New Machine Learning Methods for Lung Nodule Detection and Classification to Optimize Cancer Screening
Principal Investigator
Name
Maxine Tan
Degrees
Ph.D.
Institution
Monash University Malaysia
Position Title
Lecturer
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-522
Initial CDAS Request Approval
Jun 13, 2019
Title
New Machine Learning Methods for Lung Nodule Detection and Classification to Optimize Cancer Screening
Summary
Lung cancer is the leading cause of cancer death in many countries including the United States. We would like to download the NLST dataset to develop new deep learning and machine learning methods to improve the accuracy of lung nodule detection and classification schemes on computed tomography (CT) images. The current conventional computer aided diagnosis (CAD) schemes produce too many false positives that are distracting to radiologists. Our team aims to develop new methods that improve the accuracy of current CAD schemes for lung cancer classification, detection, and risk prediction whilst reducing the false positive rate.
Aims
1. To develop accurate lung nodule classification schemes that correctly classify lung nodules as benign or malignant at low false positive rates
2. To develop accurate lung nodule detection schemes that correctly differentiate lung nodules from non-nodule structures including vessel junctions and airway trees at low false positive rates
3. To develop risk prediction models that can accurately predict the risk of lung cancer occurring over sequential examinations
Collaborators
1. University of Malaya
2. Bioinformatics Institute, Agency for Science, Technology and Research, (A*STAR) Singapore