Authors

  1. Jantzen, Kelly J. PhD
  2. Sedita, Steve BSc
  3. Magne, Cyrille PhD
  4. Anderson, Brian ATF
  5. Steinberg, Fred L. MD
  6. Kelso, J. A. Scott PhD

Article Content

Objectives: We outline, quantify, and attempt to reduce the sources of variation that may influence prospective functional imaging approaches to studying the neuropathophysiology of mild traumatic brain injury. Hypothesis: BOLD signal variability in a single subject can result in significant between-session differences. Quantifying this variability will aid in establishing new, appropriate statistical criteria. Variability will be reduced when employing a parametric approach. Participants: Between-subject variability was assessed using a single recording session from each of 8 different college students. Within-subject variability was assessed using 2 subjects each studied over 9 sessions. Methods: A session consisted of 3 finger-sequencing tasks and an NBack memory task. F tests quantified the contribution of between-subject variance. Permutation methods provided a measure of the magnitude of within-subject differences expected by chance. Results: Within-subject variability was small requiring a slight increase in the statistical threshold for detecting true session differences. This variability was reduced when parametric measures that quantify the pattern across task levels were compared. Conclusions: The research underscores the increases sensitivity of a prospective functional imaging approach and its potential advantage over group approaches for understanding the neurofunctional basis of mild brain injury and recovery of function.

 

REFERENCE

 

1. Jantzen KJ, Anderson B, Steinberg FL, Kelso JAS. A prospective functional MR imaging study of mild traumatic brain injury in college football players. Am J Neuroradiol. 2004;25:738-745.