By Tyler Shewbert
Magnetoencephalography (MEG) has been shown to be an effective method to study the effects of deep brain stimulation in patients with chronic pain, Parkinson’s disease (PD), and Essential tremor (ET). The advantages that MEG provides over other neural imaging methods, functional magnetic resonance imaging (fMRI) and positron emission tomography (PET), are that MEG has less intense magnetic fields than fMRI, so the DBS equipment is not harmed, and the ability for temporal resolution in the millisecond range which neither PET or MRI scans can provide [1, 2]. The main disadvantage of MEG is that is does not allow for accurate localization of deeper brain activity. MEG is the technology that is best suited for studying deep brain stimulation (DBS) due to its ability for accurate temporal and spatial resolution, and its lack of negative effects on the DBS device while it is functioning [1, 3, 4].
Magnetoencephalography (MEG) is a neural imaging technique first developed by David Cohen at MIT in 1968. MEG records the activity of the brain’s magnetic fields on the outside of the head [1, 4-6]. Cohen reported that these fields are on the order of 1 pico-tesla in strength . MEG works by using a superconducting quantum interface (SQUID) which converts sub-quanta magnetic field changes into voltage changes . The SQUID is connected to superconducting coils placed as close as possible to the head of a patient . Due to the weak magnetic fields caused by the brain’s currents, the potential for outside interference is high. The external interference is reduced by magnetically shielding the room and the MEG device using Mu-metal, Al and other materials with differing magnetic properties [5-7]. Also, a gradiometer is placed on the other side of the superconducting current coil to help reduce external magnetic noise by evening out the external signals over the two coils . Software is used to filter out signal noise by using reference sensors placed in positions where they can pick up mostly external noise and this can then be subtracted out of the harmonics of the superconducting coil data . The following image shows the basic setup of an MEG recording device:
Figure 1. The basic setup of an MEG showing the superconductor, gradiometer, input coil and SQUID. (from )
The major drawback is that MEG cannot accurately locate brain functioning within the cortex. Computational methods are used for localization deeper within the brain but for those methods to work accurately there needs to be further study of how the deeper regions of the brain function . This is because the two main methods of localization, dipole fitting and minimum based approaches for spatial reconstruction, both require assumptions of how the brain works to localize activity .
Deep brain stimulation has proven successful in treating patients who have otherwise not responded to other types of treatment for various neurological disorders . However, there is a lack of understanding why DBS is successful in treating these disorders . Brain activity is difficult to accurately record and the implantation of DBS devices makes this more difficult [1, 4, 8]. The use of fMRI imaging can cause overheating or movement of DBS electrodes, or the associated pulse generator which has been implanted within the patient due to the strong magnetic fields that the fMRI machine uses . PET scans have temporal resolution in the order of minutes which does not allow researchers to observe changes in brain activity that the DBS device is causing accurately. Therefore, research has been performed on the use of MEG as a method of imaging brain activity while the DBS implants are functioning. Results and Discussion
Several studies have been conducted on the effectiveness of using MEG imaging to study the causes of why DBS is successful at treating a range of neurological disorders including chronic pain, PD, and ET [1, 3, 4]. In each of these studies the conclusion was reached that MEG allowed the researchers a valuable way of studying the mechanisms behind the success of DBS treatments [1, 3, 4]. MEG allowed researchers to use DBS in both high and low frequency situations [1, 3, 4]. MEG also allowed researchers accuracy in the order of milliseconds for temporal resolution allowing researchers to see detailed neurological changes while the DBS devices were being turned on and off with spatial resolution of ~5 mm3 [1, 3, 4]. The results of three studies will be discussed here.
Since little was known about the neural mechanisms that alleviate pain in patients with chronic pain, a study was performed in 2006 to see whether MEG would be useful in imaging a patient’s brain while the DBS implant was operating and while it was off . The selected patient had phantom limb pain which was being treated by a low frequency (7 Hz) stimulation . The patient’s brain was initially recorded using MEG while the device was on for ten minutes and then switched off for ten minutes a total of four cycles. He reported the pain as increasing in each off cycle and the pain diminishing during on cycles . The researchers found that the periods of DBS did not affect the MEG imaging . They also compared the brain activity data during the periods when the DBS device was turned off to data from fMRI scans previously taken and these were similar, showing that MEG was accurate at studying the effects of DBS at low frequencies .
