Bench Press

With smartphones becoming more sophisticated and more popular, its only natural that there are a growing number of attempts to use them as a platform for scientific inquiry (pocket ultrasound, microscopy, and astronomy for example). This is especially useful in developing countries, where a relative lack of high-end computers and fixed broadband access make smartphones a very suitable alternative to the more expensive, bulkier solutions that are used in the developed world.

It should come as little surprise, then, that doctors in India are helping to pioneer a new “telemedicine” tool using the camera and processing capabilities of Apple’s popular iPhone to do remote diagnosis of Retinopathy of Prematurity (RoP), a condition which is more likely to afflict infants born underweight. While curable, RoP needs to be treated within days of detecting it as to prevent permanent damage to a child’s eyes, something which the iPhone’s camera, mobile broadband, and robustness of software and security platform allows pediatric eye surgeons to diagnose from remote locations, hundreds or even thousands of miles away.

But, the potential of smartphones to function as a tool for tele-medicine can probably go far beyond this. At least, that’s what i2i TeleSolutions, an Indian-based startup, is betting on. They provided part of the software solution for the RoP diagnosis tool, and are aiming to provide software and services to enable further telemedicine technology – mainly:

• Security – It is important that sensitive medical information is transmitted securely in a way such that only the appropriate medical professionals see the information.
• Data compression – As fast as 3G and the new LTE networks are (and will be), network coverage and data transfer rates will continue to be a limiting factor on the adoption of telemedicine. As such, a true telemedicine solution will require lossless compression techniques.
• IT support – Medical organizations are not especially well-suited for building sophisticated IT capabilities, nor do medical professionals necessarily have the time to learn an arcane user interface. For that reason, telemedicine solutions should aim to provide web-based access methods (in addition to any non-web based methods they may choose to push) to access and react to data.

i2i provides further details on the scope of their platform on their web page, but I think it represents a strong start for a solution. Going forward, I’d like to see them (and any competitors that emerge) provide support for:

• Additional types of data – i2i’s focus seems to be primarily on images, but the full range of capabilities on smartphones is massive – GPS, accelerometer, magnetometer, and even microscopy and other medical attachments – and I would hate to think that tele-medicine would be limited only to its imaging capability
• Deployment on more phones – The iPhone is unique in the maturity of the platform, but it would be nice to see similar applications on other operating systems like Android, Symbian, and Windows Mobile.
• Interactivity – The i2i platform appears to be very unidirectional: (1) take a picture, (2) send it to a remote surgeon. I think the true promise of telemedicine is something which allows for a greater level of flexibility and interactivity on both ends (to refine the view, or make a suggestion on some other place to scan, etc).
• Ability to tack on analytics – There is a significant amount of medical data that needs to be analyzed/processed before it can be acted upon. Building some sort of open protocol or extendability (a la Firefox or Salesforce or LinkedIn/Facebook model) would do a great deal towards enhancing the potential of a telemedicine platform

Anyone else have any other ideas?

In a world with flashy distractions like YouTube and Modern Warfare 2, how the heck do you get students to be interested in monocyte recruitment?

One idea that the Federation of American Scientists is proof-of-concept-ing is the use of video games as tools for science education. To that end, the FAS developed, in conjunction with game studio Escape Hatch Entertainment, a game called Immune Attack (trailer below):

The premise of the game is pretty creative. A patient who suffers from a non-functioning immune system needs the player’s help to train her immune cells on how to fight off a bacterial infection. More detail can be found in the lesson plan on the Immune Attack website, but the game itself covers multiple phases showing:

• how leukocytes move from bloodstream to infection site
• how leukocytes are recruited by chemical signals
• how the immune system can recognize pathogen-associated molecular patterns
• phagocytosis (how white blood cells devour pathogens that they find)
• how white blood cells can recruit additional immune cells with chemokines
• how natural killers and MHC molecules can identify cells infected by viruses

While the game’s concept is original, the production value of the game is not quite up to a full-fledged professional studio. Although, to be fair, for only a ~500MB download and from an effort that wasn’t backed by a major game company, the quality was fairly impressive. The problem, though, is that the game mechanics are oriented around maneuvering about the 3D world to train the patient’s immune system how to respond to infection. The game is thus very dependent on the quality of the controls and the graphics. As I was playing on a Thinkpad T400 using a Trackpoint, it was actually fairly difficult at times to do the maneuvers necessary to move on to the next level.

