DEHA-Net: A Dual-Encoder-Based Hard Attention Network with an Adaptive ROI Mechanism for Lung Nodule Segmentation

Abstract

Measuring pulmonary nodules accurately can help the early diagnosis of lung cancer, which can increase the survival rate among patients. Numerous techniques for lung nodule segmentation have been developed; however, most of them either rely on the 3D volumetric region of interest (VOI) input by radiologists or use the 2D fixed region of interest (ROI) for all the slices of computed tomography (CT) scan. These methods only consider the presence of nodules within the given VOI, which limits the networks’ ability to detect nodules outside the VOI and can also encompass unnecessary structures in the VOI, leading to potentially inaccurate segmentation. In this work, we propose a novel approach for 3D lung nodule segmentation that utilizes the 2D region of interest (ROI) inputted from a radiologist or computer-aided detection (CADe) system. Concretely, we developed a two-stage lung nodule segmentation technique. Firstly, we designed a dual-encoder-based hard attention network (DEHA-Net) in which the full axial slice of thoracic computed tomography (CT) scan, along with an ROI mask, were considered as input to segment the lung nodule in the given slice. The output of DEHA-Net, the segmentation mask of the lung nodule, was inputted to the adaptive region of interest (A-ROI) algorithm to automatically generate the ROI masks for the surrounding slices, which eliminated the need for any further inputs from radiologists. After extracting the segmentation along the axial axis, at the second stage, we further investigated the lung nodule along sagittal and coronal views by employing DEHA-Net. All the estimated masks were inputted into the consensus module to obtain the final volumetric segmentation of the nodule. The proposed scheme was rigorously evaluated on the lung image database consortium and image database resource initiative (LIDC/IDRI) dataset, and an extensive analysis of the results was performed. The quantitative analysis showed that the proposed method not only improved the existing state-of-the-art methods in terms of dice score but also showed significant robustness against different types, shapes, and dimensions of the lung nodules. The proposed framework achieved the average dice score, sensitivity, and positive predictive value of 87.91%, 90.84%, and 89.56%, respectively.

Enhanced Lung Nodule Segmentation using Dung Beetle Optimization based LNS-DualMAGNet Model

“…Usman, M. et al developed a technique for dividing lung nodules into two stages [21]. Originally, the lung nodules in the axial slice of thoracic CT scans was segmented using a dual-encoder-based hard attention networks (DEHA-Net).…”

“…To enhance feature extraction capabilities, the pretrained Resnet34 is further integrated with the DSConv. When the kernel size is 𝐾 × 𝐾, the usual convolution's computational cost is 𝐾 × 𝐾 × 𝑀 × 𝑁 × 𝐻 × 𝑊, the computational cost of the DSConv is 𝐾 × 𝐾 × 𝑀 × 𝐻 × 𝑊 + 𝑀 × 𝑁 × 𝐻 × 𝑊 , as shown in Equation (4).…”

“…AUC also shows the likelihood that the model will accurately predict positive or negative samples; a value nearer 1 denotes superior model performance. Equations provide the specifics (18)(19)(20)(21)(22)(23).…”

“…Reducing the search space can help make the difficult work of lung nodule detection less complicated. To do this, the lung lobes are segmented prior to candidate nodule detection [4,5].…”

Integrated Deep Learning and Stochastic Models for Accurate Segmentation of Lung Nodules From Computed Tomography Images: A Novel Framework

“…Bhattacharjee et al [29] introduced ResiUNet, combining U-Net and ResNet152 to enhance segmentation accuracy. Usman and Shin [30] proposed DEHA-Net, incorporating hard attention networks and adaptive ROI mechanisms for improved segmentation precision.…”