11 replies to this topic

[solbanger]

Theory time. I was reading this article about how John Hopkins researchers discovered a way to inject sugar into patients so that the MRI could detect the cancers that are feeding off of it.

 

http://www.medicalne...cles/262999.php

 

 

Part of the difficulty with diseases is determination. I was just thinking how come our MRI scanners currently can't get to that fine level of detail? Identifying each one of our cells including position, type, age, size and density. Every last one even in the eyes, teeth and hair. It could separate cells that should be in the body vs cells that shouldn't such as viruses, cancer, mold and bacteria. We already have the profiles of thousands of pathogens, can't we feed them to an AI? We have AI that can sort through hundreds of medical texts already.

 

With an intelligent AI our super MRI could use heuristic programming to sort and categorize through the trillions of cells it picks up through raw electron shifts.

 

What is the limitation? Computational time? Is it actual scan resolution? Are our computers not sophisticated enough to count the stars in the sky? Can someone illuminate on the progress of this technology?

[niner]

None of our imaging methods have the resolution needed to look at individual cells. 

[ceridwen]

Or I'd have a proper diagnosis by now very disappointing

[solbanger]

So can someone explain what is the technological hurdle?

 

Are there tissue types that simply do not shift their electrons very well? Or theoretically does the MRI release information on all of the electrons in the body and our sensors are simply too primitive to capture it all in fine detail? Or is much of the information scattered and the MRI only can grab what's available and not random, such as with moving particles like blood cells?

 

If it is a matter of all the information of a human body being released (and theoretically their clothes and the table they are laying on at the time) then wouldn't that be a good start or are our computers simply too slow and cruddy to process the tsunami of information? I feel that the MRI machine could achieve the goal of mapping every cell in our body with quick results if it were refined. Imagine photograph quality slides coming from a computer showing snippets of your body as if someone opened you up and snapped a pic with their iphone!

 

 

Edited by solbanger, 01 October 2015 - 11:34 PM.

[ceridwen]

This would be crucial in the future for rebuilding the brain

[niner]

So can someone explain what is the technological hurdle?

 

Are there tissue types that simply do not shift their electrons very well? Or theoretically does the MRI release information on all of the electrons in the body and our sensors are simply too primitive to capture it all in fine detail? Or is much of the information scattered and the MRI only can grab what's available and not random, such as with moving particles like blood cells?

 

If it is a matter of all the information of a human body being released (and theoretically their clothes and the table they are laying on at the time) then wouldn't that be a good start or are our computers simply too slow and cruddy to process the tsunami of information? I feel that the MRI machine could achieve the goal of mapping every cell in our body with quick results if it were refined. Imagine photograph quality slides coming from a computer showing snippets of your body as if someone opened you up and snapped a pic with their iphone!

 

The MRI looks at subtle differences in the chemical environments of hydrogen nuclei.  The problem is that there are a vast number of nuclei, but they mostly all look alike, and the MRI can't read an individual atom, but rather averages over huge numbers of them.  As a result, nearly all of the detailed information is simply lost in the averages.  The small fraction of information that gets through is useful for imaging, but it can't tell you much in detail.  These are fundamental problems based on the physics involved, so it's not like we can refine an MRI to the point of getting atomic resolution from an organism.

 

There's a second flaw in this scheme, and that's the idea that if we got atomic level detail, we'd be able to interpret it in ways that would be helpful.  We can do some of that, but we are missing some fundamental links between atomic structure and biological activity.  To look at an entire body at once in atomic detail, or even cellular detail would require more storage and compute power than we have in the entire world.

[ceridwen]

Would humans ever be able to do that?

[niner]

Would humans ever be able to do that?

 

I don't think we'll ever be able to "read" our bodies at an atomic level of detail.  If by "do that", you mean the simulation/prediction part, then hypothetically, yes, in some far distant future.  However, by the time we can do cellular level simulations of entire organisms, it will be somewhat irrelevant, as we will have already developed a profound understanding of biology using more efficient methods.  

 

It shouldn't be too terrifically long (a small handful of decades) before we can have real-time monitoring of all relevant substances in the blood stream.  This would be high-value information that would allow us to optimize our health and head off most dysfunction before it becomes a problem.

[Turnbuckle]

Ultimately, MRI or other scanner types will be able to do what you want. For instance, this $270 million scanner is capable of much better resolution than typical medical scanners--

 

Medical researchers expect unprecedented resolution from 11.75-Tesla imager

 

Standard hospital scanners have a spatial resolution of about 1 millimeter, covering about 10 000 neurons, and a time resolution of about a second. The INUMAC will be able to image an area of about 0.1 mm, or 1000 neurons, and see changes occurring as fast as one-tenth of a second, according to Pierre Védrine, director of the project at the French Alternative Energies and Atomic Energy Commission, in Paris. With this type of resolution, MRIs could detect early indications of brain diseases such as Alzheimer’s or Parkinson’s and perhaps measure the effects of any methods developed to treat those illnesses. It would also allow much more precise functional imaging of the brain at work than is currently available. “You cannot really discriminate today what is happening inside your brain at the level of a few hundred neurons,” Védrine says.

 

http://spectrum.ieee...mri-takes-shape

 

 

 

PET scans are potentially capable of detecting single cells of certain types (for example cancer cells). And this technique gets you down to sub-cellular levels (for very small samples)--

 

Subnanometre resolution in three-dimensional magnetic resonance imaging of individual dark spins

 

 

 

 

 

[ceridwen]

It's a shame it's so closely allied to nuclear weapons development

[platypus]

I guess better resolution is possible with higher magnetic fields...I've seen a 45 Tesla magnet and it was the size of a small house and still I don't think a human would fit in its core. The forces involved are enormous...

[ceridwen]

It would probably be a health risk for anyone to get in that!