Posts by Dante Alighieri 75 posts • joined 20 May 2010
in NE who has a better tower with a crawler for the saturn 5, I can only regret there is not a picture he has on line
I've seen it at our scouts (figure this out if you have to) and has models in te local LEGO store.
Got t love the LUG (LEGO/Linux User Group)
I'll be at the back of the Brick Alley (hint)
My concern would be that while it may be possible to automate a stroke/ischaemia/haemorrhage ML solution, the problem is that a small number have other mimics - tumours, paraneoplastic, encephalitis to name but a few. I doubt the training set will contain enough of these to make it safe or useful.
Pretending the issue is stroke/not stroke and haemorrhage/no bleed is a failure to properly understand the issue. You need to teach it a much larger part of brain CT interpretation*
And spotting a bleed on an acute CT is trivial - white bits where they don't belong, it's not subtle.
Hyperacute ischaemia is harder, but if in thrombolysis window you are going to get to roll the dice.
This appears to be a solution looking for a problem. Stroke teams and Radiologists do this daily and have sufficient expertise to recognise the odd ones. There are other more difficult areas where computer support is more appropriate, eg lung nodule analysis and marking.
Humans being humans have to be careful not to rely on such tools and miss the unmarked ones or other pathologies
{thrombolysis is a double edged sword - it takes a normal distribution and moves cases to the ends - much better or much worse/dead}
*I mention CT and not MRI, as acute assessment with MRI is exceptional in the true sense of the word.
No - that is NHS identity
This is a permission to work system (see my response above) : are all necessary checks in place and trusting another organisation to do this for you.
There are plenty of ways to establish your bona-fides, not least of which is simply logging in to your NHS email!
NHS Identity is basically a remote access token to a live clinical system.
NHS Passports which have been discussed in my region for the last year (so they are not new or related to COVID!) are meant to bundle your Occ Health data, Mandatory Training, DBS and professional registration checks in some form of bundle that the relevant HR can determine that these are in place (but not necessarily share the actual data) - one Trust guaranteeing that the staff member is correctly checked and safe to allow on your premises (an NHS equivalent of the schemes we have for contractors on-site)
Hope this helps
It was meant to be more a comment on Oracle's billing model than the architecture - I exaggerated RISC/CISC for effect, so apologies for upsetting the technically more competent than me regarding this.
I took the original post as what's the trap and suggested a possibility.
Mea Maxima Culpa
which icon is for self flagellation? gimp or..
Ask the medical system vendors who hamstring the deployments and then the failure to specify browser agnostic delivery of service from the purchasing team - which often doesn't include appropriate IT involvement (includes all of: not invited, didn't bother to come and don't enforce vendor/browser neutral solutions)
DOI involved peripherally in some £M+ procurement of said systems
Apart from an excellent Jethro Tull track,
genuine question from a hobbyist* : is it too late to sniff traffic and reverse engineer protocol to allow an on premises re-direct of traffic to believe it is seeing the shyster server?
Is it time to do this for other IoT / smart tellies etc (PiHole recognised)
I've flirted with automation stuff as a concept for around 5mS (which as Commander Data pointed out is an incredibly long time) but unless I can run it on an open home automation server I control it is all someone else's property.
*I am a professional in a related area which absolutely relies on IT, as some will have worked out from my prior postings
We may finally implement Russell's Teapot in the form of our sovereign positioning system. There's no proof it's not there already!
Oh, nearly forgot :
Just dumped the unsupported router and glad I've avoided this.
I will confess to a 4 port GB unmanaged switch though. More impressed with the 8port+1 TP Link for my small LAN. QoS is on the router for now.
If I had bought gear limited like this
Sale of Goods - not as described, does not function correctly - full refund (by chargeback if necessary)
I see your point.
I read it differently - a simple observation of when the brakes come off night life (in many Toons) there will be a demand on the emergency services - one of which i work in.
plus ca change.
...how do I post anonymously? ;)
Surely Elbonia...
https://dilbert.com/strip/2020-05-05
one of these days I'll learn to do a linky thing...
I find the article a little odd.
It is as though Leeds is the first hospital in the UK to think of doing this.
Mine a bit further north has already ripped them all out. perhaps Leeds should ask for advice rather than pretend to be an authority.
