Skip to Main Content

COVID-19 is an emerging, rapidly evolving situation.

What people with cancer should know: https://www.cancer.gov/coronavirus

Get the latest public health information from CDC: https://www.coronavirus.gov

Get the latest research information from NIH: https://covid19.nih.gov/

Principal Investigator
Name
Albert Hsiao
Degrees
M.D., Ph.D.
Institution
UC San Diego
Position Title
Associate Professor
Email
About this CDAS Project
Study
NLST (Learn more about this study)
Project ID
NLST-706
Initial CDAS Request Approval
Sep 16, 2020
Title
3D Convolutional Neural Network for Classifying Lung Nodule Malignancy
Summary
We are a multidisciplinary team of researchers led by Dr. Albert Hsiao at UC San Diego, with both clinical and machine learning expertise. By leveraging the power of convolutional neural networks accelerated by state-of-the-art remote GPU clusters, we hope to answer the following question: can an algorithm determine the malignancy of a lung nodule or lesion based solely on location, composition, and morphology in a CT scan? If so, such an algorithm could be used to send fewer patients to biopsy with far greater confidence. Already, with a limited dataset of annotated CT scans, we have produced a 3D convolutional neural network that has shown promising results. Using the NLST dataset we will explore this simple model’s full potential, as well as test other, more complex models.
Aims

- Classify annotated lung nodules as benign or malignant
- Classify malignant lung nodules as some kind of malignancy

Collaborators

Albert Hsiao, UC San Diego
Ola Besser, UC San Diego
Samuel May, UC San Diego
Jonathan Guiang, UC San Diego
Amir Khannehad, UC San Diego