# Automated Multiscale 3D Feature Learning for Vessels Segmentation in Thorax CT Images

## TL;DR

This paper presents an extension of multiscale feature learning to 3D for vessel segmentation in thorax CT images, achieving improved accuracy over previous slice-wise methods through unsupervised learning and parallel processing.

## Contribution

The authors extend multiscale feature learning to 3D using dictionary learning and develop a parallel implementation, improving vessel segmentation accuracy in CT images.

## Key findings

- Achieved 97.24% accuracy on VESSEL12 dataset.
- Improved over previous slice-wise method.
- Developed a parallel implementation for 3D processing.

## Abstract

We address the vessel segmentation problem by building upon the multiscale feature learning method of Kiros et al., which achieves the current top score in the VESSEL12 MICCAI challenge. Following their idea of feature learning instead of hand-crafted filters, we have extended the method to learn 3D features. The features are learned in an unsupervised manner in a multi-scale scheme using dictionary learning via least angle regression. The 3D feature kernels are further convolved with the input volumes in order to create feature maps. Those maps are used to train a supervised classifier with the annotated voxels. In order to process the 3D data with a large number of filters a parallel implementation has been developed. The algorithm has been applied on the example scans and annotations provided by the VESSEL12 challenge. We have compared our setup with Kiros et al. by running their implementation. Our current results show an improvement in accuracy over the slice wise method from 96.66$\pm$1.10% to 97.24$\pm$0.90%.