Kumar, Estine and Prakash, Surya and Sharma, Anuraganand (2025) Stroke Classification Using 2-D Convolutional Neural Networks. In: Communication and Intelligent Systems. Lecture Notes in Networks and Systems . Springer Nature, Singapore, pp. 507-521. ISBN 978-981-96-5728-5
Full text not available from this repository.Abstract
A cerebrovascular accident, also known as stroke, is the second leading cause of death and the most common cause of disability globally. Stroke occurs when a blood vessel supplying the brain with oxygen and nutrients is blocked by a clot or ruptures, resulting in lasting brain damage, long-term disability, or even death. Hence, early diagnosis of such disease is a subject of great importance and extensive research among researchers. Effectively classifying stroke and non-stroke patients through clinical records necessitates a diagnosis test with a high level of accuracy and precision for a reliable prediction. In recent years, deep learning methodologies like convolutional neural networks have gained prominence due to their cutting-edge performance in various computer vision tasks such as visual object classification, detection, and segmentation. CNNs are designed for processing an ensemble of 2-D matrices whose attributes show correlation with related elements within image data. However, non-image data examples represented as sets of 1-D vectors (or tabular data) cannot be directly utilized with CNNs. In our study, we enhance an existing method called non-negative matrix factorization (NMF), which was originally used for manual feature extraction, to transform clinical data related to stroke into structured matrix formats suitable for image generation such that CNN could be used for stroke prediction thereafter. Our enhanced algorithm, called image-based non-negative matrix factorization (INMF), enables the use of convolutional neural networks (CNN) for stroke prediction. The processed and transformed data using INMF along with CNN demonstrates promising results, surpassing the performance levels achieved by other well-known machine learning algorithms on the stroke healthcare dataset.
| Item Type: | Book Chapter |
|---|---|
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | School of Information Technology, Engineering, Mathematics and Physics (STEMP) |
| Depositing User: | Ms Shalni Sanjana |
| Date Deposited: | 05 Nov 2025 00:07 |
| Last Modified: | 05 Nov 2025 00:07 |
| URI: | https://repository.usp.ac.fj/id/eprint/15191 |
Actions (login required)
 |
View Item |