Toshiba Claims Its Device Tests For 13 Cancer Types With 99% Accuracy From a Single Drop of Blood (japantimes.co.jp) 56
Toshiba has developed technology to detect 13 types of cancer from a single drop of blood with 99 percent accuracy, the company claimed this week. From a report: Toshiba developed the diagnosis method together with the National Cancer Center Research Institute and Tokyo Medical University, and hopes to commercialize it in "several years" after starting a trial next year. The method could be used to treat cancer in its early stage, it said. The method is designed to examine the types and concentration of microRNA molecules secreted in blood from cancer cells. Toray Industries and other companies have also developed technologies to diagnose cancer using microRNA molecules from a blood sample. "Compared to other companies' methods, we have an edge in the degree of accuracy in cancer detection, the time required for detection and the cost," Koji Hashimoto, chief research scientist at Toshiba's Frontier Research Laboratory, told a press briefing. The test will be used to detect gastric, esophageal, lung, liver, biliary tract, pancreatic, bowel, ovarian, prostate, bladder and breast cancers as well as sarcoma and glioma. Toshiba has developed a chip and a small device that can conduct the diagnosis in less than two hours. A blood test using it is expected to cost $182 or less, it said.
Rebirth of Theranos? (Score:3, Funny)
Re:Rebirth of Theranos? (Score:5, Informative)
Toshiba is generally credible. It's not like they're Samsung, where they'll just set your blood on fire.
Theranos could detect +200 types (Score:4, Funny)
so Elizabeth was lightyears ahead of the competition.
Re: (Score:1)
They never could detect 200+ types. All the problems with theranos revolved around her. Theranos had a board of directors that made it easy for the company to get funding and they hired some of the brightest engineers and scientists to develop a working machine. The problem with theranos was elizabeth wanted a machine that could both perform 200+ tests on a single blood AND be portable sized that everyone could own in their home. It was that arbitrary constraint on how big the prototypes could be that fucke
Re: (Score:2)
he failed at copying what made steve jobs successful. Steve jobs was about "making things that work"
Steve Jobs did indeed abandon concepts if his idea didn't work, so you're right. However, he was definitely the flair, not the substance guy. For instance, he famously critiqued the board design for the prototype Macintosh purely on the aesthetics [folklore.org]. To his credit, he agreed that if they couldn't make it work, they'd use the "uglier" layout, but his job at Apple wasn't about engineering decisions: it was about what making the product attractive to buyers.
Re: (Score:2)
Theranos was a fraud. The fact that they were a fraud, does not invalidate the possibility of performing meaningful tests on a drop of blood. Correlation is not causation.
Turtlenecks (Score:3)
Re: (Score:1)
Re: (Score:3, Informative)
Did they also issue extra apostrophes to all their possessive pronouns?
it's
means
IT IS
You effectively wrote "black turtleneck sweaters to all IT IS employees."
Say it out loud. Does it sound right? No? Then you wanted its.
Re: (Score:3)
I find it utterly mystifying that so many out there seem to be unable to grasp the elementary principles that unambiguously tell you when to use "its", rather than "it's", and vice-versa. It is DEAD simple:
If you mean "it is" or "it has" then it is "it's"; otherwise, it is "its". What's so tricky about it?
Re: (Score:2, Flamebait)
100% (actually close, you're the exception) let it pass because grammar fascism isn't very useful. Perhaps it was too much coffee, or maybe there's a touch of autism there, not sure.
We all knew what the post meant, and what it said. We glossed the typo, all of us, until you came along. There is no mystery. It was a mistake, but of us all, only you've pointed it out, enraged sufficiently to spend the musculature to point out the absence of an apostrophe. Good for you. Hope you feel better for shaming the lac
Re: (Score:2)
I absolutely know the rules for it's versus its and their versus there but, inexplicably, sometimes type the wrong one. I usually catch it in proofreading but not always. I don't think the misuse necessarily means the person doesn't know the difference.
Re: (Score:2)
I find it utterly mystifying that so many out there seem to be unable to grasp the elementary principles that unambiguously tell you when to use "its", rather than "it's", and vice-versa. It is DEAD simple:
If you mean "it is" or "it has" then it is "it's"; otherwise, it is "its". What's so tricky about it?
