Deep Learning-Based HCNN and CRF-RRNN Model for Brain Tumor Segmentation

Abstract

This paper proposes a strategy where a structure is developed to recognize and order the tumor type. Over a time of years, numerous specialists have been examined and proposed a technique in this space. A brain tumor segmentation approach is developed based on efficient, deep learning techniques implemented in a unified system to achieve the appearance and spatial accuracy outcomes through Conditional Radom Fields (CRF) and Heterogeneous Convolution Neural Networks (HCNN). In these steps the 2D image patching and picture slices of the deep-learning model is developed. The Proposed method has following steps as follows: 1) train HCNN by image patches; 2) train CRF with CRF-Recurrent Regression based Neural Network (RRNN) by means of image slices with fixed variables of HCNN; 3) fine tune with HCNN and CRF-RRNN image slices. In general, 3 segmentation models have been trained using axial-, coronary-and sagittal image patches and slices, Further assembled into brain tumor segments using a voting fusion technique and it can be examined with Internet of Medical Things (IoMT) Platform. The experimental results proved that our approach has been capable of developing a Flair, T1c and T2 segmenting model and of achieving good performance as with Flair, T1, T1c, and T2 scans.

Brain Image Segmentation in Recent Years: A Narrative Review

“…According to Deng et al [65], the segmentation performance of hybridized deep learning and machine learning methods demonstrated better precision and sensitivity for the BRATS benchmark. Meanwhile, Nema et al [40] reported better performance using hybrid deep learning-based segmentation for the Dice score.…”

“…This hybridized approach was developed to overcome the limitations of promising deep learning-based methods, and the segmentation results are increasingly aggregated using the machine learning methods in the post-processing stage. Various studies [56][57][58][59][60][61][62][63][64][65] have applied this combination of methods. Kamnitsas et al [56] proposed a dual 3D-CNN pathway to extract both local and contextual information from the 3D brain tumor images.…”

“…A unified Incremental DCNN model based on Heterogeneous CNNs (HCNN) and CRF for brain tumor segmentation was proposed by Deng et al [65]. The steps involved in the method include the following: (i) training of the HCNN using image patches, (ii) training of the CRF-recurrent regression-based neural network (RRNN) using image slices with fixed variables of the HCNN, and (iii) adjustment of the whole network with image slices.…”

“…Deep learning and clustering or classification [56][57][58][59][60][61][62][63][64][65] • High performance for intra-tumor segmentation.…”

The Application of Deep Convolutional Neural Networks to Brain Cancer Images: A Survey

“…Deng et al [ 73 ] propose a novel architecture to segment FLAIR, Tc1, and T2 images from the BraTS2013 and 2015 datasets of over 270 HGG and LGG scans. The architecture composes HCNN and CRF-RRNN models to segment.…”