Figure 1. The implant for head-fixation (A) with EEG electrode connector (B), a head-fixed rat wearing walking harness in a custom-made habituation and imaging holder for behavioral studies (C), and a habituated rat inside the 9.4T MRI scanner. Two 2 mm pins penetrating the implant (A, B) allow a robust head-fixation (C). Warm-water circulation kept the platform of the holder warm (C). The nose cone was remotely removed (compare C and D) and returned when necessary. A single loop transmitter-receiver coil was placed around the implant and below the head-fixing pins (D) to conduct MRI.

Figure 4. Activation maps (A-E) obtained during spontaneous behavior. Timings of the behavior are shown on the fMRI time series as shaded regions. The block design analysis included only a single event. In A, the rat was whisking. In B, the rat was sniffing. In C, the rat expressed increased arousal. In D, the rat tried to walk. In E, the rat tried to walk but paws slipped. The maps are overlaid on anatomical images. The slight mismatch between images and maps originates from the susceptibility-induced signal void in anatomical images that is not present in the MB-SWIFT images.

Figure 2: Example of results in a single subject

Top: BOLD fMRI showing Superior Colliculus and line position.

Middle: 2D/3D renderings of the space-time activation pattern (b=0) ([9 9] median filtering). The activation spot extends beyond the end of the stimulation.

Bottom: id with b=1800s/mm² showing 2 peaks with slightly different spatial extents, one inside the activation window and the other after the stimulation

Right: Response time courses. The response amplitude (beta) is higher for b=1800s/mm² and present 2 peaks. Both responses are statistically significant.

Figure 3: Group results

Left: Group average (n=6) for BP (15 points moving average). The amplitude increase with b value of the main peak and the shoulder are clearly visible.

Middle: Individual results (5 subjects) showing the differences in amplitude and in time-to-peak between b=0 and 1800s/mm² (p=0.016 and p=0.032, respectively).

Right: Group average (n=5) for ER (10 points moving average).

(A) Statistical BOLD activation map of a coronal section of the OB. ROIs were arbitrary chosen and represented by green frames. (B) ROIs were reported onto RARE anatomical acquisitions and (C) transferred to fUS-PD maps by means of the co-registration. (D) BOLD (red) and fUS-PD (black) time courses within each ROI: the correlation between the 2 signals is apparent.

(A) top view of the bone over the OB and the 3D-printed plastic reservoir. The shape on the OB under the bone is depicted by the black lines. (B) 3D reconstruction from a RARE acquisition of the ensemble OB + agar-filled reservoir. Under the reservoir is the nasal cavity with the fibers connecting the olfactory epithelium to the OB.

Figure 1. Laminar BOLD responses to neural stimulation and hypercapnic challenge have different origins.

A) The cortical areas (S1FL and M1) were linearized using radially projecting lines perpendicular to the cortical edges.

B) Dynamic BOLD changes were calculated with average window of 2 s duration at interval of 1 s. Sensory-evoked responses first appeared in S1FL middle layer (yellow arrow), whereas early responses by M1 excitation in upper areas of M1 and S1FL (white arrows), and hypercapnic responses started in M1 upper layer (red arrow).

Figure 3. Layers receiving synaptic inputs show the earliest BOLD responses.

A-B) The fitting curves were normalized to the peak (showing from 0 s to 3 s after stimulus onset), and the times to reach 5 % and 30 % of the peak were then measured in the S1FL layer-specific ROIs. The order of early BOLD responses was Middle → Upper → Lower for forepaw stimulation and Upper → Middle → Lower for M1 excitation and hypercapnia.

error bars in B, SEM; colored circles in B, individual animal data; * and ** in B, p < 0.05 and p < 0.01, respectively (repeated ANOVA followed by Tukey post hoc test).

Fig. 1. CBV-weighted (wCBV) activation maps evoked by visual stimuli. (A) Area 18 of cat primary visual cortex; field of view of imaging denoted by a white rectangular box. F-statistics of wCBV activation induced by 8 orientation gratings presented with blue-purple colors. The pronounced wCBV activation (P < 0.001) from the volumetric data displayed on orthogonal anatomical slices: (B) axial, (C) sagittal and (D) coronal plane.

