Deep Learning for Whole Heart Segmentation and Disease Diagnosis
Project Description
Congenital heart disease (CHD) is the most common cause of infant death due to birth defects. This project will develop deep learning algorithms for whole heart and great vessel segmentation from CHD 3D CT images. Other possible diseases can also be detected. We will use CT images and echocardiogram videos.
Supervisor
LI Xiaomeng
Quota
10
Course type
UROP1000
UROP1100
UROP2100
UROP3100
UROP4100
Applicant's Roles
1, A weekly meeting is required.
2, Applicant should have a strong desire to learn deep learning and has solid programming ability.
3, At least 10 hours is required each week.
2, Applicant should have a strong desire to learn deep learning and has solid programming ability.
3, At least 10 hours is required each week.
Applicant's Learning Objectives
1, Participate in some existing projects to develop the DL algorithms, if possible.
2, Write a paper or report, if possible.
2, Write a paper or report, if possible.
Complexity of the project
Moderate