MRI Noise source(s) and suggestions

Hi All,

We are developing a near/in MRI bore device and we are having imaging artifacts in our 1.5 t/64 MHz MRI scans. We are comfortable with shielding devices for far field EMI and have done this successfully in the past. We are using similar techniques for this product and we are not seeing noise in the 64 MHz range using our off site test equipment. We are however seeing near field magnetic and electric noise near the pcb components. We know we can shield the electrical component but will not be able to shield the magnetic component without using ferrous metal. We are not sure the magnetic component is the cause of our as the noise appears at many frequencies but not near 64 MHz. My 2 thoughts are why does this small noise make artifacts as it dissipates an inch or so from the pcb and the frequencies are different. I am wondering if the distance may not matter. As the scanner moves and encodes voxels across a region, perhaps it encodes the small magnetic noise as it encodes the pcb area. And I am wondering if the magnetic frequency not being near 64 MHz matters as the noise would be heavily affected by the eddy currents and changing gradients so in my mind, who knows what this might be doing to the small magnetic noise and vise versa. My gut tells me any small amount of broad noise, possibly changing noise (think electronic component arcing) is easily picked up by the MRI and the like arcing noise it is aggressively distributed across the MRI image…Any thoughts or experiences appreciated!

Can you specifically evaluate the oscillating magnetic field noise?
These should not be a problem beyond the sensitive field of recever coil elments, but perhaps your active components are within the imaging Field of view or immediately adjacent to receiver coils
You say there is no component near the 1.5T resonant range, are their harmonics ? remembering you may want your design to accommodate 0.55T to 3T for largest application in the market m ,but at least 1.5T and 3 T
I am leading to the question, do you actually need Magnetic shielding?
Magnetic shielding in the bore will upset static field homogeneity, Maybe it could be compensated for by individual shimming each exam.
“As the scanner moves and encodes voxels across a region, perhaps it encodes the small magnetic noise as it encodes the pcb area.”
The scanner doesn’t move. but the MR signal has small phase and frequency variations encoded to represent spatial information, So if your device is creating noise coherent with the encoding modulations of the time varying MR signal you might face a problem.
Not clear what you mean by having the noise affected by gradient switching eddy currents. Would eddy currents in your device alter the noise?
Does anything in your device generate sparks? Does anything allow for arcing from gradient pulses?
Interesting. Good luck


Hi Greg,

thanks for your great response. A couple of points; “The scanner doesn’t move” - what I meant to say is that as the MRI phase/frequency encodes across a region using the gradients it could intercept/scan the pcb and possibly it would pick up the near field magnetic noise, which does not propagate very far, but if that noise was in the fov at the moment the MRI system might still pick up the small regional noise. However we have run some additional tests since my post and we are getting noise when the pcb is at the end of the bore, 2 feet away from the coil. So we believe we have a far field component though we are not able to see anything near 64 MHz (the center frequency of our 1.5 t) on our spectrum analyzer in our lab. We do have a clock frequency at 75 MHz but its harmonics are far from the magnets 64 MHz. We are going to do more testing in the next day or so but we are just wondering what fast changing gradients and hence fast changing eddy currents might affect the pcb noise in a way to momentarily alter some of the noise frequency so that the noise frequency ends up, for a moment, at 64 MHz. I guess I could see heavy and rapidly changing magnetic flux lines interacting with the pcb’s near field magnetic noise and maybe changing the near field magnetic noises frequency…but still we are from the coil and getting noise. So more testing to do, I guess.

Category icons made by Freepik from