AI predicts lung cancer risk

Lung cancer is the leading cause of cancer death worldwide, with an estimated 1.8 million deaths in 2020, according to the World Health Organization. Low-dose chest CT is used to screen people at a high risk of , such as longtime smokers. It has been shown to significantly reduce lung cancer mortality, primarily by helping to detect cancers at an early stage when they are easier to treat successfully.

While lung cancer typically shows up as pulmonary nodules on CT images, most nodules are benign and do not require further clinical workup. Accurately distinguishing between benign and malignant nodules is therefore crucial to catch cancers early.

For the new study, researchers developed an algorithm for lung nodule assessment using , an AI application capable of finding certain patterns in imaging data. The researchers trained the algorithm on CT images of more than 16,000 nodules, including 1,249 malignancies, from the National Lung Screening Trial. They validated the algorithm on three large sets of imaging data of nodules from the Danish Lung Cancer Screening Trial.

The deep learning algorithm delivered excellent results, outperforming the established Pan-Canadian Early Detection of Lung Cancer model for lung nodule malignancy risk estimation. It performed comparably to 11 clinicians, including four thoracic radiologists, five radiology residents and two pulmonologists.

Schematic representation of convolutional neural networks (CNNs) used in the deep learning algorithm for malignancy risk estimation of pulmonary nodules detected at low-dose screening CT. Given a CT image and the coordinate of the pulmonary nodule, a three-dimensional (3D) patch that was 50 mm in size and resampled to 64 pixels (px) in each direction was extracted around the nodule. For the two-dimensional (2D) CNN, nine different views were sectioned from the three-dimensional patch. Features were extracted with a ResNet50 CNN for each two-dimensional view, and the features were combined in a fully connected layer. For the three-dimensional CNN, the entire three-dimensional patch was fed as input to an Inceptionv1 three-dimensional CNN. Both architectures had a final layer that produced a continuous output. Finally, the outputs from the two-dimensional and three-dimensional CNNs were averaged in an ensemble to compute the pulmonary nodule malignancy risk between 0 and 1. Credit: Radiological Society of North America

Examples of CT images in nodules from the Danish Lung Cancer Screening Trial (DLCST) with (a-d) high and (e-h) low agreement between the deep learning (DL) algorithm and the clinicians for malignancy risk estimation. Numbers in rings on bottom left of each image are the algorithm's malignancy score, and numbers in rings on bottom right of each image are the clinicians' median malignancy score. The extent of the color filling is proportional to the malignancy risk (on a scale of 0 to 1, where 0 represents the lowest risk and 1 represents the highest risk). Credit: Radiological Society of North America

A 15-mm part-solid malignant nodule from the Danish Lung Cancer Screening Trial (DLCST) as seen from the axial, coronal, and sagittal views (left side). The nodule was found in the right upper lobe and the bar graph (right side) shows that the nodule was not suspected to be malignant by most clinicians, except for one thoracic radiologist and a pulmonologist. The deep learning (DL) algorithm classified this nodule correctly as malignant. Credit: Radiological Society of North America