End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography

With an estimated 160,000 deaths in 2018, lung cancer is the most common cause of cancer death in the United States . Lung cancer screening using low-dose computed tomography has been shown to reduce mortality by 20-43% and is now included in US screening guidelines . Existing challenges include int...

Full description

Saved in:
Bibliographic Details
Published in:Nature medicine Vol. 25; no. 6; pp. 954 - 961
Main Authors: Ardila, Diego, Kiraly, Atilla P, Bharadwaj, Sujeeth, Choi, Bokyung, Reicher, Joshua J, Peng, Lily, Tse, Daniel, Etemadi, Mozziyar, Ye, Wenxing, Corrado, Greg, Naidich, David P, Shetty, Shravya
Format: Journal Article
Language:English
Published: United States Nature Publishing Group 01.06.2019
Subjects:
Algorithms
Automation
Cancer screening
Computation
Computed tomography
Databases, Factual
Deep learning
Deep Learning - statistics & numerical data
Diagnosis, Computer-Assisted - methods
Diagnosis, Computer-Assisted - statistics & numerical data
Humans
Imaging, Three-Dimensional - statistics & numerical data
Lung cancer
Lung Neoplasms - diagnosis
Lung Neoplasms - diagnostic imaging
Mass Screening - methods
Mass Screening - statistics & numerical data
Medical imaging
Medical screening
Model accuracy
Neural Networks, Computer
Retrospective Studies
Risk Factors
Tomography
Tomography, X-Ray Computed - statistics & numerical data
United States
United States
ISSN:1078-8956, 1546-170X, 1546-170X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With an estimated 160,000 deaths in 2018, lung cancer is the most common cause of cancer death in the United States . Lung cancer screening using low-dose computed tomography has been shown to reduce mortality by 20-43% and is now included in US screening guidelines . Existing challenges include inter-grader variability and high false-positive and false-negative rates . We propose a deep learning algorithm that uses a patient's current and prior computed tomography volumes to predict the risk of lung cancer. Our model achieves a state-of-the-art performance (94.4% area under the curve) on 6,716 National Lung Cancer Screening Trial cases, and performs similarly on an independent clinical validation set of 1,139 cases. We conducted two reader studies. When prior computed tomography imaging was not available, our model outperformed all six radiologists with absolute reductions of 11% in false positives and 5% in false negatives. Where prior computed tomography imaging was available, the model performance was on-par with the same radiologists. This creates an opportunity to optimize the screening process via computer assistance and automation. While the vast majority of patients remain unscreened, we show the potential for deep learning models to increase the accuracy, consistency and adoption of lung cancer screening worldwide.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:1078-8956
1546-170X
1546-170X
DOI:10.1038/s41591-019-0447-x