{"id":2123,"date":"2014-06-11T16:43:36","date_gmt":"2014-06-11T06:43:36","guid":{"rendered":"http:\/\/blog.panicola.com\/?p=2123"},"modified":"2014-06-11T16:43:36","modified_gmt":"2014-06-11T06:43:36","slug":"wired-ai-telling-doctors-how-to-treat","status":"publish","type":"post","link":"https:\/\/blog.panicola.com\/?p=2123","title":{"rendered":"Wired: AI telling doctors how to treat&#8230;"},"content":{"rendered":"<p>&nbsp;<\/p>\n<p>&nbsp;<\/p>\n<p>http:\/\/www.wired.com\/2014\/06\/ai-healthcare\/<\/p>\n<h1 id=\"headline\" style=\"font-weight: inherit; color: #333333;\">Artificial Intelligence Is Now Telling Doctors How to Treat You<\/h1>\n<ul class=\"entry-header\" style=\"color: #333333;\">\n<li class=\"author\" style=\"font-weight: inherit; font-style: inherit;\">BY\u00a0DANIELA HERNANDEZ, KAISER HEALTH NEWS<\/li>\n<li class=\"entryDate\" style=\"font-weight: inherit; font-style: inherit;\"><time style=\"font-weight: inherit; font-style: inherit;\" datetime=\"2014-06-02T06:30:01+00:00\">06.02.14<\/time>\u00a0\u00a0|<\/li>\n<\/ul>\n<p><a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/wp-content\/uploads\/2014\/06\/EMA-Derm.jpg\"><img loading=\"lazy\" decoding=\"async\" class=\"size-660-single-full wp-image-978961\" style=\"font-weight: inherit; font-style: inherit;\" src=\"http:\/\/www.wired.com\/wp-content\/uploads\/2014\/06\/EMA-Derm-660x440.jpg\" alt=\"Image: Courtesy of Modernizing Medicine\" width=\"660\" height=\"440\" \/><\/a><\/p>\n<div class=\"entry\" style=\"color: #333333;\">\n<p style=\"font-weight: inherit; font-style: inherit;\">Long Island dermatologist Kavita Mariwalla knows how to treat acne, burns, and rashes. But when a patient came in with a potentially disfiguring case of bullous pemphigoid\u2013a rare skin condition that causes large, watery blisters\u2013she was stumped. The medication doctors usually prescribe for the autoimmune disorder wasn\u2019t available. So she logged in to Modernizing Medicine, a web-based repository of medical information and insights.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">Within seconds, she had the name of another drug that had worked in comparable cases. \u201cIt gives you access to data, and data is king,\u201d Mariwalla says of Modernizing Medicine. \u201cIt\u2019s been very helpful, especially in clinically challenging situations.\u201d<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">The system, one of a growing number of similar tools around the country, lets Mariwalla tap the collective knowledge gathered from roughly 3,700 providers and more than 14 million patient visits, as well as data on treatments other doctors have provided to patients with similar profiles. Using the same kind of artificial intelligence that underpins some of the web\u2019s largest sites, it instantly mines this data and spits out recommendations. It\u2019s a bit like Amazon.com recommending purchases based on its massive trove of data about what people have bought in the past.<\/p>\n<div class=\"pullquote-left\" style=\"font-weight: 600; font-style: inherit; color: #000000;\">Using the same kind of artificial intelligence that underpins some of the web\u2019s largest sites, it instantly mines this data and spits out recommendations.<\/div>\n<p style=\"font-weight: inherit; font-style: inherit;\">Tech titans like <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2014\/01\/google-buying-way-making-brain-irrelevant\/\" target=\"_blank\">Google<\/a>, Amazon, Microsoft, and Apple already have made huge investments in artificial intelligence to deliver tailored search results and build <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2013\/03\/future-of-siri\" target=\"_blank\">virtual personal assistants<\/a>. Now, that approach is starting to trickle down into health care, thanks in part to the push under the health reform law to leverage new technologies to improve outcomes and reduce costs\u2013and to the availability of cheaper and more powerful computers. In an effort to better treat their patients, doctors are now exploring the use of everything from IBM\u2019s Watson supercomputer, the machine that won at <em style=\"font-weight: inherit;\">Jeopardy<\/em>, to iPhone-like pop-up notifications that appear in your online medical records.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\"><a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2014\/01\/geoffrey-hinton-deep-learning\/\" target=\"_blank\">Artificial intelligence<\/a> is still in the very early stages of development\u2013in so many ways, it can\u2019t match our own intelligence\u2013and computers certainly can\u2019t replace doctors at the bedside. But today\u2019s machines are capable of crunching vast amounts of data and identifying patterns that humans can\u2019t. Artificial intelligence\u2013essentially the complex algorithms that analyze this data\u2013can be a tool to take full advantage of electronic medical records, transforming them from mere e-filing cabinets into full-fledged doctors\u2019 aides that can deliver clinically relevant, high-quality data in real time. \u201cElectronic health records [are] like large quarries where there\u2019s lots of gold, and we\u2019re just beginning to mine them,\u201d said Dr. Eric Horvitz, who is the managing director of Microsoft Research and specializes in applying artificial intelligence in health care settings.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">Increasingly, physician practices and hospitals around the country are using supercomputers and homegrown systems to identify patients who might be at risk for kidney failure, cardiac disease, or postoperative infections, and to prevent hospital re-admissions, another key focus of health reform. And they\u2019re starting to combine patients\u2019 individual health data\u2013including genetic information\u2013with the wealth of material available in public databases, textbooks, and journals to help come up with more personalized treatments.