Algorithm development on lung medical image

Principal Investigator

Name
Jingya Liu

Degrees
Ph. D. candidate

Institution
City College of New York, City University of New York

Position Title
Research Assistant

Email
jliu1@ccny.cuny.edu

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-401

Initial CDAS Request Approval
Apr 17, 2018

Title
Algorithm development on lung medical image

Summary
Lung cancer is considered as one of the leading cancer killers around the world especially in the US and it makes the study of the lung cancer particular important. Although deep learning based method of semantic segmentation, object detection, classification brings a great benefit for medical image automatic analysis, the annotation of the medical images is very rare and expensive due to the fact that deep learning based models need the high quality and quantity of dataset. And another big challenge of the medical image processing is the large variance of anatomy across the patients. We mainly focus on the algorithm development of tumor region segmentation, recognition, and detection.

Aims

To help improve the accuracy of the model and reduce the cost of the dataset, we aim at developing an algorithm use a small portion of annotation data to train a network with a large lung image dataset, called weakly supervised learning on lung cancer image analyze.

Collaborators

Yingli Tian, City College of New York.