Navigation

DL based image reconstruction/enhancement for contrast agent minimization in cardiac MRI

Type: MA thesis

Status: running

Supervisors: Elisabeth Preuhs (former Hoppe)

DL based image reconstruction/enhancement for contrast agent minimization in cardiac MRI

Motivation:

  • Myocardial infarction (MI) can be robustly detected using contrast-enhanced MR imaging
  • High contrast agent (CA) concentrations may have adversal effects
  • Contrast mechanisms in MI are goverened by physics
  • Detecting MI while minimizing CA would be beneficial

Goals:

  • Extend an MRI simulation framework to MI imaging with different CA concentrations
  • Develop DL based methods for image enhancement/reconstruction of low-CA concentration
  • Interdisciplinary team-work (physicists, clinicians, computer scientists) in an academic/industry environment
  • Apply the developed methods on patient data acquired at clinical cooperation sites

Requirements:

  • Knowledge of MRI physics and willingness to familiarize with CA kinetics
  • Strong background in implementing DL methods (Python)