The interface was also somewhat klunky – being similar enough to a standard first-person shooter controls but with enough variations to make the controls a little awkward (the need to hold down the right mouse button while steering with the mouse and the inability of the keyboard to change the pitch of motion were annoying). The software also didn’t feel complete bug-free. Just to see what would happen, I deliberately failed a mission requiring me to identify and destroy 5 infected cells before a viral infection destroyed 5 healthy cells. When I re-started the mission, the count of destroyed healthy cells began at 5 – is it any wonder that I failed the mission, again?

With all that said, I do believe that this was a very impressive effort that just needs a little polishing. The music and graphics were a little hokey, and the lesson plan materials need to be fleshed out a bit better, but the game mechanics were designed very well to ingrain visually and physically how monocyte transmigration worked, how white blood cells are recruited, and how basic viral and bacterial pathogens spread infection. While I wouldn’t say I’m yet fully convinced that this approach will work, I am optimistic that this is a good method to help scientists convey very complicated phenomena to students.

We’ve written a couple of times about the asteroid Apophis which, while unlikely to hit Earth, will make a very near pass of 18,300 miles above the planet’s surface in 2029. NASA’s Jet Propulsion Laboratory just released an excellent animation of just how close Apophis will be when it passes by.

Wow for some reason 18,300 miles doesn’t seem quite as far anymore.
(Video Credit – Wired Science)

Image depicting ant movements. Credit: University of Granada.

Its impressive how much humans can learn from biomimicry. Soldiers, for example, may soon owe their lives to the same pesky ants living in your own backyard. Researchers from the University of Grenada(UGR), under Antonio Miguel Mora García, Professor Juan Julián Merelo Guervós, and Professor Pedro Ángel Castillo Valdivieso, have taken inspiration from how ants find trajectories from their colonies to their food sources in order to develop a simulator that can devise the “safest” trajectory in a battlefield between any two points, given the necessary parameters.

Dubbed the “ant colony optimization (ACO),” this algorithm has already allowed Antonio to employ it to the videogame, Panzer General, with promising results. Currently, the University of Grenada has received participation from the Ministry of Defense to devise new strategies for them if its success continues.

The scientists of the UGR have developed a mini-simulator in order to define the settings (battlefields), locate the unit and their enemies, execute the algorithms and see the results. In addition, the software designed by them offers a few tools useful to analyze both the initial map and the results.

To prepare this system, Mora García started from the battlefields present in the videogame Panzer General, defining later the necessary properties and restrictions to make them faithful to reality.

While the ACO has immediate benefits to saving the lives of our soldiers, I’m also excited for its applications outside of the military. By extrapolating its uses, we could potentially use the ACO to optimize shipping orders, create shortest routes, plan airplane seating, etc, all by harnessing the creativity and intelligence of Mother Nature. So the next time you see an army of ants devouring your picnic basket, take some time to marvel at the beautiful tapestry of Nature before squashing all of them with your feet.

(Image Credit)

If you’re like me, and you’ve spent endless hours programming in front of your compute , you’d agree with me that sometimes it’s not the funnest thing to do. While the finished product might be really cool, getting there is oftentimes tedious, frustrating, and hair-splitting. What usually causes these problems is that coders get bogged down with the details due to the fact that certain blocks of code require unavoidably intricate and detailed logic. However, with the new EU-funded research project, ReDSeeDS, Michal Smialek and his team of researchers hope to lighten the burden for all coders and make coding enjoyable again.