Fig. 2. Laminar iso-orientation maps in cat visual cortex obtained by fMRI. (A) a coronal slice where radial intracortical veins are seen as dark lines. The yellow lines indicate 7 slices cut orthogonal to the cortical surface by (post-processed including interpolation) of the volumetric data. (B) laminar orientation preference maps seen in these 7 slices. (C) Interpolated depth-dependent orientation selectivity index (OSI) from wCBV fMRI; individual animals (gray lines) and group average (black line, n = 7 cats). Short black vertical lines represent ±1 S.E.M.

Fig. 1. BiLS acquisition in resting-state fMRI. A) Sequential signal processing of the conventional UniLS method. B) Spatiotemporal map and percentage change map in evoked fMRI using the UniLS method. C-F) resting-state BOLD responses in the symmetric FP-S1 regions using BiLS method (n = 32 trials of 4 rats). C) Top: Representative Z-normalized fMRI time series. Bottom: normalized cortical depth maps in both the FP-S1 regions from one representative trial. D-E) Comparison of the filtered time series (D) and their PSDs from the same trial (E). F) PSDs from another representative trial.

Fig. 2. Laminar-specific coherence in rs-fMRI. A) Laminar-specific coherence (n = 32 trials of 4 rats) showing that L2/3 is significantly different at 0.08-0.1 Hz (one-way ANOVA, post-hoc: *p-value <0.05, Bonferroni correction). B) K-means clustering based grouping with L2/3 specific coherence values at 0.01-0.02 Hz (x-axis) and 0.08-0.1 Hz (y-axis). C) Independent group t-test with the L2/3 coherence values from the individual groups at 0.01-0.02 Hz (left) and 0.08-0.1 Hz (right), showing significant difference at 0.08-0.1 Hz (**p-value = 1.4127*10^-10).

Figure 1. Activation in insula and lateral sulcal areas elicited by stimulation of the basal nucleus of the amygdala.

A-L. Monkey M-P. Monkey Y. A-B. Stimulation site in the basal nucleus. C-D. Schematics of cortical areas in the lateral sulcus. E-H. Activations within the lateral sulcus. Voxels: p<0.0001 (FDR 5.5%). I-L: Mean time courses of significant voxels in E-H (mean of significant voxels). Inset: enlarged schematic of each stimulus train (red line). Stimulation: block design. Error bars: SEM. M-P. Activations within the lateral sulcus are seen in monkey Y. Voxels: p<0.001.

Fig.1: Stimulation Regimes (A) Rat visual pathway schematic. Orange - geniculate pathway; Blue - extrageniculate pathway; Black - Callosal pathway; Green: Cortical feedback projections; Light Purple: Commissural tectotectal projections. The blue dots represent the visual stimulus (either monocular or binocular). The different stimulation regimes and the implications along the pathway are represented in the different panels: (B) Binocular stimulation; (C) Monocular stimulation; (D) Binocular stimulation with V1 lesion; (E) Monocular stimulation with V1 lesion.

Fig.3: BOLD t-maps and time-courses for the binocular stimulation regimes. (A) Binocular stimulation of non-lesioned animals; (B) Binocular stimulation of V1 lesioned animals. Top: BOLD t-maps; Bottom: Percent signal change average cycle for the different ROIs at different ISIs.

Fig.3: TTA-P2 vs Baseline t-statistic maps highlighting regions with enhanced FC to the ROI encompassing all right hemisphere auditory cortex regions after TTA-P2 administration. The slice-location co-ordinates are in Paxinos space ^16,17. Left-hemisphere is on the left-hand side of the maps.

Fig.1: The evolution of the strengths of QPPs with time assessed with spatio-temporal correlation of the fMRI time-series with corresponding QPP template. Examples from (a-c) three rats after systematic administration of TTA-P2; and one rat (d) after systematic administration of Vehicle.

Figure 2. Inactivating hippocampal outputs abolishes response selectivity to aversive vocalizations in IC, MGB and AC. (A) Illustration of the atlas-based ROI definitions (Top). Averaged auditory-evoked activation maps before and after TTX infusion (Bottom). (B) BOLD signal profiles extracted from the defined ROIs of IC, MGB, and AC (error bars indicate SEM). (C) Averaged BOLD signal comparison showing the influence of TTX inactivation of vHP on response selectivity to aversive vocalizations. (Paired two-sample t-test ,* for p < 0.05, and ** for p < 0.01).