I sometimes make the mistake. It's because apostrophe s is also considered possessive.
Fred's widget is correct usage.
It's widget is not, even if "it" is the appropriate descriptor for Fred.
It's a stupid general language rule that should simply not be taught, like "i before e except after c".
https://en.wikipedia.org/wiki/... [wikipedia.org]
Re: (Score:2)
FTFYI find it utterly mystifying that so many out there seem to be spelling naziesthat don't grasp the elementary principles that people simply make a typo ....
Or do you you really thing people don't know the difference between "its" and "it's"???
If you mean "it is" or "it has" then it is "it's"; otherwise, it is "its". What's so tricky about it?
Yepp, and both are correctly spelled words that are not red underlined, for a "whole word reader" a typo he does not spot.
Re: (Score:2)
that just gives me the shits
Re: (Score:1)
Re: (Score:2)
Cost $182 (Score:2)
Cost to the hospital or to the patient?
Re:Cost $182 (Score:4, Informative)
Re: (Score:2)
...in countries where corporate lobbyists control health care.
How many of those are there? I can think of only one.
For profit healthcare (Score:2)
How many of those are there? I can think of only one.
Switzerland is also among the countries where medical insurance is handled by for-profits companies (and despite the country being a direct democracy, the for-profits manage to scare the general population into vote for the companies' best interests).
Of course, on the scale of the US and the general WTFuckery going there (the whole concept of overpriced but negotiable services in hospitals - how could this be a thing in a developed country ?!?), it's a bit ridiculous to complain.
Still it's a drawback when c
Re:Cost $182 (Score:4, Interesting)
Cost to the hospital or to the patient?
I predict that the cost to the patient will be free, for the same reason that many auto parts stores will do a free diagnostic code scan on your car.
Re: (Score:3)
I predict that the cost to the patient will be free, for the same reason that many auto parts stores will do a free diagnostic code scan on your car.
Yeah, the test only needs one drop of blood, but the costs of the prolonged patented medicine treatment will squeeze every last drop of blood out of your body.
Re: (Score:2)
I predict that the cost to the patient will be free, for the same reason that many auto parts stores will do a free diagnostic code scan on your car.
It costs less than a dollar to send a counter monkey with a crap scanner out to your car to read codes, and there's a very high chance that they'll actually convert a scan into a sale, whether you need a part or not. There's also very little liability involved. The OBD-II interface is fairly fault resistant, and there's very little potential to cause harm by hooking up to it. Every time the hospital pokes a hole in you, it puts both the worker and the patient at risk, especially since it's happening in a fi
Re: (Score:2)
If it's literally a drop of blood, there is about 0 risk, and it's a 5 second procedure. Diabetics do finger pricks a dozen times a day often in far worse places than hospitals and don't die from them.
Blood sugar testing is down to a science, and the finger prick sticks are designed to literally get 1 drop of blood out a finger, and no more. $0.10 disposable finger prick stick, alcohol swap, snap, squeeze, capillary tube, alcohol swab, and you're done.
Re: (Score:2)
I predict that the cost to the patient will be free, for the same reason that many auto parts stores will do a free diagnostic code scan on your car.
Yes who knows what they might unexpectedly find that they were not even looking for. It's like a value add! For one of you.
Seriously though, I would love to see this technology become reliable and available. Most people here will still die of cancer or heart disease but we all hope it will be later than sooner.
Re: (Score:2)
>"I predict that the cost to the patient will be free, for the same reason that many auto parts stores will do a free diagnostic code scan on your car."
Possibly, but it is more likely to be low cost because [I assume most] insurance companies want to catch cancer early while it is easier and cheaper to identify and treat. It is the same reason that routine/annual exams are typically low-cost (or no-charge to patient), because it is much cheaper to treat a whole lot of conditions if caught early.
I do loo
Re: (Score:2)
Possibly, but it is more likely to be low cost because [I assume most] insurance companies want to catch cancer early while it is easier and cheaper to identify and treat.
One would think so, but most insurance companies still resist paying for birth control or abortions even though going through pregnancy and birth will cost a couple of orders of magnitude more even without complications.
Re: (Score:1)
Re: Cost $182 (Score:2)
Sensitivity vs Selectivity (Score:5, Interesting)
A test like this has multiple attributes determining its usefulness:
1) Its false-negative rate & confidence. IE, how confidently does a negative result RULE OUT cancer.
