CXR-LC: Deep learning using Chest Radiographs to Identify High-risk Smokers for Lung Cancer Screening: Development and Validation of a Prediction Model

To appear in Annals of Internal Medicine.

Overview

Lung cancer screening with chest CT reduces lung cancer deaths. Currently, CT screening is offered to those eligible by the Centers for Medicare and Medicaid services (CMS) screening criteria. This criteria misses many lung cancers and requires detailed smoking history information that is poorly documented in the medical record. Chest x-rays are among the most common diagnostic imaging tests in medicine. We hypothesized that a convolutional neural network (CNN) could extract information from these x-rays, along with data readily documented in the medical record, to select individuals at the highest risk for future lung cancer.

CXR-LC is a CNN that predicts the risk of 12-year incident lung cancer based on a chest radiograph image, and a person's age, sex, and smoking status (current vs. former smoker). In our testing dataset, CXR-LC had better discrimination than the CMS criteria (AUC 0.755 vs 0.634, p < 0.001). When matching the size of the CMS eligible population, CXR-LC missed 30.7% fewer lung cancers.

Inverted Kaplan-Meier for Incident Lung Cancer based on CXR-LC category

This repo contains data intended to promote reproducible research. It is not for clinical care or commercial use.

Dummy data files are provided in dummy_datasets/Tabular_Data.csv and dummy_datasets/test_images/; dummy images are provided in the test_images folder.

Datasets

PLCO (NCT00047385) data used for model development and testing are available from the National Cancer Institute (NCI, https://biometry.nci.nih.gov/cdas/plco/). NLST (NCT01696968) testing data is available from the NCI (https://biometry.nci.nih.gov/cdas/nlst/) and the American College of Radiology Imaging Network (ACRIN, https://www.acrin.org/acrin-nlstbiorepository.aspx). Due to the terms of our data use agreement, we cannot distribute the original data. Please instead obtain the data directly from the NCI and ACRIN.

The data folder provides the image filenames, split of PLCO into independent training/tuning/testing datasets, and the CXR-Risk output probabilities:

• PLCO_Scaled_Probabilities.csv contains the probabilities (Scaled_CXRLC_Risk) generated by CXR-LC in the PLCO testing dataset. Probabilities were put into ordinal categories (CXRLC_Group) as described in the manuscript. filenamepng refers to the original TIF filename converted to a PNG. "Dataset" is "Tr" for training, "Tu" for tuning, and "Te" for testing.
• NLST_Scaled_Probabilities.csv contains the probabilities (Scaled_CXRLC_Risk) and ordinal probability categories (CXRLC_Group) for the NLST testing dataset. The format for filenamepng is the (original participant directory)_(original DCM filename).png

Image processing

PLCO radiographs were provided as scanned TIF files by the NCI. TIFs were converted to PNGs with a minimum dimension of 512 pixels with ImageMagick v6.8.9-9.

Many of the PLCO radiographs were rotated 90 or more degrees. To address this, we developed a CNN to identify rotated radiographs. First, we trained a CNN using the resnet34 architecture to identify synthetically rotated radiographs from the CXR14 dataset. We then fine tuned this CNN using 11,000 manually reviewed PLCO radiographs. The rotated radiographs identified by this CNN in preprocessing/plco_rotation_github.csv were then corrected using ImageMagick.

cd path_for_PLCO_tifs
mogrify -path destination_for_PLCO_pngs -trim +repage -colorspace RGB -auto-level -depth 8 -resize 512x512^ -format png "*.tif"
cd path_for_PLCO_pngs
while IFS=, read -ra cols; do mogrify -rotate 90 "${cols[0]}"; done < /path_to_repo/preprocessing/plco_rotation_github.csv

NLST radiographs were provided as DCM files by ACRIN. We chose to first convert them to TIF using DCMTK v3.6.1, then to PNGs with a minimum dimension of 512 pixels through ImageMagick to maintain consistency with the PLCO radiographs:

cd path_to_NLST_dcm
for x in *.dcm; do dcmj2pnm -O +ot +G $x "${x%.dcm}".tif; done;
mogrify -path destination_for_NLST_pngs -trim +repage -colorspace RGB -auto-level -depth 8 -resize 512x512^ -format png "*.tif"

The orientation of several NLST chest radiographs was manually corrected:

cd destination_for_NLST_pngs
mogrify -rotate "90" -flop 204025_CR_2000-01-01_135015_CHEST_CHEST_n1__00000_1.3.51.5146.1829.20030903.1123713.1.png
mogrify -rotate "-90" 208201_CR_2000-01-01_163352_CHEST_CHEST_n1__00000_2.16.840.1.113786.1.306662666.44.51.9597.png
mogrify -flip -flop 208704_CR_2000-01-01_133331_CHEST_CHEST_n1__00000_1.3.51.5146.1829.20030718.1122210.1.png
mogrify -rotate "-90" 215085_CR_2000-01-01_112945_CHEST_CHEST_n1__00000_1.3.51.5146.1829.20030605.1101942.1.png

As a final step, we used the Python Imaging Library (PIL) to resize each image to 224 x 224 using a bilinear interpolation.

python preprocessing/ImageResizer.py <path_to_512_images> <path_to_output_directory_224_224_images>

Acknowledgements

I thank the NCI and ACRIN for access to trial data, as well as the PLCO and NLST participants for their contribution to research. I would also like to thank the fastai and Pytorch communities. A GPU used for this research was donated as an unrestricted gift through the Nvidia Corporation Academic Program. The statements contained herein are mine alone and do not represent or imply concurrence or endorsements by the above individuals or organizations.