Early detection of lung cancer with deep learning and graph databases

Principal Investigator

Name
Rafael Berlanga

Degrees
PhD.

Institution
Universitat Jaume I

Position Title
Full-time professor

Email
berlanga@uji.es

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-370

Initial CDAS Request Approval
Nov 6, 2017

Title
Early detection of lung cancer with deep learning and graph databases

Summary
In this investigation, we propose a new approach that combines deep learning mix with a knowledge graph database.The knowledge graph aims at representing as a graph all relevant information around lung cancer such as symptoms, anatomical details, etc. In a second step, by using this graph, we will select subsets of images with the aim of training neural network models that help early lung cancer detection.

Aims

To improve current DL approaches on Radiography with proper examples for training.
To represent all relevant information related to decision support into a knowledge graph.

Collaborators

ActualMed, a company that manages RIS/PACS systems in the cloud (http://www.actualmed.com/)