Cheap Blood Clot Detection Device

Gearoid_Murphy writes "The BBC details the news of a cheap handheld device to detect blood clots on the surface of the brain. The device uses infrared light to penetrate 3 cm into the body; light that has passed through clotted blood changes detectably. A doctor who is testing the device in India said, 'We found a 98% accuracy for showing blood clots or haematomas.'"

Comments

To stem the statistical comments:

[cgenman]

Inevitably someone is going to say "Well yeah, that means 2% died. Rough lot of good that did them."

Before that person is you, think of the 98% that lived. I bet they're pretty happy that their chances of detection and survival went way up. And if you were sitting on an operating table in rural India with a poorly underfunded doctor wondering what's going wrong with you, wouldn't you like to take those odds too?

Re:To stem the statistical comments:

[Repton]

I'd also be interested in the false positive / false negative rates, and the overall rate of blood clots.

Eg, suppose 1 in 10 patients develop blood clots under some circumstances. You could get a 90% accuracy by making a device that just reports "No clots" every time. If you're classifying 98% of clots as clots and 98% of nonclots as nonclots, over 1000 tests you'll have 98 blood clots correctly identified, 2 missed, and 18 nonclots misclassified as clots..

(obviously I have no idea what the true rate of blood clots is)

Of course, the engineers who made the device and the scientists who test it almost certainly know all this, so I'm not being particularly insightful. If they call it a breakthrough or think it will be useful, then they're probably right. We just can't tell either way from the article...

("Mainstream news article lacks useful details: film at 11!")

Re:Company Website...

[fuego451]

You really have to wonder what the FDA's motivation is for not allowing at least experimental use of this device in emergency settings, along with other accepted practices, to measure its effectiveness. Is there a genuine concern for the patients safety? The device certainly seems harmless enough.

Ever the cynic, I would guess that the device and the procedure are relatively inexpensive and all parties involved are working out how best to monetize (god I hate that word).

Re:How it works

[nanosquid]

infrared light is non-ionizing, so it's absolutely no dangerous to use that kind of instrument continuously on a person

A kitchen broiler is also non-ionizing radiation, but I suspect using it "continuously" on someone's brain is not such a good idea.

Re:Woah..

[clickclickdrone]

Well, India, rather than England but if it does the job... Heck in the UK they still use leeches c/o the NHS and it's not uncommon to use maggots on badly infected wounds because they eat the bad stuff and leave the good stuff far better than any other treatments. Just because something sounds medieval, don't write it off.