Artificial Intelligence's next frontier could be stroke intervention

Stroke, also known as “brain attack”, is a global health crisis, with 15 million people worldwide suffering a stroke annually. Each year, at least 5 million die from stroke and another 5 million are permanently disabled. Stroke affects people of all ages and backgrounds, and their impact reverberates through families, communities and healthcare systems.

On October 29^th, World Stroke Day reminds us of the critical importance of addressing this challenge. This year, it is worth exploring the role that technology and innovation can play in helping us find new and better ways to address the problem of strokes.

The good news is that there have been many innovations in stroke treatments, ranging from portable MRIs for diagnosis to brain stimulation for rehabilitation, along with next-generation medications that are in development. Digital health tech has also shown great promise, ranging from stroke telemedicine (telestroke) to patient portals, internet of things, augmented reality and more. One up-and-coming innovation has been minimally-invasive endovascular thrombectomy (EVT) surgery, hailed as a revolutionary advancement.

The bad news is that most people worldwide do not have access to life-saving care at the right time. There’s a saying: in stroke, ‘time is brain,’ because every second counts. A stroke occurs when the blood supply to part of the brain is blocked or when a blood vessel in the brain bursts. Nearly 2 million neurons die every minute the brain does not receive oxygen from blood flow. For treatments like EVT to work effectively, they must be initiated as soon as possible.

Artificial Intelligence (AI) has shown immense promise in its potential to transform healthcare. This includes acute care conditions such as sepsis, which contributes to one out of three in-hospital deaths in the US, where organizations such as Bayesian Health and Johns Hopkins have demonstrated major reduction in sepsis mortality leveraging AI.

Viz.ai has aimed to address this opportunity in stroke by developing AI software that analyzes in real-time all the brain scans conducted on patients. Then, Viz.ai sends an immediate alert to the stroke care team on their mobile phones on an app. The care team then reviews the images on their phones and coordinates care using chat, all the while protecting private patient information so that patients can get to the right neurocare team for immediate treatment.

Diverse data is used to train the AI algorithms so that AI can detect suspected disease in real-time across all types of patients when a test is done. This test could be a brain scan for stroke or a heart test, such as an electrocardiogram, for cardiac disease. In the case of stroke, the AI algorithm is created based on computed tomography (CT) scans. Algorithms must achieve rigorous performance results to be cleared by regulatory agencies; in the United States, Viz.ai was the first to receive a de novo approval by the FDA for a computer-aided triage and notification platform to identify large vessel occlusion (LVO) strokes in CTA imaging. In 2021, the software also achieved CE Mark for use in Europe for stroke.

Entire healthcare systems are on the Viz.ai network, connecting smaller community hospitals to larger metropolitan institutions. This means that healthcare providers throughout this network of 1,500 hospitals covering 220 million lives can easily, confidentially share information about a patient and decide which hospital and care team will manage the patient’s care. Real-world clinical evidence has shown that this expedites treatment by as much as 102 minutes, resulting in a significant reduction in the amount of brain cell death. This is the difference between a stroke victim walking out of a hospital in good health or being disabled for the rest of their life.

The use of AI for stroke care coordination is considered the standard of care in the United States, according to the American Heart Association guidelines released in February 2023. However, there are still stroke centres remaining in the US that have not implemented AI software, and studies show that black and low-income patients may be less likely to receive endovascular thrombectomy. That’s why health equity is so critical as part of the rollout of AI in health systems.

While much progress has been made, health equity is a major concern. There is still a long way to go before AI-powered care coordination and access to immediate stroke treatment are available worldwide. Many developing countries face barriers such as limited infrastructure, healthcare workforce training and shortages and inadequate healthcare financing, leading to disparities in patient outcomes, including outcomes in stroke care and recovery.

The key to breaking these barriers is bringing together the broader ecosystem of stakeholders, such as healthcare providers, pharmaceutical and medical device companies, policymakers and patient advocacy groups, to affect change collaboratively. Efforts of the World Economic Forum and partners — such as the Global Health Equity Network (GHEN) and Digital Healthcare Transformation initiative — aim to do just that.

This World Stroke Day, collectively, we can increase global awareness of stroke and advance progress toward increasing access to better treatment. AI can help to reduce deaths and disability due to stroke for everyone — but only if we focus on true partnership for equitable digital transformation.