CellUniverse - guidelines for new members

About

NOTE: This page is a daughter page of: NCMIR

This page represents a set of guidelines I typed up in 2011 for people interested in joining the CellUniverse / CellSculpt project. The CellSculpt project sadly never really got off the ground, because we lacked resources and I was offered a job in industry late 2012. I still have hopes someone might take over, but I've copied this page here so I have a permanent record of the guidelines I typed up and was very useful in mid 2011 when I had three great intern students do a little work for me over the summer. This page is currently restricted, but I may make it public later in case other people like the way I've written guidelines for interns to help them get oriented, be useful and know their expectations before they commenced work for me.

• NOTE: This page is a duplicate of a restricted access page I typed up here.

This page represents a guideline for anyone joining or thinking about contributing to the CellSculpt team at the The National Center for Microscopy and Imaging Research (NCMIR) in San Diego. I started this page after finding out several students were interested in doing a 2011 summer internships on the game - however the guidelines here also apply to people thinking about joining the project longer term. The contents of this page will evolve over time.

Basic Guidelines

About CellSculpt

If you don't already know about CellSculpt I suggest you start by surfing the pages of this website... as it says on the front page:

Welcome to CellSculpt, an experimental multi-player online game where you can contribute to important research by tracing images of microscopic cells which can't be traced accurately by computers. This game isn't just fun for all ages - it's a "game with a purpose" - every line you trace helps us advance scientific knowledge! In addition to helping scientists make new discoveries in cell biology, players can earn certificates and other rewards, as they level up, learn and build their own virtual research community.

Quick Background / History

Work on the game began at the start of 2011, but the idea of crowd-sourcing the job of tracing is something Mark Ellisman, the founder and director of NCMIR, and many others have been tossing around for decades. You'll notice on the "Design Team" page a long list of people who have already contributed. Don't let this page fool you - currently (updated April 2011) it's only me (Andrew) working on this game full time and Jeff Bush sitting on the side while he finds a PhD project. Since a game of this nature requires a tonne of work (programming, web development, content creation, testing, promotion etc etc) I'm very keen to bring on more people! I created the "design team" page to help thank people like David Lee, Tomas Molina, Sean Penticoff, Rich Stoner and many others who have helped me develop ideas ever since I moved here in July 2010. I very much like soaking up ideas (especially from people with youth and enthusiasm) and giving credit where it's due. My own background is a Bachelor of Computer Science with Honours in Cairns (Australia), follwed by a PhD in diabetes research and electron tomography in Brisbane (Australia). Although my PhD thesis was biologically focussed, I spent much of my PhD developing drawing tools and interpolation techniques (see videos here) to help speed up the process of manually segmenting (tracing) huge tomograms (3D images). Even five years ago segmentation was a huge bottleneck in electron tomography, and the problem is growing exponentially as modern electron microscopes are collecting bigger and bigger datasets. These black-and-white greyscale images are too noisy (i.e: blotchy and blurry) to be segmented effectively with automatic algorithms - hence biologists usually trace their own data, or pay undergraduates to do it for them‚ with mixed results. In less than a week some of these microscopes can collect up to a hundred gigabytes of data - enough to take a single person thirty years to trace! So although the game only started in 2011 it's been a long time coming and the opportunity to work on this great project is the reason I moved from Australia to San Diego. This game creates a unique opportunity to combine biology, programming and fun.

SIDE-NOTE: As you may have noticed, this website is underdevelopment and says "coming soon" on the front page. It's my plan to make this site public towards the end of 2011... but until then don't want any of these pages vastly publicized.

The Vision

The big vision of this game is no-doubt ambitious: to create a game like no other - something which combines citizen science, education and fun. We want to create a game which (a) allows complete novice users (everyone from elementary students to parents) to contribute to cell science from the comfort of their own homes, (b) is educational enough that high school teachers will want to "enroll" and (c) is entertaining enough that gamers will want to play and keep playing and leveling up. There are several challenges in the way: probably the biggest is that tracing organelle after organelle isn't the most exciting thing to do. Fortunately citizen science projects such as Galaxy Zoo and FoldIt have already proved that tens of thousands of people out there are willing to perform tasks if they feel it contributes to science. Still more people will play if we are able to make the game feel exciting, rewarding, and lively in other ways.

Possible Jobs

With anyone who joins the project - even if only for a few weeks as a summer intern - we'll have a good chat and work out how you can best fit in and contribute! In these early stages of game development there are many jobs (working on either the website or the game) which you can help with. The work that needs to get done can be classified and broken down into three broad areas: (1) programming, (2) creating content and (3) testing / feedback / promotion.

