CAMS: Convolution and Attention-Free Mamba-based Cardiac Image Segmentation

TL;DR

CAMS-Net introduces a novel convolution and self-attention-free segmentation model using Mamba-based modules, achieving state-of-the-art results in cardiac image segmentation with reduced complexity and parameters.

Contribution

The paper presents a new segmentation network that eliminates convolution and self-attention, utilizing Mamba-based modules with novel aggregators and a LIFM block for improved efficiency and performance.

Findings

Outperforms existing CNN, Transformer, and Mamba-based methods on cardiac datasets.

Achieves linear computational complexity and fewer parameters.

Provides source code and pre-trained models for reproducibility.

Abstract

Convolutional Neural Networks (CNNs) and Transformer-based self-attention models have become the standard for medical image segmentation. This paper demonstrates that convolution and self-attention, while widely used, are not the only effective methods for segmentation. Breaking with convention, we present a Convolution and self-Attention-free Mamba-based semantic Segmentation Network named CAMS-Net. Specifically, we design Mamba-based Channel Aggregator and Spatial Aggregator, which are applied independently in each encoder-decoder stage. The Channel Aggregator extracts information across different channels, and the Spatial Aggregator learns features across different spatial locations. We also propose a Linearly Interconnected Factorized Mamba (LIFM) block to reduce the computational complexity of a Mamba block and to enhance its decision function by introducing a non-linearity between two factorized Mamba blocks. Our model outperforms the existing state-of-the-art CNN, self-attention, and Mamba-based methods on CMR and M&Ms-2 Cardiac segmentation datasets, showing how this innovative, convolution, and self-attention-free method can inspire further research beyond CNN and Transformer paradigms, achieving linear complexity and reducing the number of parameters. Source code and pre-trained models are available at: https://github.com/kabbas570/CAMS-Net.