h a l f b a k e r yLike a magnifying lens, only with rocks.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
this idea is just so crummy, yet so compelling.
How many terrorists have been prevented with airport screenings? What is the rate of the most common cancers?
if 1 per thousand persons at a particular age profile is a screening success then the multibillion annual air travellers successfully alerts
millions of people they have a treatable illness. Im sure this antiterrorist airport imaging has saved vastly less than a million lives. Medicalizing the scans to rescue 1 per thousand people provides 4 or 5 orders of magnitude greater fatality reduction than seeking terrorists.
As a technology you could use a few options, medically beneficial scan, minimum scan, refusal, or predictive software scan, noting that the vast majority of people could skip the medically beneficial scan yet might want physician programmed risk estimation without having to think about it.
all the scan data uploads automatically to the persons handyphone, or if they prefer, the cloud.
Traveloscopy
Traveloscopy
Redundant to the linked idea [theircompetitor, Jan 19 2012]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Annotation:
|
| |
What would be the additional cost from all the false
positives? |
|
| |
[swimswim] well thats among the well being opportunities, they could simply create a false alarm rate say a tenth or a hundredth of that of normal medicine, yet still find much curable illness. |
|
| |
Also, some things have very high amounts of detectability. viewing this generously lets say the thing can detect tobacco use, as well as high blood pressure (chemosniffer machine or EMpulmonagram) . Together those are notably yucky as tobacco kills 20 times more people at the US than are murdered. If they have just visited an area with tobacco while viewing a thrilling movie the machine might well tell them, yet they already know they are tobaccoless. |
|
| |
even if only 1 per thousand tobacco users with high blood pressure cease tobacco from this machine, 999 people say, "you know, I knew that" while one person says, "what, high BP, with tobacco, Im ceasing tobacco" then that could save 1/1000 of a several billion, times the actual quit rate, times the number of annual travels. Hundreds of thousands of or perhaps a million lives saved. |
|
| |
I could be wrong, but I sincerely doubt even the newest
TSA scanners are up to the task of ferreting out tumors and
other anomalous masses. Beyond the fact that they're just
not sensitive enough to show the difference between
normal and diseased tissue (sometimes a subjective art
even with far more powerful and precise medical-grade
equipment), security scanners are specifically configured
to highlight solid objects and non-biological substances
such as explosives. I read in Newsweek (making the
information somewhat suspect) that they even have a hard
time disguishing packets of heroin and cocaine in GI tracts
of drug mules from normal digestive matter, which shows
up as vague blobs, if at all. |
|
| |
// lets say the thing can detect tobacco use, as
well as high blood pressure// |
|
| |
Great idea!! Let's also say the thing can detect
early degeneration of the substantia nigra, and
pre-cancerous lesions! Wow - that was easy! Now
let's say it can detect the signs of incipient
schizophrenia too. Why didn't we think of this
saying business sooner?? |
|
| |
There's not much diagnostic information available
from any of the current security scanners
(although they could probably pick up if you had,
say, a bullet in your liver). |
|
| |
Moreover, doctors hate this sort of population-
wide screening (particularly if it's a whole-body
look-see scan), because the false positive rates
enormously outweigh the true positives. False
positives kill people because the follow-up
investigations often have a finite risk, and
because they take up doctors' time. |
|
| |
Dumb idea, but then again [theircompetitor] was
dumb before you. |
|
| |
Im thrilled, I get to ignore my qualms because other
people actually have started building it. I'm
croissanting [theircompetitor]s idea |
|
| |
Also, the terahertz scanners used by the TSA see through
clothing, but not skin, effectively displaying the person
being scanned as if they were nude -- if they *did* see
through skin, there would be much less of an invasion of
privacy issue. |
|
| |
Cancers, except for skin cancer, are fully inside the body,
and wouldn't show up on a THz scanner. |
|
| |
What [Max-B] said - general scanning of a population is a bad idea. If we simplify things and say the scan is very good and is wrong only 1 time in 1000 (and symmetrical, so one cancer-free person in 1000 will be told they have cancer, and 1 person in every 1000 with cancer will be told they are OK). We shall also assume that in the population as a whole, 1 person in 1000 people actually has cancer.
So, in a population of 10,000,000 there are 10,000 people with cancer and 9,990,000 without. These will go through the scan as follows:
* 9,990 people with cancer will be told they have cancer
* 10 people with cancer will be told they don't have cancer
* 9,990 people without cancer will be told they have cancer
* 9,980,010 people without cancer will be told they don't have cancer
So, you'll be telling 19980 people they have cancer, but only half of these people will actually have cancer. This is not a very good test, it's incredibly unethical to worry so many people unnecessarily, and no healthcare system can cope with this number of false alarms. |
|
| |
Actually on those numbers it would probably be a
good thing - most mass-cancer-screening is way
less good than that. |
|
| |
In reality, though, I'd say that you'd get a false
positive rate of at least 1%, and a false negative
rate of at least 10%. (You could change the
criteria, but you'd either increase the false
positives or increase the false negatives.) |
|
| |
On that basis:
*9,000 people with cancer will be told they have
cancer
*1000 people with cancer will be told they don't
have cancer
*99,900 people without cancer will be told they
have cancer
*8,991,000 people without cancer will be told they
don't have cancer. |
|
| |
(OK, they wouldn't be told they have cancer;
they'd be told that there's a chance they have
cancer and ought to get a follow-up
investigation.) |
|
| |
I'm splitting from the rules and bunning this as a
great idea for a Saturday Night Live skit. |
|
| |
[Max] I know - I was assuming a very good test, and even then you get terrible results. Your error rates sound more realistic. |
|
| |
It is strange to see news reports suddenly backtrack on recommending PSA tests for example, which seem to have relatively high error rates, but given that, if you're not lucky enough to live near the equator, up to 50% of people will be genuinely diagnosed with cancer at least once in their life, wouldn't annual testing make sense even with false positives between 1-10%? In some countries, technicians administer the tests, with doctors only becoming involved later if at all. |
|
| |