Tumor Longitudinal Growth Prediction
Principal Investigator
Name
Bohan Yang
Degrees
Ph.D. student
Institution
Wuhan University
Position Title
Research Assistant
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1184
Initial CDAS Request Approval
Jan 9, 2024
Title
Tumor Longitudinal Growth Prediction
Summary
My name is Bohan Yang, a PhD student focusing on medical imaging at Wuhan University. I am currently studying the imaging development pattern of tumors. Tumors usually develop from abnormal proliferation of cells and tissues and are usually difficult to detect in the early stages. Over time, tumor cells gradually and irreversibly infiltrate into the surrounding tissues, thus depleting the stored energy in the body, which considered the main reason for cancer's high mortality rate . The tumor will further malignantly evolve under the influence of specific factors such as the patient's smoking history, work history, and alcohol consumption history. Longitudinal tumor prediction is imperative as it provides information about the future development of the tumor. Doctors will provide appropriate interventions such as active surveillance, surgical interventions and medications based on important factors such as tumor type and growth rate to provide better management of tumor therapy.
This project requires image data (including CT/MRI images and annotated tumor masks) at earlier time points of each patient to make prediction of tumor growth/shrinkage at a latter time point. Medical imaging data provides morphological information and basic physiological parameters of the tumor. The NLST data contains CT images and tumor masks/contours organized based on annual-level. Specifically, three annual CT exams to screen for lung cancer are included: one at baseline (T0) and two more on the first and second anniversaries of randomization (T1 and T2). This dataset highly aligns with the requirement of my current project in that I can use the CT images from T0 and T1 to make longitudinal tumor growth prediction of CT images on T2.
I will follow the terms of the restricted license agreement. If the study is accepted by a journal or conference, I will include the proper citations of the dataset in the study. I kindly request you to consider the proposed use of the dataset, and I would be very appreciative if you can grand the access to the NLST data.
This project requires image data (including CT/MRI images and annotated tumor masks) at earlier time points of each patient to make prediction of tumor growth/shrinkage at a latter time point. Medical imaging data provides morphological information and basic physiological parameters of the tumor. The NLST data contains CT images and tumor masks/contours organized based on annual-level. Specifically, three annual CT exams to screen for lung cancer are included: one at baseline (T0) and two more on the first and second anniversaries of randomization (T1 and T2). This dataset highly aligns with the requirement of my current project in that I can use the CT images from T0 and T1 to make longitudinal tumor growth prediction of CT images on T2.
I will follow the terms of the restricted license agreement. If the study is accepted by a journal or conference, I will include the proper citations of the dataset in the study. I kindly request you to consider the proposed use of the dataset, and I would be very appreciative if you can grand the access to the NLST data.
Aims
- propose the growth estimation model to characterize the dynamic growth relationship of tumors under long-term sequences.
Collaborators
I am the only investigator who would like to access the data and I currently do not have collaborators. If my co-workers ask me to for the access of the data, I will notify them to apply on CDAS website using their own project.