Medical and IT researchers at Johns Hopkins University, healthcare application software supplier Harris Corp, and virtualization juggernaut VMware have teamed up to create a medical imaging cloud that they hope will become the central, secure repository for US citizens and the doctors who care for them. The desire to provide …
Another fool suckered by "The Cloud"
Great. Not only will hackers more easily be able to steal proprietary company info from "The Cloud" but now all my medical records will be fair game. What a bunch of maroons.
Re: Another fool suckered by "The Cloud"
So I see. "OMG! It says cloud! It must be crap!"
Have a quick look past the marketing before casting aspersions, eh? I'm sure there's plenty of real issues here, not the tired old knee jerk reaction to the word 'cloud'.
@kev99. Yeah sure. Cloud = Bad.
It's in a colo. It's managed by them on their kit. I very much doubt (unlike AWS) you'll be able to cuddle up alongside them in their infrastructure with your hacking tools. If it's well designed with security baked in it'll be far more secure than individual organisations all trying (and potentially failing) to do it themselves. Look up 'community cloud'.
My concern is the use of the thin client / VDI approach. Extremely laudable from a data loss prevention perspective (and potentially bandwidth/speed too) but will PCoIP introduce loss into the imagery which in my understanding is a big no-no for medical imaging where the tone and shape of a grey blob can be the difference between life and death.
Image quality is only an issue on the initial diagnostic read, which is done within the first day or so. After a while, lossy compression can be used as the images are for reference only. Also, Mammography images are held at least 7 years from last exam.
Re: Image quality
"Image quality is only an issue on the initial diagnostic read, which is done within the first day or so. After a while, lossy compression can be used as the images are for reference only. Also, Mammography images are held at least 7 years from last exam."
I really hope this is your *opinion* and you have no idea what you're talking about.
Because AFAIK *lossless* data compression is the *only* type that's allowed for medical imagery.
Consider that *one* white dot on chest image that turned out to be a very small (and very killable) tumor.
Erased as the compression algorithm eliminates it as "noise".
Not noticed. Patient dies. Family sues and is told "Well we're not sure what the *actual* image is because the "lossy" compression algorithm threw some pixels away".
The reply from Sporkinum confuses the compression applied to the image on disk (which may or may nto be lossy - I don't know).
Anonymous coward asked about PCOIP compression - which is the compression used by the remote desktop protocol. Read up about PCOIP - it is pretty smart. It will 'build to lossless' - ie you might see some loss when an image is rotating, depending on how much bandwidth you have.
HOWEVER when the image is stationary - ie like an X-ray image - you will have a perfect image.
Speaking as someone who worked on the first PACS system in the UK, this is really interesting stuff and a good use for PCOIP.
In the US, even though the Social Security Number (SSN) is not supposed to be used as either a general or national ID, in practice it is. Of course, this leaves yet more room for miscreants to get the 'crown jewels' of identity theft.
On the cloud front, the more sensitive data that's out there, the more vulnerability, History seems to prove that internet security is a game of leapfrog.
"an open-source clinical image and object management system."
So how much of the £15Bn did the NHS spend developing something to do this?
Note this approach is *probably* only possible because US healthcare (for those who can *afford* it) is a Terra dollar business.
In the UK it was an awesomely stupid architecture.
Re: "an open-source clinical image and object management system."
this is not (as far as I know) a medical records project like the NHS project - it is a PACS project, ie allowing remote access to medical images.
You should not confuse the two.
Re: Re: "an open-source clinical image and object management system."
"You should not confuse the two."
I guess that raises 2 further questions.
The UK PACS systems seems to have been one of the *good* parts of the colossal cluster ***k of NPfIT. Did the UK system use this software (at least as a starting point) and if not how much did the base system cost.
Could the "objects" being managed have included all the data to constitute an electronic patient record?
I wonder how it deals with loops and colour, 2 things that generally let PACS systems down when viewing ultrasound exams. This is why Acuson and later Siemens Medical had to create their own PACS-for-ultrasound solution KinetDx.
Employees vs government vs hackers
Hacking is so last year. Why not just pay an employee with direct access to the database for details of medical records? I'm sure there are plenty of low-paid healthcare employees who would not be against taking a few buck in return to do an illegal data search for the ploice / your prospective employer / the press etc
Anyways, the real problem with the security of a medical mega-database is who the government legally allows access. Expect upcoming data trawls for DNA matching, for meta scans for medical insurance fraud, etc etc etc. And what about giving Google "limited, anonymous" access for marketing purposes?
Data is power, and the more easily a government can tie all our data up together the more power they have over us.
- Vid Hubble 'scope snaps 200,000-ton chunky crumble conundrum
- Bugger the jetpack, where's my 21st-century Psion?
- Google offers up its own Googlers in cloud channel chumship trawl
- Interview Global Warming IS REAL, argues sceptic mathematician - it just isn't THERMAGEDDON
- Windows 8.1 Update 1 spewed online a MONTH early – by Microsoft