survival analysis
Principal Investigator
Name
yifan shen
Degrees
M.C.S.
Institution
Beijing University of Posts and Telecommunications
Position Title
master
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-788
Initial CDAS Request Approval
May 7, 2021
Title
survival analysis
Summary
Currently, many approaches have been proposed to incorporate with patients’ imaging data for survival analysis. But a pathological image
is of high resolution which means the existing methods based on regular image size cannot directly learn discriminative patterns from WSIs.
To solve this problem, some pioneering works try to develop an end-to-end manner to predict survival on WSIs. As the wide application of transformer variants in the CV field, the effects of multiple tasks have been improved. Inspired by it, we want to develop a model to predict survival on WSIs based on transformer, in order to improve the effect of survival analysis and provide interpretability from the attention mechanism for clinical work.
is of high resolution which means the existing methods based on regular image size cannot directly learn discriminative patterns from WSIs.
To solve this problem, some pioneering works try to develop an end-to-end manner to predict survival on WSIs. As the wide application of transformer variants in the CV field, the effects of multiple tasks have been improved. Inspired by it, we want to develop a model to predict survival on WSIs based on transformer, in order to improve the effect of survival analysis and provide interpretability from the attention mechanism for clinical work.
Aims
Improve the current convolutional neural network for survival prediction with pathological images.
Collaborators
Completed independently