AN ARTIFICIAL INTELLIGENCE APPROACH FOR PREDICTING DIFFERENT TYPES OF STROKES

AN ARTIFICIAL INTELLIGENCE APPROACH FOR PREDICTING DIFFERENT TYPES OF STROKES

Mrs. G. Ramya Asst.Prof1, Ramavath Rajesh2, Yenugu Akshitha3, Talati Sneha4, Nrusimhadri Hrushikesh5

Sreyas Institute of Engineering and Technology1 Assistant Professor (AIML)

ABSTRACT:
Stroke is a chief global health concern, necessitating precise diagnosis and treatment. This looks at introduces a modern-day synthetic intelligence (AI) approach for predicting severe stroke forms of the use of advanced gadget studying techniques. Leveraging a comprehensive dataset encompassing clinical facts, neuroimaging scans, and medical information, our multi- modal AI version harnesses deep getting to know architectures to extract hard patterns and relationships. It demonstrates awesome talent in discriminating amongst ischemic, hemor- rhagic, and transient ischemic stroke types, achieving an accuracy price of 75%. Beyond stepping forward accuracy, our method gives a promising tool for healthcare professionals, helping in early analysis and remedy desire for various stroke lessons, ultimately improving patient care and evaluation. Stroke, a prime worldwide fitness task, necessitates precise prog- nosis and treatment. This seems at introduces a present-day synthetic intelligence (AI) method for predicting numerous stroke types of the use of advanced device reading tech- niques. Leveraging a complete dataset encompassing clinical records, neuroimaging scans, and scientific facts, our multi-modal AI version harnesses deep learning architectures to ex- tract difficult patterns and relationships.

Keywords: Stroke Prediction, Artificial Neural Networks (ANN) with TensorFlow, Logistic Regression, Cross-Validation, Support Vector Machine (SVM) with Scikit-Learn.