> > >>> > >>>>> > >>>>>>> > >>>>>>>>> > >>>> When it comes to medical technology the US is clearly the leader, no
> >>>> two
> >>>> ways about that. I work in that field and that's a core reason why I=
am
>>>> here. Yes, there are European companies in dem tech. But what do the=
y
>>>> do? Siemens bought Acuson, and Philips bought my first employer ATL.
> >>>> Both US-based companies.
> >>> Probably because US medicos buy US technology. I was working on ultra=
-
>>> sound for EMI Central Research from 1976 to 1979, when ATL was making
> >>> its first splash with a mechanical sector-scanner. We had a 21-elemen=
t
>>> phased array scanner that did better, and had better techology than
> >>> the Hewlett-Packard phased array scanner available at the time.
> >> Why didn't it succeed in the marketplace then?
>
> >>> EMI went bust, so the phased array scanner never went anywhere, but a=
t
>>> that time the US was definitely a follower.
> >> Have you ever given it any thought why EMI (along most other non-US an=
d
>> non-Asian) companies threw in the towel and we at ATL, and the guys at
> >> others, didn't?
>
> > EMI threw in the towel because it ran out of money. ...
>
> Yup. That's how they usually go down.
>
> > ... It merged with
> > THORN Electrical Industries in October 1979 to form Thorn-EMI, which
> > is to say Thorn bought it for what it cost to pay off it's debts, and
> > Thorn went through the place selling off everything it could. The
> > phased array ultrasound scanner patents and technology was sold to a
> > US firm which apparently didn't get around to exploiting them.
>
> Patents? Phased array was around in Radar long before. There wasn't too
> much patenting going on in my line of work, we just engineered the
> stuff, ramped up production and then sold lots of machines.
The patented part was the scheme for controlling the delay between channels
formatting link
The crititical - analog - stuff was all Ian Green's. I'd done a theoretical study on generating the same sort of of pulse streams digitally, whose total practical use was to draw Ian's attention the fact that he could ignore signals from the outermost transducers when looking at points very close to the array - obvious enough, but he hadn't thought of it. The lawyers wanted to cover all possible implementations, so my digital stuff went into the patent and my name was added.
> One of the contributing factors was a couple of warehouses full of
> > Body-Scanners - full-body X-ray tomography machines - waiting for the
> > software team to get the patient billing program to work. Each machine
> > sold for a couple of million dollars, but nobody was going to accept
> > them without the software that kept track of which patient had been
> > scanned so that the patients could be charged for the service. ...
>
> That sounds like lacking program management.
Software program management was still pretty primitive back then. "The Mythical Man-Month" was published in 1975, but wasn't all that widely read at the time.
formatting link
> ... There
> > were other problems - the recorded music business was usually a
> > reliable cash cow, but fashion can be fickle, and the revenue from
> > that side of the business fell short at precisely the wrong moment.
>
> Huh? What do you do with tomography in the music biz?
EMI was found in 1931 by a merger of existing gramophone companies. It produced not only gramophone records but also recording and playback equipment. Alan Dower Blumlein was one of the electronic engineers who worked for them from the start
formatting link
after inventing stereo, he went on to make major contributions to 405- line television system. The EMI Central Research Lab kept on producing interesting electronic inventions for another forty years, including the first tomographic brain scanner. I got to talk to Hounsfield once in the course of my work there, and narrowly avoided having to work with C.A.G.LeMay.
> The magnetic resonance imaging work that was going on at EMI-Central
> > Research at the time did get sold to a group who continued the
> > development and made a lot of money out of selling the machines.
>
> >> To invent something or be first is one thing. Making a product that is
> >> successful in the marketplace, very different thing. Fact is, the firs=
t
>> mainstream array scanners came from ADR and ATL. Meaning scanners that
> >> doctors could actually buy without taking out another mortgage on thei=
r
>> house.
>
> > The Emisonic 4500 was certainly expensive - of the order of $50.000 -
> > but not as expensive as a doctor's house. It had continuously variable
> > focusing in the recieve channel - each channel had it own delay line
> > (a Reticon bucket-brigade delay line) clock by its own voltage
> > controlled oscillator, so the delays between each channel were
> > continuously variable.
>
> Ours had that, too, except we called it dynamic focus and it was done
> via an elaborate crosspoint matrix and delay lines. Plus some had
> vernier tricks that I am not at liberty to talk about.
The vernier tricks were crucial. At EMI our phased array machine performed better than the simulation. When Ian Green dug into the reason why, he found the simulation quantised the delays in steps of
5nsec, while the machine provided continuously variable delays. Quantising the simulated delays with smaller steps brought the simulation closer to reality.
Although, now
> water under the bridge as it's all digital these days. I looked into
> charge-coupled lines because I designed a CCD camera as my masters
> project so was familiar with the technology. But I did not like the
> business side of that technology and boy was I glad I ditched that idea.
> It all became unobtanium, product lines and whole companies died away
> left and right.
