Pulmonary Nodule Detection and Classification using large scale CT data
Principal Investigator
Name
feifei zhou
Degrees
Master Degree
Institution
Independent Researcher, not applicable
Position Title
Independent Researcher
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-283
Initial CDAS Request Approval
Feb 13, 2017
Title
Pulmonary Nodule Detection and Classification using large scale CT data
Summary
Lung Cancer is the most common cancer and the leading cause of cancer related death worldwide. The annual incidence of Lung cancer in China has grown significantly during the past decades, due to increasing risk factors such as tobacco use, environmental pollution, genetics and COPD.
The economic and social burden of lung cancer, especially late stage lung cancer, has been growing tremendously. It is hat patients who are found at early stage (stage 0-1) has a five-year survival rate of 85%, whereas late stage lung cancer patients exhibit an overall survival rate of less than 18%. However, there’s a shortage of radiologists, especially good radiologists in developing countries like China, which calls for a robust computer aided detection and diagnostics system to enhance the efficiency and accuracy of the radiologist workforce.
We have developed a nodule detection and classification system on a smaller dataset, using the state of art deep neural nets. The goal of this project, is to improve the generalizability of the current system and to build up a comprehensive predictive model using the larger scale and multi-dimensional data from NLST.
The economic and social burden of lung cancer, especially late stage lung cancer, has been growing tremendously. It is hat patients who are found at early stage (stage 0-1) has a five-year survival rate of 85%, whereas late stage lung cancer patients exhibit an overall survival rate of less than 18%. However, there’s a shortage of radiologists, especially good radiologists in developing countries like China, which calls for a robust computer aided detection and diagnostics system to enhance the efficiency and accuracy of the radiologist workforce.
We have developed a nodule detection and classification system on a smaller dataset, using the state of art deep neural nets. The goal of this project, is to improve the generalizability of the current system and to build up a comprehensive predictive model using the larger scale and multi-dimensional data from NLST.
Aims
1). Early detection of pulmonary nodules that call for clinical actions
2). Automatic classification of pulmonary nodules based on the level of malignancy
Collaborators
none