Is there an optical viral scanner yet ? (computerized detection of viral agents that can identify which viruses are in the blood ?) If not: I came up with a simple idea (maybe).. Multi angle OCR..using the different viruses as 'characters'.. Then when scanning for the viruses, you'd change the angle through each axis by 30 or 45 degrees.. First one axis until you go all the way around, then you 'd go +30 or 45 degrees on the second axis and go all around on the first again..this would yield 64 or 144 different angles of the virus, of which at least one would be OCR-able.. When blood would be scanned a stereoscopic optical picture would allow for 3D recreation of any stuff in the blood.. Since everything is either known or not (bloodcells, viruses, bacteria) it would also allow for unknown detection.. (if it's NOT known, it MUST be an unknown..) In any case, if it does not exist, seemed useful to post my idea here.. (Feel free to flame, comment, ignore or use...idc..just watched a Helix, and it always appears like those ppl are always using like 70's movie tools..)
rather than flaming you or ignoring you, I would applaud you but ask you to develop this idea further. For example viruses are very very small, much smaller than a cell or even an organelle, they are as small as the DNA of the cell, how would you distinguish where these viruses are? How wold you identify them? What kind of sensor are you planning to use? An active or a passive sensor? For example a novel technique currently in development is attaching nanoparticles to viruses, thus tagging them and than using the sensor technology to locate these tagged viruses. http://phys.org/news7264.html What are your thoughts?
Please Register or Log in to view the hidden image! nanotaggings ? But then you already know that they are there ? My idea is was a base for detection where you had to separate the infected persons from the un-infected ones.. The 'sacnning' would be done with steroscopic opticals..under a microscope.. Or rather, the machine would take a sample of the blood, clear out the larger bloodcells, so only the plain plasma remains.. Then it would take a drop and put that on a slide..then it would use a regular, but powerfull microscope (stereoscopic) to create a 3D image.. That 3D image would then be rotated through two axis..and create either 64 or 144 different 2D aspects.. Each of the 2D aspects would then be compared to the database, similar to OCR (Optical Character Recognition).. I'm thinking high speed processing of large amounts of bloodsamples... (/w powerfeeder..lol..since the actual comparison would take less time than the sifting of the blood..) (umm like a processor pre-processes ?....ah yes..pipelined processing ?..as well as autonomous viral detection, which could be useful in large scale outbreaks....) (maybe even remote scanning ? for CDC ?..possibilities on safety are endless..no human error or fraud...and with the latter I think of ppl forging their tests..) But the idea is simple no ? (also..since I came up with the idea, I also found a way to scan the samples even better..) (when the microscope is zoomed in, moving the the slide around like an old time scanner..) (but instead of mechanical movement, which is rather shaky, one could use a bimetal arm..) (the slide is held by the bimetal arm, in a constant temp...with two bimetal junctions, one could create really small motions..once again, through two axis..) Even seperate recognition routines are now very viable for each virus.... I thinks some look like spaceships..weird.. Checkout bacteriophage..looks like little lander..kinda scary...I'd almost say they are similar to nanotech..
The parameters for specificity -- especially with viruses is so marginal. It would be like finding a needle in a haystack. Your idea is plausible, yes. But the specificity isn't there -- we lack the instrumentation. I would like you to develop your idea further, because the suggestion is just flat out cut and dry impossible under normal conditions.
wait... I'm talking microscope.. I'm pretty sure they exist.. And I know there's 3D imaging... Reconstruction of a 3D image from a duoscopic image, if need be from two sides. (making it quad scopic..) And the rest is plain computer software.. So..what does NOT exist ?... Please Register or Log in to view the hidden image!
"Plain computer software" The software you describe is meant to give a caliber number of what we call a "viral load" or the amount of virus per a set volume of blood (or any other fluids). Due to the viral load we are able to give a calibrated scale that helps for viral diagnosis. Now, what you described (at least I think) is a sort of measurement of singular viruses - NOT a viral load. Now, if we know what virus we are looking for - this isn't so impossible. If we are looking for a singular strand of a mystery virus we are talking nearly impossible. I would like you to be more specific with your research parameters, if you don't mind.
Please Register or Log in to view the hidden image! Edge 3D microscope capability: Combine this 3D microscope with this technology that can see objects smaller than 50 nm (with viruses at 20nm): http://laser.mace.manchester.ac.uk/...011-08-02_11-04-27_LFW-microscopy_feature.pdf
T Try being accurate within 50nm, capability isn't everything. It would still be very difficult and impractical even with this technology. That's more or less what I meant by my earlier post. How exactly are we to determine the exact location down to 50nm -- it would be very hard would it not? If not, please explain.
well like I said earlier go with nano-taggings and after application see the fluorescence, zoom in on that.
