Reproducibility Study Of FMRI Signal Change Evoked By Acupuncture Stimulation

Objective:The study was designed to test the reliability of fMRI(functional Magnetic Resonance Imaging) signal changes by using blood oxygenation level dependent (BOLD) of echo planar imaging (EPI) sequence evoked by acupuncture stimulation and to explore the functional localization and its distributing orderliness of the central nervous system (CNS) reaction on acupuncturing acupoints of Zusanli (ST36) of stomach meridian of foot-yangming (ST) and Taixi (KI3) of kidney meridian of foot-Shaoyin(KI) and an false acupoints near the two acupoints of the meridians which were mentioned above, respectively, finger tapping task(frequency 2HZ) as a control group.Material and method:Twelve right-handed volunteers without any symptom and sign of the central nervous system (male 7, female 5; age ranged 21 to 30 years old, average 23) who were studying in campus as students or postgraduates were included in our studies. All of them had no history of drug abusing and acupuncture within one week. Female volunteers were asked to avoid acupuncture test during menses. Twelve subjects participated in five identical scanning sessions consisting of four functional scans, one for each of the four conditions: electro acupuncture stimulation at ST 36, KI 3, non-acupoint (NP), and a control task of the finger tapping. The order of test was determined randomly.Images were acquired on a 1.5-T magnetic resonance scanner (Magnetom Sonata Maestro Class; Siemens Medical Solution; gradient strength, 40 mT/m/s). In each subject, after a sagittal scout sequence was performed, 35 slices axial images of the entire area of the whole brain paralleled to the AC-PC line were obtained. Imaging sequences included spin echo (SE) T₁WI, gradient field mapping, and EPI (BOLD) were performed. The parameters of MR imaging as following: slice thickness, 3 mm; slice gap, 0.3 mm; FOV, 200 mm. T₁WI: TR=600 ms, TE=15 ms; Matrix, 256×256; FA, 70°; Bandwidth, 130Hz/Px; Scan time, 4 minutes and 54 seconds. Gradient field map: TR = 483 ms, TE = 5.19 ms; NEX, 1; FA, 60°; Matrix, 64×100; Bandwidth, 260 Hz/Px; Scan time, 1 min and 4 seconds. T₂ WI of BOLD: TR = 3000 ms, TE = 50 ms; NEX, 1; FA, 90°; Matrix, 64×100; Bandwidth, 2604 Hz/Px, Scan time, 12 minutes and 21 seconds; Measurements, 245.For acupuncture, a Multi-purpose Health Device G6805-2 (made in Shanghai, China) was used. The parameters of it consisted of wave, continuous; frequency, 2 Hz; The stainless no-magnet acupuncture needles we used were brand of "Huatuo" (made in Suzhou,Φ0.35 mm×40 mm).The block design of the acupuncture was "rest-stimulation- rest". From the first image to the 85^th image was called No.1 rest period, the 86^th image to the 165^th image was stimulation period, and the 166^th to the 245^th was called the second rest period. The stimulation with electric current started at the 86^th measurement and stopped at the end of the 165^th measurement. Electric needles were pulled out at the beginning of the 166^th measurement. The blood pressure, frequency of breath, and pulse were measured before test for every volunteer. The block design of finger tapping task was the same as acupuncture.Data were analyzed by Statistical Parametric Mapping software (SPM99; Wellcome Department of Cognitive Neurology, London, UK) implemented in Matlab 6.0 (Mathworks Inc, Sherborn, MA). The datasets were normalized to a standard space (Montreal Neurologic Institute space, MNI) within SPM99. This space was based on a template T1-weighted dataset, a close analog of Talairach space developed by the MNI. The temporal datasets were smoothed by using an isotropic Gaussian filter kernel having a full-width at half maximum of 8×8×8 mm the normalized voxel size. Statistical parametric maps were generated by using the general linear model to characterize regionally specific effects in the imaging data. Terms in the model included the activity condition (acupuncture) and the globle mean value of each temporal dataset. A boxcar reference waveform convolved with a kernel that approximates the hemodynamic response curve was used to test specific hypotheses, resulting in a t value at each voxel.To explore the reproducibility of acupuncture of fMRI,βvalues of all the ROIs for each individual subject across each individual session were extracted to further evaluate the reliability across different individuals and experimental sessions using an intraclass correlation coefficient(ICC) analysis.ResultsOf the twelve volunteers who consented into the study, ten completed all five sessions. Two subjects withdrew from the study after Session 2(two female volunteers). Only the data from the subjects who completed all five sessions were analyzed.The average brain activations evoked by the right handed finger tapping task across all subjects and sessions are as following: significant BOLD signal increases were observed in left primary motor cortex (M1), primary somatosensory cortex (S1), thalamus and putamen, bilateral medial frontal cortex (supplementary motor area, SMA), insula/operculum, and cerebellum. Activations in the middle prefrontal gyrus were also observed.Bold signal increases were observed in bilateral postcentral gyris, paracentral lobule, Contralateral thalamus, and ipsilateral cerebellar cortex during acupuncture at ST36 and KI3. In addition, left prefrontal cortex and right thalamus were also observed BOLD signal increases during acupuncture at ST36. Left insula and left temporal cortex were observed BOLD signal increases when acupunctured at KI3. Deactivation patterns during acupuncture at two acupoints were observed in bilateral prefrontal orbital gyri, cingulate gyri and ipsilateral hippocampus complex, in addition ipsilateral cerebellar cortex were observed BOLD signal decreases. During acupuncture at nonacupoint, activation pattern were observed in Contralateral thalamus, insula, bilateral cingulate gyrus, deactivation pattern were observed in generallycerebral hemisphere, and could not be localized pricisely.ICC analysis revealed that fMRI signal changes evoked by electroacupuncture stimulation were significantly more variable than those from the control finger-tapping task (p<0.05, Kruskal-Wallis test) . The relatively large variability across different sessions within the same subject was also observed.Conclusions:1. The sensation of DE-QI and of BOLD signal changes has the similar pattern during acupuncture at ST36 and KI3, which is significant different with nonacupoint stimulation.2. Stabile brain activity model was observed in finger-tapping task between subjects and between sessions; acupuncture at ST36 and KI3 and nonacupoint revealed variable brain activity model, the variability between sessions was more than between subjects.3. Despite the variability across individual sessions, the activation patterns in the group average data from each experimental session are more similar to each other fMRI is a useful tool for characterizing average behavior.4. The reliability of fMRI signals changes evoked by electroacupuncture stimulation were significantly lower than those from the control finger-tapping task. To accurately capture the activation patterns evoked by acupuncture stimulation at a particular point for a specific subject multiple sessions should be performed.