Monitoring Acupuncture Effects on Human Brain by fMRI


Your institution must subscribe to JoVE's Biology section to access this content.

Fill out the form below to receive a free trial or learn more about access:


Enter your email below to get your free 10 minute trial to JoVE!

We use/store this info to ensure you have proper access and that your account is secure. We may use this info to send you notifications about your account, your institutional access, and/or other related products. To learn more about our GDPR policies click here.

If you want more info regarding data storage, please contact



FMRI and physiological monitoring is used to study the effects of Acupuncture on the central and peripheral nervous systems. Acupuncture mobilizes a limbic-paralimbic-neocortical network, with great overlap with the default mode network, to modulate neurological activity, possibly related to its autonomic effect in the peripheral nervous system.

Cite this Article

Copy Citation | Download Citations

Hui, K. K., Napadow, V., Liu, J., Li, M., Marina, O., Nixon, E. E., Claunch, J. D., LaCount, L., Sporko, T., Kwong, K. K. Monitoring Acupuncture Effects on Human Brain by fMRI. J. Vis. Exp. (38), e1190, doi:10.3791/1190 (2010).


Functional MRI is used to study the effects of acupuncture on the BOLD response and the functional connectivity of the human brain. Results demonstrate that acupuncture mobilizes a limbic-paralimbic-neocortical network and its anti-correlated sensorimotor/paralimbic network at multiple levels of the brain and that the hemodynamic response is influenced by the psychophysical response. Physiological monitoring may be performed to explore the peripheral response of the autonomic nerve function. This video describes the studies performed at LI4 (hegu), ST36 (zusanli) and LV3 (taichong), classical acupoints that are commonly used for modulatory and pain-reducing actions. Some issues that require attention in the applications of fMRI to acupuncture investigation are noted.


Part 1: Informed consent and subject preparation

Healthy adult volunteers are recruited for the study according to Institutional Review Board approved procedures. Informed consent by the subject is signed before beginning the study.

Part 2: Setting up physiological monitoring equipment

The AD Instruments PowerLab system allows for physiological data to be synchronized with the fMRI scan by using the fMRI trigger signal recorded on an input channel. This is a laptop based system with inputs for ECG, respiration, skin conductance response, pulse oximetry, and any other physiological signal that would need to be recorded.

  1. If a respiratory belt system is being used, the transducer must be plugged into a power outlet in the room adjacent to the scanner, with the tube leading to the belt fed via inlet into the scan room and a BNC connector leading to the PowerLab in the scan control room.
  2. If a trigger is being used to synchronize the physiological data to the fMRI data, then a BNC cable must also connect the MRI scanner trigger signal to the PowerLab in the scan control room.
  3. Gather monitoring devices. Four ECG electrodes will be needed, as well as an ECG cable connected to an MR-compatible In Vivo system rollcart.
  4. If skin conductance is being measured, plug the skin conductance connectors into the breakout box in the scan room, and have skin conductance gel ready in the scan room.

Part 3: Preparing subject for MR scanning

  1. It is important to rule out undue anxiety and fatigue on the day of the scan because basic state can affect the response of the brain to acupuncture stimulation. Suggest the subject eat a light meal prior to the procedure so that weakness and untoward reactions can be avoided.
  2. Instruct the subject to lie supine in the machine wearing ear plugs to suppress scanner noise and with the head immobilized by cushioned supports. Since the signal/noise ratio of acupuncture stimulation scans is low compared with common motor or visual tasks, the data tends to be more easily corrupted by head motion. A support is placed under the knees to keep the heels form the surface of the stretcher to reduce body motion. The head coil is fastened and cushions are adjusted for comfort. For an average scanning session of 2 hours, comfort is vital for uninterrupted scanning.
  3. If physiological monitoring is desired, subject should be prepared according to the previous section before being put into the scanner. It is desirable to monitor the wakefulness of the subject with eye-tracking during the scans if available.
  4. Attach 4 ECG electrodes to the subject's left upper chest (for female subjects, take privacy precautions such as lowering the blinds on the window to the bay in the scan control room and shutting the scanner door). Attach cable to the electrodes and to the In Vivo device. For safety this cable and all other cables leading to the subject should run directly along the scan table, without forming any loops. Turn on the In Vivo device and check for ECG signal quality. Adjust electrode positioning if necessary.
  5. Fill the wells of the SCR electrodes with conductive gel. This special SCR gel should be isotonic to the skin. Perform open circuit zero in Chart software. Attach SCR electrodes to 2 adjacent distal finger pads of the subject's hand by wrapping the velcro around as tightly as is comfortable.

