Reducing False-Positives in Lung Nodules Detection Using Balanced Datasets.

Jinglun Liang,Guoliang Ye,Qifan Huang,Shaohui Zhang,Jianwen Guo

doi:10.3389/fpubh.2021.671070

Abstract

Malignant pulmonary nodules are one of the main manifestations of lung cancer in early CT image screening. Since lung cancer may have no early obvious symptoms, it is important to develop a computer-aided detection (CAD) system to assist doctors to detect the malignant pulmonary nodules in the early stage of lung cancer CT diagnosis. Due to the recent successful applications of deep learning in image processing, more and more researchers have been trying to apply it to the diagnosis of pulmonary nodules. However, due to the ratio of nodules and non-nodules samples used in the training and testing datasets usually being different from the practical ratio of lung cancer, the CAD classification systems may easily produce higher false-positives while using this imbalanced dataset. This work introduces a filtering step to remove the irrelevant images from the dataset, and the results show that the false-positives can be reduced and the accuracy can be above 98%. There are two steps in nodule detection. Firstly, the images with pulmonary nodules are screened from the whole lung CT images of the patients. Secondly, the exact locations of pulmonary nodules will be detected using Faster R-CNN. Final results show that this method can effectively detect the pulmonary nodules in the CT images and hence potentially assist doctors in the early diagnosis of lung cancer.

Highlights

Lung cancer is one of the most common cancers in the world and has the highest proportion of new cases (11.6%) and deaths (18.4%) among all cancers in 2018 [1,2,3]
ImageNet is a famous database in the ImageNet Large-Scale Visual Recognition Challenge (ILSVRC), and the majority of the pretrained networks are trained on it [21]
A simple filtering process can remove a lot of unrelated images so that the Faster R-convolutional neural network (CNN) model could detect the pulmonary nodule in a small quantity of CT images that may contain pulmonary nodules without examining all CT images

Summary

Introduction

Lung cancer is one of the most common cancers in the world and has the highest proportion of new cases (11.6%) and deaths (18.4%) among all cancers in 2018 [1,2,3]. Malignant pulmonary nodules are one of the main manifestations of lung cancer on CT images. Pulmonary nodules are a common disease, which is a small round or oval tissue growing in the lungs. It does not mean all pulmonary nodules are malignant, so it is necessary to detect the malignant pulmonary nodules in the lung. There are mainly three feasible methods to successfully apply CNNs to medical images: [1] Training CNN from the ground up, [2] conducting unsupervised CNN pre-training with the supervised fine-tuning base on off-the-shelf pre-trained CNN features, [3] transfer learning [20]. We introduce the networks that were used in this paper, Alexnet Faster R-CNN and ResNet

Methods

Results

Discussion

Conclusion