We accept radiology referrals from consultant NHS.net accounts - it is an electronic signature in its own right, for those based outside our organisation. Internally we have a request system which you can use or be told to resubmit your scrap of paper electronically. Almost completely paperless referrals now.
Re fax ct scans etc - you can email/scan/fax a report - the images not so much.
It's my day job.
my RISC PC drives just dots...
and I thinkn the AMSTRAD printers were just dot engines...
I will remember not to unpack my STAR given the comment above.
Will the olivetti BJ be susceptible to this?
It seems that you have to be seriously up the political chain of influence for your print outs to be subjected to this. My Scout program print outs are probably not going to attract the 5 Eyes.
Dr Liam Fox is a qualified medical practitioner.
The protected title in the UK is Registered Medical Practitioner
ene fule can call themselves doctor, even dieticians in white coats with stethoscope.
A very quick check of the GMC medical register shows that Liam Fox qual. 1983 is Registered WITHOUT a licence to practice - someone hasn't kept up to date, been appraised or revalidated, all policies pursued by...
politicians passeem. Do as I say....
Common usage is that for those of us with MBBS / MBCHB we are allowed the honorific Dr
Oddly at my diving club, most had PhD's, none used doctor outside of the lab/work, but muggins here used Dr without a PhD. Still do...
Hello.
I am a Radiologist.
I have an interest in radiation safety.
There is NO limit to the dose my CT scanner gives a patient.
I authorise the use of radiation. I have had patients receive >50mSv for a diagnostic test.
The UK accepted rate for cancer induction is 1:20000 / mSv.
My patient has a death risk of 1:400 to look for a possible pathology they have a <1:4000 of having.
Dose is cumulative
Induction time lag is around 15-20 years.
I'll fry your granny - but not your kids.
Comparing targetted radiotherapy doses to tiny areas is not appropriate.
Yes we run XP / 7 for our underlying systems as the suppliers will not warrant any other environment.
The business end (custom hardware) works with its own internal systems (think CNC machine or 3d printer) screwing with the dose regulators is beyond the XP bit. When that goes down I have shut several scanners over the years - including writing one off! Mr Popular...
Dose is controlled by exposure factors set at the operators console (the XP bit) and the area covered and how many times.
The first acknowledged Fukishima related cancer is at a dose of 10-20mSv.
In the larger adult this dose is not uncommon on our current scanner(s) which are not particularly low dose.
New algorithms and technology is reducing delivered dose - but better tech is leading to multi phase exams that increase dose.
We have patients with accumulated doses of >200mSv, in an age range where this is an issue and a worse than 1:100 risk of death from their radiology scans.
Re MRI : it is a giant microwave. never mind there is no ionising radiation, it is perfectly possible to induce fits from neuronal stimulation from aggressive gradients or literally cook some one with RF energy.
We cannot air gap the provider need dial in access for fault finding and remote management when required.
We run NTPc on CT that connect to GMT-8 (ish) as that that is the worldwide system clock.
Radiology runs on a private subnet the firewall and bridging is managed.
Denial of service : highly likely
Manipulation of target devices : low - too varied, custom device v operators console manipulation
This is my day job at the pointy end. I spend some time on IT projects at work.
I remain a competent 6502 assembler programmer.
I use these. Professionally.
They are pointless unless in a controlled environment <10lux for diagnosis.
Medical monitors outside this are pointless. Accurate DICOM reproduction is nice. If they are <2MP they are not diagnostic, If they are in an uncontroled environment they are not diagnostic
Outside of proper diagnostic offices they are an entirely avoidable and unncesssary expense.
DOI : I have stopped unnecessary over-specced monitors being deployed across an organisation, and also making sure diagnostic screens are available in the appropriate departments with explicit "your environment is non-diagnostic" warnings.
Many think tha the CT and MR scans are the most details - they aren't 512x512, 10 bit pixel depth. Most bog standard monitors far exceed anything necessary.
"plain film" x-rays - much more challenging!
My A420/1 emulated a 386 well enough to be competitive. Flew on StrongARM Risc PC
If we are looking at alternative os I would pay **sausages** for one with RISC OS
Some features of which are yet to be matched by both (ex) windows or my current linux boxes.
make an !Impression, use !Artworks, !Squirrel away your data
still waiting for a native box in 2017
The Register - Independent news and views for the tech community. Part of Situation Publishing
Biting the hand that feeds IT © 1998–2020
