graeham

Expensive by UK standards, but cheap vs Bay Area: Cambridge UK.

I've not been anywhere else that has the density of interesting people doing interesting stuff. London is not far if you need to break the bubble (and Stansted Airport even closer).

Downsides: salaries are generally lower than SV (esp for technical work), Brexit uncertainty, weather is fairly average, limited outdoor sport options (other than running, road cycling, rowing)

generally = sometimes it is less expensive. I guess it wasn't clear, but my point was that healthcare is on average much more expensive in the US.

"In the US food is generally cheaper than it is in Europe, medical care generally more expensive."

I'm not sure when medical care is cheaper in the US. Within the VA system, maybe?

Also, the ranking does not stack up against wikipedia figures:

https://en.wikipedia.org/wiki/List_of_U.S._states_by_GDP_per... https://en.wikipedia.org/wiki/List_of_countries_by_GDP_(PPP)... (Where PPP is already accounted for, so this guy is doing it twice).

UK would come in ~35th. Similar to Nevada or Michigan.

I did my PhD in a lab with a few people do 3d printing of organ scaffolds. Many (all?) of them used gelatin over collagen - I think the main factor was cost being much lower. There may have been something about gelatin being easier to extrude in a 3d printing process as well. I think the idea is that gelatin is a prototyping material.

For Tesla or Newfies?

Its great that the NFL is supporting biomechanics research (and will hopefully help the sport as well).

However, while the models are open source they seem to be created for LS-DYNA, which is a fairly expensive program. I'm not aware of any good open source finite element packages, but I bet this will limit uptake of their models.

I don't think it's 100% of either of these (outcome vs tech used). If its true research, the outcome isn't known. There are major benefits in developing technology for it's own sake, which then provides tools that can be used towards other applications.

To be truely useful as a diagnostic in major decision-making, it needs to be 90-95%+ accurate.

But unfortunately there isn't anything close to this yet.

High blood pressure, for example, probably more accurate than the retinal image technique here. https://www.ncbi.nlm.nih.gov/pubmed/25632496

The major goal here would be to screen for people to go for a more advanced test, like an MRI. But even that wouldn't hit the 90-95% level of predictor.

There is an interesting additional link between cardiovascular disease and bones: calcification (calcium deposits in arteries) also seem to have some link with heart attacks and strokes.

Seen similar things, with the effect of learning curve being fairly massive - particularly for the first 5-10 procedures done. Really understated in the article for being a fairly major factor.

Agreed that multiple stents and multiple vessels can be stented in one cath session. My point was more that there isn't very strong evidence or criteria for what constitutes a vessel that should be stented versus which should not. Worse, plaques are dynamic – what looks benign today could rupture and cause a complete/near occlusion. Rupture and subsequent thrombosis is more dangerous than stenosis itself.

People in a clinical trial typically don't pay for the treatment (or placebo).

Stents are an incredible piece of engineering, but probably are over-used (particularly in the USA). There are a few issues at play:

-Chest pain

-Artery narrowing

-Heart attack risk

They aren't the same, although there is correlation between each. The original idea behind stents was to stabilise vessels in the process of a heart attack. They are increasingly used in a preventative way - but the problem is there aren't very many or good tools to tell how at risk an individual is to a heart attack.

Another major problem (as the article mentions) is that stents improve the situation in one artery, but most patients that need a stent have atherosclerosis in multiple vessels. This likely explains why the chest pain remains.

Agreed. A large amount of time also goes into things like making figures, which is mostly time-consuming grunt work.

I think the key is differentiating between creative work (which is better to be done deep with fewer hours) and grunt work which is almost purely time. Both are needed for a sucessful PhD - few are good at both and it is hard to switch between the two.

In some ways 25 years is an ideal amount of time - means that key patents recently expired. A similar factor lead to the explosion of consumer 3D printers about 5 years ago.

I learned mostly from MATLAB documentation. Good if you want theory and implementation, and most capabilities can be done with open source equivalents if you don't have MATLAB.

https://uk.mathworks.com/products/computer-vision.html

I think you've amplified an out-of-context statement to try and start an argument, and in fact part of my point is that the main responsibility would go to the teacher. But I would say there is still some responsibility for a creator - duty can be held by multiple parties. As an analogy, if I were to break my leg cycling on someone else's private property, there may be some liability held by the land owner depending on the circumstances and jurisdiction. Hence the popularity of 'No Trespassing' signs in these jurisdictions. My background is in medical device design, and we are trained to obsess over the ways that users might cause harm through incorrectly using a product - even if this incorrect use is from ignoring the instructions. If an artist made a sculpture, they may be responsible if the sculpture crumbled onto someone.

As an educational purpose - my idea would be mainly to show it as a fun start to a discussion around product design evolution, computing history, and emulators. The kids I teach were born about a decade after this era of computing, and it would be interesting to see what they perceive to be similar and different. Most of these students have done very little or no coding themselves so the implementation isn't relevant for this use.

Protest here was more of 'make a comment on'...

