An anonymous reader writes "BBC news is reporting a new handheld scanner that can detect cancers in patients. In clinical trials at a hospital in Milan, the scanner was able to correctly identify 93% of prostate cancer patients whose condition was later confirmed by a biopsy operation."
What about false positives?
by
DustMagnet
·
· Score: 4, Insightful
93% sounds great. Even the 66% for breast cancer is amazing. What the article doesn't mention is false positives. I could identify 100% of all cancers if you let me have unlimited false positives.
-- 'SBEMAIL!' is better than a goat!!
Re:What about false positives?
by
DustMagnet
·
· Score: 4, Informative
The article in New Scientist has some more details, but still no information on false positives.
-- 'SBEMAIL!' is better than a goat!!
Re:What about false positives?
by
anthony_dipierro
·
· Score: 3, Funny
I could identify 100% of all cancers if you let me have unlimited false positives.
I could identify 100% of all cancers with zero false positives if you give me a stick of uranium.
Re:What about false positives?
by
Smidge204
·
· Score: 3, Insightful
Obviously, a more traditional method of diagnosis is still required.
What I didn't notice is if the 34% that weren't a success were shown to have cancer and did not, or shown to not have cancer and did.
Taking a "better safe than sorry" approach, if the scanner says you're positive for cancer, then go ahead and get a biopsy even if you're actually healthy. But if the device misses the cancer, then it really can't be relied upon.
Although 94% is pretty good. I wonder what the success rate of biopsy diagnosis is...
And am I the only one who thinks it's a bit odd that they can't get their results published in a medical journal? =Smidge=
Re:What about false positives?
by
Sgt+York
·
· Score: 4, Informative
I wonder what the success rate of biopsy diagnosis is...
Biopsy is nearly 100% effective; the false positive/negative risks are low in the hands & eyes of a good pathologist. The hard part is knowing what to biopsy. Some other test has to be done beforehand to show you where the putative cancer is.
Even tumors ID'd with this device would need to be confirmed via biopsy, MRI or CAT scan. Most likely MRI followed by biopsy.
As for publication, they may not have submitted it yet (need more data, timid researchers), or reviewers haven't finished their reviews yet, or the editor is sitting on it to publish it alongside another paper, or the BBC could be blowing the data out of porportion (the latter is the most likely. Happens all the time.)
--
There is a reason for everything. Sometimes that reason just sucks.
It is interesting to note that this machine works by means of detecting solid tumors, alas the vast majority of cancers (prostate, breast, lung, colon) have a solid tumor component, but i wonder if one of these machines could be used to pick up, say: muliple myeloma. Or leukemias...
Also it would be interesting to know if it could be adapted to detect other disorders of normal tissue via changing microwave resonance: pulmonary fibrosis, alzheimer's disease, et al.
Good ol' BBC didn't even mention the false positive rate either...
-tid242
--
With a few exceptions, secrecy is deeply incompatible with democracy and with science. --Carl Sagan
Re:Let's just hope the device isn't *causing* canc
by
GiMP
·
· Score: 3, Informative
Only at frequences that resonate with water (2.4ghz) like your microwave oven or your WiFi card. In the case of the WiFi card it isn't a big deal since the power is very low.
Re:Let's just hope the device isn't *causing* canc
by
Sgt+York
·
· Score: 3, Interesting
They are, but the doses here aren't high enough to hurt you.
This reminded me of something I heard years ago, in college. A guy from a cancer research organization (Stehlin) came by & talked to our class about cancer treatments. One of the investigative techniques focused on microwaves.
Apparently, it's more damaging to cancer than normal cells (has to do with plasma membrane composition). If exposed to low levels for a fairly long period of time (several minutes to an hour a day for several weeks), the tumor would dissolve and the local normal tissue would remain. The main problem they had at the time, he said, was that the cancer cells tended to lyse all at the same time, and the subject would go into lethal toxic shock, a prety bad way to go(this was working in mice.)
This was a while ago, 1990 or 91. I haven't heard anything from them about this stuff in a while, so it probably fell through.
--
There is a reason for everything. Sometimes that reason just sucks.
Biopsy still necessary in many cases
by
Metameme
·
· Score: 4, Insightful
IANAD, but my wife's currently on her third bout with lymphoma (NHL once, Hodgkin's twice) so I have researched this pretty heavily.
