Skip to Main Content

An official website of the United States government

Computer-aided diagnosis on lung cancer using enhanced machine learning

Principal Investigator
Name
Lu Yao
Degrees
Ph.D
Institution
School of Data and Computer Science, Sun Yat-sen University
Position Title
Professor
Email
About this CDAS Project
Study
NLST (Learn more about this study)
Project ID
NLST-351
Initial CDAS Request Approval
Sep 21, 2017
Title
Computer-aided diagnosis on lung cancer using enhanced machine learning
Summary
Lung cancer is one of the leading causes of death in both men and women. Early detection and diagnosis of lung cancer have strong impact on patient survival. Based on CT, X-ray and MRI, Computer-aided diagnosis (CAD) has greatly improved the accuracy of lung cancer diagnosis. We have collected a dataset from our own institute, Sun Yan-Sen University, China (SYSU), including CT and X-ray images, and have studied computerized nodule detection/diagnosis for many years. However, the CAD performance may be different among different populations in China and other countries. Thus, in this study we plan to test our developed CAD system in different regions population and compare their differences for further improvement and generalization of our system. Further more, we plan to develop a CAD system for lung cancer diagnosis with digital-pathology images that can be applied to pathological clinic.
Aims

1. Compare the performance of our model between SYSU Data and NLST Data on CT images.
2. Further develop a computerized detection system for lung cancer with chest X-ray images. We will train our system with SYSU date and NLST data separately and then cross-validate it with SYSU and NLST data.
3. Develop Pathology-based CAD for lung cancer diagnosis and staging.
4. With our own SYSU Data and NLST Data, we will compare the performance of our CAD system for lung cancer between populations of China and the US in terms of patient gender, occupation and family history, etc.
5. Develop an integrated computer graphic interface (GUI) that will bridge different image modalities such as CT, Chest X-Ray and digital pathology images to aid radiologists and pathologists in lung cancer detection, diagnosis and staging.

We therefore, request the access permit to acquire the NLST data for our studies.

Collaborators

Xiangsong Zhang, The First Affiliated Hospital, Sun Yat-sen University
Yaqin Zhang, The Fifth Affiliated Hospital, Sun Yat-sen University