Volumetric segmentation via 3D active shape models

Abstract

We propose a novel fully convolutional network architecture for shapes, denoted by Shape Fully Convolutional Networks (SFCN). 3D shapes are represented as graph structures in the SFCN architecture, based on novel graph convolution and pooling operations, which are similar to convolution and pooling operations used on images. Meanwhile, to build our SFCN architecture in the original image segmentation fully convolutional network (FCN) architecture, we also design and implement a generating operation with bridging function. This ensures that the convolution and pooling operation we have designed can be successfully applied in the original FCN architecture. In this paper, we also present a new shape segmentation approach based on SFCN. Furthermore, we allow more general and challenging input, such as mixed datasets of different categories of shapes which can prove the ability of our generalisation. In our approach, SFCNs are trained triangles-to-triangles by using three low-level geometric features as input. Finally, the feature voting-based multi-label graph cuts is adopted to optimise the segmentation results obtained by SFCN prediction. The experiment results show that our method can effectively learn and predict mixed shape datasets of either similar or different characteristics, and achieve excellent segmentation results.

A new deformable model for analysis of X-ray CT images in preclinical studies of mice for polycystic kidney disease

“…One of the main challenges in the 3-D application of any PDM-based segmentation approach is the practical matter of manual training by an anatomy expert who may not be accustomed to viewing volumetric data on a computer screen. A new training approach has been developed to overcome this difficulty by presenting the 3-D volumetric data in a unique way that simplifies the overall process [12]. The improved PSAM optimization approach reported here is being updated for use with this 3-D algorithm.…”

“…Hence, optimizing the objective function in the PCA subspace with respect to the more compact vector, , is a simpler task. If we substitute the new boundary representation, , into (7), the objective function then takes the form (11) where (12) and…”

Outlier Detection and Handling for Robust 3-D Active Shape Models Search

“…Specific constraints found in the model are usually imposed to limit the shapes to valid instances. Model construction [2]- [4], as well as image segmentation [5], [6] using ASM, has been generalized to 3-D and higher dimensional cases, and the method is used extensively in clinical applications [7], [8].…”

Comparison of 3D Segmentation Algorithms for Medical Imaging

“…M.M. Dickens and his friends use 3D active shape models for volumetric segmentation [13]. Sarang Lakare [1] reviewed various segmentation algorithms found in the literature and suggested a classification of algorithms into three categories: structural techniques, statistical techniques and hybrid techniques.…”