fingerprinting can accurately predict the different hospital zones relevant in the acute endovascular intervention workflow such as emergency department, CT room and angiography suite. The most accurate algorithms were random forest and support vector machine, both of which were 98% accurate. The algorithms remain robust when new data points, which were distinct from the training dataset, were tested. Conclusions ML-based RTLS technology using WiFi fingerprinting has the potential to streamline delivery of acute stroke endovascular intervention by efficiently tracking patient and staff movement during stroke calls. [ABSTRACT FROM AUTHOR]

Background and Purpose- Mobile stroke units (MSUs) are increasingly used worldwide to provide prehospital triage and treatment. The benefits of MSUs in giving earlier thrombolysis have been well established, but the impacts of MSUs on endovascular thrombectomy (EVT) and effect on disability avoidance are largely unknown. We aimed to determine the clinical impact and disability reduction for reperfusion therapies in the first operational year of the Melbourne MSU. Methods- Treatment time metrics for MSU patients receiving reperfusion therapy were compared with control patients presenting to metropolitan Melbourne stroke units via standard ambulance within MSU operating hours. The primary outcome was median time difference in first ambulance dispatch to treatment modeled using quantile regression analysis. Time savings were subsequently converted to disability-adjusted life years avoided using published estimates. Results- In the first 365-day operation of the Melbourne MSU, prehospital thrombolysis was administered to 100 patients (mean age, 73.8 years; 62% men). The median time savings per MSU patient, compared with the control cohort, was 26 minutes ( P <0.001) for dispatch to hospital arrival and 15 minutes ( P <0.001) for hospital arrival to thrombolysis. The calculated overall time saving from dispatch to thrombolysis was 42.5 minutes (95% CI, 36.0-49.0). In the same period, 41 MSU patients received EVT (mean age, 76 years; 61% men) with median dispatch-to-treatment time saving of 51 minutes ([95% CI, 30.1-71.9], P <0.001). This included a median time saving of 17 minutes ([95% CI, 7.6-26.4], P =0.001) for EVT hospital arrival to arterial puncture for MSU patients. Estimated median disability-adjusted life years saved through earlier provision of reperfusion therapies were 20.9 for thrombolysis and 24.6 for EVT. Conclusions- The Melbourne MSU substantially reduced time to reperfusion therapies, with the greatest estimated disability avoidance driven by the more powerful impact of earlier EVT. These findings highlight the benefits of prehospital notification and direct triage to EVT centers with facilitated workflow on arrival by the MSU.