{"id":494,"date":"2013-02-13T02:20:14","date_gmt":"2013-02-13T02:20:14","guid":{"rendered":"https:\/\/skepticink.com\/gps\/?p=494"},"modified":"2013-02-22T20:07:24","modified_gmt":"2013-02-22T20:07:24","slug":"can-a-computer-out-perform-your-doctor","status":"publish","type":"post","link":"https:\/\/skepticink.com\/gps\/2013\/02\/13\/can-a-computer-out-perform-your-doctor\/","title":{"rendered":"Can a computer out perform your doctor?"},"content":{"rendered":"

In short, the answer is yes. Recent research <\/a>by computer scientists Kris Hauser<\/a> and Casey Bennett at Indiana University has shown that computer-based decision making can potentially save enormous amounts of money for health-care consumers when compared to typical human decisions.<\/p>\n

Artificial intelligence framework for simulating clinical decision-making: A Markov decision process approach.<\/a><\/p>\n

Basically, the researchers trained an AI that works through two<\/a> distinct processes<\/a> to learn how to make effective clinical decisions from real-life information (obtained from electronic medical records). They then compared the AI’s decisions to decisions made by human physicians. Turns out, not only did the AI outperform the human docs, it freaking smoked them<\/em>:<\/p>\n

\"AI<\/p>\n


\n<\/a>…simulation modeling that understands and predicts the outcomes of treatment could reduce health care costs by over 50 percent while also improving patient outcomes by nearly 50 percent.<\/p><\/blockquote>\n

That’s pretty impressive savings. Perhaps the most interesting aspect of this work, though, is the ability of this AI to perform these tasks even<\/em> with missing data. This seems to be due to not just a reliance on one method of data processing<\/a>, but the combination of the two used. As the researchers write<\/a>:<\/p>\n

Given careful design and problem formulation, an AI simulation framework can approximate optimal decisions even in complex and uncertain environments.<\/p><\/blockquote>\n

The results are not very surprising to me, as numerous studies in my field (clinical psychology) have long shown that using computer-based or algorithmic decision making can greatly improve diagnosis<\/a> or even predict who will not complete therapy<\/a>.<\/p>\n

A reliance on algorithms in combination with clinical experience and accumen could prove highly beneficial to patients. Maybe clinical diagnoses would be more accurate than the 50% rate<\/a> found in one large study, or a larger portion of healthcare providers would follow best-practice guidelines<\/a>. It could also address a major plaguing health-care systems worldwide – cost increases with no significant gain in quality.<\/p>\n

\"Bones\"Here’s hoping that this is another step towards building a real-life medical tricorder<\/a>, able to make accurate medical decisions with the wave of a hand!<\/p>\n","protected":false},"excerpt":{"rendered":"

In short, the answer is yes. Recent research by computer scientists Kris Hauser and Casey Bennett at Indiana University has…<\/p>\n","protected":false},"author":25,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[104,29,12,16],"tags":[105,106,20,102,30,18,25,7],"class_list":["post-494","post","type-post","status-publish","format-standard","hentry","category-medicine","category-mental-health","category-psychology","category-science-2","tag-algorithm","tag-computers","tag-critical-thinking","tag-health-2","tag-mental-health-2","tag-psychology-2","tag-public-policy","tag-science"],"_links":{"self":[{"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/posts\/494","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/users\/25"}],"replies":[{"embeddable":true,"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/comments?post=494"}],"version-history":[{"count":0,"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/posts\/494\/revisions"}],"wp:attachment":[{"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/media?parent=494"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/categories?post=494"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/skepticink.com\/gps\/wp-json\/wp\/v2\/tags?post=494"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}