# A Partially Reversible U-Net for Memory-Efficient Volumetric Image Segmentation

## TL;DR

This paper introduces a partially reversible U-Net architecture that significantly reduces memory usage in 3D image segmentation, enabling deeper networks and full field-of-view processing without high memory costs.

## Contribution

The novel partially reversible U-Net reduces memory consumption in 3D segmentation, allowing for deeper networks and full FOV processing, inspired by RevNet principles.

## Key findings

- Substantial memory savings demonstrated on BraTS dataset.
- Deeper networks improve segmentation accuracy.
- Memory efficiency enables processing entire volume instead of patches.

## Abstract

One of the key drawbacks of 3D convolutional neural networks for segmentation is their memory footprint, which necessitates compromises in the network architecture in order to fit into a given memory budget. Motivated by the RevNet for image classification, we propose a partially reversible U-Net architecture that reduces memory consumption substantially. The reversible architecture allows us to exactly recover each layer's outputs from the subsequent layer's ones, eliminating the need to store activations for backpropagation. This alleviates the biggest memory bottleneck and enables very deep (theoretically infinitely deep) 3D architectures. On the BraTS challenge dataset, we demonstrate substantial memory savings. We further show that the freed memory can be used for processing the whole field-of-view (FOV) instead of patches. Increasing network depth led to higher segmentation accuracy while growing the memory footprint only by a very small fraction, thanks to the partially reversible architecture.