Stroke is one of the leading causes of death and disability worldwide. It is the third leading cause, following heart diseases and cancer. Stroke is considered a medical emergency and can cause permanent neurological damages, complications related to mobility, cognition, sight or communication, and often death.
The majority of strokes are classified as ischemic, which has two types: thrombotic and embolic. In a thrombotic stroke, the blood clot forms in one of the arteries, while the embolic stroke occurs when a blood clot that forms somewhere else in the body breaks loose and travels to the brain through the bloodstream. Hemorrhagic stroke is another type of brain stroke that happens when an artery in the brain leaks blood or ruptures.
The increase in stroke incidence imposes a huge economic burden on individuals and society because stroke patients have longer hospital stays, higher re-admission rates, and medical expenditure than patients with other chronic diseases. Therefore, awareness of stroke warning signs and appropriate actions in the event of a stroke improves outcomes.
One main technique used to diagnose the clot is the brain computed tomography scan or brain CT scan, which uses x rays to take clear, detailed pictures of the patient brain. It is mainly done immediately after a stroke is suspected. Magnetic Resonance Imaging (MRI) is the second popular test to examine brain strokes. MRI is based on magnets and radio waves used to produce pictures of the organs and structures in the patient’s body.
Is it possible to predict and detect a stroke in advance? Thanks to artificial intelligence (AI) machine learning (ML), companies today have developed sufficient sophisticated ML techniques to leverage, integrate, and optimize patient data amassed in distributed electronic health record (EHR) databases and voluminous clinical, imaging, and laboratory datasets, among others to predict disease incidence and prognosis.
The main objective of these machine learning algorithms is to develop computer software that can adapt and learn patterns from large and multidimensional medical datasets, including clinical, biological (genetic, immunological, and serological markers), in addition to imaging information.
These huge, ever-changing datasets are difficult to interpret and often lack value. Still, they are evidently important for ML models to define stroke characteristics and predict stroke prognosis.
The majority of the models use a decision tree and k nearest neighbor (KNN) algorithms to accurately classify the strokes and better understand too many variables and avoid causal factors. Timely prediction of the stroke helps neuro-physicians identify high-risk patients and guide treatment approaches, leading to decreased morbidity.
Several AI-based techniques have been utilized to develop automated platforms to predict prognosis and functional outcome. Eunjeong Park and others have proposed a Bayesian Network Model for predicting post-stroke outcomes with available risk factors. They also introduced an online “Yonsei Stroke Outcome Inference System” for predicting functional independence at 3 months and mortality within 1 year in patients with stroke using the Bayesian Network Model.
Since a large number of clinical practice guidelines for IS treatment necessitates the requirement of a unified model for stroke treatment to assist in the clinical decision-making process, Alexa Love and others have developed a unified treatment model derived through a review of existing clinical practice guidelines, meta-analyses, and clinical trials using a Bayesian belief network.
Joshua Sarfaty Siegel and others have developed an AI-based model to predict behavior parameters such as attention, visual memory, verbal memory, language, motor, and visual in stroke patients using resting Functional Connectivity (FC) and lesion topography. The visual memory and verbal memory were better predicted by FC, whereas visual and motor impairments were better predicted by lesion topography. The study provides direct connections between key organizational features of brain networks to brain-behavior relationships in stroke.
Jane M Rondina and others applied a Gaussian Process Regression tool to develop a suitable methodology for predicting long-term motor outcomes from early post-stroke structural MRI.
AI and ML techniques in stroke imaging could markedly change the milieu of stroke diagnosis and management in the near future. AI’s ability to provide clinically relevant output information solely depends on the correctness of the input data and the machine learning method used to train the AI model. Yet, the timely diagnosis of stroke is vital for functional recovery and to minimize mortality.
In addition, foreseeing the degree of post-stroke recovery in advance and informing patients/family members of the prognosis may enhance the treatment rapport and rehabilitation processes.
New advancements in imaging technology for stroke diagnosis have led to the availability of a large volume of scattered neuroimaging information. Here, artificial intelligence and machine learning have been employed in several ways to extract the most coherent information, which can be used as an identifier or marker for stroke diagnosis and for analyzing its severity.
