Research and Application of Deep Learning Models for Detecting Brain Abnormalities Based on CT Images

This paper studies a method to improve the accuracy of a deep learning model for detecting brain abnormalities based on computed tomography images. The process begins with image preprocessing using the Histogram Equation algorithm and Gabor filter. Then, features are extracted from the fully connect...

Full description

Saved in:
Bibliographic Details
Published inComputer Science and Interdisciplinary Research Journal Vol. 1; no. 2
Main Authors Cong Tung, Dinh, Van Linh, Mai, Borisov, Dimitar
Format Journal Article
LanguageEnglish
Published 26.02.2025
Online AccessGet full text
ISSN3033-1218
3033-1218
DOI10.70862/CSIR.2024.0101-06

Cover

More Information
Summary:This paper studies a method to improve the accuracy of a deep learning model for detecting brain abnormalities based on computed tomography images. The process begins with image preprocessing using the Histogram Equation algorithm and Gabor filter. Then, features are extracted from the fully connected layer of the AlexNet model. To optimize feature extraction, we use the MIFS algorithm to identify the most important features. Finally, the SVM machine learning model is deployed to detect brain abnormalities. To confirm the effectiveness of the proposed model, we compare it with other popular deep learning models, including AlexNet. Experimental results show that our method achieves an accuracy of over 93%, higher than pure deep learning models.
ISSN:3033-1218
3033-1218
DOI:10.70862/CSIR.2024.0101-06