Convex Shape Prior for Deep Neural Convolution Network based Eye Fundus Images Segmentation

Abstract

Convex Shapes (CS) are common priors for optic disc and cup segmentation in eye fundus images. It is important to design proper techniques to represent convex shapes. So far, it is still a problem to guarantee that the output objects from a Deep Neural Convolution Networks (DCNN) are convex shapes. In this work, wepropose a technique which can be easily integrated into the commonly used DCNNs for image segmentation and guarantee that outputs are convex shapes. This method is flexible and it can handle multiple objects and allow some of the objects to be convex. Our method is based on the dual representation of the sigmoid activation function in DCNNs. In the dual space, the convex shape prior can be guaranteed by a simple quadratic constraint on a binary representation of the shapes. Moreover, our method can also integrate spatial regularization and some other shape prior using a soft thresholding dynamics (STD) method. The regularization can make the boundary curves of the segmentation objects to be simultaneously smooth and convex. We design a very stable active set projection algorithm to numerically solve our model. This algorithm can form a new plug-and-play DCNN layer called CS-STD whose outputs must be a nearly binary segmentation of convex objects. In the CS-STD block, the convexity information can be propagated to guide the DCNN in both forward and backward propagation during training and prediction process. As an application example, we apply the convexity prior layer to the retinal fundus images segmentation by taking the popular DeepLabV3+ as a backbone network. Experimental results on several public datasets show that our method is efficient and outperforms the classical DCNN segmentation methods.

A New Regularization for Deep Learning-Based Segmentation of Images with Fine Structures and Low Contrast

“…To combine deep learning framework and spatial regularizations, Liu et al [ 23 , 28 , 29 ] have given variational explanations for some widely used activation functions in deep learning including softmax, ReLU, and sigmoid [ 23 , 28 , 29 ]. For example, softmax operator can be written as and it can be represented as the solution of the following optimization problem [ 23 ] when the parameter is 1: where is the domain of the whole image, C is the number of classes, is the feature map input, and is the output of the softmax operator.…”

SSP-Regularizer: A Star Shape Prior Based Regularizer for Vessel Lumen Segmentation in Oct Images

“…For the segmentation of objects in specific classes, the shape prior is usually incorporated into a given traditional solution scheme to exploit the shape information. Liu et al [19] proposed a deep CNN with a convex shape constraint to improve the segmentation performance for convex objects. Shape prior based CNN (SP-CNN) [9] was designed by inte-RoI c p q Fig.…”

Medical Image Segmentation With Limited Supervision: A Review of Deep Network Models

“…Other types of anatomical priors such as star shape prior [108], [118]- [121], convex shape prior [122], topology [123]- [127], size [128]- [130], etc., have also been introduced to improve the segmentation robustness and anatomically accuracy.…”