The biomedical world is flooded in data. We have a lot of genomic information from mouse to human, precious health measurements from clinical tests, and a large amount of so-called real-world data from insurance companies and drugstores. Using powerful computers, scientists have carefully researched it with some fine results, but it has become clear that we can learn much more with an assist from artificial intelligence. Over the next decade deep-learning neural networks will likely transform how we look for patterns in data and how research is conducted and applied to human health. This special report explores the future of this new transformation.
Right now the biggest assumptions are being placed in the field of drug discovery, and for good reason. The average cost of bringing a new drug to market nearly doubled between 2003 and 2013 to $2.6 billion, and because nine out of ten fail in the final two periods of clinical tests, most of the money goes to waste. Every large drug company is working with at least one AI-focused start-up to see if it can raise the return on investment.
Machine-learning algorithms (算法) can get through millions of chemical compounds, narrowing the choices for a particular drug target. Perhaps more exciting, AI systems—free by leading theories and biases—can identify entirely new targets by spotting tiny differences at the level of cells, genes or proteins between a healthy brain and one marked by Parkinson's—differences that might avoid or even puzzle a human scientist.
That same sharp-eyed ability is also being used to explain medical scans. Some systems can already discover early signs of cancer that might be missed by a radiologist or see things that are simply beyond human capacity—such as evaluating cardiovascular (心血管) risk from a retinal scan. The Food and Drug Administration is approving imaging algorithms at a rapid click. Other AI applications lie a bit further down the road.
Will the inefficiencies of today's electronic health records (EHRs) be solved by smart systems that prevent prescribing mistakes and provide early warnings of disease? Some of the world's biggest tech giants are working on it.
Despite fears that machines will replace humans, most experts believe artificial and human intelligence will work cooperatively. The bigger concern is a shortage of people with both biomedical knowledge and algorithm proficiency. If this human problem can be solved, the key to creating successful AI applications may depend on the quality and quantity of what we provide them with. "We rely on three things," says the CEO of one deep-learning start-up. "Data, data and more data."
CP: Central Point P:Point Sp: Sub-point (次要点) C:Conclusion
B . 
C . 
D . 
