Machine learning for brain-stroke prediction: comparative analysis and evaluation

• Published in: Multimedia tools and applications Vol. 84; no. 21; pp. 24025 - 24057
• Main Authors: Bhowmick, Rahul, Mishra, Soumya Ranjan, Tiwary, Sanjeeb, Mohapatra, Hitesh
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2025; Springer Nature B.V
• Subjects: Algorithms; Blood pressure; Brain; Brain research; Cholesterol; Computer Communication Networks; Computer Science; Data analysis; Data Structures and Information Theory; Electronic health records; Feature selection; High density lipoprotein; Hypertension; Investigations; Lipoproteins; Machine learning; Medical records; Mortality; Multimedia Information Systems; Neural networks; Prevention; Special Purpose and Application-Based Systems; Stroke; Track 2: Medical Applications of Multimedia; Veins & arteries; Stroke diagnosis; Neural networks; Biomarkers; Predictive analysis; K-nearest neighbors; Machine learning models; Support vectors
• ISSN: 1573-7721; 1380-7501; 1573-7721
• DOI: 10.1007/s11042-024-20057-6

This study focuses on the intricate connection between general health, blood pressure, and the occurrence of brain strokes through machine learning algorithms. To achieve this, we have thoroughly reviewed existing literature on the subject and analyzed a substantial data set comprising stroke patients. Implementing a combination of statistical and machine-learning techniques, we explored how general health indicators, including overall well-being and blood pressure, influence the risk of strokes. The findings of this study hold substantial implications for stroke prevention, treatment, and the development of novel diagnostic tools and therapies. Our ultimate aim is to gain fresh insights into the intricate interplay of general health and blood pressure, aiding in identifying individuals at risk of future brain strokes. This study entails a data-driven analysis of various algorithms across multiple datasets. Within this scope, we have thoroughly examined the behaviours and accuracy of diverse machine learning algorithms, assessing their interrelationships. This research aims to assist novice researchers in comprehending the performance of different machine learning algorithms in the context of brain stroke prediction.