Lung Segmentation and Nodule Detection in Computed Tomography Scan using a Convolutional Neural Network Trained Adversarially using Turing Test Loss

TL;DR

This paper presents a two-stage deep learning framework for lung segmentation and nodule detection in CT scans, utilizing adversarial training with Turing test loss to improve accuracy and efficiency.

Contribution

It introduces a novel adversarial CNN training approach with Turing test loss for lung segmentation, enhancing nodule detection accuracy in a computationally efficient manner.

Findings

Achieved a dice coefficient of 0.984 on LUNA16 dataset.

Demonstrated high accuracy in lung segmentation and nodule classification.

Validated the method's effectiveness with cross-validation results.

Abstract

Lung cancer is the most common form of cancer found worldwide with a high mortality rate. Early detection of pulmonary nodules by screening with a low-dose computed tomography (CT) scan is crucial for its effective clinical management. Nodules which are symptomatic of malignancy occupy about 0.0125 - 0.025\% of volume in a CT scan of a patient. Manual screening of all slices is a tedious task and presents a high risk of human errors. To tackle this problem we propose a computationally efficient two stage framework. In the first stage, a convolutional neural network (CNN) trained adversarially using Turing test loss segments the lung region. In the second stage, patches sampled from the segmented region are then classified to detect the presence of nodules. The proposed method is experimentally validated on the LUNA16 challenge dataset with a dice coefficient of $0.984\pm0.0007$ for 10-fold cross-validation.