Moisture gradient vector calculation as a new method for evaluating NMR images of corn (Zea mays L.) kernels during drying

Magnetic resonance imaging (MRI) was used to study the moisture migration inside single corn kernel during artificial drying. MRI data were taken every half an hour during drying while the intact kernel was kept inside the magnet, The moisture distribution was not uniform inside the kernel during drying, with the fastest moisture loss being detected in the endosperm, whilst the scutellum maintained the water. Resultant gradient vectors were visualized, as a new evaluation method, from each pixel of the subtraction of segmented MRI images showing the movement of proton density inside the kernel, The data matrix containing the gradient vectors can be stored for latter use as input parameters and to cheek the mass transfer models. (C) 1999 Elsevier Science Inc.