Cancer research involves measuring tumors. This measurement is presently done using calipers to measure two perpendicular widths through the center and assuming that the tumor has a known ellipsoid shape.
This procedure is neither accurate, nor precise. In order to mitigate the lack of precision, labs often require a single person to do all the measurements for some experiment, reasoning that any idiosyncrasies in their method would remain consistent throughout.
Companies have introduced scanners that use known light patterns and computational retrofitting to the observed pattern to obtain accurate 3d profiles of the tumor
Unfortunately, these solutions are expensive (definitely >1000$ for the machines; we figured our solution was significantly cheaper than that lower bound, so didn't bother looking further.) Plus, they are quite complicated, and we wondered whether something simpler would work.
We've developed a simple solution that uses individual needles to take point measurements of the height of the tumor at multiple points in a grid, and then multiplies them by the area of each grid square, giving the tumor volume along with a maximum error that is well understood from the study of approximations of definite integrals.
Pictures of the 3d models, as well as the schematic for the circuit to poll the keys in the measurement needle are given below. This page will be updated with more detailed notes and videos soon!
