Automatic quantification of traumatic brain injuries

It is difficult to predict what will happen in the long-term to somebody with traumatic brain injury (TBI) after a car or motorcycle crash, a fall, sports injuries or an assaults. TBI is common, it occurs to 2.5 million people each year in europe, and of those, 1 million are admitted to hospital and 75.000 die. TBI is the leading cause of death and disability in young adults and the incidence in elderly patients is increasing.

On magnetic resonance imaging (MRI) it is possible to identify small bleedings, and it is hypothesized that the amount and location of such bleedings may be used as a prognostic markers. Delineating such lesions, and assigning them to a brain region, is a tedious and cumbersome process, and not done in routine practice. We are therefore developing AI tools to automatically perform this quantification.

This project is part of a larger study, CENTER-TBI.

Funding

People

Kevin Koschmieder

Kevin Koschmieder

PhD Candidate

Anke van der Eerden

Anke van der Eerden

PhD Candidate

Rashindra Manniesing

Rashindra Manniesing

Assistant Professor

Publications

  • K. Koschmieder, A. van der Eerden, B. van Ginneken and R. Manniesing, "Brain Extraction in Susceptibility-Weighted MR Images using Deep Learning", Annual Meeting of the Radiological Society of North America, 2018. Abstract/PDF
  • T. van den Heuvel, A. van der Eerden, R. Manniesing, M. Ghafoorian, T. Tan, T. Andriessen, T. Vyvere, L. van den Hauwe, B. ter Romeny, B. Goraj and B. Platel, "Automated detection of cerebral microbleeds in patients with Traumatic Brain Injury", NeuroImage: Clinical, 2016;12:241 - 251. Abstract/PDF DOI PMID Cited by ~37