Interactive Boundary Detection for Automatic Definition of 2D Opacity Transfer Function

In computer assisted diagnostics nowadays, high-value 3-D visualization intake a supporting role to the traditional 2-D slice wise visualization. 3-D visualization may create intuitive visual appearances of the spatial relations of anatomical structures, based upon transfer functions mapping data values to visual parameters, e.g. color or opacity. Manual definition of these transfer functions however requires expert knowledge and can be tedious. In this paper an approach to automatizing 2-D opacity transfer function definition is presented. Upon few parameters characterizing the image volume and an user-depicted area of interest, the procedure detects organ surfaces automatically, upon which transfer functions may automatically be defined. Parameter setting still requires experience about the imaging properties of modalities, and improper setting can cause falsely detected organ surfaces. Procedure tests with CT and MRI image volumes show, that real time structure detection is even possible for noisy image volumes.