.png){kind=icon}
.png){kind=icon}
Return
BDST: Boundary-dependent deviating segmentation technique using concurrent layer learning
.png){kind=icon}
OA
AI
DOI：10.1016/j.eij.2026.100896.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Brain tumor segmentation and detection are eased by magnetic resonance imaging (MRI) and computer-aided processing. Segmenting tumor and non-tumor regions rely on the boundaries identified based on pixel distribution. In this article, a Boundary-dependent deviation segmentation Technique (BDST) is projected for leveraging the tumor detection accuracy. This technique identifies the pixel deviations for defining boundaries between even and uneven distributions using contrast and skew features. The identified boundaries are segmented for the presence and size of the tumor to improve the precision. The conventional deep learning paradigm with two hidden layers for even and uneven distributions is designed for training boundary characteristics. The boundary characteristics for deviation and constant pixel distributions until the end of the image size are accounted for in this training. Therefore, the learning process is concurrent over differentiating pixel distribution easing segmentation. The segmentation is performed between deviating boundaries for less complex detection. Finally, the tumor incurred boundaries are segmented regardless of the size/ presence of the tumor.
Keywords:
Boundary Detection
Brain Tumor
Deep Learning
Pixel Concentration
Segmentation
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4.3
Papers:
768
Citations:
1.4K
View more organizations