We were well aware that digital was the way to go. The technology wasn't quite there at the time, but we were happily digitisng our signals and dumping them into a video frame store. Ian invented his own vector to raster transformation scheme, which dumped the vector scans into store as vectors, and played them back into a raster scan, by digitally combining several vector points into one raster point in the near field of the transducer. At the outer end of the far field each point on the vector scan matched a single point on the raster scan, but the raster points representing information from closer to the source were intelligent combinations of adjacent vector points.
With those bucket-brigade chips you wouldn't have stood a chance. Plus
> their dynamic range is usually very inadequate for Doppler, and at least
> in the 80's an ultrasound machine without Doppler was a non-starter in
> the marketplace.
We had a Data General minicomputer for doing Doppler. In 1978 the chips we could buy would have needed a lot of board space to do the computational work involved - the BodyScanner being developed elsewhere in the building at the same time used lots of AMD bit-slice chips for its heavy-duty computation. Motorola had an ECL bit-slice system that was twice as fast, but offered a a much more restricted range of chips. The BodyScanner people thought about going that way but didn't need the extra speed badly enough - and most digital engineers were frightened of ECL and transmission line terminations back then. I was primarily analog, and had already used ECL.
formatting link
> The first commerical examples weren't as good as they should have been
> > - I sweated blood on the clinical trials machine to provide the
> > facility to check each of the 21 channels in isolation, but when we
> > got to play with the first commercial machine, all of our hardware was
> > in place, but not accessible from the maintenance software.
> > Unsurprisingly, only seven of the 21 channels were actually working.
>
> Our 1st machine had 32 channels and they always worked. No alignment
> needed whatsoever, it was all active laser trimmed hybrids.
Our's didn't need alignment or laser trimming. Why the first commercial machine had so many duff channels pobably has to do with their front end amplifier board - they'd done a printed circuit layout early on and gave me a copy of the artwork to use in the clinical trials machine. When I looked at it I could see that they'd made a mess of the layout around the uA733 amplifier - output track running next to input - corrected it, and told them that the layout was useless. They still have some 220 boards - ten machines worth - made and loaded, then had to take out the 733's (which, predictably oscillated) and replace them with a couple of transistors hand- soldered onto the board.
The rest of the layout was fine, and my boards all worked reliably for the year or so I was responsible for keeping an eye on the clinical trials machine.
It seems you know nothing about the US system. At Kaiser the doc decides
> and that's that.
Sure, but if he makes too many expensive decisions, he's going to be let go.
> But they still have to justify the decision. You don't get to see the
> > economic arguments at work, but they still influence the treatment
> > that you get.
>
> They do, but the doc can always deviate. In the case I mentioned
> (country with more or less socialized medicine) that was not allowed.
> And that's wrong.
It was not forbidden, but the patient got to pay for the doctor's extravagance, rather than the doctor. That's not nice, but it doesn't strike me as making "more or less socialised" medicine morally bankrupt.
> Medical malpractice cases and insurance do drain money form the US
> > medical system, and cause it waste money on unnecessary tests and
> > excessively defensive treatments, but objective studies put most of
> > the blame for the extra cost on a top-heavy administration, most of it
> > in the medical insurance companies, some which ended up paying for the
> > lobbying campaign against Obama-care.
>
> Wrong again, sorry. A core reason for large administration _and_
> excessive medical attention is to shield and brace against lawsuits.
Since a whole lot the extra money is spent on lobbying, which doesn't seem to be at all interested in tort reform, it would seem that the medical insurance community doesn't share your opinion. If they could stop wasting money on avoiding malpractice suits, you'd think that they would be out there lobbying like fury for tort reform.
Tort reform is the only way out of that, everybody knows it, but of
> course right now nothing will move.
>
> >>> You may have ideological objections to the various flavours of health
> >>> care available in Europe, but the American system is over-priced and
> >>> under-performs. Hanging on to it is cutting off your nose to spite
> >>> your face.
> >> No, I hang on to my HMO plan because it works. Simple. We just receive=
d
>> notice that we are grandfathered in for the full coverage it offers bu=
t
>> that if we ever dropped out we can't get back in, and that no new
> >> applicants can get in. So far for reform, I guess.
>
> > It works, but it costs you a lot more than my medical insurance - with
> > the CZ group in the Netherlands - costs me.
>
> Have to look it up but IIRC it's about $650/mo for the two of us, and I
> am not a spring chicken no more.
Mine is 113.85 euro per month - about $150. I've got to pay the first
165 euro of what is spent in any year, which effectively pushes it up to 127.6 euro per month, since my various medications use it all up. My wife has separate policy via her employer, and I don't know how much she pays - except that it is more than I do.
Salient fact: After moving to the US I found out that our coverage in
> Germany was more expensive than here, a lot. I paid over 800
> Deutschmarks in 1996 and that was through an engineer's health
> organization which had slightly better rates than the rest (but only
> took engineers, scientists, technicians, et cetera).
1.95583 Deutsch Mark =3D =801. 800 DM is 409 euro or $537, appreciably less than you are paying at the moment You've got a high deductible on your current policy - how high was the comparable "eigen risico" on your German policy? I presume your German policy covered your wife as well.
If your original US coverage came via your employer, was there an element of employer subsidy in that cover?
-- Bill Sloman, Nijmegen