.png){kind=icon}
.png){kind=icon}
Return
The supervised learning algorithm for MRI images using the Bayesian method and the kernel-based density function
.png){kind=icon}
PRE
AI
DOI：10.1007/s42081-025-00326-w.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This research introduces a novel supervised learning algorithm for MRI images, delivering significant improvements in performance. First, the images are normalized and segmented using a fuzzy cluster analysis algorithm. This step amplifies the distinctions between pixels belonging to different groups. Leveraging this enhanced dataset, the study extracts texture features for each image across various orientation angles and distances, and estimates the probability density function corresponding to each cluster. Subsequently, the research calculates the prior probability for each classified image and employs the Bayesian method for classification. By integrating these advancements, an effective supervised learning algorithm for MRI images is developed, termed the Fuzzy-Bayesian Classifier (FBC). The FBC can be efficiently implemented using numerical examples through a MATLAB procedure. When applied to cancer image datasets, the FBC demonstrates exceptional performance, achieving an accuracy of 97.21% and an F1-score of 98.96%, outperforming various classification algorithms, including convolutional neural network variants and transfer learning models such as ResNet, AlexNet, VGG16, and MobileNet.
Keywords:
Bayesian method
Brain tumor
Classification
Fuzzy clustering algorithm
Segmentation
Journal
J
IF:
1
Papers:
33
Citations:
0