Image Database Idea

[MrPink]

Purpose:
To allow users to access X-ray diffraction images

(pictures of x ray patterns - you can think of it like going to the hospital and getting an xray of a broken bone and want it to be down loadable by others)

Need:

Start of a bulletin board discussion the topic

Quotes:

Garib
"I do not know any reason (apart from technical one - size of files)
why images should not be deposited and archived. I think this problem
is very important."

Herbert
"This is a good time to start a major crystallographic image archiving
effort."

Kay
"...access to images sometimes may help to answer a scientific question."

Graeme
"It does seem clear to me though that we cannot set up the perfect repository in the first instance, but we do have to start
somewhere. Perhaps we do not need the right answer, but one which is less wrong
that not making available the data at all?"

etc...
Is it fastlane?
Scalable:
The current rate at which X-ray diffraction images are being generated.
In short, the number of deposits are exponentially growing every year.
Although not shown the user number is also greatly increasing.

Solution: (?)
Using Amazon S3 to store the estimated 318 terabytes of data (assuming I can get it all).

Design an easy way to search the images. The depository already has a method for generating a unique ID number for each dataset so this may not be that hard.

Cost:
(the costs become cheaper for more data, but am simplifying them here)
0.15 GB stored/month
0.10 GB transfer in
0.17 GM transfer out

Currently, if you had all 318 terabytes you would looking at:
7500/month in storage.

I do not how much the transfer in and out would be
Therefore set up a payment that would work regardless

Income:
Free: deposition and transfer in
(this would allow the database to grow quickly)
I would follow this up by directly contacting users to deposit their information

Transfer out: 6 month fee of 19.95 for 10 transfer outs, which should be enough for most users

Profit:
Hypothetical if I had all currently available x-ray images (50k)
318 terabytes: -45000/ 6 months
Transfer in: 0.10*5000 = -5000
Transfers out: 2600 users 0.17*10 = -4420
Total: 54420

2600 users at 19.95 = 51870
----
I am totally making up the fee structure off of the top my of head (which makes a huge difference), but wanted to 'put it out there' for any feedback.

Thanks as always.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[MJ DeMarco]

I like the idea and it seems the need exists -- however, not having intimate knowledge into the industry operational nuances prevents me from feeling confident in any diagnosis. The best revenue stream would be B2B -- income earned from the medical provider -- like a yearly membership fee which gives their patients access to the records for free (plus other medical providers.) Also, if this system is perceived to add free time to medical practitioners, they will pay for it.

 

[MJ DeMarco]

Also, data storage is really cheap so I'm not sure if you'd want to outsource that -- having control over these files would seem to be important.

 

[MrPink]

Time:
I did not address this, but yes it would save time. The current system involves emailing the author of a paper that has the images, sending them a DVD to copy the images and having them mail it back.

Revenue:
I initially thought of revenue as allowing individual users to sign up with a set number of downloads (such as 10). The idea could be expanded to B2B in a sense by allowing Universities or companies to purchase a large download number.

however, not having intimate knowledge into the industry operational nuances prevents me from feeling confident in any diagnosis.

I have been learning that being in a niche can be a good thing

Also, data storage is really cheap so I'm not sure if you'd want to outsource that -- having control over these files would seem to be important.

I am definitely hurting as far as knowledge about data storage. The reason for outsourcing (using Amazon S3) is initially very low cost compared to buying my own data storage devices (this is based on looking at various websites for an hour so could be off on this).

Thanks for the input, I think the next step is to write out how the site should function and what it should include.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[LightHouse]

Yeah doing it on S3 would be an option is you had serious transfers going on. You could store the images for cheaper than that i imagine. with only 2600 users i can't see your data transffers being super high all the time which is what you would use amazons servers for because whatever else you had would probably crash.

Take a look around at other options, also see if you can get interest on differetn price points and see if there is a money value on the actual need. Then you can better estimate the sales vs the expenses.

 

[MrPink]

Yeah doing it on S3 would be an option is you had serious transfers going on. You could store the images for cheaper than that i imagine. with only 2600 users i can't see your data transffers being super high all the time which is what you would use amazons servers for because whatever else you had would probably crash.

I would be providing data that results in the protein data bank, which according to their website is used by "more than 150,000 scientists from more than 150 countries per month". I believe there will be a drop off in the number of scientists needing the images. I will have to think of how to best determine the actual number.

Take a look around at other options, also see if you can get interest on differetn price points and see if there is a money value on the actual need. Then you can better estimate the sales vs the expenses.

I will work on this, thanks.

 

[TC2]

I just had this thought, but not sure if it will apply to your niche.

With my understanding, Amazon is pretty friendly with non-profit organization. If you can somehow make your business as non-profit organization, Amazon may be able to donate a lot of bandwidth and storage space for your business.

Just a thought!

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[MrPink]

I believe there will be a drop off in the number of scientists needing the images. I will have to think of how to best determine the actual number.

Too bad I didn't have a way to poll other crystallographers (get an idea of a head count of users)... oh wait, my blog

 

[WildFlower]

You know what is going to come.. it's just a matter of time.. they already have the technology for facial recognition.. which is currently being used for security... imagine a search engine for facial recognition.. you put in a facial picture of someone and it searches the images on the internet for that person.... jeez.. I hope you didn't post naked pictures of yourself for the engine to find! LOL LOL

 

[Jonleehacker]

Seems to me there would be privacy issues having patient x-rays on the cloud. Besides that it's a great idea especially if the patients need to have the images looked at by more than one provider.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[WildFlower]

It's been around for a couple years now. Facebook has tested it, there are a couple Flickr apps, Google is working on it with Picassa, there are several mobile phone apps that do this, dating sites have tried it out, etc.

I remember seeing some beta technology two years ago from a company called ActiveSymbols that was actually quite good...I wouldn't be surprised if the technology were pretty turnkey and ubiquitous throughout social networking sites in the next 2-5 years...

That's funny.. I've been saying that for over 10 years.. I saw all these people.. some I knew.. posting pics.. with their faces in them.. sigh... Also... video cam.. HELLLLOOOO.. you really think that is private and will stay private!! To those close to me like my friends.. I would warn them and then let it go... I said my peace. LOL ... It's sad when I see teenagers posting pics they will soon regret.

 

[Inphinity]

Hmm how much competition is there in this area in the US at the moment? How hard would it be to break in to?

I don't know how many companies here are doing this but it has become pretty popular in a short space of time among medical practitioners, and at this stage it is limited to medical access not patients / public.

 

[MrPink]

Funny to see this post come alive again. I figured out that there was a way to covert the X-ray images into text files, which could be emailed. The need for long term storage of these images is still there, but felt that the transfer component would have been the way to have generated profits.

The result being that it killed the business idea.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.