What Smialek and his team discovered was that since most coders start from scratch when beginning a project, these programmers often need to code entire programs from the ground up despite the fact that other people have probably previously coded programs which accomplished similar tasks. As a result, programmers often re-write code simply because working on a project usually means starting from a blank screen. In order to avoid this, Smialek aims to create a repository which will house previously written code stored with a list of the program’s aims and requirements. Thus, when a user searches this database with a query of requirements, the database will return previously written code that is expected to produce similar outputs. These functionally equivalent snippets of code are called “artefacts,” and Smialek’s database is essentially a library of artefacts.

In a project, you may produce several artefacts which are design blueprints and then an artefact which is the code that tells the system how to work. The final program is also an artefact which is served by the other artefacts – that is the design and the code.

While most programs obviously will not completely overlap in terms of requirements, the fact that most programmers will not need to “re-invent the wheel” will greatly enhance productivity and allow them to not get so bogged down with the details. Not only does this allow for a more rapid rate of software releases, it vastly decreases the amount of time needed to fix problems within the code simply because the code within the database will be (we’d like to hope) error-free. Smialek notes that:

What it will do as a commercial product is to reduce considerably the amount of work required to develop a new software application, and that means the ability to develop more and larger systems using the same human resources.

(Image Credit)

Dust can be a pain if you’re an astronomer. In the same way that clouds obscure a view of the night-sky, interstellar dust can distort an astronomer’s view (even if through the Hubble telescope) of interesting astronomical phenomena. This problem is compounded when you consider that stars and planets tend to form in dense interstellar dust clouds.

The distortions caused by spacedust are caused by radiative transfer – a process of light absorption and scattering which also explains why the sky is blue and why sunsets/sunrises look red. Astronomers have built highly sophisticated models to understand radiative transfer across a wide range of different dust backgrounds. These models have enabled researchers to build very cool simulations, such as this one of two galaxies colliding:

Interestingly, greater sophistication in our understanding of spacedust and a greater desire for precision and resolution in the modeling has meant that more and more of the computational processing of these radiative transfer models has been spent on calculating dust grain temperatures rather than the math behind the actual radiation transfer (a product of the fact that you need to calculate across many points in the “dust cloud”, across many types/sizes of dust particles, and because you need to iterate many times to find an equilibrium) – something on the order of 10¹¹-10¹² exponentials per simulation!

That traditional processors are not well suited for calculating exponentials and that there were simply so many calculations which needed to be done in parallel convinced researchers to turn to NVIDIA’s CUDA as a potential solution. As we’ve noted before with using raytracing as a means to accelerate radiotherapy dosage calculations, NVIDIA’s CUDA is a standard programming toolset which lets programmers more easily use the power of (NVIDIA) graphics cards for calculations. Because the calculations needed to do high-performance graphics for a game of Modern Warfare 2 are similar to the calculations that supercomputers crunch through, NVIDIA’s CUDA has been demonstrated to be able to accelerate calculation speed by orders of magnitude!

In the case of dust grain temperature calculation, the results were equally impressive. Not only were the researchers able to accelerate dust grain calculation using a NVIDIA Tesla C1060 (with 4 GB of memory) over an 8-core Intel Xeon E5420 processor (with 32 GB of RAM) alone by a factor of 55, they were able to do this despite:

• the fact that 17% of processing time on the GPU solution was dedicated to data transfer (something the CPU-only solution has to worry about less)
• the maximum theoretical capacity of the GPU was only 6 times greater than that of the CPU, highlighting a big difference between the CUDA philosophy (crank up performance) and the CPU compiler philosophy (abstract but flexible)

Amazingly, the researchers found that even if the CPU were to run an interpolation scheme (requires less processing power, but introduces a little more error and makes it harder to do more sophisticated calculations vs. the equilibrium calculations done here), the GPU solution is still faster by a factor of 16 times!

So: spacedust – 0. GPU – 1. Now let’s see if they can tackle the flexible dust temperature problem…

Paper: “Accelerating Dust Temperature Calculations with Graphics Processing Units”, submitted to New Astronomy; ArXiV link