Figure 1. (A) Illustrations of TTX infusion into right vHP. (B) Experimental protocol of auditory fMRI experiments. Standard block paradigm (20s ON and 40s OFF) was used to present vocalizations to the left ear. Forward and reversed vocalizations were interleaved during each auditory fMRI session. (C) The temporal waveform of forward (left) and temporally reversed (right) aversive vocalizations. (D) The spectrograms of forward (left) and temporally reversed (right) aversive vocalizations.

Figure 4. ZTE and EPI functional responses to forepaw electrical stimulation. a) ZTE-fMRI and b) BOLD-weighted fMRI activation maps and averaged timcourses. Time courses were extracted from the mean of 3 voxel³ ROIs in contralateral S1. ZTE-fMRI exhibited a 67 % greater CNR than that of BOLD-weighted EPI ( t(52)=4.80,P=1.37x10^-5).

Figure 2: Effect of reconstruction algorithm on ZTE functional data. Coronal (scanner coordinate system) views of raw ZTE-fMRI data (a) following online reconstruction, nufft and l1-Wavelet regularization and (b) the corresponding tSNR maps. Comparison of CNR of evoked response and SNR of functional data during rat forepaw electrical stimulation (c). Effect of l1-wavelet regularization parameter on CNR and SNR of functional ZTE data. For both c) and (d) n=3 subjects, 27 trials.

Figure 1: Representative Spin Echo EPI (top) and Fully/Non-Fully-Refocused SPEN (bottom) acquisitions used in this study. The latter included a 180˚ chirp pulse acting in the presence of a gradient that encodes the more artifact-prone, low bandwidth dimension, and lasts half the duration of the readout acquisition train Ta. This is preceded by a pre-encoding delay; if set to Ta/2 full-refocusing is achieved, and T2* effects are largely attenuated (red); if set to a shorter time (blue) a T2* weighting is partially reintroduced.

Figure 5: Box-whiskers plots describing the maximal signal activation observed for the mice examined by SPEN in this work.

Figure 3: Correlation matrix of rs-fMRI data. This figure shows an adjacency matrix of the 144 nodes where red indicates a positive correlation and blue indicates a negative correlation. Nodes from 1-72 are located on the left side of the brain while nodes 73-144 are located on the right side of the brain. Correlations can be seen generally split by hemisphere. Positive correlations tend to appear within hemispheres while negative correlations appear across hemispheres.

Figure 2: Activation map of filtered and corrected data highlighting the areas of activation, informed by the anatomical ROI identification. (A) all rs-fMRI signals detected (red = higher intensity), whereas (B) displays only the areas of highest intensity after removal of residual noise. (C) activation data overlaid on original EPI with the areas of activation to identify anatomical regions of interest.

Figure2. (A) show SNR and tSNR maps from 2D/3D-EPI (blue/green squares) and ERASE (red circles) in the color scale range between 0~60 and 0~30. As shown in (B) where total voxel values of tSNR and SNR were scatter plotted. SNR: signal-to-noise ratio, tSNR: temporal SNR, error bars by ± standard error of mean (SEM).

Figure1. ERASE sequence diagram (A) and schematic description of its sequential and local excitation and refocusing mechanism in the SPEN direction (B). In the ERASE sequence, the excitation duration of the chirp pulse (T_e) is set to be same as total acquisition duration (T_a) and, with a re-phasing gradient between them, all spins experience constant TE across an object (B). SPEN is applied for slab encoding in the slab-selective direction. G_e: excitation gradient, G_a: acquisition gradient, G_r: re-phasing gradient, RO: read-out, PE: phase-encoding, FM: frequency-modulation.

Fig. 4 In-vivo experiment: comparing denoising performances for VST-SVS vs Gaussian smoothing (Gauss) vs MPPCA in terms of BOLD percent change. Note that the proposed VST-SVS method outperformed the conventional Gaussian smoothing when both used to denoise a single run, increasing activation areas to a level visually comparable in size to what was achievable with Gaussian smoothing but using all 20 runs.

Fig. 5 In-vivo experiment: laminar BOLD profiles are shown for both hemispheres and are displayed for noisy data (black lines), Gaussian smoothing (red lines), and VST-SVS (α=0, patch averaging) (blue lines) derived from either single run (ind, solid lines) or concatenated 20 runs (mul, dashed lines). Note that the use of VST-SVS to denoise a single run gave rise to a laminar profile comparable to that achievable with 20 runs using conventional Gaussian smoothing.

