Aside from the questionable nature of patents on what is essentially mathematics, this just means that either (1) you didn't care about execution speed, or (2) your algorithms work at a sufficiently high level that it doesn't matter what the algorithm is in as long as the primitives are fast (e.g. lots of matrix math).

If you want to create active contours, find subnetworks in sparsely-connected graphs where all nodes are connected to a high out-degree node and also at least one other subnetwork node, or even implement a bitset that keeps track of which bytes have been encountered, Python is a poor choice.