Visualizing the detection area of a unilateral NMR sensor using deconvolution and back-projection.

Understanding the detection volume of a unilateral NMR sensor is crucial to interpret acquired data appropriately. Whereas this is easily done in the sensor's axial dimension by running a depth profile on a well-defined sample, the lateral dimension is commonly characterized with very small samples, where each position along a regular grid is scanned individually, typically resulting in measurement times of several days and a resolution that is limited to the dimensions of the sample. Here we apply two mathematical procedures known from image processing that employ samples larger than the pixel size to characterize the lateral detection area. One procedure uses deconvolution algorithms to account for blurring effects caused by a larger sample while the other utilizes back-projection of radial field profiles. Both approaches are demonstrated with a Profile NMR-MOUSE® (PM5). They yield field maps in good agreement with those acquired with pixel-size test samples but save about one order of magnitude in scanning time.