Follow Slashdot stories on Twitter

 


Forgot your password?
Close
typodupeerror
×
Java Software Programming Linux Technology

Preventing Epidemics with STEM 114

Posted by Zonk from the stop-em-where-they-live dept.
Anonymous Coward writes "IBM has released a Linux based technology enabling spatiotemporal modeling of infectious agents across the United States, providing scientists and public health officials with a powerful tool for understanding, and potentially preventing, the spread of infectious diseases. The new STEM technology provides Geographic Information System (GIS) data for every county in the United States supplied by TIGER files."
This discussion has been archived. No new comments can be posted.

Preventing Epidemics with STEM More

Preventing Epidemics with STEM

Comments Filter:

  • Intelligence an asset (Score:4, Funny)

    by skomes ( 868255 ) on Sunday April 24, 2005 @03:57AM (#12327755)
    Anybody who understands 100% of the article should get an automatic 15 points.
    • From what I understand, it's a framework for building models of infectious diseases.

      It has some geographical data (US only), and uses this to plot or guess the spread of a disease, given the some parameters about the disease and the model.

      It's the type of thing we constantly see in movies and TV shows about people researching diseases and disease control.

    • Spatio (Spatial) = Location
      Temporal = Time

      IBM has released a Linux based technology enabling the visualization of infectious agents across the United States at any point in time, providing scientists and public health officials with a powerful tool for understanding, and potentially preventing, the spread of infectious diseases.

      Being able to model the data using different visualization/modelling techniques has the potential to allow officials to identify areas of interest. For more info look at ESRIs [esri.com]

  • Linux... (Score:4, Informative)

    by 0x461FAB0BD7D2 ( 812236 ) on Sunday April 24, 2005 @03:59AM (#12327758) Journal
    it is not only immune from viruses, it stops them too!

    Note: this technology is based on Java2. It should run on any Java-supported platform, although IBM only lists Win2000 and Linux as supported platforms.
    • Except they only provided instructions for windows users, as far as I can see.

      Linux:
      1) unzip archive
      2) export STEM_HOME=/location/of/stem
      3) Obviously the batch file isn't gunna work, but if you've exported STEM_HOME, this should work fine "java -Xmx768M -Duser.dir=$STEM_HOME -jar $STEM_HOME\bin\stem.jar"

      Just got it running in FC3, lickety split... and it already looks intimidating!

      • So I just ran the Flu1, and this is actually cool, despite being mostly over my head.

        Does someone know where we can find the properties of other contagions so we can punch those in and see the visuals?
      • That's because you are on a Linux machine. From my Windows box, I see the instructions for Windows.

    • Re:Linux... (Score:2, Funny)

      by aussie_a ( 778472 )

      it is not only immune from viruses, it stops them too!

      It's virii damnit! Not viruses.
    • Anyone have any ideas for how to run it on a Mac? I tried doing it how it would be run in Linux, and I tried launching the app with the Java Launcher, but it doesn't seem to want to work.

      Any thoughts?
      • Oops. This was the first release and we didn't have quick access to a MAC....and...well...we forgot. Ok, I forgot. Sorry. Next release. Basically, the system needs to find its "home" directory as the value of "user.dir"

        "java -Xmx768M -Duser.dir=$STEM_HOME -jar $STEM_HOME\bin\stem.jar"

  • Awesome, but there are some roadblocks (Score:5, Interesting)

    by Dancin_Santa ( 265275 ) <DancinSanta@gmail.com> on Sunday April 24, 2005 @04:02AM (#12327765) Journal
    This type of tracking system is sorely needed in order to prevent little outbreaks from turning into pandemics. Knowing where the hotspots are and using that information to take measures to prevent the spread of a pathogen is of increasing importance as we become more and more mobile as a species.

    However there isn't a way to track some problematic pathogens. For many diseases, there is a political aspect that prevents authorities from excercising their ability to contain the germ. HIV is the most obvious example of a virus that is so inherently related to a specific set of behaviors that it ought to be a simple matter of monitoring infected persons and preventing the disease from spreading. Unfortunately, the bearers of the virus claim that their right to these behaviors trumps the public health and safety risk posed by the virus.

