I will occasionally use this page to explain and reflect on public presentation of my own academic works. Here is one about gradient coil efficiency, published last month in MRM (here).
One would think that there is nothing much left to publish on how to make static magnetic field coils, such as those for gradient or higher-order harmonic fields that are much used in MRI. This publication addresses something closely related -- how to characterize and compare different gradient coils -- and does show that there is still something to be learned about magnetostatics of MRI field coils.
One interesting finding is that the efficiency of a gradient coil, when normalized to account for different number of turns (that is, discounting efficiency gains obtained just by increasing number of turns for a given design), very well follows the power law: [efficiency] ∝ [diameter]-2.5.
The following figure demonstrates this. This is from an earlier version of the work, which was eventually turned into a figure in the paper. Each of the 24 markers corresponds to a distinct gradient coil in the public record, with all sorts of sizes and designs, and they line up surprisingly well along the predicted 2.5-th power-law dependence.
