Can we create a world without disease?

<span property="schema:name">Can we create a world without disease?</span>
IMAGE CREDIT:  http://www.michaelnielsen.org/ddi/guest-post-judea-pearl-on-correlation-causation-and-the-psychology-of-simpsons-paradox/

Can we create a world without disease?

    • Author Name
      Andre Gress
    • Author Twitter Handle
      @Quantumrun

    Full story (ONLY use the 'Paste From Word' button to safely copy and paste text from a Word doc)

    Is it possible to have a disease free world? Disease is a word most (if not all) people feel uncomfortable hearing when either they or someone they know has one. Fortunately, Max Welling, a professor of machine learning at the University of Amsterdam and member of Canadian Institute of Advanced research, and his team of entrepreneurs have created a data analysis system for the diagnosis of diseases for patients.  Fun fact: He directs AMLAB (Amsterdam Machine Learning LAB) and co-directs QUVA Lab (Qualcomm-UvA Lab). Here we shall see how this wonderful man and his team of entrepreneurs (Cynthia Dwork, Geoffrey Hinton and Judea Pearl) made some incredible breakthroughs to rid the world of disease.

    Max Welling’s concerns

    Some of the facts that Welling points out during his TEDx talk does bring attention to the fact that there are times when a doctor might miss something during a diagnosis of a patient. For example, he says that “half of medical procedures have no solicit scientific evidence.” That diagnosis is done primarily through their own practice and knowledge acquired in school, whereas Max is saying that there should be some form of an analytical prognosis towards other possible diseases. He goes on to explain that some patients can be misdiagnosed and end up back in the hospital, in which he specifies that they are 8 times likelier to die. The most intriguing thing is this is an issue that has always existed. The reason is as simple as mistakes are bound to be made which unfortunately could cost someone or several people their lives. Not only that, as Welling says, there are 230 million medical procedures every year costing half a trillion dollars. Like any industry trying to provide a service to help others, it costs money; furthermore, that means hospitals and those in charge of funding medical centres need to listen to innovators trying to further the industry in a better direction. Nevertheless, being frugal is always beneficial.

    Preserving Privacy

    Welling stated that he and his team have made 3 breakthroughs. One of which is a computer that can preserve privacy within a hospital; furthermore, computers can also analyze a plethora of data to further improve diagnosis for patients whom are quite ill. This software is named Machine Learner. Essentially, the computer sends a query to the hospital database, which answers the query then the machine learner will change the answer by “adding some noise to it.” For further details please click here (Max Welling explains it more closely between minutes 5:20 – 6:06). In other words, as Max explains it, the computer wants to “better itself” via diagnosis and “build a better model of data”. All of this is thanks to Cynthia Dwork, whom is a distinguished scientist from Microsoft Research.  She focuses on preserving privacy based on a mathematical foundation. For more on her and what she has done, click here. In short, this first breakthrough not only shows that Max wants to be respectful of patients’ personal information but also wants to provide hospitals with a more solid foundation for diagnosis.

    Deep Learning

    The second breakthrough was brought to light by Geoffrey Hinton. Yann Lecun, Yoshua Bengio and Geoffrey have explained that: “Deep learning discovers intricate structure in large data sets by using the backpropagation algorithm to indicate how a machine should change its internal parameters that are used to compute the representation in each layer from the representation in the previous layer.” In layman terms, it helps a machine understand itself better through its complex layers via its deepest parameters (for further details please read the rest of the review the three gentlemen wrote).

    Causality vs Correlation

    The third and final breakthrough is more of a collaborative idea to further differentiate causality from correlation. Max feels that Judea Pearl’s tools can help distinguish these two concepts and get organized. Essentially Judea’s role is to help give more structure to data which can be done if patient files are digitally transferred into a database. Pearl’s work is quite complex so if you’d like to further understand what his “tools” are click here.

    Max’s Wish

    Welling summarized at the end of his TEDX Talk that he wants to preserve privacy through the machine learner. Secondly, to engage data minors and scientists in order to further improve diagnosis to save money and lives.  Lastly, he wants to revolutionize health care by better serving hospitals, doctors and patients through technology that can help shorten hospital visits and utilize money more efficiently. This is a beautiful vision on health care because not only does he want to be respectful of the medical industry, he also wants to help save lives while thinking of hospitals' and medical centres' budgets.

    Tags
    Category
    Topic field