Permutohedral Attention Module for Efficient Non-Local Neural Networks
This addresses the need for efficient non-local context in medical image processing, particularly for segmentation tasks where organs and tissues share similar appearances, though it appears incremental as it builds on existing attention and non-local methods.
The paper tackles the problem of limited receptive fields in CNNs for medical image segmentation by proposing the Permutohedral Attention Module (PAM) to efficiently capture non-local information, demonstrating its effectiveness in vertebrae segmentation and labeling.
Medical image processing tasks such as segmentation often require capturing non-local information. As organs, bones, and tissues share common characteristics such as intensity, shape, and texture, the contextual information plays a critical role in correctly labeling them. Segmentation and labeling is now typically done with convolutional neural networks (CNNs) but the context of the CNN is limited by the receptive field which itself is limited by memory requirements and other properties. In this paper, we propose a new attention module, that we call Permutohedral Attention Module (PAM), to efficiently capture non-local characteristics of the image. The proposed method is both memory and computationally efficient. We provide a GPU implementation of this module suitable for 3D medical imaging problems. We demonstrate the efficiency and scalability of our module with the challenging task of vertebrae segmentation and labeling where context plays a crucial role because of the very similar appearance of different vertebrae.