Eeeeek! In this project I implemented a parallel processing system for analyzing cardiac motion in medical imaging sequences using optical flow analysis. What you see below are image sequences from cardiac MRI scans overlayed with vectors showing the motion of the heart contractions – the direction and magnitude. To improve the efficiency of this model, I implemented a parallel processing technique (both in Python and Go) which achieved significant performance improvements, with the best-case speedup being 2.31x on large datasets – yippee!

The vectors may look a little all over the place sometimes, but I am as only as good as the data provided >_<