Add patient size scaling

[Jonathan Cole]

Add an option for patient phantom scaling with size.

[Ed McDonagh]

Would that allow describing the fields/dose on a surface other than at the reference point?

[Jonathan Cole]

The current calculation on the flat phantom uses the table position to calculate skin dose. A mattress correction could be applied but hasn't at the moment.

Any surface can be used for the phantoms with some limitations. Rays within 90 degrees of the normal from the skin surface are assumed to be visible, outside 90 degrees are not. This works for a plane or a cylinder but would fail for complex concave surfaces.

I have a design for a 3D phantom in mind.

[Ed McDonagh]

I don't quite picture what you have at the moment then. Is the skin surface assumed to be the table - how does the flat phantom work?

I was thinking that the dose at the reference point was always 15 cm from the isocentre towards the source, therefore describing some sort of sphere around the isocentre as the C-arm rotates in two planes, with complication of the body moving around the isocentre in the z and x-y directions as the table moves.

Am I on the right track?

[Jonathan Cole]

The current phantom is a flat plane, so yes, effectively the table. Table height, lateral and longitudinal shift and x-ray tube rotation are all mapped onto the flat phantom. Ignores lateral or top-down beams.

Looks like this:

The current plan for the 3D model is to include lateral and top-down beams. I'm planning a cuboid with semi-cylinders, which looks like this:

[Ed McDonagh]

Cool. How exciting :)

It is difficult to not start thinking of 'stretch goals'! Like

• guessing a starting lateral offset from isocentre of the phantom based on the protocol
• (just for fun/education?) Sped up recreation of the procedure with the dose indication building up :)

[Jonathan Cole]

That second one was on my list.

I'll add it as an issue.

[Jonathan Cole]

Included in the 3d phantom class. Commit 35e431a