Figure 1.
Details of the electrode set used in the current study.
(A) Photograph of the electrode set used in the current study. (B) Schematic diagram with details of the structure inside the electrode set. Two tungsten wires were welded to the copper tip of the connector, and extended approximately 5 mm through the PE tube. The connector and PE tube were fixed together with dental cement, and the extended tungsten wires were bent 90o around the wood stick and fixed using epoxy resin. The final length of the electrode wires, extended from the wood stick, was approximately 7 mm, and the distance between the 2 electrodes wires was approximately 300–400 µm. (C) The electrode was bent 90° to allow the placement of the MRI surface receiver coil over the rat head.
Figure 2.
(A) Protocol for MRI scanning. For dexmedetomidine (Dex) based fMRI, the animal was first anesthetized using 5% isoflurane (Iso), then received a bolus of 0.025 mg/kg Dex subcutaneously. Iso was stopped 15 min later, and a continuous subcutaneous infusion of Dex (0.05 mg/kg/h) was initiated 15 min later for maintenance throughout the MRI experiment. The first session of the fMRI test was performed 60 min after the first bolus of Dex application. Each session of the fMRI test was separated by an inter-session interval of at least 5 min. (B) Stimulation protocol for DBS-fMRI session. Monophasic constant electrical current pulses with 0.4 ms duration, and the intensity and frequency adjusted in variable values, ranging from 100 to 300 µA and 6 to 12 Hz, were given by the block design, including 5 DBS blocks (stimuli on) and 6 control blocks (stimuli off), each block lasted 20 s.
Figure 3.
Evaluation of image distortion caused by tungsten electrode.
(A) An example of the identification of the electrode target site by cresyl violet stained brain section. The lesion at the electrode tip (black arrow) was produced by a constant electric current of 30 µA for 30 s before euthanizing the animal. The electrode lesion was located within the forepaw region of the VPL thalamus. (B) The tungsten electrode caused limited distortion to the MR signal around the electrode in T2 anatomical images (upper panel) and in EPI images (lower panel). In each panel, 3 serial scans from caudal (left) to rostral (right) are shown. The white arrows indicate the electrode track. (C) SNR in different brain areas in control rats and rats implanted with a tungsten electrode. The ROIs were defined as shown in the left panel: VP thalamus = red, globus pallidus (GP) = blue, caudate putamen (CPu) = yellow and S1 = green. In all 4 brain areas, no significant difference between the tungsten electrode compared with the control was observed. (D) Representative DBS-fMRI results from 3 individuals. White arrows indicate the electrode tracks. Positive BOLD response can be clearly observed in the VP thalamus and ipsilateral S1.
Figure 4.
Amplitude and frequency dependency of BOLD response in dexmedetomidine-anesthetized rats.
Averaged time courses within the S1 of 6 animals are plotted for 3 stimulus currents in the left panel, and 3 frequencies in the right panel. Red bars under the time courses indicate the 20-s stimulation period. The insets show the response amplitude versus the stimulus current or frequency.
Figure 5.
Stability of DBS-induced BOLD response under isoflurane and dexmedetomidine anesthesia among a series of sessions within a one-day experiment.
Here shows the examples from one animal under isoflurane- and dexmedetomidine-based fMRI, tested on different days respectively. (A) The BOLD activation maps (P<.05, voxels >10) from a consequent of 5 repeated fMRI sessions (inter-scan interval: 10 min) with the same DBS parameter (0.4 ms, 200 µA, 9 Hz). When under dexmedetomidine anesthesia, a stable activation pattern was evident in S1 among sessions. Session 5 was eliminated because of an unexpected animal motion. When under isoflurane anesthesia, although a conserved positive BOLD response was observed in S1, the response pattern varied among sessions. In sessions 2, 3, and 5–7, different stimulation parameters were used, hence they are not shown in this figure. (B) The time courses within S1 for the 5 repeated scans were plotted together. The red trace is the average time course of all 5 scans, and the gray bars under the time courses indicate the 20-s stimulation period. Inset shows the S1-forelimb ROI used to obtain the signal time course. Variability was high under isoflurane anesthesia, compared with that under dexmedetomidine anesthesia.
Figure 6.
Reproducibility in repeated (separated by 5 days) fMRI tests of DBS-induced BOLD response under dexmedetomidine anesthesia.
(A) Original data of longitudinal BOLD responses from 3 representative experiments. Each rat was scanned in 2 repeated sets of sessions with a 1-week interval. Gray bars under the time courses indicate the 20-s stimulation period. The stimulation parameter used was 200 µA with 9 Hz. Similar dynamic BOLD responses in S1 were observed between the session in Week 1 (blue trace) and the session in Week 2 (red trace). The activation map shows the spatial distribution of the BOLD response, the left and right maps show the results from the Weeks 1 and 2, respectively. (B) The response amplitude (left panel) and activated voxel number (right panel) in Weeks 1 and 2 were plotted against each other. All the stimulation parameters from each animal (n = 5) were pooled. Five individual rats were labeled as following symbols: •▴♦–▪. Different stimulation parameters were labeled as different colors (red: 9 Hz/300 µA, yellow: 9 Hz/200 µA, green: 9 Hz/100 µA). The cortex response amplitudes and activated voxel number in the 2 weeks showed high consistency. In the pooled data of different stimulation parameters: ICC = 0.9169 (p<.0001) for activation amplitude; ICC = 0.9569 (p<.0001) for activation voxel number. In each stimulation parameters: ICC = 0.82 (p<.05) and 0.73(p<.05) when using the stimulation parameters of 200 µA and 300 µA (9 Hz) for activation amplitude, and ICC = 0.89 (p<.05) and 0.82 (p<.05) when using the stimulation parameters of 200 µA and 300 µA (9 Hz) for activation voxel number. (C) Reproducibility of statistical parametric maps in different stimulation parameters. Overlay ratios ranged from 0.50 to 0.88 (mean = 0.67, SD = 0.14) and 0.59 to 0.66 (mean = 0.61, SD = 0.03) when using the stimulation parameters of 9 Hz/200 µA, 9 Hz/300 µA respectively. Size ratios ranged from 0.76 to 0.99 (mean = 0.88, SD = 0.04) and 0.84 to 0.96 (mean = 0.91, SD = 0.02) when using the stimulation parameters of 9 Hz/200 µA, 9 Hz/300 µA respectively. (D) The S1 BOLD signal amplitudes and activated voxel numbers under different stimulation parameters. The BOLD response amplitudes and activated voxel numbers were significantly different between stimulation parameters, but not between week 1 and week 2. (*p<.05, **p<.01, ***p<.001).