Part 4: Functional MR imaging sequences tailored for imaging acupuncture stimulation.

  1. We use a 1.5 or 3.0 Tesla Siemens Trio MRI system (Siemens, Erlangen, Germany) equipped for echo planar imaging with a standard quadratic head coil.
  2. Our studies over the years consistently demonstrated that acupuncture had extensive action over functional systems at multiple levels of the brain. The structures located at the base of the brain such as the amygdala, hippocampus, ventral medial prefrontal cortex and subgenual areas are particularly susceptible to susceptibility artifacts. It is important to check each EPI functional scan to rule out signal loss in these regions as we found differences in loss of signals between two models of a Siemens scanner with similar field strength and with a similar MRI sequence for data collection.
  3. Thinner brain slices are preferable for reducing the susceptibility artifacts. Slices 3 mm thick with 0.6 mm or 0.75 mm gap in the sagittal or axial orientation are used to cover the entire brain including the brainstem and cerebellum depending on the memory capacity of the imaging system. Other orientations and slice thickness can be used for partial brain imaging. The special attention we pay to these issues contribute to the more consistent findings of acupuncture effects on the structures at the base of the brain.
  4. The functional scans are acquired with a T2*-weighted gradient echo sequence (TR/TE 4000/30 ms, FOV 200 mm, matrix 64x64, flip angle 90°, slice thickness 3.0 mm, gap 0.6 mm). We acquired a total of 150 time points in 10 min. Image collection is preceded by 4 dummy scans to allow for equilibration of the fMRI signal. The shorter TE reduces susceptibility artifacts. The relatively long TR permits whole brain coverage with high spatial resolution.

Part 5: The initial localizer and anatomic scans

  1. Before functional scans with acupuncture are run, a localizer scan, a high resolution anatomical scan, and the T1 weighted 3D MP-RAGE are acquired to assure correct positioning and anatomical soundness. General observation at this point can assure that unnecessary scans are not run. The localizer, a very short scan, assures that the whole brain is included in the scan and allows you to line it up with the AC-PC line. If there are any suspect anomalies, researchers will send images to a trained radiologist but should not make any claims about what is on the scan to subjects.
  2. The initial scans take about 20 minutes total. Brief viewing can tell the investigator if it will be useful to continue scanning. This is a good time to speak with the subject and make sure that they are still comfortable and aren't experiencing any adverse effects from being in the Magnet.

Part 6: Acupuncture with fMRI monitoring

Of the two forms of treatment (manual needle manipulation and electric stimulation), this video will be limited to manual acupuncture, which has a much longer history of clinical use. Acupuncture stimulation is administered in duplicate runs to classical acupoints LI4 on the hand, LV3 on the foot or ST36 on the leg, employing sterile, one-time use only acupuncture needles.

