The Naked Scientists

The Naked Scientists Forum

Author Topic: Could FMRI, PET, CT, EEG, and MEG be more advanced, but it would take more Cost.  (Read 478 times)

Offline Nicholas Lee

  • Full Member
  • ***
  • Posts: 70
  • Thanked: 1 times
    • View Profile
Are these current neuroimaging technologies, at their most advanced level, or is the cost to make them more advanced just too much money, and there is not enough funding.
It would help a lot with neuro disease, and how the human memory is understood, like how memory in encoded, because it is not fully understood.
If these technologies were at the most advanced level.
I am grateful for your help, anything helps even a few words.
« Last Edit: 07/08/2016 05:27:06 by Nicholas Lee »


 

Offline alancalverd

  • Global Moderator
  • Neilep Level Member
  • *****
  • Posts: 4704
  • Thanked: 153 times
  • life is too short to drink instant coffee
    • View Profile
CT, MRI and DSA seem to be adequate for locating and to some extent characterising anatomical pathology, including some very subtle recent discoveries in the flow of cerebrospinal fluid. Incremental developments make these techniques easier to use but I don't see a need for or possibility of a quantum step in performance. To increase signal/noise ratio or spatial resolution you need to increase radiation dose or primary magnetic field, which introduces other problems (death or chemical shift artefacts, respectively). And there's not much point in resolving below the level at which we can reliably treat.

Functional imaging is limited not only by hardware but also by our limited understanding of what we are looking at. To a large extent FMRI and PET are mapping blood flow to active parts of the brain, and it's a bit like trying to find a lighted match in a well-lit room by mapping the oxygen flow towards it, or even the aircon that tries to counter the heat input. MEG does provide some fairly high resoluton images of electrical activity, but the simile here would be looking for a faint radio beacon in a thunderstorm - except that you still don't know whether what you have found is an airport, a seaport, or a police car!  EEG will identify gross malfunctions like stroke, epilepsy or death, but I can't see its inherent spatial resolution getting much better: God or evolution seems to have put an insulating  skull and a conductive scalp in the way of the electric field.  As of now, I think we can unequivocally detect and roughly locate the response to a given stimulus, but whilst functional imaging greatly helps clinical characterisation of disease, it's not a stand-alone diagnostic like CT, or ever likely to be.

Happy to be proved wrong!
« Last Edit: 06/08/2016 16:02:14 by alancalverd »
 
The following users thanked this post: Nicholas Lee

Offline evan_au

  • Neilep Level Member
  • ******
  • Posts: 4109
  • Thanked: 245 times
    • View Profile
More compute power and more memory never go astray when you are doing complex calculations on huge gobs of data.

The other major advance is in the area of DNA sequencing, which seems to be doubling in performance about every 9 months (last I heard). Again, lots of compute power, access to huge databases and efficient algorithms are just as important as improvements in the hardware which actually reads the DNA sequence.

In both cases, the main shortcoming is in automated understanding of what is being seen, so that the computer can identify patterns that jump out to the experienced eye.
 
The following users thanked this post: Nicholas Lee

Offline Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8655
  • Thanked: 42 times
    • View Profile
This thread seems to assume that progress is not being made.
Is there any reason to suppose that is the case?
 
The following users thanked this post: Nicholas Lee

The Naked Scientists Forum


 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums