Abstract #2886

Gibbs P, Beach T, Turnbull L, Pickles M

Textural analysis, and in particular the spatial gray-level dependence matrices method, is becoming increasingly utilised in MRI. However, there is little published data investigating either the optimal degree of gray-level reduction or the most appropriate pixel separation to utilise in co-occurrence matrices, and subsequent textural parameter calculation. This work aims to optimise these parameter choices in the application of textural analysis to breast MRI.