Global lung cancer leader explores capabilities of deep learning and AI to improve screening and save lives
DENVER – Evidence proves that screening saves lives, with screening for breast, colon, prostate and cervical cancers, for example, being accepted in many parts of the world as a routine part of medical care. While generally caught in the later stages, lung cancer survival curves show that the has much better outcomes when it is caught early, so developing a lung cancer screening protocol could lead to dramatic improvements in patient care and ultimately reductions in mortality. And yet, despite lung cancer killing more people than breast, prostate and colon cancers combined, there is no global consensus or standard for lung cancer screening.
It is within this context that the International Association for the Study of Lung Cancer (IASLC) is committed to publicizing the results of a pilot project of the Early Lung Imaging Confederation (ELIC). The goal of this four-month pilot project was to develop an innovative and effective technique for improving early lung cancer detection through a network of shared computed tomography (CT)-based images. The application of quantitative imaging approaches could improve the accuracy and the efficiency of lung cancer screening implementation.
“The IASLC ELIC pilot project represents what global collaboration, cutting edge science and technological advancement can achieve,” said Giorgio Scagliotti, President of the IASLC, commenting on the project. “The initial results we’re seeing from ELIC support our view that such a novel approach can assist in improving lung cancer screening and reversing the trend where lung cancer is overwhelmingly detected in a later and more deadly stage.”
As envisioned, the ELIC project will create a globally-accessible, privacy-secured environment for the analysis of large collections of quality-controlled CT lung cancer images and associated biomedical data from around the world. A live demonstration at the IASLC 19th World Conference on Lung Cancer showcased ELIC’s capabilities.
While the project is still in its early stages, the IASLC perceives great potential in ELIC. Having a repository of scanned images and clinical data could support the development of deep learning methods or artificial intelligence (AI) approaches to detect early lung cancer and other diseases using thoracic CT images, as well as measure the responsiveness of those conditions to therapeutic interventions. ELIC has functionality to facilitate the implementation of low-dose CT screening by disseminating validated software tools for the early detection of thoracic diseases.
Early detection by low-dose CT screening can decrease lung cancer mortality by 20-61% among high risk populations (1), but currently only 16% of lung cancer patients are diagnosed with stage I disease (2). When lung cancer is discovered before it spreads, the chances of surviving for more than five years increases to 55% (3).
Two major studies presented at the IASLC 19th World Conference on Lung Cancer reported impressive results in 10-year follow-up of lung cancer screening that revealed even greater reduction in lung cancer deaths than previously reported in the U.S. National Lung Cancer Screening Trial (NLST).
The ELIC pilot program consisted of a central server, hub and 10 globally distributed servers running on the Amazon cloud. Each spoke server contained a de-identified and publicly-available CT image collection and was programmed to run one of two quantitative CT lung image measurement algorithms upon request by the hub server. Its success indicates the exciting potential for this project to conduct important quantitative analysis on lung cancer imaging cases without needing to move datasets from their geographical location.
“It is my hope that ELIC will serve to accelerate progress with lung cancer screening and early detection,” said Jim Mulshine, MD, Chair of the IASLC’s Prevention and Early Detection Committee and Professor of Internal Medicine, Rush Medical College, Chicago, Ill., USA. “Those of us who are involved with ELIC see it as eventually housing millions of images so that its automated screening tools can continually improve in detecting early stage lung cancer, improving treatment decisions and saving more lives.”
(1) IASLC WCLC 2018. PL02.05. Effects of Volume CT Lung Cancer Screening: Mortality Results of the NELSON Randomised-Controlled Population Based Trial. De Koning HJ, Van Der Aalst CM, ten Haaf K, Oudkerk M. and U.S. Preventive Services Task Force. Screening for Lung Cancer. U.S. Preventive Services Task Force Recommendation Statement. AHRQ Publication No. 13-05196-EF-3. http://www.