Fig. 3 – Group conventional fMRI results (n=3). (A) Group BOLD Fourier maps computed from data of 9 individual scans without denoising (top) and after MP-PCA_10 denoising (bottom). (B) Difference of spectral amplitude maps at the fundamental frequency between MP-PCA_10 and undenoised data. (C) Regions of interest (ROIs) of the mouse visual pathway delineated on the Allen Reference Atlas.

Fig. 4 – Ultrafast fMRI results. (A) Individual (top) and group (bottom, n=3, 9 scans) BOLD Fourier maps from ultrafast data, before and after MP-PCA_250 denoising. (B) Difference between spectral maps. (C) Amplitude spectra from the stimulation paradigm and from the right SC’s signal, using undenoised and MP-PCA_250 data from an individual scan (top) or all scans (bottom). (D) ROIs in the oblique brain slice. (E) Variation of maximum spectral amplitude in 9 individual maps with denoising.

Fig. 1. wCBV fMRI results improved by NORDIC denoising. (A) wCBV activation map (T-statistics (T ≥ 3), sagittal slice, single subject) induced by stimuli. (B) The mean time course from the region-of-interest (ROI) denoted by white rectangles in (A); stimulus-induced wCBV signal changes shown during stimulus on epochs (blue/gray rectangles). The raw EPI signal times series (black) is shown together with the GLM fit (blue). (C) Effect of NORDIC on the average T-stats in the region-of-interest (ROI), R-square, and functional CNR; the error bars are SEM.

Fig. 4. Orientation preference mapping and statistics improved before and after NORDIC denoising. (A) Orientation preference mapped by NORDIC denoised wCBV responses to 8-orientation (single subject); colors indicate the orientation preference in degree. (B) Representative tuning curve of a single voxel before- (red) and after-NORDIC denoising (blue). (C-D) The standard deviations of multiple wCBV responses repeatedly measured and goodness-of-fit of curve fitting (root-mean-square-error) were compared between before and after NORDIC denoising.

fig. 5: Color coding for number of measurements that showed oscillations for resting state. Regions in which at least 20 % of the data showed oscillation are shown for the right hemisphere at bregma 1.0. For group 1 and 2 (A and B), data was separated into early (first 3 hours) and late. Only late datasets showed oscillations in all regions. In group 3 and 4 (C and D), oscillation occurred not under isuflurane, but under medetmidine in cortical regions. Stimulation data showed similar results.

fig. 1: Exemplary time courses without and with BOLD oscillations for both, (A) resting state and (B) electrical paw stimulation. Stimulation phases are indicated by a grey bar.

Figure 1. Averaged Global Signals for Muscle and Brain. A.) Shows the frequency power spectrum of the averaged brain global signal of 49 scans, each 10 minutes in length, taken across 8 rats. B.) Shows the frequency power spectrum of the averaged muscle “global signal” of 49 scans, each 10 minutes in length, taken across 8 rats. It is important to note the relative difference in magnitude between the two signals with the muscle global signal displaying much lower power.

Figure 2. Voxel-wise Correlation to Brain and Muscle Global Signal. Panels A – C show a voxel-wise map of Pearson correlation coefficients for 3 individual rats for which the timecourse of each voxel was compared to the global signal of the brain. Panels D – E show the same analysis comparing voxels to the “global signal” of the muscle. Panels D – E display the same rat and slice number as the corresponding panel in the first row (A – C, respectively). Correlation values for muscle to brain global signals for each rat, from left to right, were R = 0.7313, R = -0.0839, and R = -0.1345, respectively.

Visual outline of analysis methods

Figure 1: Primary visual cortex is specified with reference to Paxinos and Franklin Mouse Brain Atlas (Figure 1A), V1 is marked with orange arrows in T2-weighted structural scan (Figure 1B). Primary visual cortex indicated by the arrows. (C) BOLD and (D) VASO functional maps of mouse primary visual cortex. Row 1 (Top) and Row 2 (Bottom) are the results acquiried from two different mice.

Figure 2: VASO signal time course of averaged across all stimulation blocks (shaded area) of 4 runs for 2 different mice, errorbars represents the standard error of mean.

