Similar patch image retreival for cancer diagnosis
Principal Investigator
Name
Moti Moravia
Degrees
M.E
Institution
Datability
Position Title
CTO
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-954
Initial CDAS Request Approval
Sep 6, 2022
Title
Similar patch image retreival for cancer diagnosis
Summary
Research the ability to retrieve similar image patches from a query image patch. The user will mark an ROI (Region of interest) around the desired zone in the image. The AI algorithm will analyze the patch and retrieve the most similar image patches from other CT scans. In case the doctor has marked an anomaly (for example lung nodule), the retrieved patches suppose to include the same anomaly as a reference for the doctor to have the most precise diagnosis. In case no anomaly is present in the patch, the system retrieves the most similar patches with no anomaly present. Clinical data is to be analyzed to research the relevancy and correlation with this method.
Aims
* Successfully retrieve similar image patches to the query image patch.
* Check the correlation between the retrieved images and the clinical data for each patient.
Collaborators
NAN