MRI Imaging Platform

Client work through my consultancy, and the most interesting thing that came through it: a platform for testing experimental algorithmic transformations on MRI images, with an OCaml back end and a Java front end.

Why two languages

The split is the point, and it was deliberate rather than accidental.

The transformations were the research — the part being changed, re-run, and compared against previous results. That work rewards a language with an honest type system, where a change that breaks an assumption fails at compile time rather than three hours into a batch. I had just spent a year writing OCaml daemons at CSUN and knew exactly what it was good for.

The front end was the part a researcher sat in front of. It needed to be conventional, unsurprising, and unremarkable — everything the research code was not.

The platform belonged to the client, and it went with them.