• (1) Programming - To work in this domain you'd have to know how to program, so it's not for everyone! The game itself is being programmed in C# using a versatile game engine called Unity (see highlight reel)... however this is just the front end of the game! Another large programming component is creating a database for which we've setup a PostgreSQL database with PostGIS enabled - within this database contours we will store users, contours and so on. Between the game and database we'll also need servers and/or server pages, which will probably be programmed in something like PHP and/or Python. In addition to the game logic (checking what contours overlap etc), the website will also require lots of dynamic pages in PHP where players can register, login, contact us, view high scores and view their profile page. Ideally there will even be a way to chat to and/or add other players as friends. Hopefully there are some good PHP libraries which will help expedite the code needed for this, but it's still a big job! To summarize possible jobs:
  • Helping code CellSculpt (in Unity/C#). (++)
  • Helping design and create the database (starting with schema diagrams). (+)
  • Making server pages to help the game link to the database (requires the database setup).
  • Making PHP pages for player login, high score tables, etc (which ties in with the database). (+)
  • Find a way to tie the game into with Facebook applications (not high priority just yet).
  • Helping port the game to other platforms (mobile phones, xbox, wii etc).
  • Helping code other games in the CellUniverse - such as CellExplorer... and we've also come up with ideas for a CellViewer or CellNanobotMaker.

...

• (2) Creating Content - This is something almost anyone can contribute to. As an educational game, there will need to be a lot of text content, images, artwork.. and even music! One of the biggest jobs I'll push early on is creating content for the in-game multiple-choice* quizzes, where players can earn extra research credits if they pass, and the in-game "library", where each organelle type will need its own page. I like the idea of short-term summer intern students doing this job because it something they can get stuck into almost immediately (without needing lengthy training), and learn a lot about biology as they write the content. Don't forget this is a game, so you'll have to make the text short (concise) and fun. Game players don't want to read a textbook - think more like a comic book or playful "story boards"… minimal text, with lots of analogies and pictures. If you have a flare for art you may be able to create your own images to include in the game and/or website, but (as an alternative), you may prefer to find appropriate images on the web and would have to either request or make sure we have permission to use each of these images. For lots of organelles I already have 3D models I've made in a program called "Cinema 4D", so I can quickly render images for that - you can just draw a placeholder. We also need content for the website. You'll notice some pages on this site are empty, some need more images and most could use improvements and spell checking. There are many more pages I'd like to add talking about what different people in the group are working on - this would give you a great chance to meet other biologists in the lab and offer to type up a profile on their work. Your challenge would thus be to sit down with these senior researchers, get them to explain what they do, then write it up in such a way that visitors to the website (mostly non-science people) can understand. To summarize possible jobs:
  • Create quizzes. (++)
  • Create library pages. (++)
  • Create artwork for game.
  • Gather together royalty free image and music for the game.
  • Create wiki pages to profile other people's projects. (++)
  • Help create storyboards for the game.

...

• (3) Testing / Feedback / Promotion - I've grouped together testing, feedback and promotion because I believe they belong together. If you were to work here and your role was just "game testing" it sounds a bit wishy-washy - to make it sound better your role should involve fostering collaborations and surveying/collecting feedback from our target audience. One of the most important groups of testers will be the biologists who are interested in "using" our system by pushing their datasets through CellSculpt, and thus have a vested interest in making sure CellSculpt players are given sufficient introduction and training so they do a good job segmenting the data. These biologists will help refine the game, and we definitely want their support and approval on content before we launch CellSculpt! To launch this game our plan is to start with high schools, and so an important job will be forming ties and collecting feedback from these high schools. High school students and teachers will make great beta testers and I have no doubt effectively collecting feedback from them will lead to improvements… and if you were clever about it you might even get a paper. For CellSculpt to be successful, it has to be good, but what's just as important is promotion. To promote online we can make video trailers, post on forum, and I'd also like people to help take the Facebook and twitter accounts I've setup (and listed on Resources if I've given you a login) and do something with them! Even more valuable will be any ties with any media we can get a hold of (think local news, science shows and magazines) - and once the game is developed enough I believe the idea of "helping cure disease from your living room" and "if you can trace, you can help science" will be enough to attract attention - which will translate to more players and more funding. The media should be especially interested in the idea that we are forming ties with high-schools and giving students the chance to earn credits and certificates in cell biology. To summarize possible jobs:
  • Play the game yourself and give detailed feedback (not quite enough meat as your only role though).
  • Help collect feedback from biologists.
  • Seek out high-school teachers to incorporate CellSculpt into their biology programs (we have a couple of high-schools in mind).
  • Sit with high-school student or other volunteers and collect feedback though surveys.
  • Help make "CellScupt News" by posts to mailing list, Facebook and twitter accounts (requires us to get a following first!).
  • Help blog about CellSculpt online.
  • Help seek out real-world media connections.
  • Help catalog and apply for funding opportunities (especially educational grants).