    If there were an outbreak of Capt. Tripps, would the government have the political will to actually put all those infected under quarantine, dooming them and essentially stripping them of all human rights, in order to prevent the spread of the disease? They have shown in the past that they are not willing to contain infected groups. What is to say that they will be in the future?
    • The problem is that many diseases now are coming from Asia and Africa, more than the US. And this application seems to support the US as a model only.

      As for diseases like AIDS, many like Thabo Mbeki refuse to acknowledge the link between AIDS and HIV. And President Mbeki is seen by many as the voice of Africa. His views on AIDS are more disturbing than those who put politics over safety.

      • Actually, when I unpacked it, I noticed all the data was just xml. If you can compile the source data for other countries, the program should work with that.
      • Yeah, whatever. You can't trust Java for anything remotely numeric, it's broken, it's a fact. Anyone who's read professor Kahan's (Berkeley) How JAVA's Floating-Point Hurts Everyone Everywhere [berkeley.edu] papers, and followed the issue for a little time will run like the devil from Java...
        I mean, modelling disease involves differential equations. You have to be crazy to use Java. C is much more trustworthy in that respect.
        In fact, this knowledge should be widespread.
        It's sad to see so many people take for granted that

        • And why would you think that C is trustworthy? It uses the underlieing FPU, which means that you can end up with all sorts of definitions of a float. Yes there is a floating library IFF there is no FPU. Well this is 200x, not 198x. The only systems without FPU's is embedded, and few of them are doing computational type programs.

          Of course, with that said, I am a C/C++/Perl hacker. But there are advantages to Java such as consistency

          • So? Enlighten me: what does that say about FP computations in Java, since you rely on a virtual machine? Have you read the paper? I guess not, huh? Oh, by the way, it's by an expert from Berkeley in Numerical Computing. Also, let me remind you that a correct implementation of FP in Java was /pulled out/ of the JCP.
            Have you read the follow ups? Probably not.
            Consistency in Java's VM? You've got to be kidding. Some do tail-recursions, some don't.
            Java is not so well-designed. It's a fact. If it were not so, why
        • Anyone who's read professor Kahan's (Berkeley) How JAVA's Floating-Point Hurts Everyone Everywhere papers

          That paper is from 1998. Got any follow-ups?

      • Yes, this was my first thought too.

        Even if a disease starts in the US, it seems to me that people from the area where an outbreak starts will be travelling abroad long before the infection is detected.So, while it is cured in one place, it may develop and spread in parallel elsewhere.

        So then, the problem is that worldwide flights and other forms of human travel must be taken into account, not to mention trade of food, materials, animals, etc.

        This software is probably a great help, and I welcome it.

        I

      • The system supports arbitrary geographic areas, the current release includes models for the lower USA and Thailand. Check out the configuration files, there's one for Thailand.
    • "so inherently related to a specific set of behaviors" HIV is not currently restricted to any particular sub-section of sexual behaviour. The comment might, possibly, statistically, nearly, approach partial correctness in parts of the US, but not in the world. And ZA's governmental attitude to HIV and anti-HIV drugs has changed recently. (One of the drivers of it was rapacious capitalism and "intellectual property" pressures in the US)
      • "However there isn't a way to track some problematic pathogens. For many diseases, there is a political aspect that prevents authorities from excercising their ability to contain the germ. HIV is the most obvious example of a virus that is so inherently related to a specific set of behaviors that it ought to be a simple matter of monitoring infected persons and preventing the disease from spreading. Unfortunately, the bearers of the virus claim that their right to these behaviors trumps the public health an

    • Umm, there are many ways to get HIV: irresponsible orgies, making love to your committed gay partner of the last nineteen years, being born with it, providing emergency medical aid without proper precautions, or even having it cross a species divide from your pet monkey. The incubation period is long, so you may not recognise the people who are carriers.