  1. Instruct subjects to relax, keep their eyes closed and refrain from motions during the scanning. To avoid noxious stimulation, we specifically instruct the subject to raise one finger if any of the acupuncture sensations approaches a score of 8, and raise two fingers in case of sharp pain of any degree.
  2. After inserting the needle, pretest the subject's sensitivity and tolerance to needle manipulation to help choose the depth of insertion and the force of manipulation for data collection.
  3. During the 10 min functional imaging, the needle is rotated for 2 min during M1 and M2, and left in place during the rest periods R1 (2min) ), R2 (3min), R3 (1 min). The acupuncturist makes sure not to approach or leave the subject directly before or after the stimulation period, as even that motion could affect the results. (Paradigm, Fig 1)
    Figure 1
    Figure 1. Acupuncture fMRI Paradigm.
  4. Upon the completion of acupuncture stimulation, the acupuncturist records the force with which the needle resists manipulation. At the end of each scan, another staff investigator asks the subject to grade the sensations experienced on a scale of 1-10. The list, aching, soreness, pressure, fullness, heaviness, numbness, tingling, dull pain, sharp pain, is based on the experience of the acupuncturist community and previous experimental data rather than pain questionnaires in the literature. The psychophysical response is categorized as a typical deqi group and another group of deqi mixed with sharp pain based on these records.
  5. Speak to the subjects between scans to make sure they feel comfortable and they stay awake. If any part of their body begins to cramp or feel uncomfortable, the research staff on site should take appropriate measures to relieve the discomfort. Movements generated by the subject would cause data corruption. Occasional subjects may fall asleep during scanning, which would affect the hemodynamic response.

Part 7: Data Processing

  1. Multiple levels of analysis can be run. A general linear model uses rest periods as a baseline to determine the % BOLD MR signal change at every voxel of the brain during the task. It also determines the statistical significance of these changes and removes BOLD data that do not reach 3 contiguous voxels at a specific p-value.
  2. Perform seed-based cross correlation analysis to show the functional connectivity between one neurological region and the rest of the brain. This is done by selecting a reference regions and comparing its time-course with that of every other voxel in the brain. Other analyses such as independent component analysis, fuzzy cluster analysis, structural equation modeling, Granger Causality can be applied.
  3. Physiological data can be analyzed by averaging response within windows related to acupuncture needle stimulation. For instance, ECG data can be annotated for QRS complex timing followed by analysis of heart rate and heart rate variability.

Part 8: Results

Figure 2
Figure 2. Hemodynamic response in acupuncture deqi vs. tactile stimulation at right LI4, ST36 and LV3 (52 runs/37 subjects). The clusters of deactivated structures at the MPFC (medial prefrontal cortex) (1), MPC (medial parietal cortex) (2), MTL (medial temporal lobe) (3), were more prominent in acupuncture (Fig A) than in tactile stimulation (Fig B). Extensive deactivation of the right BA22/21 (4) in acupuncture was not observed in tactile stimulation. The sensorimotor BA43 (5) and association cortex (6) showed more activation in tactile stimulation than in acupuncture (not shown). Interestingly, the right anterior insula,(7) was activated in acupuncture and not in tactile stimulation. (p<0.0001).

Figure 3
Figure 3. Hemodynamic response in deqi (52 runs/37 subjects) versus deqi mixed with sharp pain (52 runs/29 subjects) during acupuncture stimulation at right LI4, ST36 and LV3, The prominent deactivation of the MPFC (1), MPC (2) and MTL (3) seen with deqi absent pain (Fig A) was attenuated in the presence of pain (Fig B). With pain, activation of the sensorimotor and association cortices (4) became more prominent, and a subset of the limbic and paralimbic regions such as the midC/SMA (5), postC_BA23 (6), Amy (7), and cerebellar vermis (8) became activated. The right anterior insula (9) was markedly activated in deqi while smaller areas of anterior and posterior insula (10) were activated bilaterally in pain. P<0.0001.

Figure 4
Figure 4. Overlap of limbic-paralimbic-neocortical network (LPNN) in acupuncture with the default mode network (DMN); The BOLD response of the LPNN by GLM analysis (Fig.A) and by model-free Fuzzy Cluster analysis (Fig B) during acupuncture stimulation at LI4, ST36 and LV3 show virtually identical patterns with the DMN (Fig.C, adapted from Shulman et al, 1997). Clusters of regions with coherent signal decreases appear in the medial prefrontal cortex (1), postero-medial cortex (2) and temporal lobe (3). p<0.0001.

Subscription Required. Please recommend JoVE to your librarian.


