SUPR
Radiology image analysis using deep learning
Dnr:

sens2022001

Type:

SNIC SENS

Principal Investigator:

Ida Häggström

Affiliation:

Chalmers tekniska högskola

Start Date:

2022-02-01

End Date:

2024-02-01

Primary Classification:

20603: Medical Image Processing

Webpage:

Allocation

  • Castor /proj/nobackup at UPPMAX: 7000 GiB
  • Cygnus /proj/nobackup at UPPMAX: 7000 GiB
  • Castor /proj at UPPMAX: 3000 GiB
  • Cygnus /proj at UPPMAX: 3000 GiB
  • Bianca at UPPMAX: 20 x 1000 core-h/month

Abstract

We will perform medical image analysis in collaboration with the Sahlgrenska University hospital, to analyze large scale CT, MR and PET images. We will use large data cohorts (>1,000 cases per sub-study), including the national SCAPIS cohort (30,000 cases). Our aims are to design, train, and test deep learning models for automatic diagnosis, outcome prediction, and outlier detection in these images.