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">For now, the recommendations from Modernizing Medicine are largely based on what is most popular among fellow professionals\u2013say, how often doctors on the platform prescribe a given drug or order a particular lab test. But this month, the system will display data on patient outcomes that the company has collected from its subscribers over the past year. Doctors will also be able to double-check the information against the latest clinical research by querying <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2012\/10\/watson-for-medicine\/\" target=\"_blank\">Watson, IBM\u2019s artificially intelligent supercomputer<\/a>. \u201cWhat happens in the real world should be informed by what\u2019s happening in the medical journals,\u201d said Daniel Cane, CEO of Florida-based Modernizing Medicine. \u201cThat information needs to get to the provider at the point of care.\u201d<\/p>\n<h3 style=\"font-weight: inherit; font-style: inherit;\">\u2018Quick and Seamless\u2019<\/h3>\n<p style=\"font-weight: inherit; font-style: inherit;\">Using homegrown systems, doctors at Vanderbilt University Medical Center in Nashville and St. Jude\u2019s Medical Center in Memphis are getting pop-up notifications within individual patients\u2019 electronic medical records. The alerts tell them, for instance, when a drug might not work for a patient with certain genetic traits. It shows up in bright yellow at the top of a doctor\u2019s computer screen\u2013hard to miss. \u201cWith a single click, the doctor can prescribe another medication. It\u2019s a very quick and seamless process,\u201d says Vanderbilt\u2019s Dr. Joshua Denny, one of the researchers who developed the system there.<\/p>\n<div class=\"pullquote-right\" style=\"font-weight: 600; font-style: inherit; color: #000000;\">\u2018Computers are notoriously bad at understanding English. It\u2019s a slow haul, but I\u2019m still optimistic.\u2019<\/div>\n<p style=\"font-weight: inherit; font-style: inherit;\">Denny and others used e-medical records on 16,000 patients to help computers predict which patients were likely to need certain medications in the future. Take the anti-blood clot medication Plavix. Some people can\u2019t break it down. The Vanderbilt system warns doctors to give patients likely to need the medication a genetic test to see whether they can. If not, it gives physicians suggestions on alternative drugs.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">Doctors heed the computer\u2019s advice about two-thirds of the time, figuring in, for example, the risks associated with the alternative medication. \u201cThe algorithm is pretty good,\u201d says Denny, referring to its ability to predict who\u2019s going to need a certain drug. \u201cIt was smarter than my intuition.\u201d<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">So far, computers have gotten really good at parsing so-called structured data\u2014information that can easily fit in buckets, or categories. In health care, this data is often stored as billing codes or lab test values. But this data doesn\u2019t capture patients\u2019 full-range of symptoms or even their treatments. Images, radiology reports, and the notes doctors write about each patient can be more useful. That\u2019s unstructured data, and computers are less savvy at handling it because it requires making inferences and a certain <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2013\/10\/sentiment-analysis-deep-learning\/\" target=\"_blank\">understanding of context and intent<\/a>.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">That\u2019s the stuff humans are really good at doing\u2013and it\u2019s what scientists are trying to teach machines to do better. \u201cComputers are notoriously bad at understanding English,\u201d said Peter Szolovits, the director of MIT\u2019s Clinical Decision Making Group. \u201cIt\u2019s a slow haul, but I\u2019m still optimistic.\u201d<\/p>\n<h3 style=\"font-weight: inherit; font-style: inherit;\">The Challenge Ahead<\/h3>\n<p style=\"font-weight: inherit; font-style: inherit;\">Computers are getting better at reading unstructured information. Suppose a patient says he doesn\u2019t smoke. His doctor checks \u2018no\u2019 in a box\u2013structured data, easily captured by a machine. But then the doctor notes that the patient\u2019s teeth are discolored or that there are nicotine stains on his fingers\u2013a clue that the patient in fact does smoke. Soon a computer may be able to highlight such discrepancies, bringing to the fore information that otherwise might have been overlooked.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">In recent years, universities, tech companies, and venture capital firms have invested millions into making computers better at analyzing images and words. Companies are popping up to capitalize on findings in studies suggesting that artificial intelligence can be used to improve care. \u201cArtificial intelligence\u2013ultimately that\u2019s where the biggest quality improvements will be made,\u201d says Euan Thomson, a partner at venture capital firm Khosla Ventures.<\/p>\n<div class=\"pullquote-left\" style=\"font-weight: 600; font-style: inherit; color: #000000;\">The data is often stored in servers at individual clinics or hospitals, making it difficult to build a comprehensive reservoir of medical information.