Difference between revisions of "Talk:Open Source Medicine"
From AdCiv
(→Open Medical AI) |
|||
| (11 intermediate revisions by 2 users not shown) | |||
| Line 1: | Line 1: | ||
| − | + | ==The biggest killers== | |
| − | : | + | Cancer, cardiovascular disease and stroke are already dealt with in the article. It would be nice to get some mention of these: |
| + | *'''AIDS''' - | ||
| + | **Vaccines: Maybe create a new section on vaccine design and production. HIV is trickier than any other virus to find a cure or vaccine against. It mutates more than any other virus and though there are antibodies against it, they are effective against only a few strains. Vaccines have been proven that protect against 40% of strains; this is not clinically useful, but proves that a HIV vaccine is possible. 'Broadly neutralizing antibodies' that are effective against 91% of strains have been found [http://singularityhub.com/2010/07/22/antibody-neutralizes-91-of-hiv-strains-strong-step-towards-a-vaccine/], and a cocktail of antibodies could be 100% effective. We have identified a site on the virus that doesn't mutate - the viral spike glycoproteins - the antibodies that latch onto that site, and the genes that generate these antibodies. The trick now is to develop something like a bacterium that resembles the viral spike glycoproteins yet elicits a strong immune system response. This will train the immune system to attack the spike, hence to attack all strains of HIV. Attaching it to a bacterium would mean it could be cultured quickly and cheaply. (Incidentally, the research methodology pioneered in this quest for an AIDS vaccine might also be used to develop a universal flu vaccine [http://singularityhub.com/2009/03/05/new-antibody-breakthrough-paves-the-way-for-near-universal-flu-vaccine/]. This might be relevant to the [[Pandemic preparedness]] page.) | ||
| + | **NK cells | ||
| + | **Proteomics??? The proteome of HIV has been mapped, but I think no one yet knows whether it could be controlled with proteomic interventions. --[[User:Balatro|Balatro]] 10:31, 18 June 2011 (CEST) | ||
| + | **Education is probably the real answer, at least until a cure or vaccine is found. | ||
| + | **Nanobots could, of course, be programmed to search-and-destroy, but they are rather far off in the future | ||
| + | *'''Malaria''' - One of the most interesting ways to control malaria, and certainly the cheapest, is with permaculture. A bat-house will control mosquitos<sup>[http://www.batcon.org/index.php/media-and-info/bats-archives.html?task=viewArticle&magArticleID=397]</sup>, planting plants rich in citronella oil around human habitations will repel mosquitos and growing the ''Polyporus umbellatus'' mushroom is effective against the parasite itself. | ||
| − | = | + | == Interesting links == |
| − | + | * http://www.nytimes.com/2010/08/13/health/research/13alzheimer.html?_r=1 | |
| − | + | *http://www.openclinical.org/ | |
| − | + | ||
| + | ===Proteomics=== | ||
| + | *http://www.ted.com/talks/david_agus_a_new_strategy_in_the_war_on_cancer.html | ||
| + | *http://www.ted.com/talks/danny_hillis_two_frontiers_of_cancer_treatment.html | ||
| + | ===Regenerative medicine=== | ||
| + | *http://www.youtube.com/watch?v=tijEl8I38mo&feature=related | ||
| + | ===Imaging=== | ||
| + | *http://www.youtube.com/watch?v=pqTkuaAykBw | ||
| + | ===Vaccines=== | ||
| + | *http://www.ted.com/talks/seth_berkley_hiv_and_flu_the_vaccine_strategy.html | ||
| + | |||
| + | ==Open Medical AI== | ||
| + | *Communicates with patient in natural language | ||
| + | *Turns every patient into a data point - bridges gap between clinic and research. | ||
| + | **Update, Feb 2013: This is happening! "Watson has the power to sift through 1.5 million patient records representing decades of cancer treatment history, such as medical records and patient outcomes, and provide to physicians evidence based treatment options all in a matter of seconds." [http://www.kurzweilai.net/watson-provides-cancer-treatment-options-to-doctors-in-seconds] | ||
| + | *Accesses medical journals (with Natural Language Processing) | ||
| + | *Interprets scans (with machine vision) | ||
| + | *Analyzes test results, like proteomics, genomics, blood tests | ||
| + | *Simulations of biochemistry, proteomics | ||
| + | *Analyzes small, wireless sensors | ||
| + | *Makes decisions (with Bayesian logic, expert systems, machine learning) | ||
