Western medicine has a somewhat morbid predilection with the technological. Some of that is attributable to the perverse financial incentives of the participating institutions; however, at a deep level we all want to believe that modern technological advancements are going to deliver us from disease and suffering.
A wildfire of highly complex machine learning investigations has swept through the medical literature, and one can hardly open a medical journal without being slammed in the face by outlandish predictive claims supported by equally preposterous AUCs (Area Under the Receiver Operating Characteristic Curve). The West has long worshiped at the altar of our big medical journals offering god-like veneration to its published work; however, the diversity of experience necessary to critically interpret machine learning work is mostly absent from the review process. We at point85 know this because we review this type of work for a variety of journals.
The modern editorial process is not yet equipped to contend with the complexities of machine learning in healthcare, and the only the only possible response we can afford is a more aggressive and open dialogue about the capabilities and limitations of the field.
Whether machine learning is going to be the Deus ex Machina that rescues us mortals from the morass of disjointed, expensive, imprecise care delivery in the West is yet to be seen. In the meantime point85 will shine some light into the black boxes that are making headlines, while simultaneously helping readers become more adept at navigating the topic.
Need my attention?—> firstname.lastname@example.org