ROC analysis improves CMRO2 estimation for rehabilitation subjects

B., Glielmi C.; A., Schuchard R.; S., LaConte; Hu, X. P.

by Glielmi C. B., Schuchard R. A., LaConte S., X. P. Hu

Abstract:

Longitudinal rehabilitation studies require fMRI measures and analysis techniques that provide consistency between scanning sessions such as cerebral blood flow (CBF) and cerebral metabolic rate of oxygen (CMRO2). CBF and sub-optimal BOLD are often simultaneously acquired to estimate CMRO2. This work utilizes ROC analysis to select (i) sub-optimal BOLD datasets with the highest detectability and (ii) statistical thresholds for optimal ROI determination. The results demonstrate that both of these tactics improve the likelihood of reasonable CMRO2 estimation in low vision subjects, presumably improving the accuracy of CMRO2 estimation for future studies involving rehabilitation patient populations.

Reference:

abstract Glielmi C. B., Schuchard R. A., LaConte S., X. P. Hu. ROC analysis improves CMRO2 estimation for rehabilitation subjects. In Proceedings 15th Scientific Meeting, International Society for Magnetic Resonance in Medicine, Berlin, page 3473, 2007. [bibtex]

Bibtex Entry:

@inproceedings{Berlin3473,
   Author = {Glielmi C. B. and Schuchard R. A. and LaConte S. and Hu, X. P.},
   Title ={{R}{O}{C} analysis improves {C}{M}{R}{O}2 estimation for rehabilitation subjects},
   BookTitle = {Proceedings 15th Scientific Meeting, International Society for Magnetic Resonance in Medicine, Berlin},
   Pages = {3473},
   Abstract = {Longitudinal rehabilitation studies require fMRI measures and analysis techniques that provide consistency between scanning sessions such as cerebral blood flow (CBF) and cerebral metabolic rate of oxygen (CMRO2). CBF and sub-optimal BOLD are often simultaneously acquired to estimate CMRO2.  This work utilizes ROC analysis to select (i) sub-optimal BOLD datasets with the highest detectability and (ii) statistical thresholds for optimal ROI determination. The results demonstrate that both of these tactics improve the likelihood of reasonable CMRO2 estimation in low vision subjects, presumably improving the accuracy of CMRO2 estimation for future studies involving rehabilitation patient populations.},
 Keywords = {Berlin3473},
   Year = {2007} }