Large vessel acute ischemic stroke (AIS) is the leading cause of disability and fourth leading cause of death with approximately 250,000 per year in the United States alone. Although treatments including interventional (stentriever) and pharmaceutical (tPA) exist for large vessel occlusion (LVO), there is not a rapid prehospital tool for diagnoses which guides the use of the correct therapy. Because the time to treatment for stroke is critical, there is a large need for a device which can be used in ambulances and other prehospital environments so that patients are rapidly provided the correct care. Currently, computed tomography angiography (CTA) imaging is the standard of care but its use is limited to the hospital or in some cases multimillion dollar stroke on wheels ambulances which are not practical. There is a significant delay that occurs between stroke onset and diagnosis by the time CTA is performed and read resulting in unnecessary death or disability. More delay occurs when patients need to be transferred to a stroke center because the hospital they are at does not perform interventional therapy. Rapid diagnosis in a prehospital setting will enable patients to be taken to the correct hospital and provide LVO location to surgeons in the operating suite. As the critical factor affecting cerebral anatomy in AIS patients is the limitation of blood flow to the brain the most likely technology to assist in diagnosis of LVO is one that can measure flow directly. Transcranial Doppler (TCD) has been used to directly measure blood flow velocity since 1982 but it relies on a trained operator skilled in both finding the affected vessel as well as interpreting a complicated waveform. Guidelines already exist for how to diagnose LVO with TCD and hospitals with a trained operator currently use it. However, it?s not used in smaller hospitals, urgent care or ambulance settings due to the training requirements. This work will revolutionize TCD for stroke diagnosis because it removes the need for a trained operator; key for mass adoption of the technology. TCD is already small and portable so it can easily fit in an ambulance similar in size to an automated external defibrillator. This work will use machine learning techniques we have developed to analyze the morphology of TCD waveforms and provide diagnosis information. The key is the ability to remove the operator variability by understanding how waveform morphology of stroke changes based on depth relative to occlusion so that the need to measure in a precise location is eliminated. Our preliminary results show 95% accuracy for LVO diagnosis but we need to broaden our dataset while updating our algorithms. Phase I will use our FDA Cleared Lucid M1 Transcranial Doppler Ultrasound System to measure confirmed stroke subjects after CTA imaging and update our algorithm. Phase II will use algorithms developed in Phase I with a fully automated TCD system we have developed with investor funding that does not require an operator or trained sonographer. We currently have Western IRB approval and are upgrading usability features of the device for ease of use with minimal training.
Large vessel acute ischemic stroke (AIS) is the leading cause of disability and fourth leading cause of death with approximately 250,000 per year in the United States alone. This work will reduce the time to treatment for stroke by developing necessary algorithms to diagnose AIS with a portable ultrasound technology which measures cerebral blood flow velocity. Successful completion of this project will result in a portable diagnostic device suitable for use in prehospital settings similar to automated external defibrillator for heart attack.
