|This article is of interest to the following WikiProjects:|
|Ideal sources for Wikipedia's health content are defined in the guideline Wikipedia:Identifying reliable sources (medicine) and are typically review articles. Here are links to possibly useful sources of information about Magnetoencephalography.
- 1 Infinite solutions
- 2 MEG images produced
- 3 Infinite solutions
- 4 Electromagnetic source imagin
- 5 MEG machine types, manufacturers and history
- 6 Possibly free images
- 7 Health risks, current guidelines?
- 8 Current dipoles are not the standard physical dipoles
- 9 Undetermined and Overdetermined systems
- 10 Imaging Technique
- 11 Sensor types
- 12 Weasel words
- 13 Relative influence of magnetic field measurements on current density estimation
- 14 Shielding
- 15 Business view missing: manufacturers, service providers
Is wrong I don't have the time to edit it now but the VARETA algorithm developed by E.R. John was designed to allow for source localization, which is later z normed. —Preceding unsigned comment added by Androm (talk • contribs) 17:23, 6 May 2008 (UTC)
Although the grammar was wrong, the idea is correct. There is an infinite continuum of solutions. A weighted sum of two solutions (where the weights sum to 1) is a new solution. This means that there cannot be "several" solutions: there is one, or there is a continuous manifold. (I am an expert in the field.) Barak (talk) 20:13, 9 December 2011 (UTC)
MEG images produced
What type of images do these MEG's produce? Can anyone in the field post one?
MEG's don't produce images directly. They generate data points which can be plotted on a DICOM image such as an MRI. Here is an example from the 4D Neuroimaging's website (they manufacture these devices) http://4dneuroimaging.com/images/MEG_Work_4.jpg These images represent brain functions such as language centers or motor/sensory function areas in the cortex. It is also valuable for localizing epileptiform (i.e. seizure foci) activity. With this information, a neurosurgeon has a better view of which portion of cortex is functional (i.e. should be preserved) and which is diseased and should be resected (seizure focus, tumor, etc.). The difference in location may be millimeters.
Actually, some methods attempt try to produce the image of the current in different parts of the brain. One example is http://www.kolumbus.fi/kuutela/mce/mce.gif However, there is no general "format" for this kind of images, and every application uses their own way of visualizing the estimated currents.
KUutela 20:26, 7 February 2007 (UTC)
I've read about other inverse problems having an infinite number of solutions, but this is often extremely misleading, such as the failure to include in the model critical domain constraints. In this case, a domain constraint could be that the solution corresponds to a physical possible vascularization although in this case MEG doesn't seem to be much coupled to blood flow patterns.
Another case is that the "infinite solutions" correspond to a probability distribution produced by the same model which defines a finite entropy.
A final case is that the infinite solutions correspond to fussy details no one actually cares about, while features of primary concern are fixed uniquely.
What the lay person rarely understands when they encounter the phrase "infinite solutions" is that this generally translates to "unfortunately, we weren't able to come up with an equation that entirely spares us from the [socially contentious] application of common sense". MaxEnt 15:32, 6 January 2007 (UTC)
Unfortunately in the case of MEG/EEG, some of that space of potential solutions is of practical import. For instance, Irina Gorodnitsky showed some time ago that, given finite resolution etc, a more focal deeper source produces the same measurements as a less focal more shallow source. Barak (talk) 21:20, 18 November 2007 (UTC)
Electromagnetic source imagin
MEG machine types, manufacturers and history
I request a section on what companies have manufactured MEGs, characteristics of each machine (e.g., # sensors, sensor types, etc.), how many, and whether they are still alive.
I know 4D Neuroimaging recently went out of business.
I know Elekta's MEG division is still in business, but has had a complicated story of mergers and acquisitions.
Why is a picture of the door to the shielded room shown rather than a picture of the MEG detector helmut, etc. Weird. —Preceding unsigned comment added by Quadruplecheck (talk • contribs) 18:10, 7 May 2010 (UTC)
Possibly free images
Health risks, current guidelines?
I don't want to stir the pot, but is there anyone in the field who can comment on MEG's guidelines for use? (for example, no pregnant mothers/babies/people with heart conditions,etc.)? After reading the article I was curious about the safety of the procedure. —Preceding unsigned comment added by 22.214.171.124 (talk) 22:56, 22 January 2010 (UTC)
- I don't see why there would be safety issues with MEG -- it records magnetic fields but doesn't generate them. Are you perhaps confusing it with MRI, which involves very strong magnetic fields and has definite safety issues? Looie496 (talk) 23:14, 22 January 2010 (UTC)
Current dipoles are not the standard physical dipoles
A current dipole is something different as either an electric dipole (two charges of opposite sign near eachother), or a magnetic dipole (a circular currents). It is a sink and a source of current close to each other. It should therefore not refer to the "Dipole" page, which does not discuss it. Perhaps a new page needs to be written for it?
