Technical debt: probably the main roadblack for machine learning to medicine
andrewtmckenzie.comanother barrier is that clearcut value-generation or employment opportunities don't exist. you say that you'd like to see open source software- how is that profitable? by the way, check out the latest kaggle competition: "Transforming How We Diagnose Heart Disease"
Great point -- there's a huge disconnect between what is needed in terms of the data and algorithm development and the profit motive. This is why I think some portions of this, especially the data generation and curation portions, may need to be driven by non-profits and/or the government.