I wouldn't be so picky...I'd attach a whole database of viruses...if it would come up with something that is NOT in the database, it could also be used to find new viruses and/or mutations.. There's one thing a computer can do...and that is comparing lots of stuff with lots of other stuff.. Also..if the 3D microscopes you have there are NOT strong enough..well..I'd stick with several electron microscope images.. (at diff angles..but fixed each scanning at their own angle..) While they'd scan, the computer starts processing the data..I'm thinking Linux chain (remember the data does not have to be processed locally, with each detector high bandwidth availability would be neccesary in that case)..or maybe by the time this would be applied maybe a multiprocessor machine, or maybe a dedicated card..(in order to create 3D images from the performed electron scan..) With each viral particle you'd get a count of how many of each in said drop....and also a count for 'unknowns'.. 'unknowns' would be the variations, mutations and new viruses..(or do you say virae when it comes to plural..I dunno..) Do you know how an old timer optical scanner works ?..combine that with the electron microscope..(remember the double bi-metal jointed arm..) You'd scan the ENTIRE specimen..the computer will determine what is and what is not..(plain fluid won't matter..preprocessing of data by the local machine could lessen the flow of data, if needed..but also..remote operation would allow for alternate handling of any specimen and/or it's data..like a 'raw' mode..) In any case..all that in concert with power loader..place in samples on one end...computer spits out results...it could even sticker the vials for later reference.. Like I said..there's a lot of stuff a computer can do for you...specially when it comes to searching for stuff... If that still wouldn't be enough you could isolate the sample and STILL check by hand if need be.. When you leave the database open, you'd also be able to add new entries when found, thus removing the 'unknowns' more and more with use.. ofc.starting with an empty database would be an option, but is not recommended..the work alone would kill you I thinks.. (and sorry for the mixing of both reply posts..)
On a personal note..I don't know anything much about virology or micro biology....I'm more of a physics kinda guy.. You tell me what it is that you'd think might be the problem, and I could think up a way of either going around or just fixing the problem.. I tend to see more practical solutions..I like solving problems with those... (also..I'm a decent mnemonic programmer..yes old style..but in some cases way better than what ur average standard programming language spits out as 'useful' code...single environment is so good..) Anyways..here's hoping it helps.. If not..post back here plz..I'll see if I can add more detail.... - LaterZ..
I have also come up with a way to make the bimetal joints react faster..so resetting the microscopes won't take too long..) First it's more algorithmic than actual 'return' to position zero.. One sample scans in positive temp..(the bimetal going one way...) The second scan is in reverse..so..negative temp.. don't try and get it to move faster by using coolants..as a result it might lessen it's lifespan.. one important factor..any tremors ?...both the scanning part and the slide oughta be mounted on a single mount..that way any distortion of the image would be kept to a minimum..individual electron sweep image comparison and error correction can be left to the software.. - LaterZ.. Also..don't be disturbed by the fact that the data provided in it's first stage would be as much as a TeraByte.... (roughly 1.100.000.000.000 x 8 bits of data..)..it would be completely unprocessed.. The first thing the computer would have to do is to pack together the relevant data.. (omitting the liquid itself maybe..?..) I need to know one thing though... What is the actual amount of viral particles that are usually found in a slide ? Do you guys still use droppers ? or for that matter..slides ? Or do you guys use like a dot of specimen ? (like 0.01 milliliter or something ?) - is clueless...
