We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Brain tumor recognition from multimodal magnetic resonance images using wavelet texture features and optimized artificial neural network.
- Authors
Yadav, Ashwani Kumar; Yadav, Vaishali
- Abstract
Magnetic resonance (MR) images are commonly used for quantitative analysis and diagnosis of various diseases. An optimized wavelet-based textural feature model for the classification of brain MRI images for two different image modalities has been proposed in this work. Image-based features are extracted from MR images with the help of the proposed feature model. Feature dimensionality is reduced with Principal Component Analysis (PCA). Selected features are used to train various classifiers and validation is done with the 10-fold cross method. The experimental results achieved show that the optimized Artificial Neural Network (ANN) with PCA for T2w modality has the best performance for tumor diagnosis in MR Images with a maximum of 98.78% accuracy, and K-nearest neighbor (KNN) is fastest with a training time of 0.8372 sec.
- Subjects
ARTIFICIAL neural networks; FEATURE extraction; MAGNETIC resonance imaging; TUMOR diagnosis; K-nearest neighbor classification
- Publication
Multimedia Tools & Applications, 2024, Vol 83, Issue 29, p72975
- ISSN
1380-7501
- Publication type
Article
- DOI
10.1007/s11042-024-18489-1