Lung nodule and lung disease detection algorithms on CT and X ray images.
Principal Investigator
Name
Joseph Chui
Degrees
M. Sc.
Institution
Ferrum Health Inc
Position Title
Computer Vision Engineer
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-434
Initial CDAS Request Approval
Aug 1, 2018
Title
Lung nodule and lung disease detection algorithms on CT and X ray images.
Summary
Our algorithms use imaging data of CT and X Ray, diagnosis, patient demographics, and image acquisition and processing information to train inference models to:
Detect lung nodules and lung diseases
Measure their properties and characteristics.
It is important for our algorithms to train and validate the models, not only with data from patients of cancer or other diseases, but also with data from patients of negative diagnosis. The vast amount of data of negative findings in NLST can become very valuable to the proposed project.
Our proposed project aims to train and validate statistical models using both positive and negative data from NLST as well as other datasets together to improve the inference quality of the models. Another aim of the project is to analyze the correlations of inference results from CT alone and from X Ray alone. Patents with both CT and X Ray scans can provide a non-bias way to access these correlations. The results of the analysis might provide an insight on how to combine the inferences of different models.
Detect lung nodules and lung diseases
Measure their properties and characteristics.
It is important for our algorithms to train and validate the models, not only with data from patients of cancer or other diseases, but also with data from patients of negative diagnosis. The vast amount of data of negative findings in NLST can become very valuable to the proposed project.
Our proposed project aims to train and validate statistical models using both positive and negative data from NLST as well as other datasets together to improve the inference quality of the models. Another aim of the project is to analyze the correlations of inference results from CT alone and from X Ray alone. Patents with both CT and X Ray scans can provide a non-bias way to access these correlations. The results of the analysis might provide an insight on how to combine the inferences of different models.
Aims
Aim 1: Improving the quality of the inference models by including NLST data in the their training, and to validate the models with NLST data held out from the training.
Aim 2: Assessing the correlations between our inference results between models using CT alone and models using X Ray alone.
Collaborators
Ferrum Health