Class Balanced PixelNet for Neurological Image Segmentation

TL;DR

This paper introduces Class Balanced PixelNet, a CNN-based method for brain tumor segmentation that uses hyper-column features and class-balanced sampling to improve accuracy on medical images.

Contribution

It proposes a novel pixel-level CNN with hyper-column features and class-balanced sampling to address class imbalance and spatial redundancy in medical image segmentation.

Findings

Achieved promising results on brain tumor datasets

Effectively handled class imbalance in training data

Utilized hyper-column features for improved segmentation accuracy

Abstract

In this paper, we propose an automatic brain tumor segmentation approach (e.g., PixelNet) using a pixel-level convolutional neural network (CNN). The model extracts feature from multiple convolutional layers and concatenate them to form a hyper-column where samples a modest number of pixels for optimization. Hyper-column ensures both local and global contextual information for pixel-wise predictors. The model confirms the statistical efficiency by sampling a few pixels in the training phase where spatial redundancy limits the information learning among the neighboring pixels in conventional pixel-level semantic segmentation approaches. Besides, label skewness in training data leads the convolutional model often converge to certain classes which is a common problem in the medical dataset. We deal with this problem by selecting an equal number of pixels for all the classes in sampling time. The proposed model has achieved promising results in brain tumor and ischemic stroke lesion segmentation datasets.