I think its a cool site. I was thinking of showing it to high school students I teach summer courses to, but its probably not appropriate for them. That's why labels are helpful - nude images weren't ok in schools or work in the 'old culture' either. As art or entertainment (as this website in general was intended), I think there is wide license to the creator, but I think there is still some duty to a website creator to consider what else people might want to use it for.

I was going to protest the full Lena image without a NSFW warning, but hadn't realised the full story of its history[1]...

The site in general is a beautiful work of art, a great blend of attention to detail with comedy of computing in that era.

[1]https://en.wikipedia.org/wiki/Lenna - tl;dr is this iconic test picture for computer imaging was a cropped Playboy centerfold from 1972. I've just finished a PhD which included a fair bit of image processing, but I was unaware of the story behind this iconic image.

Oil prices are essentially capped at $60/barrel for the foreseeable future due to unconventional extraction techniques -> was not long ago oil production was thought to have peaked. Solar power can be economical without subsidy. Robotic and minimally invasive surgery. Battery cost becoming viable for mass-market cars. Gene sequencing. Sensors for pennies. New materials (cheaper carbon fibre, metal alloys, semiconductors, graphine/carbon nanotubes). Most of the things mentioned as 'first half of 20th century' coming to the next 5 billion people.

For aerospace, how about affordability of a ticket over time (Ryanair vs Concorde)? Space travel: number of journeys per vehicle or time astronauts are able to spend in space (was minutes in early space flight, unlimited now). Transportation speed is a straw-man, the industry optimises for cost not speed.

I think the incentives of different investor types should be considered.

If you're a speculative investor and plan to own for <1 month or so, who cares about voting rights.

If you're a long term investor, say >2 years, non-voting shares means the company doesn't have to burn itself for quarterly or annual results.

Activist investors won't like it (but there's not really that many of them).

Will probably work well as long as things are going well for the company. If things go badly, management will need to be very self-aware or the recourse will be a large discount to the share price.

There is different math for each system.

In the USA:

if riskFromTreatment <= riskFromNoTreatment: doTreatment

In most public systems:

if (treatmentCost <= someThresholdPerBenefit) && (riskFromTreatment <= riskFromNoTreatment): doTreatment

In practice, treatment outcomes are pretty similar, but more treatments are preformed in the USA system. Being a for-profit system is one cause, but so is the litigation culture. Not doing every possible treatment and test can lead to the physician being accused of being negligent.

This would be a pretty big step - right now there is a big incentive to be first to publish. Often the researchers who would be doing the replication are competitors, who are racing to publish their results first.

Part of publishing is to put it out for discussion. It should be the best effort and thought to be true, but science is about the evolution of knowledge. Very few things are completely understood, especially in biological and micro/nano fields.

Usually, if you want the full details you look/ ask for the thesis or full report that the journal article is the condensed version of. Most journals also allow supplementary sections for more detailed methods. If its too long, people simply won't read it - already most will just read the abstract and look at figures.

And U of Toronto has one of the top Neural Networks groups anywhere. UBC is also a great university in a nice city.

You work from the key stops (decimal and comma), then go backwards to the only steps that can get you to these stops.

Probably not intended, but the implication this argument leads to is for the development of regulation in use and censorship - like the FCC for radio.

So what's the take-away? Make many high risk bets in hoping one will pay off?

Newton is an exception as well. While his biographers down-play his failures, he is also credited (at least in High School Physics) with things that were the work of dozens or even hundreds of scientists - or giants you might say.

There is work going on in this sort of arena, and I've wondered why there hasn't been more. I think probably there are a few reasons.

First is medical imaging tends to have relatively low resolution and high noise compared to optical images. So there is a danger of over-processing.

Second is, somewhat related, that I think there is distrust in clinical practice and regulation of black boxes where processing happens in the background. You're taking responsibility away from a doctor and putting it on your system, so there is a greater burden of validation.

Third is cultural. Doctors get a ton of training to interpret the signal in these images, and I think there is both a badge of honor in this as well as a barrier to adjust to something new. In my Masters, one of the medical device design instructors talked about a prototype of a digital stethoscope with huge improvements to acoustics. Doctors hated it - even though the sound was better, it wasn't the sounds they were used to. Although this was before the Eko Stethoscope, which is a digital stethoscope and seems to be doing well. Maybe an example of market timing.

This is a great post, thanks for writing and sharing.

I'm a biomedical engineer and found both posts interesting.

There are really exciting things happening right now where the murky area between expensive, slow medtech and fast-moving, lower cost consumer products are being explored.

There have been some specatular implosions of approaches straying to far towards the consumer tech-startup approach.

Like most things, the best path is probably somewhere in between. Medicine and health is hard because the systems involved are complex and non-linear. There is a higher burden for doing things right than other industries. That said, people also suffer from innovation being slowed by red tape and leading to increased costs.

One great but unlikely solution would be for an improved market for used equipment, where the device loses its certification for medical use and obsolete software becomes open source. Little incentive for manufacturers to do this, but would be great for hobbiests, more 'wellness' rather than medical applications, and improved use in developing countries (where used medical devices like ultrasound machines often are donated but aren't used, because they can't be adapted to local needs).