The role of artificial intelligence (AI) in stroke management

October 29 was World Stroke Day. Stroke is a medical emergency that affects nearly one person every 5 seconds worldwide according to the WHO and the World Stroke Organization. This is why the WHO does not hesitate to speak of Pandemic and expects a gradual increase in the incidence of strokes in the world from 16 million cases in 2005 to nearly 23 million in 2030.[1].

A CVA (cerebrovascular accident) results in a sudden neurological deficit, caused by an interruption of blood circulation, either by bleeding (hemorrhagic stroke), or by the obstruction of an artery resulting in a deficit of oxygen and glucose, fuels necessary for the proper functioning of the brain. ischemic strokes account for approximately 80% of strokes.

Artificial Intelligence (AI) has the potential to improve lesion detection, reduce risk, and even improve the quality of life of stroke patients. Fortunately, there are already simple technological measures that allow us to anticipate, treat, and even avoid the consequences of this dramatic, disabling and fatal condition.

AI and stroke prediction models

About 9% of stroke patients are misdiagnosed, which is associated with a poor outcome. An article published in October 2020 entitled “Incorporating Artificial Intelligence into Stroke Care and Research”,[2] the authors highlighted a set of factors that make AI a major advance in the field of diagnostic and therapeutic management of stroke. This has the ability to predict the onset of a stroke, based on the patient’s clinical and biological data, to detect early on a large number of people at risk of stroke and above all to provide immense support to emergency physicians, in analysis and interpretation of radiological images. For example, if a radiologist has a hundred X-rays to interpret, the AI ​​algorithm would be able to classify them in order of severity, in order to quickly take charge of the most urgent cases.

The Canadian PLAKK project brings hope. Its ultrasound imaging technology combined with an artificial intelligence system would better assess the degree of blockage of the artery by more accurately measuring the composition and size of the plaque.

In the same vein, AI has been shown to shorten the time to diagnosis, it allows neurologists to quickly diagnose ischemic stroke, and provides physicians with a great ability to detect and prioritize even small lesions that may go unnoticed in emergency rooms. AI will never replace the human mind and its transversality, and above all its ability to detect nuances and identify hidden risks. But, it can be an invaluable ally in improving the prognosis of a stroke. This is particularly important because the earlier a stroke is detected, the better the prognosis for the patient. [3]

AI and the therapeutic management of stroke

Using AI algorithms, an acute ischemic stroke can be treated by removing the blood clot blocking the brain vessel using mechanical thrombectomy devices, the latter being advanced from the groin (femoral artery) and ascended to the brain along a rail called the guide. To improve the patient’s prognosis, the clot should be removed as soon as possible. Today, without knowing the characteristics of the clot, doctors can only remove it on the first try in 1/3 of cases. Thanks to AI, a connected guide coupling impedance micro-sensors with machine learning algorithms, would be able to instantly identify biological tissues with unparalleled predictive reliability and provide doctors with real-time critical information on the clot.[4]

In recent years, disruptive innovation has made major breakthroughs in healthcare, enabling great improvement in the diagnostic and therapeutic management of chronic, serious, disabling and costly diseases. However, further work is needed to perform the decision process, and the validity of AI and improve the acuity of its algorithms. Going forward, research and application of AI in stroke is expected to keep pace with growing clinical demand. More rigorous research is needed to assess the clinical applicability of AI systems and to assess their performance, and explore their impact on the quality of therapeutic management, improving access to care, preventing stroke, survival and quality of life of patients.



[3] Lee EJ, Kim YH, Kim N, Kang DW. Deep in the Brain: Artificial Intelligence in Stroke Imaging. AVC J. 2017 Sep;19(3):277-285. doi: 10.5853/jos.2017.02054. Published online September 29, 2017. PMID: 29037014; PMC ID: PMC5647643

[4] Al Saiegh, F, Munoz, A, Velagapudi, L, et al. Theofanis, Thana, Suryadevara, Neil, Patel, Priyadarshee, et al. Selection of patients and procedures for mechanical thrombectomy: towards personalized medicine and the role of artificial intelligence. J Neuroimaging. 2022; 32: 798–807.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *