CFCM: Segmentation via Coarse to Fine Context Memory

TL;DR

This paper introduces CFCM, a novel deep encoder-decoder segmentation model using convolutional LSTMs for better multi-scale feature integration, demonstrating superior performance on medical image segmentation tasks.

Contribution

It proposes a deep residual encoder-decoder architecture with convolutional LSTMs for feature fusion, a novel approach in medical image segmentation.

Findings

Outperforms existing methods on lung segmentation dataset

Effective in surgical instrument segmentation

Demonstrates the benefits of coarse-to-fine feature integration

Abstract

Recent neural-network-based architectures for image segmentation make extensive usage of feature forwarding mechanisms to integrate information from multiple scales. Although yielding good results, even deeper architectures and alternative methods for feature fusion at different resolutions have been scarcely investigated for medical applications. In this work we propose to implement segmentation via an encoder-decoder architecture which differs from any other previously published method since (i) it employs a very deep architecture based on residual learning and (ii) combines features via a convolutional Long Short Term Memory (LSTM), instead of concatenation or summation. The intuition is that the memory mechanism implemented by LSTMs can better integrate features from different scales through a coarse-to-fine strategy; hence the name Coarse-to-Fine Context Memory (CFCM). We demonstrate the remarkable advantages of this approach on two datasets: the Montgomery county lung segmentation dataset, and the EndoVis 2015 challenge dataset for surgical instrument segmentation.