Skip to main content

Together we are beating cancer

Donate now
  • Science & Technology

Help us beat cancer – with just a few clicks of your mouse

by Henry Scowcroft | Analysis

24 October 2012

9 comments 9 comments

Computers

Help beat cancer with a few clicks of your mouse

Imagine a world where millions of people are helping to find ways to control and cure cancer, from the comfort of their own homes.

Imagine these people taking part in a project which might – one day – become a candidate for the Nobel Prize in medicine.

That’s the vision behind our new ‘citizen science’ project. Launching today, http://www.clicktocure.net/ allows the general public to directly take part in our life-saving work, by actually increasing the pace of cancer research.

So, what’s it all about?

Radical advances in technology have brought new challenges, in the form of huge amounts of biological data. Labs are now generating terabytes of research data which requires urgent analysis; analysis that our scientists are doing every day.

It takes human intuition and the human eye to spot patterns, defects and anomalies – computer algorithms aren’t yet good enough. So our progress is being hindered by the speed with which we can analyse data.

How can we move things forward?

Citizen Science

In recent years, various initiatives have attempted to use ‘the crowd’ to drive research forward, be it research on how proteins fold, or the search for extra-terrestrial life.

The trail blazers in this field are an inspirational organisation called the Citizen Science Alliance (CSA).

We were inspired by a CSA astronomy project  – Galaxy Zoo – which was trying to classify different galaxies spotted in deep space.

Faced with a deluge of NASA data, a group of astronomers uploaded it for public analysis. Galaxy Zoo’s users made 60,000,000 classifications; the equivalent of employing a single analyst for more than 110 years!

So we wondered whether this model could be used in cancer research.

Hacking to beat cancer

Since February, we’ve been working with the CSA pathologist Professor Andrew Hanby, Cancer Research UK-funded epidemiologist Professor Paul Pharoah, and their teams – along with a team of volunteer web developers who attended a ‘hackathon’ earlier this year – to develop a solution to the challenge faced by huge amounts of biological data.

The result, launched today, is Cell Slider.

Hosted on CSA’s website, Cell Slider presents users with real images of tumour samples for analysis, in the form of a simple game of snap.  It presents the data in a way that is accessible to everyone – while maintaining the anonymity of the original sample.

CellSlider screenshot

A screenshot of the Cell Slider website. Click to visit the site

Users are guided through a tutorial that explains which cells to analyse and which ones to ignore. Once cancer cells have been spotted, users are asked to record how many have been stained yellow, and how bright the stain is, by simply clicking on another image that closely matches the sample they are viewing.

This information will be used to look for trends between types of cells and a patient’s response to treatment. Initially, the programme will use breast cancer samples, and the yellow stain will indicate levels of the oestrogen receptor.

Each image of tumour samples will be analysed by several people to ensure that any accidental clicks can be discounted. And, to test how accurate the programme is, researchers will link the samples that have been flagged up by Cell Slider players with anonymised data on treatment and survival to see if it gives results they expect.

Beating cancer from your armchair

We want everyone to be able to get involved, and play their part in beating cancer from the comfort of their own homes. Cell Slider is free to do, and with just a few spare minutes and a few clicks of the mouse, citizen scientists will be accelerating our research.

This truly is a world first. And we hope it will change the face of cancer research as we know it.

Henry

    Comments

  • Mike Biddell
    11 April 2013

    I believe some worked examples depicting cell boundaries and the correct interpretation would help us to “tune in/hone” our judgements. If all samples were taken at identical magnification, so that absolute dimensions were accurate, slide to slide i.e. not zoomed in any way. This recognition could almost certainly be automated. It’s very similar tech/software to Optical Character Recognition (OCR).

  • Lisa
    7 March 2013

    This is a great idea, I’ve tried to go to the website but it just ends up as a blank page and nothin loads. Are there any browser specifications for it to work?

  • S bullock
    25 January 2013

    Just started using this. Glad to see some response to Laura’s questions. I think a skip button would be useful, sometimes it is really hard to make a judgement.

    Maybe a couple of extra “test” slides would be good too with the “correct” decisions. The examples given are very different to a lot of the slides.

  • Amy Carton
    29 October 2012

    Firstly, sincere thanks to you all for getting involved in Cell Slider and taking the time to provide this useful feedback. With your help we have already reached over 90,000 classifications!

    To answer your first query Laura, healthy cells, such as the tissue and blood cells you might be seeing, will have minimal staining-and you may see this from time to time. However, the staining is typically found in cancer cells, and the brighter the stain the more serious the cancer.

    Thomas, thank you for your positive comment; we’re glad you like it.

    Several of you have mentioned the lack of reset button, which we were really interested to note. This was a decision based on research provided by the Citizen Science Alliance (the team we worked with to create this Project) and was Beta tested to good effect. The reason is that we’re keen for people to keep going through the site – the more classifications the better, and we have all sorts of activities that take place behind the scenes to ensure that we get the best science possible for our pathologists.

    That’s not to say that this won’t change in future, and your feedback has been added to our Project documentation. I hope this helps and if you have any more suggestions by all means continue to post comments here.