Biopsied tissue is not subjected to a pass/fail test for cancer. Often the excised tissue is tested against different combinations of chemotherapy drugs to see which the tumor(s) will respond to best. In addition, a biopsy can help determine what specific type of cancer cells are present; knowing that a patient has lymphoma says very little about rate of growth, type of treatment indicated, etc. There are several types of lymphomas, both Hodgkin's and non-Hodgkins. Even knowing the difference between those two categories is helpful as non-Hodgkin's can be treated with monoclonal antibodies whereas Hodgkin's cannot.
This device, if it lives up to the hype, can still help.
First of all, we've had a lot of failed biopsies because they took tissue from areas that were benign and had to go back for more tissue in different areas. If the device could determine *where* the cancer is, it would make a biopsy more effective the first time around and prevent a lot of needless suffering and expense.
Secondly, many patients go through a staging surgery to determine the progression of the disease after the initial diagnosis. In lymphomas, the stages are ranked as I-IV, with an additional A or B tag. Simply knowing the cancer is both above and below the diaphragm in multiple sites will tell you that the patient is stage IV. The A/B tags represent whether there is organ involvement (B) or none (A).
Additionally, remission is often diagnosed through the results of CT and PET scans, which are not entirely conclusive but much less traumatic than surgery for additional biopsies, especially considering the effects of chemotherapy on the immune system and the healing ability of the body.
Hope this helps clear it up; anyone in the know please correct me if I'm wrong on any of this...
Slashdot Mirror
Damnit Jim, I'm a doctor, not a bricklayer!
93% sounds great. Even the 66% for breast cancer is amazing. What the article doesn't mention is false positives. I could identify 100% of all cancers if you let me have unlimited false positives.
'SBEMAIL!' is better than a goat!!
Also it would be interesting to know if it could be adapted to detect other disorders of normal tissue via changing microwave resonance: pulmonary fibrosis, alzheimer's disease, et al.
Good ol' BBC didn't even mention the false positive rate either...
-tid242
With a few exceptions, secrecy is deeply incompatible with democracy and with science. --Carl Sagan
Only at frequences that resonate with water (2.4ghz) like your microwave oven or your WiFi card. In the case of the WiFi card it isn't a big deal since the power is very low.
This reminded me of something I heard years ago, in college. A guy from a cancer research organization (Stehlin) came by & talked to our class about cancer treatments. One of the investigative techniques focused on microwaves.
Apparently, it's more damaging to cancer than normal cells (has to do with plasma membrane composition). If exposed to low levels for a fairly long period of time (several minutes to an hour a day for several weeks), the tumor would dissolve and the local normal tissue would remain. The main problem they had at the time, he said, was that the cancer cells tended to lyse all at the same time, and the subject would go into lethal toxic shock, a prety bad way to go(this was working in mice.)
This was a while ago, 1990 or 91. I haven't heard anything from them about this stuff in a while, so it probably fell through.
There is a reason for everything. Sometimes that reason just sucks.
IANAD, but my wife's currently on her third bout with lymphoma (NHL once, Hodgkin's twice) so I have researched this pretty heavily.
Biopsied tissue is not subjected to a pass/fail test for cancer. Often the excised tissue is tested against different combinations of chemotherapy drugs to see which the tumor(s) will respond to best. In addition, a biopsy can help determine what specific type of cancer cells are present; knowing that a patient has lymphoma says very little about rate of growth, type of treatment indicated, etc. There are several types of lymphomas, both Hodgkin's and non-Hodgkins. Even knowing the difference between those two categories is helpful as non-Hodgkin's can be treated with monoclonal antibodies whereas Hodgkin's cannot.
This device, if it lives up to the hype, can still help.
First of all, we've had a lot of failed biopsies because they took tissue from areas that were benign and had to go back for more tissue in different areas. If the device could determine *where* the cancer is, it would make a biopsy more effective the first time around and prevent a lot of needless suffering and expense.
Secondly, many patients go through a staging surgery to determine the progression of the disease after the initial diagnosis. In lymphomas, the stages are ranked as I-IV, with an additional A or B tag. Simply knowing the cancer is both above and below the diaphragm in multiple sites will tell you that the patient is stage IV. The A/B tags represent whether there is organ involvement (B) or none (A).
Additionally, remission is often diagnosed through the results of CT and PET scans, which are not entirely conclusive but much less traumatic than surgery for additional biopsies, especially considering the effects of chemotherapy on the immune system and the healing ability of the body.
Hope this helps clear it up; anyone in the know please correct me if I'm wrong on any of this...