Texture analysis of MR images of patients with mild traumatic brain injury

Kirsi K Holli1, Lara Harrison, Prasun Dastidar

  • 1Medical Imaging Center, Tampere University Hospital, Tampere, Finland. kirsi.holli@tut.fi

BMC Medical Imaging
|May 14, 2010
PubMed
Abstract

Insights

Texture analysis (TA) detected significant microstructural changes in cerebral tissue of mild traumatic brain injury (MTBI) patients. These subtle changes, invisible to the eye, can aid in early MTBI diagnosis.

Area of Science:

  • Neuroimaging
  • Medical Physics
  • Radiology

Background:

  • Mild traumatic brain injury (MTBI) can cause subtle microstructural changes in cerebral tissue.
  • These changes may not be apparent through visual inspection of MRI scans.
  • Texture analysis (TA) offers a potential method to detect these otherwise invisible alterations.

Purpose of the Study:

  • To investigate the impact of trauma on texture features in cerebral tissue in MTBI patients.
  • To determine if TA can identify microstructural changes not visible on standard MRI.
  • To evaluate TA as a tool for early MTBI diagnosis.

Main Methods:

  • 42 MTBI patients and 10 healthy controls underwent 1.5 T MRI within three weeks of trauma.
  • Texture analysis (TA) was applied to specific brain regions, including the mesencephalon, cerebral white matter, and corpus callosum (CC) segments.
  • TA data from patients were compared to controls, analyzing interhemispheric and CC segment differences.

Main Results:

  • Statistically significant differences in TA parameters (co-occurrence and run-length matrix based) were observed between cerebral hemispheres and CC segments in MTBI patients.
  • These differences were particularly noted in the mesencephalon, corona radiata, and CC segments.
  • Healthy controls showed considerably fewer differences in TA parameters.

Conclusions:

  • TA successfully revealed significant changes in cerebral tissue texture parameters in MTBI patients.
  • These findings suggest TA can detect conventionally invisible microstructural changes in MTBI.
  • TA shows promise as an additional diagnostic tool for early MTBI detection.