<\/div>\n<p style=\"font-weight: inherit; font-style: inherit;\">But many challenges remain, experts say. Among them is the tremendous expense and difficulty of gaining access to high-quality data and of developing smart models and training them to pick up patterns. Most electronic medical record-keeping systems aren\u2019t compatible with each other. The data is often stored in servers at individual clinics or hospitals, making it difficult to build a <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2014\/03\/internet-things-health\/\" target=\"_blank\">comprehensive reservoir of medical information<\/a>.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">Moreover, the systems often aren\u2019t hooked up to the internet and therefore can\u2019t be widely distributed or accessed like other information in the cloud. So, unlike the vast amount of data on Google and <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2013\/12\/facebook-yann-lecun-qa\/\" target=\"_blank\">Facebook<\/a>, the information can\u2019t be mined from anywhere by those interested in analyzing it. From the perspective of privacy advocates, this makes some good sense: A researcher\u2019s treasure trove is a hacker\u2019s playground. \u201cIt\u2019s not the greatest time to talk about\u201d health records on the web, given security scandals such as the <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2014\/05\/snowden-addresses-critics\/\" target=\"_blank\">Edward Snowden<\/a> leaks and the <a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.wired.com\/2014\/04\/heartbleedslesson\/\" target=\"_blank\">Heartbleed bug<\/a>, says Dr. Russ Altman, the director of Stanford University\u2019s biomedical informatics training program.<\/p>\n<h3 style=\"font-weight: inherit; font-style: inherit;\">Drawing the Line<\/h3>\n<p style=\"font-weight: inherit; font-style: inherit;\">Also standing in the way are concerns about how far computers should encroach on doctors\u2019 turf. As artificial intelligence systems get smarter, experts say, the line between making recommendations and making decisions could become more murky. That could cause regulators to view the systems as a medical devices, subject to the review of the U.S. Food and Drug Administration.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">Wary of the time and expense required for FDA approval, companies engineering the systems\u2013at least for now\u2013are careful not to describe them as diagnostic tools but rather as information banks. \u201cThe FDA would be down on them like a ton of bricks because then they would be claiming to practice medicine,\u201d says MIT\u2019s Szolovits.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">At the moment, he said, the technology isn\u2019t good enough to tell doctors with 100 percent certainty what the best course of treatment for a patient may be. Others agree. \u201cIt\u2019s going to be a long road,\u201d says Michael Matheny, a biostatistician at the Vanderbilt School of Medicine.<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\">Back at her clinic in Long Island, Dr. Mariwalla is thankful for the information that the artificial intelligence system can provide. For the patient with that blistering skin condition, she took the machine\u2019s suggestion for an alternative medication. The patient has recovered, Mariwalla says, but she\u2019s careful to add that she made the call herself\u2014based in part on her conversation with her patient. \u201cThat\u2019s where medical judgment comes in,\u201d she says. \u201cYou can\u2019t [just] rely on a system to tell you what to do.\u201d<\/p>\n<p style=\"font-weight: inherit; font-style: inherit;\"><em style=\"font-weight: inherit;\"><a style=\"font-weight: inherit; font-style: inherit; color: #00aeef;\" href=\"http:\/\/www.kaiserhealthnews.org\/\" target=\"_blank\">Kaiser Health News<\/a> is an editorially independent program of the Henry J. Kaiser Family Foundation, a nonprofit, nonpartisan health policy research and communication organization not affiliated with Kaiser Permanente.<\/em><\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>&nbsp; &nbsp; http:\/\/www.wired.com\/2014\/06\/ai-healthcare\/ Artificial Intelligence Is Now Telling Doctors How to Treat You BY\u00a0DANIELA HERNANDEZ, KAISER HEALTH NEWS 06.02.14\u00a0\u00a0| Long Island dermatologist Kavita Mariwalla knows how to treat acne, burns, and rashes. But when a patient came in with a potentially disfiguring case of bullous pemphigoid\u2013a rare skin condition that causes large, watery blisters\u2013she was &hellip; <a href=\"https:\/\/blog.panicola.com\/?p=2123\" class=\"more-link\">Continue reading <span class=\"screen-reader-text\">Wired: AI telling doctors how to treat&#8230;<\/span> <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5,9],"tags":[],"class_list":["post-2123","post","type-post","status-publish","format-standard","hentry","category-data-saving-lives","category-healthcare"],"_links":{"self":[{"href":"https:\/\/blog.panicola.com\/index.php?rest_route=\/wp\/v2\/posts\/2123","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.panicola.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.panicola.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.panicola.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.panicola.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2123"}],"version-history":[{"count":1,"href":"https:\/\/blog.panicola.com\/index.php?rest_route=\/wp\/v2\/posts\/2123\/revisions"}],"predecessor-version":[{"id":2124,"href":"https:\/\/blog.panicola.com\/index.php?rest_route=\/wp\/v2\/posts\/2123\/revisions\/2124"}],"wp:attachment":[{"href":"https:\/\/blog.panicola.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2123"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.panicola.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2123"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.panicola.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2123"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}