The employment of proper techniques, the experience of the acupuncturist and a large sample size to permit separation from confounding effects of inadvertent noxious stimulation are important in the acquisition and interpretation of acupuncture fMRI data. Our studies of several classical acupoints indicate that acupuncture modulates the limbic network, an important intrinsic regulatory system of the human brain. The network shows significant overlap with the default mode of the resting brain. These associated autonomic nervous function changes can be demonstrated by physiological monitoring.

Subscription Required. Please recommend JoVE to your librarian.


The work was supported by the NIH/National Center for Complementary and Alternative Medicine (R21AT00978) (1-P01-002048-01) (K01-AT-002166-01), the National Center for Research Resources (P41RR14075), the Mental Illness and Neuroscience Discovery Institute (MIND) and the Brain Project Grant (NS 34189).


  1. Hui, K. K. S., Liu, J., Marina, O., Napadow, V., Haselgrove, C., Kwong, K. K., Kennedy, D. N., Makris, N. The integrated response of the human cerebro-cerebellar and limbic systems to acupuncture stimulation at ST 36 as evidenced by fMRI. Neuroimage. 27, (3), 479-496 (2005).
  2. Hui, K. K. S., Marina, O., Claunch, J. D., Nixon, E. E., Fang, J., Liu, J., Li, M., Napadow, V., Vangel, M., Makris, N., Chan, S. T., Kwong, K. K., Rosen, B. R. Acupuncture mobilizes the brain's default mode and its anti-correlated network in healthy subjects. Brain Res. 1287, 84-103 (2009).
  3. Napadow, V., Dhond, R. P., Purdon, P., Kettner, N., Makris, N., Kwong, K. K., Hui, K. K. Correlating acupuncture fMRI in the human brainstem with heart rate variability. Conf Proc IEEE Eng Med Biol Soc. 5, 4496-4499 (2005).
  4. Shulman, G. L., Corbetta, M., Buckner, R. L., Raichle, M. E., Fiez, J. A., Miezin, F. M., Petersen, S. E. Top-down modulation of early sensory cortex. Cereb. Cortex. 7, (3), 193-206 (1997).



  1. This study is missing a control: they should have also monitored the effects of needling points of the body that are not associated with traditional acupoints. It is interesting to note that, "during acupuncture stimulation at LI4, ST36 and LV3 show virtually identical patterns with the DMN." Would needling in non-acupoints produce the same effects?

    Posted by: Jacqueline G.
    April 9, 2010 - 11:14 AM
  2. The existence of needling points not associated with traditional acupoints, so called control points that have zero acupuncture effect is, I believe, a mythical , misplaced and unfortunately a firmly ingrained belief. The meridian system is a schematic which suggests which needling points have the strongest acupuncture effect. It is a leap to conclude that needling points outside the traditional acupoints would then have no acupuncture effect. Any analogy is in order. If an artery is cut, it will bleed profusely. It would be a logical leap to conclude that if tissue outside the artery is cut, it must not bleed.
    The important idea that came out of of Hui's study is that acupuncture manipulation engages an enormous brain network, not just localized brain responses. Needling "control" points should naturally also engage large areas of the brain network. Since there is only so much real estate in the brain matter, it won't be surprising that there will be overlaps in the two brain networks. There should be differences which have however yet to be published. However, it would be rash to label whatever differences that will show up as brain regions that have zero acupuncture effect. It should likely be labeled as regions of different acupuncture effect, just like the differences observed with the stimulation of LI4, ST36 and LV3 despite a huge overlap of responses at part of the DMN. It is not surprising for the large overlap in brain responses to manipulation at different acupoints. Often one single acupoint, say LI4, is supposed to be multi-tasked, able to relieve pain, stress and anxiety, sharing basically many of the acupuncture functions of other acupoints. Differences in brain responses to manipulation of different acupoints do exist in smaller portions of the brain network which are being investigated for the specificity of each acupoint. Would any "control" point be able to show some general and vague acupuncture effect even though it may not be able to deliver as strong an acupuncture effect if specific clinical outcome is desired, for example an improvement of heart function? While I do not believe that one can find a "control" point with no acupuncture function, I do agree that imaging the brain responses of "control" point will clarify the relationship with traditional acupoints and non-traditional points.

