Investigators interested in neural basis of fMRI and resting state connectivityLow-frequency fluctuations in resting state functional MRI (rsfMRI) signals have been linked to changes in spontaneous neuronal activity [1], but their specific relationships have not been well established. Anesthesia is known to suppress neuronal activity, so by examining the effects of different levels of anesthesia on changes in inter-regional functional connectivity and signal power spectra, we are able to investigate the neural origins of rsfMRI signals. Recent studies have demonstrated that different isoflurane dosages affect connectivity strengths between cortical regions [2]. Extending this line of research, we acquired resting state MR images from live anesthetized squirrel monkeys and measured how the fractional amplitude of low frequency fluctuations (fALFF), amplitude of low frequency fluctuations (ALFF), and inter-regional functional connectivity metrics varied within a small local network (subregions within the primary somatosensory cortex, S1) as isoflurane levels were altered. Two male adult squirrel monkeys have been studied to date. MR images were acquired on a 9.4T magnet using a 3-cm diameter surface transmit-receive coil centered over S1 cortex. Four 2-mm-thick oblique image slices were centered over the central and lateral sulci. High resolution T2*-weighted anatomical images were also acquired using a gradient echo sequence (TR/TE=200/16 ms, resolution of 0.068x0.068x2 mm3). BOLD-sensitive functional images were acquired in both resting and tactile stimulation conditions, using a T2*-weighted GE-EPI sequence (TR/TE=1500/19ms, 1 shot, resolution of 0.547x0.547x2 mm3, 300 volumes). Isoflurane was maintained at three different levels: 0.5%, 0.875% and 1.25%. At each isoflurane level, multiple runs of the functional data were acquired. 8 Hz vibrotactile stimulation of individual distal finger pads, as alternating 30 seconds off/on blocks, was used to elicit cortical activations in S1 subregions. The strongest stimulus-responsive voxels were identified as seeds in areas 3b, 3a and 1 for subsequent resting state analyses, along with reference regions in adjacent surface muscle. Resting fMRI data were corrected for slice-timing, motion and then spatially smoothed in-plane with a full width at half maximum of 0.8mm in spm8 with Matlab. Motion regression was then performed as well as linear detrending. fALFF and ALFF maps were computed by transforming each voxel time series into its power spectral density via the Fourier Transform, with emphasis on the low frequency range (0.01-0.08Hz). The functional responses of fMRI seeds were confirmed with later electrophysiology recordings and those voxels were used for both power and inter-ROI correlation analyses. Representative time series and their respective power spectral densities at different anesthesia levels revealed diminishing BOLD signal fluctuations when isoflurane level was increased (Figure 1A). Power amplitudes in the time series under low anesthesia were also greater than the ones at higher anesthesia level (Figure 1B). Boxplot group analyses of fALFF and ALFF in the S1 subregions further showed that low frequency oscillations decreased linearly as anesthesia level increased (Figure 2) whereas control muscle regions showed no trends (data not shown). Furthermore, functional connectivity strengths between seed regions of interests for different isoflurane levels showed trends similar to power analyses. In particular, the number of highly correlated voxels to the area 1 seed is significantly reduced (Figure 3, thresholded at r>0.6) as anesthesia level was increased from 0.5%, 0.875, and 1.25%. Moreover, pair-wise quantification of the inter-regional functional connectivity strengths (r values) at the group level supports this observation (Figure 4A). Correlation strength between area 1 and area 3b decreased as isoflurane levels increased (left three columns of Figure 4A), while the correlation strengths remained rather constant between area 1 and the control regions (right three columns of Figure 4A). A matrix plot of the mean inter-regional correlation strengths as a function of different anesthesia levels (Figure 4B) further highlights the same phenomenon. In summary, quantification of power changes and correlation strengths at the group level indicate that increased isoflurane level decreased the fluctuation of BOLD signals and the measured correlations within a small local circuit. Variations in neural activity by changing the anesthesia levels of squirrel monkeys have been reflected in both the power analyses and apparent functional connectivity of resting state fMRI signals in specific regions of the somatosensory region. Given the known suppressive effects of anesthesia on neuronal activity, our observation of decreased functional connectivity to increased level of isoflurane supports the neuronal origins of rsfMRI signals, though further measurements of global changes in SNR and their potential influence are being evaluated.