Testing and promotion is something exciting to look forward to, but in these early stages of the game we're not quite ready for either yet, so you can ignore (#3). The two big pressing things are (#1) getting the game working and (#2) creating content! If you see something above (or think of something else) which you think you'd be good at and feel passionate about let me know ahead of time and it probably won't be too hard to convince me. Having said that - don't be disappointed/surprised if you don't get assigned the job you want - in most science institutes interns usually don't get given any choice! Here I hope to give you some choice, and you may take on a couple of different jobs (to keep things more interesting) if you're feeling keen. :-)

Work Ethic and Rewards

Something every good scientist should do is give credit and acknowledgement (especially in papers, presentations etc) where it's deserved. CellSculpt currently doesn't have dedicated funding that I can pay anyone - so I sadly can't yet offer to pay students for their work. One day I'm hoping I'll be able to find good students and/or programmers full time, but that day could be a year away, and as a general rule our lab doesn't ever pay high-school students - since the understanding is that high-school students are doing it for the work experience. Instead what I'd like to do is to offer you a good reference letter from myself and also from the director of our institute - Professor Mark Ellisman - who's very well known in our field and should be a powerful addition to your resume. And of course there's the reward of gaining experience, making contacts and helping in the early days of a big project which we hope will become the next "Galaxy Zoo" (but focussing on cells instead of space)!

PS: With your permission, I'd also love to put your name and photo on the website (but may need reminding to update such things).

Working Hours

Something I don't want to do is give away reference letters like candy. A good reference is something you have to earn! For summer students, two weeks is the bare minimum time you'll want to work (less than that and it's very hard to contribute anything), and you'll treat it with the same respect you would a normal job. Hopefully you'll find it a fun job though - you'll probably sit in the "Biomedical Sciences" building in the area of desks near me, but that's just because it's the logical place to sit and because I want to be around if you need help. Normal working hours are 9-5, but if you're a high-school summer student on "summer-time" I think that's pretty scary and could lead to you burning out.… and since my own schedule is all over the place, lets try working hours of 10:00am-4:00pm. And yes you can have a big lunch break. If you work hard and have something to show I'll even let you disappear early or take a day off (in addition to the weekend obviously), but only if you let me know and promise to explore this beautiful university campus and/or city, and to come back with fresh ideas. :-)

As a disclaimer I've never had intern students before (I've done tutoring work - but that's different), however other scientists have warned me they've had periods where they had plenty of work for their students, but other periods when they struggled to find appropriate jobs. Fortunately there is a big list above, and you may (if I'm lucky) even think of your own ways to help and add to the list!

Before Arriving

If your reading this page, chances are you've already contacted me via e-mail (anoske AT ncmir.ucsd.edu) and/or met me at NCMIR so I can show you around a little. If not, don't hesitate to ask for a little chat-and-tour before you decide to commit! The most important requirement however: if you are under eighteen, make sure you contact our main office administration lady Jennifer Santos (jsantos AT ncmir.ucsd.edu) as you will need to sign a couple of forms and get a signed letter from your parents. I believe this can all be done via e-mail attachments, but I'd strongly encourage you to start these forms a month before your arrival in case of delays. Jennifer can also help you with many other issues, such as getting access cards - should you need one for after-hours.

Room Access

The main NCMIR lab is in the Biomedical Sciences Building (BSB) at the University of California, San Diego (UCSD). To see directions and more information about visiting NCMIR visit this page I've made: NCMIR - visiting. We're located in room 1000, which is the wooden door immediately on the left after you go through the automatic glass doors at main west entrance into the building google map. After you open the wooden door I sit immediately on the left. My cell is (8five8)-349-2901 and NCMIR's number is (8five8)534-0276. Be warned that room 1000 is only unlocked between 8:00am-5:15pm - after that it will lock you out (even when you step outside to go to the restrooms) unless you've asked for the code. The glass doors are locked after 7pm weekdays and all weekend, so you'll only be able to get past this if you ask for a side key or swipe card from office administration.

