Class Balanced PixelNet for Neurological Image Segmentation

In this paper, we propose an automatic brain tumor segmentation approach (e.g., PixelNet) using pixel level convolutional neural network (CNN). The model extracts feature from multiple convolutional layers and concatenates them to form a hypercolumn where samples a modest number of pixels for optimization. Hyper-column ensures both local and global contextual information for pixel wise predictor. The model confirms the statistical efficiency by sampling few number of pixels in 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 the certain classes which is a common problem in the medical dataset. We deal 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.