      Now, sure... you could lock up all those people who you consider a threat. But then... what about smokers? Aren't they a threat to themselves, and peop

  • Interesting Project (Score:2, Insightful)

    by ianturton ( 655126 )
    I haven't had a chance to play with this yet, but it certainly seems interesting. This sort of tool is becoming ever more useful as a way of dealing with natural and man made epidemics. While of no real use during an epidemic, they do provide a useful tool to help emergency planners decide what strategies will be best when it does happen. It has to be better to carry out these experiments before hand rather than during an emergency.

    Ian

    • Re:Interesting Project (Score:1, Interesting)

      by Anonymous Coward
      I just tried registering with them to download the software, and got the following:
      This product is subject to strict US export control laws. Prior to providing access, we must validate whether you are eligible to receive it under an available US export authorization.
      Your request is being reviewed.
      Upon completion of this review, you will be contacted if we are able to give access. We apologize for any inconvenience.

      • If you're using firefox, go get the bugmenot plugin. Worked fine for me. Just right click > bugmenot > bang, you're ready to go.
    • I haven't had a chance to play with this yet

      And if you're not in the U.S., don't expect to get a chance to. Though the site at first glance seems to allow you to download it, it first requires you to create an IBM user account, which requires that you provide a fair amount of information about you and your company (guess they can't imagine, nor care, that private individuals might perchance be interested in it).

      Once you've done that, and you actually go to the download page for the product, it requires

      • Hmm I must have clicked through that bit with out reading it! I'm in the UK and downloaded it with no problems.

        I've now had a chance to play with it and it looks nice. Just how useful it is remains to be seen until I can find some time to try to put one of my models into it.

        Ian

  • Modelling != understanding (Score:3, Insightful)

    by EmbeddedJanitor ( 597831 ) on Sunday April 24, 2005 @04:06AM (#12327779)
    Of course I didn't RTFA, but this package is, it seems, a modelling tool. Models don't always help understanding because they're based on assumptions. If the assumptions don't fit reality then they break down. The bad thing is that often the people start believing the models more than reality and if the field evidence starts to disagree with the models (which look very scientific with graphs etc) people start to doubt the field data.

    While the models (assumptions) hold true, they can provide some nice "what if" input, but they can never replace field data.

    • It was once said, all models are wrong, some are just useful. ;)
    • "Of course I didn't RTFA ..."

      ... but then again, who the heck does? ;)
    • Of course, we have models inside our heads anyway. That's how we work out any problem. Computer modelling just makes complexity easier to cope with, and more accessible to people who don't grasp the entirety of the model. I agree that they can be dangerous. However, when done well, with a well understood problem -- I think pandemics are well understood, just hard to control --, they are enormously useful. Even if incomplete, they can be used well when the flaws are kept in mind. Just think weather ;)

  • Brief Tiger Explanation (Score:5, Interesting)

    by ShyGuy91284 ( 701108 ) on Sunday April 24, 2005 @04:07AM (#12327780)
    For those that don't know, Tiger is a geographic information systems package that allows the management of geograpical information that can be based on a variety of different statistical data. I haven't done too much with it, but when I was in early HS, my uncle (Who is an imaging science person, and did mapping related stuff at the time for surrounding counties) had me work for him one summer updating address ranges and directions that would be used by various county functions including emergency dispatch services. I remember I used Tiger GIS software. Interesting how much of that stuff works for geographical information management.