    Amy Carton
    Cancer Research UK

  • Sam Smith
    28 October 2012

    I agree, we need a re-set or re-do button and a “skip” button would also be nice as Laura suggested.

    Sam

  • Thomas Zolotor
    28 October 2012

    I like some of Laura’s suggestions and wondered why cell slider do not have a re-set button on it like galaxy Zoo 4 does…if one made a mistake half way thur one image can one redo the image from the start by using a re-set button…Galaxy Zoo 4 uses one.

  • Thomas Zolotor
    27 October 2012

    This is a very good thing to help people out with cancer.

  • Laura
    25 October 2012

    Is there somewhere to ask questions/suggestions about the cell-slider site? I have started using it but the faq doesn’t help with some questions. For example, if cells are stained yellow, does that mean that they are, by default, irregular or does the yellow stain also stain blood cells sometimes? Also, it would be helpful to have a ‘skip’ button for slides which you are not sure about (or ones which occasionally do not load properly).

  • George Nicholas
    25 October 2012

    I note your recent campaign highlighting cancer research. In my late wife’s case it took the hospital one month to diagnose recurrent cancer, regardless of obvious signs such as bleeding. And even after that there was no hurry to operate as the tumour was said to be ‘localised’. But as it turned out at the operation almost two months after being admitted to hospital, it was not localised – in fact, the cancer had spread significantly and wrapped itself around a major artery. As the aim of cancer research is to benefit patients, for that potential to be fulfilled there needs to be a change in culture in hospitals like St George’s (Tooting) with regard to the urgency and competency in diagnosing cancer or recurrent cancer and for doctors to be sufficiently knowledgable to accurately interpret scans.

    Comments

  • Mike Biddell
    11 April 2013

    I believe some worked examples depicting cell boundaries and the correct interpretation would help us to “tune in/hone” our judgements. If all samples were taken at identical magnification, so that absolute dimensions were accurate, slide to slide i.e. not zoomed in any way. This recognition could almost certainly be automated. It’s very similar tech/software to Optical Character Recognition (OCR).

  • Lisa
    7 March 2013

    This is a great idea, I’ve tried to go to the website but it just ends up as a blank page and nothin loads. Are there any browser specifications for it to work?

  • S bullock
    25 January 2013

    Just started using this. Glad to see some response to Laura’s questions. I think a skip button would be useful, sometimes it is really hard to make a judgement.

    Maybe a couple of extra “test” slides would be good too with the “correct” decisions. The examples given are very different to a lot of the slides.

  • Amy Carton
    29 October 2012

    Firstly, sincere thanks to you all for getting involved in Cell Slider and taking the time to provide this useful feedback. With your help we have already reached over 90,000 classifications!

    To answer your first query Laura, healthy cells, such as the tissue and blood cells you might be seeing, will have minimal staining-and you may see this from time to time. However, the staining is typically found in cancer cells, and the brighter the stain the more serious the cancer.

    Thomas, thank you for your positive comment; we’re glad you like it.

    Several of you have mentioned the lack of reset button, which we were really interested to note. This was a decision based on research provided by the Citizen Science Alliance (the team we worked with to create this Project) and was Beta tested to good effect. The reason is that we’re keen for people to keep going through the site – the more classifications the better, and we have all sorts of activities that take place behind the scenes to ensure that we get the best science possible for our pathologists.

    That’s not to say that this won’t change in future, and your feedback has been added to our Project documentation. I hope this helps and if you have any more suggestions by all means continue to post comments here.

    Amy Carton
    Cancer Research UK

  • Sam Smith
    28 October 2012

    I agree, we need a re-set or re-do button and a “skip” button would also be nice as Laura suggested.

    Sam

  • Thomas Zolotor
    28 October 2012

    I like some of Laura’s suggestions and wondered why cell slider do not have a re-set button on it like galaxy Zoo 4 does…if one made a mistake half way thur one image can one redo the image from the start by using a re-set button…Galaxy Zoo 4 uses one.

  • Thomas Zolotor
    27 October 2012

    This is a very good thing to help people out with cancer.

  • Laura
    25 October 2012

    Is there somewhere to ask questions/suggestions about the cell-slider site? I have started using it but the faq doesn’t help with some questions. For example, if cells are stained yellow, does that mean that they are, by default, irregular or does the yellow stain also stain blood cells sometimes? Also, it would be helpful to have a ‘skip’ button for slides which you are not sure about (or ones which occasionally do not load properly).

  • George Nicholas
    25 October 2012

    I note your recent campaign highlighting cancer research. In my late wife’s case it took the hospital one month to diagnose recurrent cancer, regardless of obvious signs such as bleeding. And even after that there was no hurry to operate as the tumour was said to be ‘localised’. But as it turned out at the operation almost two months after being admitted to hospital, it was not localised – in fact, the cancer had spread significantly and wrapped itself around a major artery. As the aim of cancer research is to benefit patients, for that potential to be fulfilled there needs to be a change in culture in hospitals like St George’s (Tooting) with regard to the urgency and competency in diagnosing cancer or recurrent cancer and for doctors to be sufficiently knowledgable to accurately interpret scans.