    Posted by: Kenneth K.
    April 11, 2010 - 1:02 PM
  3. "The existence of needling points not associated with traditional acupoints, so called control points that have zero acupuncture effect is, I believe, a mythical , misplaced and unfortunately a firmly ingrained belief."

    IE Just poke anywhere! You can't miss!

    "Needling "control" points should naturally also engage large areas of the brain network. Since there is only so much real estate in the brain matter, it won't be surprising that there will be overlaps in the two brain networks"


    "It is not surprising for the large overlap in brain responses to manipulation at different acupoints"

    So why bother to do the study then? If you knew that poking anywhere will light up similar or the same areas of the brain, why show that? What credence dŒs it lend the idea that poking one area will light up one region of the brain in a healing way, but poke another area and it will light up the same area, but not in a healing way? How can you draw a conclusion that they&#x²019;re different if they&#x²019;re indistinguishable by your own admission?

    Why not have controls that act in a very similar but non-acupuncture fashion? Stimulate with a fine plastic bristle or something that produces a similar physical sensation. If acupuncture is effective shouldn't it show a vast difference in brain activity, as touching with a plastic tine is vastly different to inserting a fine metal needle in acupuncture terms? One should be active, the other inactive.

    Totally omitting to carry out negative controls proves nothing further than confirming that your treatment is as effective as a placebo. Nothing more. It cannot show any greater effect as there is nothing to compare it to.

    Posted by: Anonymous
    April 20, 2010 - 1:19 AM
  4. "poking anywhere will light up similar or the same areas of the brain, why show that?"
    "if they&#x²019;re indistinguishable by your own admission?"
    There is a big difference between lighting up "similar" areas and lightening up the "same" areas. There was no admission of lighting up only the same areas.

    "Why not have controls that act in a very similar but non-acupuncture fashion? Stimulate with a fine plastic bristle?"
    The sensory control of touching with a fine plastic bristle was done and reported half a dozen times in the literature by MGH alone in the last 10 years. There was vast difference in brain activity.

    A reiteration of the very simple idea (to be verified) presented in the previous response: Traditional acupoints are chosen presumably due to their strong acupuncture effect while non-traditional points yield weaker effect (there is no need to postulate a null effect). If I were a clinician, it makes sense to always go for the stronger effect and not just to poke anywhere.