The same research team later tested their hypothesis that the success of MEG imaging was only possible due to the low, 7 Hz frequency that the patient’s DBS device had been using in the previous experiment [1, 4]. They proceeded to examine a patient who was using 7 Hz and 180 Hz DBS stimulation for treatment of cluster-headaches. The researchers assumed that the electromagnetic noise produced by the high frequency stimulation would possibly interfere with the MEG imaging . The researchers were able to accurately image the areas of the brain that had been reported in earlier studies using fMRI as activating when the DBS stimulation was on and off with MEG imaging . However, they found activity in the periaquaductal grey (PAG), which is deep within the cortex, and while consistent with fMRI studies of pain and pain relief, the researchers believed that the PAG measurements were not as locally accurate fMRI imaging due to the limitations of MEG localization within the deeper cortex . Even with that, the researchers concluded that MEG was a still a reliable method of imaging neurological impacts of DBS while using high frequency stimulation .
The third study was published in 2013. The researchers examined if MEG imaging would be useful in studying the motor tremors that Parkinson’s patients suffer from and how DBS devices reduce these tremors . Prior to 2013 researchers had found that MEG imaging would be successful in studying patients using high-frequency DBS to treat PD . The researchers were successfully able to use MEG imaging to investigate the motor tremors of the patients when the DBS device was on and off .
All three of these studies and several others not discussed here have reached the same conclusion: MEG imaging is an accurate way to study the functioning of the brain in while a DBS device is on . This is a powerful tool for researchers since it is not as electronically disruptive as fMRI scans and can be performed safely on patients while the device is on and allows for detailed temporal and spatial resolution of ~5 mm3. The use of MEG as a research tool for DBS is still in its early phase therefore further studies will be needed and improvements made to the methodology.
Outlook and relevance of work
MEG is essential to furthering the study of why deep brain stimulation works. Being able to temporally resolve brain functions on the scale of milliseconds will provide researchers insight into how the devices are working. The spatial resolution of 5 mm3 is like that of an fMRI scan. There is currently a lack of information about how DBS works and this complicates its use as a reliable treatment. The ability to study the brain’s response while DBS is occurring is the main advantage of MEG imaging and this will help expand knowledge of the device’s impact on neurological conditions.
MEG has the potential to be the best way to study DBS in the future but improvements are needed. The drawback of not being able to spatially localize the activity deep within the cortex can be improved as general knowledge of the deeper cortex is gathered from PET and fMRI scans is applied to the MEG localization algorithms. Broader study is needed. In each one of these studies only one patient was studied. They were performed as a proof-of-concept. To gain further knowledge on how to properly implement MEG imaging as a method of studying DBS, large studies with many participants will be needed. This will allow researchers to have a better foundation of what to look for and what errors are occurring in their studies.
The necessary improvements to MEG imaging for DBS studies will be made. The potential for helping patients whose only option is DBS treatment is too great. However, to improve those treatments doctors need a better understanding on how DBS is working in the brain. Improved MEG techniques will allow this to be accomplished.
 Kringelbach ML, Jenkinson N, Green AL, Owen SL, Hansen PC, Cornelissen PL, et al. Deep brain stimulation for chronic pain investigated with magnetoencephalography. Neuroreport. 2007;18:223-8.
 Ray NJ, Kringelbach ML, Jenkinson N, Owen SLF, Davies P, Wang S, et al. Using magnetoencephalography to investigate brain activity during high frequency deep brain stimulation in a cluster headache patient. Biomedical Imaging and Intervention Journal. 2007;3:e25.
 Bajwa J, Connolly A, Johnson M. Magnetoencephalography of Deep Brain Stimulation in a Patient with ET/PD Syndrome (P06.089). Neurology. 2013;80:P06.089.
 Ray N, Kringelbach ML, Jenkinson N, Owen S, Davies P, Wang S, et al. Using magnetoencephalography to investigate brain activity during high frequency deep brain stimulation in a cluster headache patient. Biomedical imaging and intervention journal. 2007;3.
 Barnes GH, A; Hirata, M. Magnetoencephalogram. Scholarpedia. 2010;5:3172.
 Cohen D. Magnetoencephalography: evidence of magnetic fields produced by alpha-rhythm currents. Science. 1968;161:784-6.
 Cohen D. Magnetoencephalography: detection of the brain’s electrical activity with a superconducting magnetometer. Science. 1972;175:664-6.
 Kringelbach ML, Jenkinson N, Owen SL, Aziz TZ. Translational principles of deep brain stimulation. Nature Reviews Neuroscience. 2007;8:623-35.