GLFNET: Global-Local (frequency) Filter Networks for efficient medical image segmentation

TL;DR

A novel transformer-style architecture called Global-Local Filter Network (GLFNet) for medical image segmentation for medical image segmentation is proposed and its state-of-the-art performance is demonstrated.

Abstract

We propose a novel transformer-style architecture called Global-Local Filter Network (GLFNet) for medical image segmentation and demonstrate its state-of-the-art performance. We replace the self-attention mechanism with a combination of global-local filter blocks to optimize model efficiency. The global filters extract features from the whole feature map whereas the local filters are being adaptively created as 4x4 patches of the same feature map and add restricted scale information. In particular, the feature extraction takes place in the frequency domain rather than the commonly used spatial (image) domain to facilitate faster computations. The fusion of information from both spatial and frequency spaces creates an efficient model with regards to complexity, required data and performance. We test GLFNet on three benchmark datasets achieving state-of-the-art performance on all of them while being almost twice as efficient in terms of GFLOP operations.

Figures (1)

• Figure 1: The proposed architecture is a UNet-like architecture with the difference that each block is a GLFNet block. To boost performance we utilize a multi-scale input for scale invariance as well as deep supervision . The encoder and decoder are symmetric.