Calcium recordings show different brain states in S1HL depending on anesthetic condition. (A) Under ISO anesthesia, calcium transients show frequent UP-DOWN transitions (green) while changing to a persistent state at 25 (purple), 45 (red), and 100 (pink) minutes after switching to ISO/MED anesthesia. (B) Fourier-transformed calcium transients show an emerging peak between 0.5Hz and 1.0 Hz 15 to 100 minutes after switching anesthetic regimens. (C) Frequencies of calcium transients between 0.1-0.5Hz decreased (p=0.002), while frequencies from 0.5-1.0Hz increased (p=0.02).

Changes in functional connectivity were mostly related to a decrease in ISO concentration. (A) Circular network representation. Each dot represents a brain region, color codes functional groups, and lines represent functional connectivity based on pearson’s correlation coefficient. (B) Averaged networks for each group. (C) Differences in networks, obtained from statistical analysis of ISO vs. later time points (left panel) and ISO/MED 45 min vs. ISO/MED 25min and ISO/MED 100 min (right panel).

Fig.1. GE-EPI of the mouse brain. Susceptibility artefacts under different combinations of fiber diameters (200/100/60 µm) and dental cements (Meliodent/C&B), with (A) no covering materials, (B) toothpaste, or (C) kwik-cast.

Fig.3. Calcium and BOLD responses under visual stimulation. (A) BOLD activation map (top panel), stereotaxic coordinates of targeted regions for injecting GCaMP6f (bottom panel). (B) LGd (n=3 scans), and (C) V1 (n=6 scans) activations, (D) Transfer functions calculated from the averaged BOLD and calcium signals.

Figure 3. (a) The correlation coefficients of 7 centroids represented 7 states of dynamic functional connectivity of Pre-FUS, FUS-35min, and FUS-3hr groups. (b,c) The bar charts and spider chart illustrated the probability of 7 states occurring in each group.

Figure 1. The flow chart of computing correlation coefficients of functional connectivity maps. (a) The determined 36 ROIs of rat brain. (b) Extracting BOLD signal from each ROI. (c) To evaluate dynamic FC maps, a sliding window method was performed to calculate Pearson’s correlation coefficients between 36 ROIs. (d) 36 ROIs were divided into 9 groups on FC map.

A rendering of the cradle assembly with coil holder, coil, acrylic pieces, and X-shaped head-bar. A rendering of a mouse skull is inserted to show scale.

Renderings of the I-shaped (left) and X-shaped (right) head-bar designs. Both head-bars are fabricated out of carbon fiber.

Histogram of mice recovery duration

Animal recovery timeline. A description of the animal recovery process from the end of fMRI scans

Fig 1: The FC of light sedated animals (top, N = 102) and awake (bottom, N = 10) were compared with the FDR-corrected studentized permutation test, *p<0.05; medial prefrontal cortex (mPFC), motor cortex (M1M2), somatosensory cortex (S1S2), visual cortex (V1V2), auditory cortex (AU), retrosplenial cortex (RSc), hippocampus (CA), striatum (CPu), nucleus accumbens (Nacc), ventrolateral thalamus (VLTh), medial thalamus (MTh), hypothalamus (HTh).

Fig 2: The FC of on-site measured light sedated animals (top, N = 100) and data from awake rat database (bottom, N = 159). ROI annotations are the same as in Fig 1 with the addition of the cingulate cortex (cG). The bilateral ROIs are denoted by L and R for the left and right sides respectively.

Figure 4 Retinotopic Maps. A: Projection of the RF estimated phase (θ) obtained for the SC for all the animals tested. B: Phase map obtained for the VC of rat 4. C: RF maps of two voxels located in the VC of the left and right hemispheres. D: Phase map obtained for the LGN of rat 4.

Figure 5. RF size profile across cortical depth. A: Back-projection of the RF size property averaged across animals. The colorbar corresponds to degrees of visual angle. B: Two RF maps obtained for voxels located in Layer LII/III and Layer IV. C: Profile of the RF size across cortical depth, the dashed blue area corresponds to the 95% CI.

Figure 1. Red-green-blue color coding of common and discrete CBVw-fMRI activations for each odor with baseline EPI underlay. Top 5% p-value threshold with a family-wise error correction of 30 minimum voxels in a cluster. (D – dorsal, V – ventral, R – right, L – left)

Figure 2. Motion characteristics of all awake fMRI scans. A) Histogram of all framewise displacement points with values exceeding a cut-off threshold of 25 μm (1/4 of a voxel) being highlighted in orange. B) Mean displacement per scan before and after motion censoring.