- It sounds like what I would recognise as a current source, between two spatially defined points. You seem to be familiar with the subject - would you add some detail to this article? (Be bold!) GyroMagician (talk) 11:18, 5 February 2010 (UTC)
Undetermined and Overdetermined systems
This article asserts that the MEG inverse problem can potentially be overdetermined and that this occurs when the number of measurements exceeds the number of unknown parameters. This is simply not true. Whether the problem is overdetermined also depends upon the model for the current density source. —Preceding unsigned comment added by Quadruplecheck (talk • contribs) 05:26, 19 February 2010 (UTC)
This article states that "Magnetoencephalography (MEG) is an imaging technique used to measure the magnetic fields ..." This sentence is nonsensical at best. But if we put aside the nonsensical nature of this sentence and focus on its principle assertion that MEG is an imaging technique there in lie claims that are far from justified. Simply put it is not an imaging technique. An imaging technique would have an associated resolution and a well defined space of functions that it is capable of imaging. It has neither. Furthermore it is well known that the measurement of magnetic fields external to a current density provides very little information about the current density. So as an imager of current density - which is what this article implies - the technique is a complete and well understood failure.—Preceding unsigned comment added by Quadruplecheck (talk • contribs) 05:09, 19 February 2010 (UTC)
- There's a tone of disparagement in that comment that worries me a bit, but in any case you should feel free to fix any errors you find in the article. As I understand it you're basically right, but it's also true that MEG does pick up magnetic fields generated by neural activity, and that point shouldn't be lost. Regards, Looie496 (talk) 21:00, 19 February 2010 (UTC)
Nobody denies that some neuronal activity produces measurable magnetic fields. If you are implying that it is the "silent currents" that are the sole source of difficulty when attempting to invert the magnetic field data to obtain the current density then you are equally wrong. Consider the inverse gravimetric problem: The gravitational field about a body is measured and from this measurement one tries to invert for the field source density (ie mass density). This problem is, as well, horribly ill-posed yet there are no silent mass densities at all. —Preceding unsigned comment added by Quadruplecheck (talk • contribs) 01:55, 5 May 2010 (UTC)
I've removed this section for now -- it seems irrelevant, seeing as none of these detector types are currently in use for MEG.
- There are at least three types of different measurement devices in use to detect magnetic fields.
- Magnetometer. Such a device detects magnetic fields directly using a simple induction loop.
- Axial gradiometer. Such a device consists of two magnetometers placed in series, (typically one above the other.) The result coming from the device is the difference in magnetic flux at those points in space (formally the "first spatial derivative.")
- Planar gradiometer. Such a device consists of two magnetometers placed next to each other. The result coming from the device is the "difference in the differences" in flux between the two points, thus the "second spatial derivative."
- Each sensor type responds differently to specific spatial signals.
This section needs expansion. You can help by adding to it. (March 2009)
I've removed these sentences from the inverse problem section -- please reinsert with citations if available.
- It is believed by some researchers in the field that more complex source models increase the quality of a solution. However this may decrease the robustness of the estimation and increase the effects of forward model errors. Many experiments use simple models, reducing possible sources of error and decreasing the computation time to find a solution.
I've tried to be bold -- hope the article is better as a result. soundray 18:02, 28 April 2010 (UTC)
Relative influence of magnetic field measurements on current density estimation
The following paragraph is repetitious and full of mistakes. It also appears to contain original thought and a half-hearted reference. If the paragraph is essential, please proofread and add proper citations.
- The MEG inverse problem is one that attempts to determine the electric current density within an volume of space from measured magnetic fields external to that volume. The extent too which this problem is ill-posed can not be overemphasized. If one's goal is to estimate the current density within the human brain with say a 5mm resolution then it is well established that the vast majority of the information needed to perform a unique inversion would come not from the magnetic field measurement but from any additional constraints applied to the problem. This conclusion is easily deduced from published works (see for example Journal Of Applied Physics. vol 94, no 8, 2003). For a 5mm resolution within a spherical volume of 30 cm it is safe to say that at least 90% of the information needed to determine the current density would come from the additional constraints needed to provide the unique solution leaving the magnetic field measurements of MEG with little influence upon the calculated current density.
soundray 10:48, 29 April 2010 (UTC)
Excuse me! Where are the mistakes? Very little information about an arbitrary current density (divergence free) can obtained from MEG. Just a bunch of snake oil.
You don't necessarily need to shield the earth's dc magnetic field. Many rooms use thick aluminum shielding. You do need to shield the earth's Schumann resonances. —Preceding unsigned comment added by 126.96.36.199 (talk) 22:15, 7 February 2011 (UTC)
Business view missing: manufacturers, service providers
I think there should be some info helping the reader to find manufacturers and service providers also. And a way for patients to find labs providing MEG scans. Some kind of estimate on how many systems there are in the world and that kind of stuff also. Regards Andyt80 (talk) 13:34, 26 March 2014 (UTC)