anyways..for the comparisons I picked a random 2D image from the net.. Please Register or Log in to view the hidden image! Like so.. Given a guesswork number of a million viral particles in the slide..that would make either 64x6 Kb or 144x6 Kb per particle.. So...384 GB's or 864 GB's..for images like you can see above.. I'm thinking for actual comparisons the computer has much better vision.. So it could be (practical unknown) that the actual needed info could be stored in smaller images.. For a picture one quarter of this size, it would take much less.. 24 Gb's (for the 64 angles) or 54 Gb's (for the 144 angles..) In any case..for each million of viral particles multiply the info times these numbers.. (in case there's less..well..GUUUD..) FYI..that amount of comparison could be done a roughly 10 secs or less, I thinks.. Depending on which algorithm used..(certainly not standard software.. that kinda sucks when it comes to speed.... So..if the scanning sequence takes up to 10 secs per specimen, you'd have roughly the same amount of output as input...and rough throughput of 6 specimen per minute.. But for now that's guesswork.. : D
crap..I forgot one factor.. The images that are checked need to be checked vs. EVERYTHING in the database... But also..classification from most common to the most uncommon, can quickly speed up that.. Since most stuffs found it the specimen will be common, it will be found relatively early on.. How many diff types are there anyway ? - clueless again.. I'm reading stuff like 1 million different ones ? That puts a whole new dent in my plan... But still even if that is so..(and how the hell do you guys memorize 1 million different viruses ?!?) - has broad new respect for the homo encyclopedeus.. (The human encyclopedias..) ..omg.. ..umm.. I hope there aren't that many common ones in each person... If that is true..well..plan scrubbed..unless you got a linux chain of 10.000.. In which case I'm sorry for wasting ur time.. Cheap secondhand Cray maybe ? I really need some factual numbers.. Questions: How many different viruses are there in most human bodies at any given time ? How many different viruses are there in all that are possibly be in ANYONE ? For the common ones there would be processors with only a couple of the common ones.. For the more exotic ones there would be a few machines with lots of those each.. And although the they would take longer to produce actual results, it wouldn't matter.. since they'd have much less frames to compare.. In any case..this machine would be a tool ofc...any really important results would have to be checked after the conclusions of the computer.. (like HIV positives and stuff like that, if and when located by the computer..I don't trust computers to think up perfect results..not because they can't, but simply because they can't do anything we don't program to...but even in such a case..still helpful I thinks..)
The human body has a total of 100 trillion cells (human body cells + microbes + etc) Only 10 trillion cells in human body are human cells (containing 21,000 genes number total) There are 10,000 species of microbes in the human body (containing 8 million genes total) There are more than 100 trillion viruses of one type in a human body during an infection (flu for example) A virus is made of RNA or DNA and is surrounded by protein/lipid/host cell...thus complicating its identification within the body (let alone the afact that it is 20nm) Normal bacteria flora of human body: http://textbookofbacteriology.net/normalflora.html Pathogenic bacteria flora of human body: http://en.wikipedia.org/wiki/Pathog...thogenic_bacteria_by_clinical_characteristics Human pathogenic viruses list: http://viralzone.expasy.org/all_by_species/678.html Current problem: Majority of viruses and bacteria genomes have not been mapped. Another problem: Viruses mutate their DNA and RNA, these mutations have to be identified and either the future simulation of such mutations need to constructed by a computer to predict these mutations or closest match to its original mutation state has to be found. Solution: The database of all viruses and bacteria genomes (pathogenic and non-pathogenic) as well as human genome would require the use of cloud computing for server access to a database on the go as soon as the results of the 3D microscope come in. References: 1) http://www.npr.org/blogs/health/2012/06/13/154913334/finally-a-map-of-all-the-microbes-on-your-body 2) http://www.livescience.com/topics/virus/
Problem is that lots of viruses would be difficult to detect without mapping parameters, which is what I was trying to say. You bring up another point I hadn't thought about in mutations, PM would make it difficult to tag their genetic signature for definition. Information would have to be updated almost constantly, or, like you said, imported into an ever changing cloud. I think that this technology is new and incredibly interesting - but doesn't our current method seem more logical for the technological barriers? I will say, though, that this will be very interesting for virological research. Certainly an exciting idea
Sorry for the late reply...kinda busy... ..but... I think nowadays there's still only the visual cortex that can discern one virus from another.... What I was thinking of, that of instead the manual (human eyesight) doing the discerning..a computer eye..(put basically..) And about the different virae, a database...with shape of each virus.. Correlating would consist out of turn a stereoscopic view into a 3D image, and as such comparing it to that database.. (making a lot less work for the cortex..as well as speeding up detection (pos/neg) on certain ones..) (but I thought I said that in the start ??..) (the way to fill the database with any item instead of a full 3D image, there'd be 2 stereoscopic still images..allowing for image 3D restoration..instead of storing the entire atomic structure of said virus..) (well..in case of multi aspected ones, it would ofc need more than a single pair..but even so..much less data then a full atomic breakdown of each of said virae..)
Viruses are ever changing, the database would have to be massive and changing constantly. This, coupled with the specificity of location would make this method very shaky. I understand the application for virology -- but I can't see any sort of practical use for it beyond that.
