# GridNet with automatic shape prior registration for automatic MRI cardiac segmentation

## TL;DR

This paper introduces a fully automatic deep learning method for MRI cardiac segmentation that incorporates shape priors and a novel registration approach, achieving high accuracy without manual preprocessing.

## Contribution

The proposed CNN integrates an embedded shape prior and an automatic registration module, extending U-Net with multi-resolution features for improved cardiac MRI segmentation.

## Key findings

- Achieved an average Dice coefficient of 0.90
- Segmented ventricles and myocardium in 0.4 seconds
- Reduced Hausdorff distance to 10.4 mm

## Abstract

In this paper, we propose a fully automatic MRI cardiac segmentation method based on a novel deep convolutional neural network (CNN) designed for the 2017 ACDC MICCAI challenge. The novelty of our network comes with its embedded shape prior and its loss function tailored to the cardiac anatomy. Our model includes a cardiac centerof-mass regression module which allows for an automatic shape prior registration. Also, since our method processes raw MR images without any manual preprocessing and/or image cropping, our CNN learns both high-level features (useful to distinguish the heart from other organs with a similar shape) and low-level features (useful to get accurate segmentation results). Those features are learned with a multi-resolution conv-deconv "grid" architecture which can be seen as an extension of the U-Net. Experimental results reveal that our method can segment the left and right ventricles as well as the myocardium from a 3D MRI cardiac volume in 0.4 second with an average Dice coefficient of 0.90 and an average Hausdorff distance of 10.4 mm.