Commentary on Gastric Emtying Macro by Mike Avison

This is a relatively simple macro and you might question why it merited a place in the user group meeting. In all honesty it wouldn’t have if there was a surplus of talks offered.

However, since its here, let me talk you through a few of its design aspects.

 

Design Specification

 

Look and Feel Aspects

The macro runs like a very free set of tools. At first the operator can only do a few things such as click buttons to signal intentions … solid or liquid , ant only or ant-post views to be processed. However once ROIs have been drawn or loaded from file, the operator is able to move back and forth generating curves from the anterior view only or anterior and posterior (geometric mean). If no posterior view is available, the anterior may be processed on its own.

When ROI's are drawn there is an obligatory step to save them, here lies a known bug; Pixie reglosto should allow the operator to cancel an operation, but if the cancel button is pressed, the macro crashes. Operators are advised to store ROIs under a name such as "trash" if they will not be needed again. ROIs can be overwritten.

The cine with ROI is an ugly thing, there being no such command in Pixie, it is effected by loading all the frames into image buffers, then displaying them in quick(ish) succession with the ROI being displayed over each frame. The cine speed depends on hardware, its about right on an LX but rather tedious on an FX.

For reference purposes the ROI is drawn over each frame of a compressed dynamic view, 15 frames per image. If there are 60 frames in all, this shows the whole dynamic in 4 images. If more some will not be displayed, however graphs are still correctly created from a study with any number of frames.

Finally there are two kinds of curve display. The economy ones are made quickly and show the whole study in cell F3 and just the user-defined downward sloping section of the curve, with exponential fit in cell F4. The deluxe curve (in F3) is a composite of the whole curve with the exponential fit between user-defined limits superimposed. In both cases a written report is generated indicating the plateau duration as well as the t-half of the exponential fit.

 

Use in the presentation on day 2 (general session) of the UGM

This macro was very useful in estimating the "error" caused in calculating T50s from just the anterior image. Once ant and post ROI's were defined, it took only minimal processing to generate ant-only and ant-post results. So have a look at that presentation too maybe.

 

[I'll put some screen grabs on in due course to illustrate the content.]