Figure:

Diagram demonstrates the differences between, A, traditional institutional research and, B, the model of research promoted by image-based machine learning competitions. While traditional model follows a serial sequence of events (hypothesis formulation, institutional review board approval, data acquisition, knowledge discovery, and manuscript submission), image competitions promote a parallel approach to knowledge discovery and dissemination in which multiple participants of varying disciplines find solutions to the same problem and openly share their discoveries on the web or manuscripts in a short time period.