  • Zombie Infection (Score:2, Funny)

    by Anonymous Coward
    Can it stop Zombie Infection? [kalleboo.com]
  • Epidemics should not be prevented with the stem, they should be prevented from the ROOT! :)

  • Comment removed (Score:4, Funny)

    by account_deleted ( 4530225 ) on Sunday April 24, 2005 @04:10AM (#12327789)
    Comment removed based on user account deletion

  • other uses (Score:3, Insightful)

    by sfcat ( 872532 ) on Sunday April 24, 2005 @04:39AM (#12327854)
    Not that disease modeling isn't a good use of technology, but Tiger is a modeling tool that integrates geographic and temporal data (spacio-temporal is just a fancy word for that). But I think there are alot of other good uses for it. Some more popular and some much less. Let's see here:

    • Modeling historic data about
      • economic development
      • spread of religons
      • spread of political organizations
    • Modeling the spread of specific ideas
    • Modeling company sales data
    • Modeling battlefield deaths
    • Modeling a crime spree

    Basically any thing that spreads over time can be modeled and viewed this way. Kinda neat.

  • Change to red level alert! IBM is developing a tool for terrorist to know the range of their biological attacks!

    It could be an IBM-AlQaeda alliance or not, by this time...

  • Stem Cells (Score:4, Funny)

    by Soul-Burn666 ( 574119 ) on Sunday April 24, 2005 @07:31AM (#12328256) Journal
    So if this new "STEM" technology runs on IBM's CELL processors, will we get STEM CELLs?
  • I'm stunned. This is great! Very similar to my PhD work using GIS to help create predictive models to use agroecological zones to predict and track plant disease development across large areas in row crops. I'm using Kansas wheat production as my system currently, hopefully it will be something possible to deploy in any cropping system any where that you can get the necessary GIS layers and other data. I'm only in my second year of working on this, but things are starting to come together and more stuf
    • Have you been following the soybean rust deal from last years hurricane Ivan? It's a natural for this one.

      With that said, any sort of open human epidemic might be hard to model, given that there's a possibility of on-purpose spreading. That data might be pretty hard to input, just way too many variables.

      We'll probably get to see this though, bird flu in particular could easily take off this summer and spread even wider around the planet, now that they have found out some humans can be carriers and not sho
      • Soybean rust, uggh. Glad I don't work with soybeans anymore. ;) Yes, I've been following that.

        I was doing work with soybean in Nebraska a couple years ago. We were working on fungicide application techniques for the anticipated Soybean Rust event.

        I have the advantage with plants that they don't move like humans. Well, row crops don't at least, hort crops get shipped all over. That's how Phytophthora ramorum was spread recently.
        • I looked it up but now I remember reading about it, with theoak trees. bummer. I remember dutch elm disease when I lived in new england, within one decade all those old huge trees were mostly gone. Pity.

          hey, jeeps! way cool! We have an 80 cj7 here now (needs a carb that works, need it *badly*, currently have a carter bbd which sucks rubber donkey...whatevers, can't get it right no matter how many times I rebuild it, tweak it, curse at it, love on it, it just won't cooperate) and I've had a 59 window wagon
    • The system is designed from scratch to be extensible. Modelling crop disease should be a "straightforward" application. The hardest part right now would be getting the GIS data into a form the system recognizes. Something we're working on.
  • ... and anyone who has USED Tiger files know how terribly innacurate they are. Greeeeat. Glad to know we're using free, lowest-common-denominator geodata for our public safety.
    • The accuracy of the Tiger files is the least of the factors that affect the results of the simulation. When someone creates a model they make many assumptions about the parameters that affect disease propagation. In fact, the Tiger files are probably the most accurate part of the simulation.

  • of why IP laws should be returned to what our founders originally concieved.

    The TIGER files is produced by the gov. on our dollars and has help numerous projects. In addition, it has helped launch numerous companies that used these files for data.

    Likewise, the Weather data from the gov. has helped citizens all over the world, and created numerous companies including accuweather.

  • Autism is a mysterious epidemic that could provide a good test for this sort of tool.

    I've done some modeling of the autism epidemic at the state level and found that the most promising explanation is some interaction between recent immigrants from India and people of Finnish ancestry.

    When you look at the scatter plot [laboratory...states.com] it is quite graphic with a correlation of 60% at 49 degrees of freedom for a very high statistical significance (p Indeed, you can take hundreds of biologically relevant variables, includin

Slashdot Top Deals

You know you've landed gear-up when it takes full power to taxi.

Close