What to Expect on Arrival

So what's the plan when you arrive on your first day of work? I'll hope to spend most of the first day showing you quickly around the lab, show some of my own research and CellSculpt progress, then we'll discuss what job you want to do, and what deliverables we want to set. If you meet me ahead of deciding on an internship (always a smart idea if you live locally!), I've probably already taken you around, so you'll already have a good idea. You may even impress me by setting your own goals for yourself. I'm also hoping to have groups of interns at a time, so I might get you to work as a team. Finally I'll make sure you have a computer. Ideally you'll have your own laptop, but let me know if you don't and I'm sure I can find something. The rest of the first day I may even get you to play games such as "CellCraft" (one of my favorite games listed on our "Biology Games" page) for ideas and inspiration. And of course I'll double check you've read this document (if not already!).

Guidance and Group Meetings

Some people are very independent workers - people who require minimal instructions and often prefer the challenge of working things out themselves - while others need structure and guidance - meaning that if they're not told what to do and given deadlines they'll quickly become lost and demotivated. I've always liked to be independent and make my own motivation and my own decisions, but even people who are generally independent can find it a shock to transition from high-school and/or jobs in industry (including real world jobs like fast food), where you have set deadlines, to academia, where very few supervisors have the time to spoon-feed you and almost all group leaders are away on conferences most of the time and too busy to give specific instructions (let alone a well defined project). Starting a job in research and discovering you have to fend for yourself is fairly normal in academia - some people can handle it and some can't…. and if you find you can't handle it that's totally fine, since these people eventually realize they want to work in a place with more structure and better pay. For student interns, my role will be to ask and try to work out how much guidance you need and trust that if you ever do need extra assistance you feel comfortable asking that question which might otherwise take you hours to work out! I'm a friendly Australian and an understanding guy, so you'd be silly not to ask for help whenever you feel lost. Whatever variety of lost you may feel - intellectually, emotionally, motivationally, geographically, biologically and/or technologically... or if you're unhappy about anything at all - it's important to let me know as-soon-as-possible so I can try to keep everyone happy. In fact, let's make this a golden rule!

Demonstrating Progress

Other advice I've been given is to set some form of deadline system for interns to make sure you stay focussed - and so if you do happen to chose to work on a quiz, by the end of each week I'll want to have a meeting and see how you're doing. With people already familiar with academia it's my hope you can already work (mostly) independently. Something that always motivates me is knowing that at the end of each week I've accomplished something - something I can show to people to prove that I've not been a slack. I'm relatively laid back, but I don't believe in time wasters so if you can't show tangible progress at the end of the week we'll need to chat. I believe in taking regular breaks to stay focussed, and I know it takes a while to get settled, but once you're setup on a computer I don't want to look over and see you on Facebook all day… not unless you're working a Facebook application for CellSculpt! If, however, I see great work from you, I'm going to make sure I tell my boss you've been terrific and deserve a glowing review. :-)

Weekly Team Meetings

Every Friday at 3:00pm I'll hope to have a small "CellSculpt progress meeting" - a very small and informal group meeting (possibly just me and any summer interns!) where we (yes even me) can show off and talk about the progress we've made, and what we might need help/guidance with. If it's your first week we can skip you - it's very informal - so don't stress over it. :-)

In addition to these meetings there are several other meetings (particularly meetings listed here) you might like to come along to, such as "dangerous ideas" and "neural seg" - going to such meetings is a good way to meet and see what other people around NCMIR are working on.

Summary

This document is pretty long, but I hope useful in answering any questions you might have. Something you've probably already figured out is that very few cell biologists are lucky enough to work on something like a computer game - and so the job I'm doing isn't representative of a normal research project. Our goal is to make the game fun, but I also want to make sure the game is as useful and educational as possible. Furthermore we want to involve as many biologists as possible from our lab (more and more as the project reaches deployment), so if you're clever about how you work you should be able to get a chance to meet some of these biologists and see a bit of how our microscopes and wet labs work - because that's where the real science happens and I hope to never loose focus of the fact we're making this tool to advance science.

I look forward to working with you if you do decide to join the team!

Regards,

Andrew Noske (your enthusiastic team leader!)

Extra Information

If you're interested in a summer student position you can probably stop reading here, but if you're interested in learning more about CellSculpt and the "meat" behind it you should definitely read on!