Fig. 2 (A) Schematic illustration of the automated odor stimulation system. A drop of odorant solution is placed within a syringe pump, and the syringe becomes saturated with vapor. The saturated vapor is infused into the animals by operating the syringe pump and the electromagnetic valve. (B) Independent component analysis. Activations were detected as periodically occurring responses, even in the presence of large noise signals.

Fig. 3 (A) Number of muscone-evoked activations detected in the olfactory bulb and the olfactory cortex under different levels of medetomidine anesthesia. (B) Representative activation map detected at the 0.01 mg/kg/h medetomidine concentration, in the olfactory bulb, the piriform cortex, the olfactory tubercle, and the nucleus accumbens. In each slice, the distance from the bregma is shown at the bottom right of the panel.

Fig.2: BOLD t-maps. TOP: Normal reared animals (N=7, 15 runs); BOTTOM: Dark reared animals (N=6, 16 runs). A significance value of 0.001 (FDR corrected) with minimum cluster size of 20 voxels was used in the generation of the maps. Blue arrows are pointing to the contralateral V1B region while yellow arrows are pointing to the contralateral lateral geniculate nucleus of the thalamus.

Fig.3: Cortical average cycle. Percent signal change (mean ± s.e.m) of (A) Contralateral VC; (B) Ipsilateral VC; (C) Contralateral V1B; (D) Ipsilateral V1B. Orange curves represent results from normal reared animals while black curves represent results for dark reared animals. Blue shaded region represents the stimulation period.

Figure 1 Group analysis of BOLD signal changes by the somatosensory stimulation in mice (n=7, p<0.05, FDR-corrected). Yellow rectangle shows the horizontal plane in right panel. Color bar indicates the t-values. SCbf, somatosensory cortex of barrel field; S1; primary somatosensory cortex.

Figure 2 Averaged time-course of BOLD signal changes in (A) contralateral primary somatosensory cortex (contra-S1), (B) ipsilateral primary somatosensory cortex (ipsi-S1), (C) contralateral primary somatosensory cortex barrel field (contra-Scbf), (D) ipsilateral primary somatosensory cortex barrel field (ipsi-Scbf). (E) contralateral ventral posterolateral nucleus (contra-VPL), (F) ipsilateral ventral posterolateral nucleus (ipsi-VPL). Dotted line shows 5 seconds after the end of stimulation. Data are mean ± SEM. *p < 0.05, compared to averaged baseline.

Fig3. Detrended time-courses of prominent auditory structures (mean ± s.e.m.) ROIs were manually drawn according to an atlas²⁶ and time-courses were detrended with a polynomial fit to the resting periods. ICi and STc show clear negative responses with a sharp post stimulus overshoot.

Fig2. GLM maps show positive BOLD activation along the auditory pathway, Cochlear Nucleus (CNi), Contralateral Inferior Colliculus (ICc), Ventral Lateral Lemniscus (LLc)) and Medial Geniculate Body (MGBc), while Ipsilateral Inferior Colliculus (ICi) and Striatum (Stc) show negative responses. A minimum significance level of 0.001 (FDR corrected), minimum cluster size of 20 voxels was used, with an HRF peaking at 1s was convolved with the paradigm.

Figure 2 | A) Excitatory pyramidal activity during 2 Hz (peak dF/F = 11.9, mean dF/F = 8.8 with SEM = 2.8) was higher than for 10 Hz stimulation (peak dF/F = 6.9, mean dF/F = 4.7 with SEM = 3.5). B) Inhibitory somatostatin+ neurons showed a similar trend during 2 and 10 Hz stimulation (2 Hz: peak dF/F = 29.4, mean dF/F = 15.7 with SEM = 2.5; 10 Hz: peak dF/F = 28.2, mean dF/F = 17.9 with SEM = 3.1). C) The comparison between the area under the curve (AUC) for the somatostatin+ dF/F for the 2000 – 5000 ms interval is higher during the 2 Hz than the 10 Hz stimulation (Arbitrary value = 205.6 and 341.7 respectively).

Figure 1 | A) Awake mouse installed in MR cradle with custom developed PEEK headbars. B) MR image of fiber placement on top of a glass cortical window. C) Schematic of the multi-color photometry setup. 488 and 561 nm laser (L) are combined through a dichroic mirror (DM). After combining the excitation lasers the beam is coupled into a fiber. The end of the fiber gets positioned over the cortical window to record Excitatory and Inhibitory cells. The emission gets projected through a tube lens (TL) onto the beam-splitting device (BSD), and recorded on a sCMOS chip (CAM).

