Standalone assessment of a CAD system for lung nodule detection
Principal Investigator
Name
Luca Bogoni
Degrees
PhD
Institution
Siemens Healthcare
Position Title
Program Manager
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-136
Initial CDAS Request Approval
Apr 27, 2015
Title
Standalone assessment of a CAD system for lung nodule detection
Summary
A CAD system for detection of lung nodules in thoracic CT data has been developed. In this evaluation we want to assess the performance of the CAD system on a very large set of unseen data. A small portion of the data, selected randomly, will be used for training, whereas the majority will be used to perform a large assessment (evaluation set).
On the evaluation set, the CAD system will be assessed for different size ranges, different types of nodules, and different acquisition parameters.
If required, the training set will be used to retrain the classifiers of the CAD system to improve robustness and to better account for patient variability.
On the evaluation set, the CAD system will be assessed for different size ranges, different types of nodules, and different acquisition parameters.
If required, the training set will be used to retrain the classifiers of the CAD system to improve robustness and to better account for patient variability.
Aims
1. Standalone assessment of a CAD system for lung nodules
2. Quantitative evaluation on a large dataset stratified into different groups
3. Improvements of the CAD system