“Dark” Learning May Better Describe Deep Learning.

In AI-Artificial Intelligence, Brain Technology, Deep learning by Brainy Days Ahead

“Dark” describes the processes by which artificial intelligence (AI) system algorithms are learning. Dark to us humans, that is. Recently, scientists at Stanford University demonstrated that a “deep learning” algorithm can diagnose potentially cancerous skin lesions as well as a board-certified dermatologist. This cancer study added to a stream of research reports in which AI software aids, or competes with, doctors. Business entities are responding to advance the progress. Late in 2017, Alphabet’s life sciences division and Nikon agreed to develop algorithms to detect causes of blindness in diabetics. The problem emerging from these coalitions, however, is that researchers are not sure exactly which evidence is being used by the algorithms to reach their conclusions. In deep learning the algorithm finds the rules itself, but often without leaving an audit trail to explain its decisions. In the meantime, the FDA has, over the past 20 years, approved “a number of image analysis applications that rely on a variety of pattern recognition, machine learning, and computer vision techniques.” The agency is seeing more software powered by deep learning and is allowing applicants to keep the details of their algorithms confidential. For example, the FDA in January , 2017, cleared for sale software developed by Arterys, a privately held medical-imaging company based in San Francisco. Its algorithm, “DeepVentricle,” analyzes in less than 30 seconds the MRI images of the interior contours of the heart’s chambers and calculates the volume of blood a patient’s heart can hold and pump. The FDA required extensive testing to make sure the results from its algorithm were on par with those generated by physicians. Eventually, as with DNA testing, people are going to want to scan themselves to evaluate their own problems. Some non-AI cellphone apps, like Mole Mapper, already allow people to track suspicious moles and record any changes over time. But knowing that is not the same as knowing what to do about it. Doctors may not be out of date just yet.