Figure 2 Averaged functional connectivity matrices under (left panel) ISO anesthesia (lower left) and awake state (upper right), and (right panel) MED + ISO anesthesia (lower left) and awake state (upper right). Color bar represents the Pearson correlation coefficients.

Figure 3 Significant changes of correlation coefficients (A) under ISO anesthesia and (B) under MED + ISO anesthesia compared with awake state. Right upper triangle panel shows the significant changes (p < 0.05, network-based statistic). Left lower panel shows the t-values. Figures are reconstructed significant changes in functional connectivity. Color bar shows t-values.

Figure 2 Brain areas showing significant changes in FC values（p<0.05） (a) Schematic diagram of brain areas showing significant changes in the deepest and lightest level of anesthesia （b）Schematic diagram of brain areas showing significant changes in ANOVA at different levels of anesthesia

Figure 3 FC values that show different change patterns as anesthesia decreases

Figure 1. Images of functional brain networks by developmental stage The red points “node” represents anatomical regions, and the lines “edge” shows measurement of functional connectivity (FC) between pairs of nodes. The number of nodes and edges was getting larger with age. The scale bar was put on right side. Binary threshold was set 0.1 in all months of age.

Figure 2. FC matrix by developmental stage The brain was parcellated into 51 labels, and FC matrix was composed by the connection of each labels. The scale bar on the right side shows the strength of positive and negative connectivity. The upper and lower images represent averaged and coefficient of variation FC matrix for each. The development of FC and large variation were indicated as marmosets grow.

Fig.2 Group functional correlation maps. p<0.01, two-sample t-tests, uncorrected.

Fig.3 Two examples of the correlation between connectivity and behavior indices. (A,B) The upper matrix shows the functional connection which significantly correlated with either number of shocks or time to the first entrance in the probe test. p<0.01, uncorrected. The lower left shows the anatomical location of the connection. The lower right shows the scatter plot between connectivity change (Z-score) and behavior indices.

One-sample T-test statistic maps, derived from its occurrences in the combined (top), WT (middle), and HET (bottom) image-series, for two prototype CAPs with significantly reduced duration and/or occurrence in the HET group at the 10-month time point. At the 3 and 6-month time points, only one of these two patterns showed significantly reduced occurrence and duration in the HET group respectively. Left panel CAPs show simultaneous co-deactivation & co-activation of default mode-like and latero-cortical networks in mice while right panels show the reverse pattern.

Classification accuracy (blue, mean +/- SEM) using duration and occurrence (top panels), and BOLD signals of voxels with significant activations (bottom panels) of each CAP within a partition as a function of partitions of the combined image series at the 3-month time point (left panels), 6-month time point (middle panels) & at the 10-month time point (right panels). The grey curve shows the corresponding chance-level accuracy (mean +/- SEM) and the red asterisk indicates significantly higher mean accuracy than the chance level, after correcting for multiple comparisons.

Changes in Quasi-periodic patterns in 4M old TgF344-AD rats. A) One-sample T-test maps of representative QPPs (rQPPs) for wild-type (WT) and transgenic (Tg) rats. Colors represent T-values (FDR corrected, p<0.05). B) Two sample t-test, reveals spatial differences between WT and TG animals, mainly in the basal forebrain (BFB) and Cingulate cortex (Cg). The colorbar indicates differences in T-values (FDR corrected, p<0,05). C) Average bold signal during QPP1 in the Cg (red), Somatosensory cortex (SS)(blue) and BFB (green). Both graphs show the mean +/- SEM (****p<0.0001)

Figure 1: Representative slices for 6 RSNs in coronal (top), sagittal (mid) and axial (bottom) planes of an ICA trained permutation test result. Red color indicates voxels with increasing functional activities (D7>D1), while blue color indicates voxels with decreasing functional activities (D7<D1). Yellow represents the six corresponding RSN atlas.

Figure 3: Bar plots of the average of percentages of the significantly decreasing (blue, D7<D1) and significantly increasing (red, D7>D1) voxels in EX and DMN network and their selected subnetworks over 100 voxels for sDL and ICA results. EX4, EX6, DMN2, DMN3, DMN5 and DMN6 are not shown as these atlases did not contain enough voxels to reach significance.

