AttentNet: Fully Convolutional 3D Attention for Lung Nodule Detection

TL;DR

This paper introduces AttentNet, a 3D fully convolutional attention framework for lung nodule detection in CT scans, emphasizing computational efficiency and improved attention mechanisms for large 3D medical data.

Contribution

The paper proposes two novel 3D fully convolutional attention blocks and demonstrates their effectiveness in lung nodule detection, offering a more efficient alternative to traditional attention models.

Findings

Fully conv. attention blocks outperform 2D generalized attention in 3D tasks.

AttentNet improves false positive reduction in lung nodule detection.

The approach is computationally more efficient for large 3D medical images.

Abstract

Motivated by the increasing popularity of attention mechanisms, we observe that popular convolutional (conv.) attention models like Squeeze-and-Excite (SE) and Convolutional Block Attention Module (CBAM) rely on expensive multi-layer perception (MLP) layers. These MLP layers significantly increase computational complexity, making such models less applicable to 3D image contexts, where data dimensionality and computational costs are higher. In 3D medical imaging, such as 3D pulmonary CT scans, efficient processing is crucial due to the large data volume. Traditional 2D attention generalized to 3D increases the computational load, creating demand for more efficient attention mechanisms for 3D tasks. We investigate the possibility of incorporating fully convolutional (conv.) attention in 3D context. We present two 3D fully conv. attention blocks, demonstrating their effectiveness in 3D context. Using pulmonary CT scans for 3D lung nodule detection, we present AttentNet, an automated lung nodule detection framework from CT images, performing detection as an ensemble of two stages, candidate proposal and false positive (FP) reduction. We compare the proposed 3D attention blocks to popular 2D conv. attention methods generalized to 3D modules and to self-attention units. For the FP reduction stage, we also use a joint analysis approach to aggregate spatial information from different contextual levels. We use LUNA-16 lung nodule detection dataset to demonstrate the benefits of the proposed fully conv. attention blocks compared to baseline popular lung nodule detection methods when no attention is used. Our work does not aim at achieving state-of-the-art results in the lung nodule detection task, rather to demonstrate the benefits of incorporating fully conv. attention within a 3D context.