Presented By O’Reilly and Intel AI
Put AI to work
Sep 4-5, 2018: Training
Sep 5-7, 2018: Tutorials & Conference
San Francisco, CA

Lung cancer detection and segmentation using deep learning

Daniel Golden (Arterys)
2:35pm-3:15pm Thursday, September 6, 2018
Interacting with AI, Models and Methods
Location: Yosemite BC
Secondary topics:  Health and Medicine
Average rating: *****
(5.00, 1 rating)

Who is this presentation for?

  • Deep learning practitioners, particularly those focused on developing software in a regulated industry

Prerequisite knowledge

  • Familiarity with deep learning
  • Experience with detection and segmentation (useful but not required)

What you'll learn

  • Understand the inherent complexity of practical deep learning solutions
  • Explore applications of deep learning in radiology
  • Learn how to prove model safety and efficacy in a regulated environment


For those of us accustomed to life’s modern automated conveniences, diagnostic radiology can seem shockingly unsophisticated. Lung cancer screening via computed tomography (CT) is an example of a common radiological procedure that is critical in ensuring that cancers are detected early so that patients have the best chance of receiving timely treatment. However, the radiological procedure for lung cancer screening is still an entirely manual affair. A given lung CT exam consists of a 3D volume of data composed of a stack of hundreds of 2D slices. During screening, clinicians manually scroll through the data slice by slice, searching for tiny nodules that can be indistinguishable from blood vessels and other structures under most viewing conditions. Not only is this process time consuming and tedious, but inter-reader variability among clinicians means that most patients are not receiving the best possible care.

Building on the successes of its previous deep learning-based tools in cardiac MRI, Arterys has developed a deep learning-based system that that can automatically detect and segment lung nodules in CT exams. Using the open LIDC-IDRI dataset of detected and segmented lung nodules in 1,018 thoracic CT exams, Arterys has developed a pipeline that consists of three connected models: a nodule proposal system (2D U-Net-based segmentation network), a nodule classification system (2.5D ResNet-based classifier), and a nodule segmentation system (3D ENet-based segmentation network). These models operate together as a complete lung nodule detection and segmentation system.

The resulting system has the potential to greatly improve the speed and effectiveness of lung cancer screening. For nodules with a diameter larger than 6 mm (the lower limit for clinical significance), the recall of the detection model is 94%, with four false positives per scan. For nodule segmentation, the mean dice coefficient is 0.83±0.10, comparable to the mean dice coefficient of expert radiologists, which is 0.79±0.09. Both models operate with clinicians in the loop, requiring that clinicians review and optionally modify the initial automated results before accepting them. These deep learning-based models form the backbone of Arterys’s FDA-cleared, cloud-based Oncology DL software product.

Daniel Golden details the deep learning technologies behind the lung nodule detection and segmentation system and discusses the method for determining that the system is as accurate as expert radiologists in order to obtain FDA clearance.

Photo of Daniel Golden

Daniel Golden


Dan Golden is the director of machine learning at Arterys, a startup focused on streamlining the practice of medical image interpretation and postprocessing. Previously, he founded a machine learning team at CellScope that used the then-nascent field of deep learning to diagnose ear disease and streamline the process of recording ear exams at home and was a postdoc at Stanford, focusing on using machine learning to predict outcomes and disease characteristics in cancer patients. He holds a PhD in electrical engineering from Stanford.