Figure 1. Experimental setup for optogenetic stimulation and histological characterization of viral expression in OB excitatory neurons. (a) Illustration of the location of viral injection (left) the fiber implantation in (middle) T2-weighted anatomical image and the CamKIIα::ChR2 expression in excitatory neurons of OB. (b) Schematic illustrating the optogenetic stimulation paradigm: consisting of four 20s-on and 60s-off blocks (1Hz, 10% duty cycle; 5, 10, 20, 40Hz, 30% duty cycle; 40mW/mm²).

Figure 2. Robust brain-wide activations upon low frequency optogenetic stimulation of OB projection neurons. (a) Regions-of-interests (ROIs) definition based on atlas. (b) Averaged activations maps of optogenetic stimulation in OB at 1Hz and 5Hz (n=7; t>3.1, corresponding to p<0.001). (c) The respective BOLD signal profiles extracted from ROIs at 1Hz and 5Hz. Error bars indicate ± SEM.

Figure 2. Brain-wide activation at cortical and subcortical regions upon optogenetic stimulation of SVN excitatory neurons. (A) Illustration of atlas-based region of interest (ROI) definitions in the vestibular, midbrain, cerebellum, and cortical regions. (B) Averaged BOLD activation maps at 20-Hz optogenetic stimulation at SVN (n = 8; P < 0.001; asterisk indicates the stimulation site). (C) BOLD signal profiles extracted from ROIs defined in A (error bars indicate ± SEM).

Figure 1. (A) 3D illustration of the four vestibular subnuclei (top-left) and illustration of CfR2-transfected MVN and SVN excitatory neurons (top-right). (B) Illustration of the stimulation site in ChR2 SVN (bottom-left) and ChR2 MVN (bottom-middle) neurons during optogenetic fMRI experiments (bottom-right). T2-weighted anatomical MRI image showing the location of the implanted optical fiber (asterisk, stimulation site; right).

Figure 2. Brain-wide fMRI revealed activations at regions mediating circadian rhythms upon optogenetic stimulation of MVN excitatory neurons. (A) Illustration of Paxinos atlas-based ROI definitions in the vestibular and midbrain, thalamic, cortical, and hippocampal formation regions. (B) Averaged BOLD activation maps at 20Hz optogenetic stimulation (n=18; P<0.05; asterisk: stimulation site). Notably, these activations include SCN, OPT, ARC, hypothalamus and LC, regions that are known to drive and regulate the circadian clock.

Figure 1. Illustration of CaMKIIa::ChR2 viral expression in medial vestibular nucleus (MVN) excitatory neurons and optogenetic fMRI stimulation setup. (A) ChR2-mCherry expression in MVN. Lower-magnification (left) and higher-magnification (right). Overlay of images revealed colocalization of mCherry and CaMKII in the cell body of MVN neurons (indicated by white arrows). (B) Illustration of the stimulation site in ChR2 MVN excitatory neurons during optogenetic fMRI experiments (asterisk, stimulation site).

Figure 2. BOLD responses to NAc-DBS occurred separately between the NAc core and shell. Averaging 9 animals with NAc core stimulation (A) and NAc shell stimulation (B) respectively. All subjects were scanned with 200 μA DBS. (p < 0.05, voxel size > 40)

Figure 1. The Positioning Validation of electrode implantation and a stimulation paradigm. (A) Representative image from one subject for electrode positioning. (n = 9) (B) Stimulation paradigm. A single stimulation was applied to each channel during a single scan session. After the first 30 seconds of rest, stimulation was given 3 times for 10 seconds each, and 120 seconds of rest was given between stimulations. These 3 stimuli were given to channels 1, 8, and 16 respectively, and repeated 5 times per one animal.

Group averaged T value maps comparing hypercania to baseline (One sample t-tests) at different time points during Long Days (LD) and during Short Days (SD). Arrows indicate statistically significant changes within the hypothalamus.

Figure 5. Immunohistochemistry data. A. Example of analysis in the arcuate nucleus (ARC) of the sheep hypothalamus near the 3rd ventricle showing the selection of a voxel of interest. Vessels (red surfaces) and stained neural stem cells (NSC) (SPOTS) can be seen. B. This analysis allowed statistical evaluation of the median distance between blood vessels (BV) and Sox2 stained NSCs demonstrating larger median distances during short day periods (SP) than long day periods (LP) in two different regions of the hypothalamus of sheep.