| + | *Integrated with an Electronic Medical Record system | ||
| + | *All this done by cloud computing | ||
| + | |||
| + | http://www.kurzweilai.net/ibm-to-collaborate-with-nuance-to-apply-watson-analytics-technology-to-healthcare | ||
| + | |||
| + | Unfortunately, most of the development in this field is being done by private companies; there is not yet a dynamic open-source project. There was EgaDSS, but it seems to have stalled out. There is an [http://www.xprize.org/press-release/x-prize-foundation-and-qualcomm-join-forces-develop-competition-enhance-integrated-digital X-Prize] with a $10 million purse to stimulate medical A.I, but this will lead away from open-source. | ||
Latest revision as of 22:23, 11 February 2013
Contents
The biggest killers
Cancer, cardiovascular disease and stroke are already dealt with in the article. It would be nice to get some mention of these:
- AIDS -
- Vaccines: Maybe create a new section on vaccine design and production. HIV is trickier than any other virus to find a cure or vaccine against. It mutates more than any other virus and though there are antibodies against it, they are effective against only a few strains. Vaccines have been proven that protect against 40% of strains; this is not clinically useful, but proves that a HIV vaccine is possible. 'Broadly neutralizing antibodies' that are effective against 91% of strains have been found [1], and a cocktail of antibodies could be 100% effective. We have identified a site on the virus that doesn't mutate - the viral spike glycoproteins - the antibodies that latch onto that site, and the genes that generate these antibodies. The trick now is to develop something like a bacterium that resembles the viral spike glycoproteins yet elicits a strong immune system response. This will train the immune system to attack the spike, hence to attack all strains of HIV. Attaching it to a bacterium would mean it could be cultured quickly and cheaply. (Incidentally, the research methodology pioneered in this quest for an AIDS vaccine might also be used to develop a universal flu vaccine [2]. This might be relevant to the Pandemic preparedness page.)
- NK cells
- Proteomics??? The proteome of HIV has been mapped, but I think no one yet knows whether it could be controlled with proteomic interventions. --Balatro 10:31, 18 June 2011 (CEST)
- Education is probably the real answer, at least until a cure or vaccine is found.
- Nanobots could, of course, be programmed to search-and-destroy, but they are rather far off in the future
- Malaria - One of the most interesting ways to control malaria, and certainly the cheapest, is with permaculture. A bat-house will control mosquitos[3], planting plants rich in citronella oil around human habitations will repel mosquitos and growing the Polyporus umbellatus mushroom is effective against the parasite itself.
Interesting links
- http://www.nytimes.com/2010/08/13/health/research/13alzheimer.html?_r=1
- http://www.openclinical.org/
Proteomics
- http://www.ted.com/talks/david_agus_a_new_strategy_in_the_war_on_cancer.html
- http://www.ted.com/talks/danny_hillis_two_frontiers_of_cancer_treatment.html
Regenerative medicine
Imaging
Vaccines
Open Medical AI
- Communicates with patient in natural language
- Turns every patient into a data point - bridges gap between clinic and research.
- Update, Feb 2013: This is happening! "Watson has the power to sift through 1.5 million patient records representing decades of cancer treatment history, such as medical records and patient outcomes, and provide to physicians evidence based treatment options all in a matter of seconds." [4]
- Accesses medical journals (with Natural Language Processing)
- Interprets scans (with machine vision)
- Analyzes test results, like proteomics, genomics, blood tests
- Simulations of biochemistry, proteomics
- Analyzes small, wireless sensors
- Makes decisions (with Bayesian logic, expert systems, machine learning)
- Integrated with an Electronic Medical Record system
- All this done by cloud computing
Unfortunately, most of the development in this field is being done by private companies; there is not yet a dynamic open-source project. There was EgaDSS, but it seems to have stalled out. There is an X-Prize with a $10 million purse to stimulate medical A.I, but this will lead away from open-source.
