“Designing for visualisation systems in endovascular surgery” may be the hardest thing to say with a mouthful of Maltesers, but it also frames a presentation around how surgeons need very specific visual information during incredibly complex procedures.
Tim began the chat with this ‘minimal’ slide:
And if you’re like me you are probably thinking “eh?” or “cool battery indicator bro!”; I get that. But what is it?
Endovascular surgery – put simply – is an operation that provides support structures for collapsing arteries in patients. This support takes the form of ‘stents’ (short pieces of mesh) which surgeons manoeuvre into place using a particular set of tools.
During the operation certain visualisation methods assist the physician. Typically an X-ray machine will ‘scan’ a patient and a series of systems will plot a representation of the artery on screen in front of the operating team.
As Tim explains, the set-up looks nothing like this:
Well, it does kinda, but much less ‘romantic’.
More like this:
This kind of ethnographic research discovered operating theatres are complicated environments; but at the heart of it all is a surgeon and the surgeon needs access to certain things. The patient is one (hahaha!) but the visualisation of an artery via an ‘interface’ is another.
During various interviews Tim was able to describe exactly was going on with this interface:
It consists of a fluorescent overlay and 3D representation of the artery, along with some peripheral data.
Further research discovered that some data was more important than others:
Opportunity: redesign the interface so that overlay structures were more visible “I can see these things better”.
And this is where the project got even more interesting: it became apparent that the state of things on screen was super important.
During the operation the system uses data from the X-ray machine and the position of various vertebrae to determine how the computer generates overlays.
But! If the patient moves the position of the vertebrae is thrown off and the overlay becomes inaccurate meaning …. the machine has to recalibrate itself.
During recalibration the system works through four different states: ‘in position’; ‘get position’; ‘verify position is correct’; ‘confirm position is correct’. With the old system surgeons had no way of determining which state the system was in and this could cause problems.
Opportunity: redesign the interface so the system showed states better “I know what the machine is doing so I can do x,y,z while I wait”.
After many iterations a design was developed showing the transitions through the four states:
The holy grail of UX is to recommend something really simple …. like a single button perhaps. And I think this case study came pretty darn close, don’t ya think?
Thanks to Tim Caynes for all the slides.