How can the resolution of magnetic resonance imaging (MRI) be increased 1000x?
If you like this page kindly support. Just subscribe or like the Page!
It depends on whether you just want to only increase resolution (zoom in on a specific region) or increase resolution and keep the field of view the same (same image but at higher resolution).
• Zoomed in image: A zoomed in image is definitely possible. As I discuss in detail here, it is possible to image a single neuron with a standard scanner with standard pulse sequences by using a really, really small receiver coil.
This is unlikely a become a clinical technique any time soon. The coil itself is difficult to construct and won’t be found in a hospital or even a normal research lab. Also this is an image of a tissue slice, not of a living brain, although it is possible to image single living neurons in solution. MRI researchers are also working on an implantable MRI coil for in vivo studies, though an implanted MRI coil will be have to be limited to non-human use.
• Same image with higher resolution: This is a lot more difficult. My best guess is that you can get a 2–3 fold improvement in clinical MRI scans using techniques used in research MRI. The main obstacle in high resolution imaging is time. The resolution of an image is set by the number of phase encoding points. Each phase encoding point requires a separate acquisition, so increasing the resolution 1000 times will require 1 million (1000x1000) times longer acquisition time for a 2D scan or 1 billion times longer for a 3D scan. Another problem is the signal strength decreases as the resolution is increased because the volume giving off the signal becomes smaller. Reducing the volume 1000 times also decreases the signal by billion times in each individual scan for a total decrease in the signal to noise ratio of
SNR= ___________=31662 times.
• There are ways to increase the resolution of MRI but none of them gets anywhere near the 1000 mark. Echo planar imaging is a standard technique in clinical MRI that acquires multiple phase encoding points for each scan and can give a ~8 fold speed up without excessive artifacts. This translates to a 8–√=2.88=2.8increase in resolution. Spiral imaging takes this concept further by sampling all the points in a spiral, skipping the points that are not essential for imaging.
• Finally, both methods can be combined with Compressed sensing,which constructs the full image from a smaller number of randomly sampled phase encoding points, and super-resolution MRI, which combines several low resolution images into a single higher resolution image.
Compressed sensing can give a 3 fold speed up and superresolution gives an approximate 1.6 improvement is resolution. My rough estimate is that combining superresolution, compressed sensing, and spiral EPI you can get a ~2-3 fold improvement in resolution over a standard EPI sequence.
By: Jeffrey Brender