use active learning to do the tumor segmentation.

Principal Investigator

Name
Sijie Wang

Degrees
B.S.

Institution
the university of Tokyo

Position Title
Researcher

Email
bragiwang@gmail.com

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-595

Initial CDAS Request Approval
Oct 31, 2019

Title
use active learning to do the tumor segmentation.

Summary
in my study, i'll combine active learning with HCI techniques. aiming at when the dataset is limited, the oracles and the machine learning model can generate a well-labeled dataset. i'll use active learning especially, which will train limited data first , and send the most uncertain data to the oracle for a manual label. then the labeled data will be fed back to the dataset. After several generations, i hope we can get a solid database.

Aims

get a solid dataset when the dataset is limited.

Collaborators

Igarashi takeo