Grouping multiple conditions

Hi guys - really tricky one here, and it's bugging me big time, so hoping you can help me out.

I have a dataset called 'PAH Activity' which lists all activity going through a business unit.  I need to split this data up and show on a matrix table under two main heading - 'Obstetric' and 'Non-Obstetric'.  Within each of these main groups are the individual areas, so for example Obstetric 1st Trimester; Obstetric 2nd Trimester etc etc.

For most of these areas, I can use a simple lookup table to apportion an individual exam with each of these areas.  The problem I'm having is that whilst ALL of the Obstetric areas use exam to lookup the values.  The Non-Obstetric areas are primarily looking up against a room but also one of them is an exam, and the other is a combination (sum) of 7 separate rooms, with a caveat of counting only a specific exam!!

I've tried to tabulate this below to show you what I mean.  I would ideally like to show all of this on a matrix table grouped under Obstetric or Non-Obsteric, but because some of the areas overlap using exam AND room, I'm not sure how to go about doing this.

I wonder if you're able to point me in the right direction, so I can apply a specific methodology to split this into the correct groups and respective areas within the groups:

The exam / room column is there to show you which area of the activity data I'd get the information from and you can see that some overlap, whilst there is one particular area (Routine Gynae Sonographers) which uses 7 different rooms but only one exam to form the count.

Really appreciate your suggestions guys.  Thanks a lot.