Showing posts with label Government. Show all posts
Showing posts with label Government. Show all posts

Tuesday, December 8, 2009

Is it Ok to Click Here?

Introducing Animoto as a Web 2.0 tool from a personal perspective. It is very easy and quick to create a video. The web interface is really intuitive. Congrats to the Animoto team! I really enjoyed it. It is a simple, 10-minute process but !

Did you ever wonder about the effort put into those amazing documentaries? Lucky for me, I know a very talented individual who works multimedia for a living. He will spend countless hours removing the sound made when someone inhales each time a new bite of narration is read. It’s one of those things where the absence of evidence is not noticed until someone inhales very loudly.

There are times I wished I was the one collecting the images. A good final product requires hours of research for rich image contents, sometimes exposing you to very interesting or, even dangerous environments. That very difficult part hinges on the creativity of the individual. Music selection also has a great part to play in the effectiveness of media delivery.

For me, this exercise reminded me of the TED video introduced by a classmate regarding "Optical illusions show how we see" In his presentation, Beau Lotto demonstrates that even at this very fundamental level, the brain's capacity to detect differences in light heavily depends on context. This Animoto project certainly has emphasized the reality of his assertion. Thanks for the experience Lyr Lobo. I think I found a new passion.

Here's my rendition of the "Good and Evil" of technology and the Wild WWW. Images are courtesy of Communications of the ACM via screen scraping. Got your membership yet? *grin*

Saturday, December 5, 2009

Consensus Methods - Who To Believe? And Why?

While studying various consensus methods as part of our curriculum, a classmate departed from main stream discourses looking at the recently controversial news about breast cancer screening. In my opinion, her analysis merits a post on this blog.

While the full version of the US Preventive Services Task Force is quite comprehensive, it gives no indication as to the type of deliberation and consensus method used. The contents of the article I am referring to can be found here: "Screening for Breast Cancer: U.S. Preventive Services Task Force Recommendation Statement. "

As a scientist, I think of myself as one of them too *smile*, I find that tit-bit somewhat disturbing. Why? Because while there is no way for all to know everything about every thing, there are definitely many reasons to validate or, invalidate such reports. Should we trust such a panel of experts? My personal verdict is still out.

Interestingly, I found out about the "Screening Policies won’t change - US Officials say" article shortly after the initial debate on this question. When I saw the title of the article, my mind put a "case closed" stamp on this rather quickly but, it probably shouldn’t be so in hind sight.

Whether the data is scientifically valid or not, a physician worth his/her salt will find time to review such reports to make informed decisions about recommending what and, to whom. So many factors will influence a decision for or, against mammograms in this case. It heavily depends on the patient. As an example, the DELPHI method, modified or not, will likely never bring any one group to consensus on this item, not yet anyway. This is in fact, one of the biggest challenges that face the migration to EMR - Electronic Medical Records. You can read more about those problems here: "Implementing Electronic Medical Records ". I have first hand testimony on this effort. I attempted to help an Internal medicine friend with her small-practice EMR project. It was already a disaster by the time I was asked to contribute. Reading the article above from my hard copy, I ripped it out and gave it to her. I recommended she held off for a bit longer. The frustrations she experiences just aren’t worth the trouble at this time.

There is great merit to moving to such a platform nationwide, even world wide! It could become a very meaningful source of data, like in the case of the breast cancer debate. There would be some source of data to analyze. However, there are also significant draw backs.

It’s evident that the importance of breast cancer screening remains high since it is the number TWO leading cause of CANCER deaths among women. I read that lung cancer is the number one but, I'm told ovarian has the highest mortality rate. This is because its discovery is nearly always in stage three, the point of no return, usually. Ovarian gets no press because it affects a very small community of women in spite of its close-to-no-mercy status. Another fact to remember is that there are also instances of cancer in patients whose family history shows no trace of cancer at all. I know this first hand as well, unfortunately. Would there not have been reason to find this out sooner? Maybe some day, enough well-analyzed data will lead to such conclusions. For now, those relatively few cases will remain victims of unknown circumstances.

To me, this is a case scenario where a process like
"Christakis's SDP" could truly make a difference. It would remove emotions and, hopefully politics, allowing for factual contributions from various sources of experts, and non-experts alike, to prevail.

Tuesday, November 10, 2009

Cloud-based Computing & the Confusion Within.

You know? Some concerns about cloud computing from a doctoral peer are very much a big deal to me as well. Recently, however, I bumped into an article which gave me new hope, albeit with the expectations that we do well implementing what we know works in terms of IT security.

Cloud computing does offer very easily accessed, mammoth quantities of both CPU power and storage. It offers scaling like never before and this, at a very inexpensive investment price. I’m hoping that the Horizon Report 2009 is on target when it states that cloud computing might open “doors to wholly different ways of thinking about computers, software, and files.” We need it in a ghastly way.

Some of the draw backs outlined in the said report include the requirement to entrust your work and applications to the cloud. Who’s in the cloud? Who will supervise this same cloud? Do we not already have enough problems with our current cloud - the Internet? It just isn’t under any particular supervision. Cloud computing might afford us better chances of survival in the WWWW, wild-world-wide-web. The needed trust becomes widely shared with not so questionably reliable entities. This still sounds like a scary proposition to me.

I’ll try to address my peer's two questions with information I recently acquired. It raised my hopes for survival. Mind you, I still think some of our methods are too reactionary.

Q1. When will the government upgrade out infrastructure?

A1: The government seems to have a jump start on this. At least, that is the impression I get from the article I read in the October 2009 edition of ComputerWorld. Here’s the link about it:

CIA building secure cloud-based system
Posted using
ShareThis

The article seems to show that the government is in fact doing what it can to protect critical infrastructures while using a cloud computing approach. Not that they have much of a choice now, do they? As they state, it's been in the making for some time. We just gave it different names. It’s only a partial solution to me though. You’ll note that they do recommend that the cloud be managed very tightly and within the confines of a firewall. Still, that’s a big step forward. Fiber optic, as Barcus suggests, is a good idea since it is virtually impossible to eaves drop on that but, it is enough?

Q2. Will congress past a bill that regulates the trust factor of Cloud Computing host vs. users?

A2: I think that’s a very good question. The only down side to this is that once we create regulation, we then must enforce it. Good side effects come of this too! That is where employment is created to accomplish such tasks. Of note however, is that government regulation about network computing has been chasing its tail. Best be pro-active about everything in every way we know how, today. Wouldn't you agree?

Personally, I find the idea of trusting outside parties with my information scary no matter how it is done. If some are afraid of tagging becoming a part of the birth process, this is not going to help us one iota. I'm left with a few catch-22 questions:

  • Is anything different a realistic choice?
  • If not, how do we build litmus tests and comfort into such technology?