Agreement Between MRI and NIRS Based Assessments of Skeletal Muscle Oxygenation During Reactive Hyperemia

Skeletal muscle oxygenation is a commonly reported variable in human physiology research, and is often assessed using near‐infrared spectroscopy (NIRS). In recent years, a novel magnetic resonance imaging (MRI) sequence, termed PIVOT (Perfusion, IntraVenous Oxygenation, and T2*) has been used to quantify skeletal muscle oxygenation during reactive hyperemia protocols. While the skeletal muscle oxygenation responses appear to be similar between investigations using the MRI PIVOT sequence and NIRS, the agreement between these two indices has yet to be explicitly reported. Purpose The purpose of this study was to determine the level of agreement between MRI PIVOT and NIRS based assessments of the skeletal muscle oxygenation (SmO 2 ) response to a reactive hyperemia protocol. Methods Fifteen subjects (6 males, 25 ± 3 yrs, 170.8 ± 11.9 cm, 73.6 ± 11.3 kg, 24.8 ± 1.9 kg/m 2 ) participated in a single laboratory visit, which was divided into two separate experiments. In one experiment, three consecutive reactive hyperemia protocols (2 min baseline, 5 min arterial occlusion, 5 min reactive hyperemia) were performed on the dominant leg, each separated by ~10 minutes of rest. Arterial occlusion was applied to the upper thigh with a pneumatic pressure cuff, and SmO 2 was recorded from the medial gastrocnemius (MG) and tibalis anterior (TA) via NIRS. Next, subjects were transitioned to the MRI facility, where the experiment was repeated inside the bore of a 3T MRI scanner. In this experiment, SmO 2 was collected using an adaptation of the MRI PIVOT sequence, and was represented as the intensity of the T2* signal. The order of experiments (MRI vs. NIRS) were counterbalanced between subjects, and the three reactive hyperemia trials within each experiment were averaged into a single response curve. From those response curves, a Pearson's correlation analysis was used to examine the linear relationship between the NIRS Tissue Saturation Index (TSI) response and the MRI T2* response during reactive hyperemia. Likewise, the areas under the curve (AUC; 2 min) for TSI and T2* were calculated for each subject and were compared between experiments using Bland‐Altman Plots. Linear Regression Analyses were also used to assess any potential measurement bias. Results Pearson's correlation analyses indicated a significant and moderate relationship between the reactive hyperemic responses collected via NIRS and MRI (R=0.760, P <0.001 and R=0.537, P <0.001 for the MG and TA, respectively). Bland Altman plots demonstrated a strong agreement between the TSI and T2* AUC values (data fell within the 95% confidence interval limits), and linear regression analyses were non‐significant (b= ‐0.753, P =0.068 and b=0.089, P =0.75 for the MG and TA, respectively), indicating no significant measurement bias. Conclusions These data suggest that there is a moderate level of agreement between the MRI PIVOT and NIRS based assessments of the SmO 2 response to reactive hyperemia.