Skip to Main Content

An official website of the United States government

Government Funding Lapse

Because of a lapse in government funding, the information on this website may not be up to date, transactions submitted via the website may not be processed, and the agency may not be able to respond to inquiries until appropriations are enacted. The NIH Clinical Center (the research hospital of NIH) is open. For more details about its operating status, please visit  cc.nih.gov. Updates regarding government operating status and resumption of normal operations can be found at OPM.gov.

Automated lung diseases analysis using deep learning techniques

Principal Investigator
Name
Peter M.A. van Ooijen
Degrees
Ph.D.
Institution
University Medical Center Groningen
Position Title
Associate Professor
Email
About this CDAS Project
Study
NLST (Learn more about this study)
Project ID
NLST-552
Initial CDAS Request Approval
Aug 13, 2019
Title
Automated lung diseases analysis using deep learning techniques
Summary
Lung cancer, as one of the most severe cancers with high incidence, has a devastating effect on human lives. It has been predicted to be one of the greatest single cause of mortality among the European population in 2019. Early dianosis of lung cancer is a crucial step since it could improve chances of survival. Compared with the traditional techniques, deep learning based on neural networks is able to detect and classify nodules without any additional information such as nodule segmentation or nodule size. It is fully automated on processing, which reduces a huge amount of work for radiologists in lung cancer screening. Our goal is to study CT image features of lung cancer and to develop lung cancer malignancy prediction models using NLST data.
Chronic obstructive pulmonary disease (COPD) is a respiratory disorder that is mainly caused by exogenous factors like tobacco smoking and air pollution. COPD is compose d of two main components: emphysema (airspace enlargement and tissue destruction) and bronchitis (airways disease). Low-dose computed tomography (LDCT) is an effective modality for early detection of COPD mainly emphysema. Hence in this study, our aim is to consider all the clinically quantifiable features and image features to develop a prognosis and classification model for COPD using NLST data.
Aims

1. To develop models based on deep learning techniques for classification of lung cancer
2. To validate our own models for lung nodule detection and classification based on the NLST dataset
3. To develop and compare different deep learning based classification model for COPD
4. To validate the deep learning model using NSCT for reproduciability
5. Prediction of lung cancer
6. Validation of the algorithms developed on the different data settings.

Collaborators

Sunyi Zheng, University Medical Center Groningen, s.zheng@umcg.nl
Yeshaswini Nagaraj, University Medical Center Groningen, y.nagaraj@umcg.nl

Related Publications