An Artificial Intelligence-Based Bio-Medical Stroke Prediction and Analytical System Using a Machine Learning
R Pitchai1, Bhasker Dappuri2, P V Pramila3
1Department of Computer Science and Engineering, B V Raju Institute of Technology, Narsapur 502313, Telangana, India.
Computational Intelligence and Neuroscience
|October 24, 2022
Summary
This study introduces a machine learning algorithm using electromyography (EMG) data for stroke prediction. The developed system offers a more accessible and accurate method for early stroke detection compared to traditional imaging techniques.
Area of Science:
- Biomedical Engineering
- Machine Learning
- Neurology
Background:
- Stroke significantly impacts economic well-being and can be fatal if untreated.
- Abnormal biosignals are common in stroke survivors, indicating potential for monitoring.
- Current stroke diagnosis relies on expensive and complex imaging like CT or MRI.
Purpose of the Study:
- To develop a machine learning algorithm for brain stroke prediction using real-time electromyography (EMG) data.
- To offer a cost-effective and user-friendly alternative to traditional stroke diagnostic methods.
- To enhance the accuracy and efficiency of stroke detection through advanced data analysis.
Main Methods:
- A support vector machine (SVM) classifier was trained using synthetic EMG data.
- Data augmentation techniques were employed to generate extensive training datasets for improved accuracy.
- The model was tested using real-time EMG samples to evaluate its predictive performance.
Main Results:
- The proposed machine learning model demonstrated higher classification accuracy than existing methods.
- The SVM classifier achieved superior rates of precision, recall, and F-measure.
- Simulations confirmed the efficacy of the developed algorithm in predicting stroke from EMG data.
Conclusions:
- Machine learning analysis of real-time EMG signals presents a viable approach for stroke prediction.
- The developed SVM-based system offers improved accuracy and efficiency over conventional diagnostic tools.
- This method facilitates prompt patient therapy through precise, real-time biosignal assessment.
