Detection and prognosis of lung cancer with deep learning
Principal Investigator
Name
Daniel Truhn
Degrees
M.D.
Institution
RWTH Aachen University
Position Title
Resident in Radiology
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-500
Initial CDAS Request Approval
Apr 16, 2019
Title
Detection and prognosis of lung cancer with deep learning
Summary
Lung cancer screening requires radiologists to read a lot of CT slices of many exams in a short amount of time. This repetitive task is error prone. Certain tasks (such as detection of new lung nodules) can be assisted by machine learning tools. Furthermore, 'big data' may allow the extraction of hitherto unrecognised image features and allow the stratification of patients into high and low risk groups based on image features alone. Hence the aim of this project is to develop new methods and evaluate deep learning algorithms for early lung cancer detection and survival prediction on the requested data set.
Aims
- Develop deep learning network for classification and detection of lung cancer from the screening population
- Develop methods for survival prediction
Collaborators
Daniel Truhn, RWTH Aachen University
Christoph Haarburger, RWTH Aachen University