Resources - Programs and Technology

Something I should quickly mention is what programs we use! I do most of my programming on my own laptop, but in our little work area there is a terrific new Mac computer I may be able to lend if you need the extra processing power, and a much older Mac opposite me‚ and then on the other side of the hall is a larger area with a mix of shared Linux workstations. What works best, however, is if you bring your own laptop (if you own one) - that way you have extra flexibility to work anywhere. Summer students may even decide they want to catch some of the summer sunshine by working outside for a bit. If you're a Window's person you may feel disconcert you that I only have Macs around me, but don't be put off. Modern Macs have all the same applications as Windows, they're very popular in biology labs, and you're doing yourself a favor by learning to use both‚Ķ I was once a PC-only person, but now I quickly become comfortable with both (which I mention on my own resume!) - I have a PC at home, but do most of my work now on Mac. Some of the applications and technologies we use on this project are listed below:

• For the website:
  • programming: PHP (for all dynamic webpages) & PostgreSQL (for user database and also contours using PostGIS functions).
  • content & art: Adobe Dreamweaver (for main webpages), Atlassian Confluence (for the wiki part of site - very easy to learn and add too), Adobe Fireworks & Photoshop (for most 2D art), Adobe Flash (for 2D animations), Cinema 4D (for 3D art and movies for science and/or education).
• For the game:
  • programming: C# using Unity game engine.
  • content & art: Cinema 4D (for 3D models).

For anyone who may have a flair for web design, or art - maybe even 3D graphics, let me know! :-)

For most interns, all you'll probably need to use is a web browser and Microsoft Word for all your work. I believe it's very valuable to learn to use new technologies when you have a chance…. but something even more valuable is learning about the biology... and that can be a bit trickier to learn!

For more information on the details of how CellSculpt is setup, including architecture diagrams, I've created several useful pages in the "Developer Area" of this wiki - in particular the page "File Formats" and "Class Diagrams". These are restricted pages, however, so you'll have to get a "CRBS confluence account" before you can see these pages... and I'll see if I can't setup a "CellSculpt guest" account for any interns.

Analyzing Data from Crowdsourcing

Something I haven't listed under "possible jobs" is what we do with the data we collect. This is a very unique crowd-sourcing game, and citizen science is all the rage nowadays, so I definitely think there's plenty of scope in CellSculpt for meaty projects such as PhDs. The term "wisdom of the crowd" refers to the proven idea that large groups of ordinary people can produce better work that a single expert, with Wikipedia being a classic example. There are many publications on wisdom of the crowd in the context of creating and editing text, however there appears to be relatively little on how this applies to drawing 2D polygons - investigating the best way to take multiple user tracings and average them to get the best result for least work. Very few supervisors give students well defined projects - that's just how it goes... however I do have a few good ideas which might help - some of them projects I'd love to work on myself, but I know I don't have the time so I'd love to find a good student who'd be interested in hearing more. Once the contours are all segmented satisfactorily (which is a big project in itself) we use a program like IMOD (which many of us use) to automatically generate a 3D mesh and then finally we are able to analyze the data (surface area, volume, shape, number of organelles and relatively position between organelles - and morphological differences between different cell conditions) in a biological context.

Combining Computer Science with Biology

A rare and valuable commodity in a biology lab is a programmer who knows his biology! If you're a short-term summer student this probably doesn't apply to you, but if you happen to be a enrolled in a biology or neuroscience program I'll tell you some words of gospel I was given when I first started in this field. No software developer gets taken seriously as a cell biologist without a biology driven project - a project that answers a solid biological question. I can't stress enough how critical it is that you have a biological driver to focus on and which will lead to publications in biological journals. For my PhD in Australia, I ended up in a fairly hardcore biology institute, the Institute for Molecular Bioscience, with an undergraduate degree in computer since, but had zero experience in cell biology. I was very fortunate to have a supervisor, Brad Marsh, who insisted I sit with biologists and learn to use Electron Microscopes, so I could (a) learn a valuable new skill, (b) be able to say I collected my own data and (c) develop a full understanding of and appreciation of the intricacies of how the processes work and the many problems involved. In addition to using the microscopes I also had to read and understand a lot of papers on beta-cell biology. Through hard work I found and helped answer a biology question - "determining structure function-relationships and key conformational changes of the key compartments of the insulin secretory pathway following glucose stimulation"…. in English (using words a *little* more appropriate for a CellSculpt audience) my question was: "what happens inside a beta cell - the cells in your pancreas which release insulin and keep your blood sugar down when they are exposed to high levels of sugar". This is the dumbed-down short version, but serves as a nice example of a "biological question" every good biologist should have…. for full details you'd have to download my PhD thesis (secret link at bottom of page). My project required me to segment several whole cells at Electron Tomography resolution - each many gigabytes worth of data - hence I had to develop new techniques and create software tools to expedite the segmentation process. I had a pretty tough and skeptical crowd, and it was only at the end of the project I had enough data and tight PowerPoint presentation to win their respect. I tell you this as a warning - there is great reward in combining computer science with cell biology, but a project such as this isn't for everyone!

