Pooling-Free Fully Convolutional Networks with Dense Skip Connections for Semantic Segmentation, with Application to Brain Tumor Segmentation

The semantic segmentation of a brain tumor is of paramount importance for its treatment and prevention. Recently, researches have proposed various neural network-based architectures to improve the performance of segmentation of brain tumor sub-regions. Brain tumor segmentation, being a challenging area of research, requires improvement in its performance. This paper proposes a 2D image segmentation method, BU-Net, to contribute to brain tumor segmentation research. Residual extended skip (RES) and wide context (WC) are used along with the customized loss function in the baseline U-Net architecture. The modifications contribute by finding more diverse features, by increasing the valid receptive field. The contextual information is extracted with the aggregating features to get better segmentation performance. The proposed BU-Net was evaluated on the high-grade glioma (HGG) datasets of the BraTS2017 Challenge—the test datasets of the BraTS 2017 and 2018 Challenge datasets. Three major labels to segmented were tumor core (TC), whole tumor (WT), and enhancing core (EC). To compare the performance quantitatively, the dice score was utilized. The proposed BU-Net outperformed the existing state-of-the-art techniques. The high performing BU-Net can have a great contribution to researchers from the field of bioinformatics and medicine.

Download Full-text

Fully Connected Pyramid Pooling Network (FCPPN) – A Method For Brain Tumor Segmentation

International Journal of Engineering and Advanced Technology - Regular Issue ◽

10.35940/ijeat.a1658.109119 ◽

2019 ◽

Vol 9 (1) ◽

pp. 7036-7041

Keyword(s):

Brain Tumor ◽

Automatic Segmentation ◽

Semantic Segmentation ◽

Tumor Segmentation ◽

Performance Measurements ◽

Brain Tumor Segmentation ◽

Hybrid Combination ◽

Fully Automatic ◽

Fully Connected ◽

The Brain

The procedure of separating the tumor from ordinary cerebrum images is called as brain tumor Segmentation . In segmenting the tumor it allows us to visualize the size and position of tumor within the brain.In Manual segmentation there is less accuracy so there is a need for fully automatic segmentation. A fully automatic segmentation called Semantic segmentation is a technique that classifies all the pixels of an image into meaningful classes of objects. Semantic Segmentation is mainly used in the area of medical imaging. It is mainly used for the doctors to identify the tumor in a clear and exact way. In this paper, we propose a new way of semantic segmentation technique to separate the tumor from the brain . The methods like Segnet, FCN, PSPNET are used for fully automatic segmentation and are used to predicate all types of Tumor. These methods are used to predicate the tumor.Our paper proposes a new architecture called FCPPNET which is a hybrid combination of FCN and PSPNET. Our proposed strategy is assessed utilizing Performance measurements, for example, the Dice coefficient, Accuracy, Sensitivity, and the outcomes appear to be more productive than the current strategies.

Download Full-text