Precision of Cerebrovascular Reactivity Assessment with Use of Different Quantification Methods for Hypercapnia Functional MR Imaging

Abstract

Tools for noninvasive mapping of hemodynamic function including cerebrovascular reactivity are emerging and may become clinically useful to predict tissue at hemodynamic risk. One such technique assesses blood oxygen level-dependent (BOLD) MR imaging contrast in response to hypercapnia, but the reliability of its quantification is uncertain. The aim of this study was to prospectively investigate the intersubject and interhemispheric variability and short-term reproducibility of hypercapnia functional MR imaging (fMRI) in healthy volunteers and to assess the effects of different methods of quantification and normalization. Sixteen healthy volunteers, (7 women and 9 men) underwent hypercapnia fMRI with a clinical 1.5T scanner; 8 underwent scanning twice. We determined BOLD amplitude changes using a visually defined block design or automated regression to end-tidal (ET) carbon dioxide (CO2). Absolute percent signal intensity changes (PSC) were extracted for whole-brain, gray matter, and middle cerebral artery territory, and also normalized to ETCO2 change. Intersubject and intrasubject (between hemispheres and sessions) coefficients of variation (COV) were derived. We assessed the effects of different quantification methods on reproducibility indices using the t test and U tests. The mean change in ETCO2 was 7.8 +/- 3.3 mm Hg. Averaged BOLD increases varied from 2.54% to 2.92%. Short-term reproducibility was good for absolute PSC (4.8% to 10%) but poor for normalized PSC (range, 24% to 27% COV). Intersubject reproducibility varied between 11% and 23% for absolute PSC and, again, was poorer for normalized data (32% to 39%). Interhemispheric reproducibility of absolute PSC was excellent ranging between 1.24 and 2.16% COV. In conclusion, quantification of cerebrovascular reactivity with use of hypercapnia fMRI was found to have good between-session and very good interhemispheric reproducibility. The technique holds promise as a diagnostic tool, especially for sensitive detection of unilateral disease.