    Posted by: Kenneth K.
    April 20, 2010 - 3:34 AM
  5. The design of controlled experiments for acupuncture
    studies is fraught with difficulties, and few attempts
    have satisfied Western scientific standards [Richardson
    and Vincent, 1986; Vincent and Richardson,
    1986]. One reason is the difficulty of finding a true
    &#x²018;&#x²018;sham point&#x²019;&#x²019; to use in comparison with the acupuncture
    point. According to traditional Chinese acupuncture,
    focal stimulation of a needle at relatively specific
    points with the evocation of the deqi sensation is
    important in generating therapeutic acupuncture effect.
    The specific location of acupuncture points along
    each meridian is derived from thousands of years of
    clinical experience. It is known that needling at points
    other than those indicated on acupuncture charts also
    can have therapeutic effects. For instance, Ah Shi and
    trigger points are focal tender points that are used in
    therapeutic acupuncture to treat chronic pain. An ideal
    &#x²018;&#x²018;sham point&#x²019;&#x²019; would be an area of skin located some
    distance from any known acupuncture point or trigger
    point. However, some of the physiological effects of
    acupuncture have been observed when such supposed
    &#x²018;&#x²018;sham points&#x²019;&#x²019; were stimulated [Margolin et al., 1993b].
    This may be due to fact that the acupuncture &#x²018;&#x²018;point&#x²019;&#x²019; is
    actually an &#x²018;&#x²018;area&#x²019;&#x²019; overlying densely innervated muscle
    [Melzack, 1989].
    We chose to use a number of controls including
    imaging during &#x²018;&#x²018;needle at rest&#x²019;&#x²019; and matched scans
    during tactile stimulation of the overlying area. Our
    data showed that neither &#x²018;&#x²018;needle at rest&#x²019;&#x²019; nor tactile
    stimulation produced the marked signal decrease in
    deep structures associated with needle manipulation.
    As can be seen from the time courses shown in Figures
    ² and 4, insertion of the acupuncture needle caused at
    most transient signal changes (increases or decreases),
    never the sustained signal decreases observed during
    needle manipulation. The tactile stimulation paradigm
    used for the quantitative comparison was exactly
    matched to the acupuncture paradigm (² epochs of ²
    min stimulation separated by 4 min of rest and stimulus
    delivery 1²0 times per min) because the magnitude
    of the fMRI signal change is affected by these parameters.
    We tried another experimental design, needle
    manipulation within the subcutaneous tissue just beneath
    the corium (n 5 ²) (data not shown). This method
    did not yield significant signal decreases in the deep
    structures described for acupuncture at LI 4.
    Our results indicate that insertion of the needle into
    the muscle and its manipulation are necessary to
    produce the deqi sensation and the prominent fMRI
    signal changes in the deep structures of the human
    brain. Additional evidence in support of this is provided
    by the results of an earlier preliminary study
    conducted in our center. They reported similar signal
    decreases in several of the limbic structures described
    above during acupuncture needle manipulation at
    acupuncture point Stomach 36 (ST 36, Zusanli) on the
    leg, but not during the control paradigm, pricking the
    skin over the ST 36 acupuncture point [Wu et al., 1997].
    We did additional control scans (data not shown) in
    which several of the subjects were imaged for up to 8
    min during needle insertion without manipulation.
    There were no accompanying signal decreases even
    during this prolonged needle insertion.

    Posted by: Tara S.
    April 20, 2010 - 9:30 AM
  6. The design of an optimal control for
    acupuncture stimulation remains unresolved (Ernst and White,
    1997). It is known that acupuncture performed to sites that are not
    located on meridians can have varying degrees of physiological
    and clinical effects. Therefore, we chose to deliver superficial
    tactile stimulation to the skin surface as a non-invasive control at
    the acupoint rather than invasive needling at a non-acupoint. Lying
    supine in the magnet bore, the subjects could not see the devices
    being used to conduct the tests on their extremity. They were not
    aware which test was being performed, the sensory control or the
    real acupuncture. When control stimulation is given in an
    environment where the subject&#x²019;s view of the site of stimulation
    ubiquitous, the use of a nearby non-point (not on any
    meridian) for control cannot be viewed as inert or inactive
    (Dincer and Linde ²003, Complement Ther Med; Lund and
    Lundeberg, ²006; Kong et al., ²009). There is no consensus as to
    the proper selection of a non-point, for example, how far from
    the classical acupoint in use, the minimal distance from
    another meridian point in the neighborhood, and how
    different in histological components and segmental innervations.
    In our exploration for a non-point control in our early
    acupuncture fMRI studies, the response of the point chosen
    was very variable, ranging from minimal to pronounced
    response. Although successful use of non-points as control is
    purported by many authors, reports are emerging that fail to
    find significant differences between non-points and real
    acupoints (Fang et al., ²008). The common practice of placebo
    needles for sham acupuncture is also problematic (Tsukayama
    et al., ²006, White et al., ²003, ²008). We, therefore, have opted
    to deliver superficial tactile stimulation to the acupoint, not as
    an inert control but as a means to help understand how
    acupuncture compares with a conventional form of sensory
    stimulation (Hui et al., ²000, ²005; Napadow et al. ²005, ²007). A
    recent study suggests tactile stimulation to be a &#x²01C;better control
    paradigm to extract the acupuncture specific brain responses&#x²01D;,
    which is in support of our findings (Ho et al., ²008).

    Posted by: Tara S.
    April 20, 2010 - 9:34 AM

Post a Question / Comment / Request

You must be signed in to post a comment. Please or create an account.

Usage Statistics