Early functional neuroanatomic studies by Broca described the association of the left cerebral hemisphere with language function. More recent investigations have included invasive optical techniques applied in conjunction with neurosurgery and, to a lesser degree, noninvasive optical techniques based on near-infrared spectroscopy (NIRS). Functional MR (fMR) imaging studies with blood oxygenation level–dependent (BOLD) contrast provide the advantage of high spatial resolution and association of the functional data with anatomic structure. These techniques have elicited widespread interest and found application in the investigation of normal brain function as well as a range of disease processes and are increasingly employed in the clinical setting for surgical planning. The case report by Murata et al published in this issue of the AJNR combines BOLD fMR imaging with noninvasive optical imaging and reports unexpected findings that may motivate further investigation into the assumptions that underlie these techniques.
To understand the potentially synergistic combination of NIRS and fMR imaging, it is useful to recall that the basic couplings between neuronal and vascular responses to stimulation are not well understood. In light of the fact that hemodynamic responses can be precisely and statistically correlated with functional stimuli, however, the connection can be studied. NIRS is designed to detect changes in cerebral perfusion produced by functional stimulation through measurement of changes in the concentrations of total hemoglobin (Hb), oxyhemoglobin (oxy-Hb), and deoxyhemoglobin (deoxy-Hb). An increase in regional cerebral blood flow, induced by stimulation and detected by NIRS, is reflected in an increase in the concentration of oxygenated hemoglobin and a decrease in the concentration of deoxygenated hemoglobin. Total hemoglobin also increases because of the increased cerebral blood flow. As noted by Obrig and others (1), NIRS is quite specific to hemoglobin and offers good temporal resolution with a simple and patient-friendly experimental setup. fMR imaging, on the other hand, provides exquisite anatomic detail to overlay the areas of detected activation. The signal intensity in fMR images reflects the concentration of (paramagnetic) deoxygenated hemoglobin but does not respond to the other two parameters obtained by NIRS. Thus, activation in fMR imaging is typically described as a decrease in paramagnetic deoxyhemoglobin, an increase in T2*, and an increase in the BOLD signal intensity.
In the case report, the classic BOLD response accompanied a physiologically consistent evoked cerebral blood oxygenation (CBO) change as a response to stimulation presurgery. That is, with functional activation of the brain through motor activity, there was a decrease in concentration of deoxhemoglobin accompanied by an increase of oxyhemoglobin and total hemoglobin concentrations. Postsurgery, an apparent atypical response occurred manifested by an increased regional cerebral blood flow and a negative BOLD signal intensity. Analysis of the NIRS data revealed that the deoxy-Hb concentration increased along with the oxy-Hb and total Hb concentrations instead of decreasing as expected with stimulation. The measured BOLD signal intensity was decreased (negative activation), consistent with the deoxy-Hb change.
The authors previously demonstrated similar evoked CBO changes associated with cerebral ischemia, and they suggest in the report that surgical injury could produce a relative ischemia due to decreased oxygen delivery in the surgical bed. This could result from a disproportionately large regional blood flow increase required to compensate for the reduced ability to deliver oxygen in the surgical bed postoperatively. That would produce a relative ischemia and increased deoxy-Hb concentration consistent with the observation. Other possibilities that could produce this atypical response are an increased oxygen extraction ratio (OER), recirculation, and variable production of deoxy-Hb. The OER can increase as a result of capillary dilatation. If this occurred, one would expect an increase in total Hb and an increase in deoxy-Hb, and there might be an associated increase in oxy-Hb concentration. Further, the NIRS volume measures total Hb and rCBF in an area much larger than the fMR imaging voxel, and this might overshadow small rCBF changes actually occurring in the voxel of brain depicted by BOLD MR imaging. This, however, is considered a less likely possibility than the increased OER and other physiologic changes previously described.
Despite the fact that this report describes a single patient, the study illuminates the potential benefit of obtaining NIRS studies with their physiologic hemoglobin measures to complement our understanding of the functional brain activation depicted by fMR imaging. The challenge of relating two techniques with different spatial resolutions should be not underestimated and may actually be a contributor to the present data. The increasing role of fMR imaging in surgical planning in particular motivates interest in characterizing the hemodynamic responses to surgical procedures. This interest may extend to intraoperative (invasive) NIRS measurements supporting the interpretation of the postsurgical NIRS. Studies that relate these evoked CBO changes to the fMR imaging signal intensity may provide surgical guidance in specific cases as well as further insight into brain organization and reorganization and may help in management of nonsurgical disease processes that affect cerebral blood flow. Interpretation of changes in BOLD contrast should include consideration of the potential for regional cerebral blood flow to increase in conjunction with either an increase or a decrease in deoxy- hemoglobin concentration. Experimental animal studies with more invasive physiologic monitoring may help to further refine and define these techniques and their application to human brain evaluation and treatment.
Reference
- Copyright © American Society of Neuroradiology