2) Its false-positive rate & confidence. IE, how confidently does a positive result INDICATE cancer.
3) How consistent are its results? If you have 5 labs, each one tests each sample 3 times, and you do 8 rounds of sampling a week apart, how consistently do the results of the 3 tests performed by one lab on one sample produce the same result? Ditto, for tests performed by one lab vs tests on the same sample performed by other labs? And from sample batch to sample batch?
4) Cost and time-to-results. A test that's cheap & quick can be performed multiple times at multiple sites on multiple samples. A test that's expensive to perform had BETTER be at least 2 out of 3 (specific, selective, consistent). A cheap & easy test gets a partial pass.
Re: (Score:1)
I can tell you're not a doctor. You actually understand what's going and seem concerned about how this affects patients.
I've seen mass murderers with more compassion than I've seen from doctors.
Re: (Score:3)
Thanks for raising false positive vs. false negative.
If a disease is found in one in 10,000 people, and a test is 99% accurate with respect to false positives, then 99% of positive results will be false. (Assuming that the test is applied to the general population, and not after pre-screening.)
Re: (Score:2)
Thanks for highlighting the issues with false positives and false negatives. Generally referred to as Sensitivity vs Specificity (definitions below).
Generally your specificity decreases the more sensitive your diagnostic / screening test so there's a trade off between missing actual cases and misdiagnosing healthy people.
Best measures for evaluating tests are generally Likelihood ratios (LR) (probability of the person having the disease given the test result) so considers both of the above. LR for a positiv
A Labrador may be cheaper (Score:1)
Re: (Score:2)
Dogs can smell **some** cancers, not very many. IIRC they're mostly skin cancers, which may be emitting volatiles.
Cool, but worrisome (Score:2)
Though only 99% accuracy was mentioned, I assume the diagnostic test's sensitivity and specificity [nih.gov] are about as high, which is much higher than current diagnostic tests available for e.g. breast cancer [getthediagnosis.org] and lung cancer [getthediagnosis.org], where different test methods may be good at either keeping the false negatives or the false positives low.
However, judging by the article's photo of the device [2xx.jp], I find it very worrisome that it's running Windows XP.
Re: (Score:2)
However, judging by the article's photo of the device [2xx.jp], I find it very worrisome that it's running Windows XP.
That's only really an issue if it is connected to a network -- not a problem when air-gapped.
Re: (Score:2)
Re: (Score:2)
Pretty much every medical device that is more than 7 or 8 years old will be running XP, because you're not going to toss a million dollar scanner just because the OS is no longer supported. If modern equipment is using XP it's probably because they need hardware drivers that won't run under the new security protocols of Win 10 or Server 2012. XP was much more forgiving of direct hardware access (more so than NT 4.0 even). If they plan on selling a gazillion of these they'll have the pull to make the hard
I initially misread that. (Score:2)
I thought it meant Toshiba device tested positive for 13 types of cancer.
How is that even possible?
What about the other types? (Score:1)
Re: (Score:1)
Re: (Score:2)
Half empty, huh?
+1
Re: What about the other types? (Score:1)
Re: (Score:1)
Well, it is a step in the right direction, and considering the list of cancers contains at least one that I seem to recall is extra important to detect early on (one could argue that all forms are important to detect as early as possible of course) since the symptoms don't really show until it's almost too late to do anything about it... well, that's a really good thing, right?
If I want to check for the possibility of some sort of cancer because of vague reasons and patient fears, most of the time they will
Re: (Score:2)
What about them? Current methodologies and procedures for determining when to test for them, how to test for them, and so on, haven't all magically disappeared.
This just means that there's now 13 types of cancer that will wind up being easily and routinely tested for, which will result in better outcomes for people that happen to have those types.
Meanwhile, research and development into new tests and treatments continues.
Seriously, what is it you're trying to get at with that comment?
Re: (Score:2)
I think you left a 0 off the number of cancers not tested for.
99%? (Score:2)
So would that be 1 wrong one in 100 healthy people? Or 13 wrong ones in 100 healthy people?
Does not sound very useful if it is the 2nd case....
Riiiight (Score:2)