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

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 ± 0.0007 for 10-fold cross-validation.

Evaluation of Deep Learning-Based Automated Detection of Primary Spine Tumors on MRI Using the Turing Test

“…Rather than assessing the performance of our DL model, this study aimed primarily to evaluate whether our DL model for the automated detection of primary spine tumors was as good as that of standard manual annotation methods using the Turing test ( 18 , 19 ). Although it is doubtful whether AI will ever pass the Turing test for various complex clinical scenarios, it is easy to misunderstand the role of AI in future medical development.…”

“…( 13 ) made an evaluation of auto contouring in clinical practice using the Turing test, and Sathish et al. ( 18 ) compared lung segmentation and nodule detection between convolutional neural network and humans using the Turing test. Using the choice monitor, the respondents assumed the human’s label as the golden standard; hence, they tried to judge the best labels as objectively as possible.…”