There are several reasons I have set myself up in the basic science building instead of the fancier CalIt2 building (which I'll show you - the StarCave is very cool). I came here wanting to work on CellSculpt, but before I stared on the game I wanted to find a core biological question which I could tie into the game. As my current "biological driver" I'm going to work on helping characterize the recycling pathway of ganglion retinal cells with the exert help of Dr Eric Bushong - a seasoned neuroscientist who's been at NCMIR longer than just about anyone! This project will require segmenting and number crunching on a huge volume of 3D data, and the only way to segment it all will be via crowd-source - hence justifying why I spend most of my time working on CellSculpt…. but hopefully will spend enough time working with biologists to slowly learn neuroscience. So no, I am not yet a neuroscientist! My background is diabetes research, but if you want to learn neuroscience you'll have to couple yourself to someone in the lab who is a neuroscience expert and willing to help teach and mentor you. I can help on the computer science side, but I haven't published anything in neuroscience - hence if you want to end up with some type of qualification/degree in neuroscience I'd want to help you find someone who's qualified for that.

For interns, what I hope to do is introduce you to a few different biologists (unfortunately I don't yet know them all), but it's a pretty busy lab so you'll have to be prepared for the fact that not everyone will have time to chat and/or show you how things work. The higher up they are, the busier - with the busiest people being group leaders who are typically away half the time at conferences. I have my own frequent distractions too, so if there's anything specific you want to see or learn how to do I hope you'll approach and ask me! Currently I'm not doing any microscopy work myself, but hopefully I might be able to set you up with colleagues to see what they do in the lab. If I'm not doing a great job introducing you to biologists please call me out on it.

How CellSculpt fits into other NCMIR Projects

Still reading?! If so I'm impressed. I'll keep this breif, but here I wanted to describe quickly how CellSculpt fit into the NCMIR projects. You can read about how NCMIR ties in with other software projects on the "featured software" page. Notice for example we want to tie this in with the Cell Centered Database, which was championed by Maryann Martone, and you'll also notice CellSculpt ties tightly with Flash program called WIB (Web Image Browser) designed by Steve Lamont, and we use the "Web Image Manager" to load image tiles. In terms of biology projects, I'm hoping CellSculpt will include data from many different projects, some of which I've listed on the "featured biology" page, which I hope grows over time. About 95% of NCMIR is focussed on neural biology - imaging brain cells - but I will always stubbornly insist the game should be broader than just tracing neurons; hence I want to incorporate dataset from other groups.

CellSculpt within the "SLASH" grant

Something not mentioned on the above pages is that CellSculpt is part of a "SLASH" grant which was put forward by a small group informally know as the "NeuroSeg" group. SLASH stands for "System for Large Area Segmentation using a Hybrid Approach". This is NIH grant I helped with back in Nov 2010 with Rick Guily, a Masters student who writes automated algorithms and runs various meetings; David Lee, who's doing a PhD in crowd-sourcing (but sadly left back to Michigan); Steve Peltier, the executive director of NCMIR and has a background in electrical engineering and two researchers, Tolga Tasdizen and Elizabeth Jurrus, who reside at the University of Utah. Hopefully you'll get to meet some of these people (the ones at NCMIR anyway). This grant should come though in May 2011 (pending), and a crucial part of this proposal was how we'd combine crowd-sourcing and auto-segmentation. Something you should know is that even if CellSculpt were able to attract tens of thousands of dedicated users, this wouldn't even come close to manually segmenting the vast amounts of data being collected on the thousands of electron microscopes around the world. The only way to permanently fix this problem is if computers are able to automatically segment the datasets. Automatic segmentation has been around for a long time, but is still no substitute for the human eye - especially on these types of noisy dataset. The goal of SLASH, however, is to demonstrate that hybrid methods - mixing automatic segmentation with manual "proofreading" (correcting any bad lines from the automatic segmentation) and tracings - can offer big improvements. In Utah, Liz and Tolga are working on "learning algorithms" which will hopefully take sets of data from our CellSculpt users, and by inputting these subsets of "training data", be able to improve automatic segmentation on the larger dataset. You can read more about SLASH here.