Colocalisation in ImageJ : correcting noisy (poor quality) images

Поделиться
HTML-код
  • Опубликовано: 18 дек 2024

Комментарии •

  • @senecarus_whitur
    @senecarus_whitur 2 года назад +2

    Hello Craig!
    Thank you so much for taking the time to put this information together! I will try this method as soon as I have access to my files at home :)
    This guide is very helpful!

    • @CraigDaly
      @CraigDaly  2 года назад

      Thanks for watching. Craig.

  • @remrblu
    @remrblu Год назад

    Godd day Dr. Amazing vídeo, i am working with colocaliztion i have problems to standarize and have sequential results in my experiments so i fear that noise is given me problems. I have a question with this methods, the time diference with Time 1 and Time 2 is in seconds or Minutes, i am using a Z stack so i will want to try this method. Thanks in avance

    • @CraigDaly
      @CraigDaly  Год назад

      Hi, sorry for the delay. The time difference would be seconds in this case, or as quickly as you can collect two images or z-series.

  • @TheDigitalis23
    @TheDigitalis23 2 года назад

    Hello Craig. Thank you for explaining how to tackle noisy images in colocalisation analysis! This is very helpful.
    Question- is this technique better than using denoising algorithms before conducting colocalisation analysis? Is there a tutorial covering how those algorithms work and the right way to use them?
    Thank you!

    • @CraigDaly
      @CraigDaly  2 года назад +1

      Hi. Good question. Denoising might work ok, but as you indicate, you would want to know how much noise is being removed and how. The advantage of the presented technique is YOU are calculating the degree of noise and can easily describe the method. I would need to look at the mechanisms behind the denoising but that would make a good video. I may make that video if there isn’t already one out there. Thanks for the idea.

    • @TheDigitalis23
      @TheDigitalis23 2 года назад

      @@CraigDaly Thank you, Craig! And yes, it would be very helpful if you could make one on the various denoising algorithms as well :)

  • @wendykim5
    @wendykim5 2 года назад

    Hello Craig, thank you for making the video! This is so helpful and informative.
    I have a small question: when there are several images, which I took different spots of one sample,
    do I need to set the denoising parameters equally?
    Although the images came from one sample, I feel like the level of noises (or background) are all different between them.
    Is it possible to quantify the colocalization ratio of several images taken in different denoising processes?
    (The images were all taken at an equal microscopic condition - gain, offsets, and laser intensity.)
    Thank you so much in advance.

    • @CraigDaly
      @CraigDaly  2 года назад +1

      Hi Julia, that’s a good point. I think the noise is probably sample dependent rather than detector variation (happy to be corrected on that). So I think you may need to have a different calculation for each sample. The Q of different regions of the same sample I think would be down to the individual user to decide. Hopefully though the variation is not that big in different regions. As ever, just be clear in your methods section when writing up. Great question. Thanks for watching. C

  • @jessicachedid6622
    @jessicachedid6622 3 года назад

    Hello, thank you for the helpful content.
    I have one question regarding this technique, just to make sure it is relevant for my work. When you say timepoint 1 and timepoint 2 are you referring to live cell imaging (at 2 timepoints) or imaging the same fixed sample twice and looking at the differences which would than be the noise?
    I am working with fixed samples and one of my channels is always noisy.
    thank you!

    • @CraigDaly
      @CraigDaly  3 года назад +2

      Hi Jessica. Yes, fixed cells/tissues. two images taken in sequence (time 1 and time 2) should be identical - assuming they are taken very quickly. The only difference is the noise in the system. So a Pearson's correlation of both images will give an indication of how much noise (i.e. how far away from 1 and how fat is the pixel fluorogram). Craig.

  • @lirisnareswari8662
    @lirisnareswari8662 4 года назад +1

    Apologize, but I don't see any noise reduction in both picture?

    • @CraigDaly
      @CraigDaly  4 года назад +7

      Hi, the process does not remove noise from the images but in the final calculation it takes account of the noise that it has estimated to be present. Thank you for watching 👍

    • @lounablc8885
      @lounablc8885 9 месяцев назад

      Hello, I had the same question as @lirisnarescari8662. After calculating the coefficient which takes account of the noise, can we remove the noise from the pictures thanks to that calculation ? If not what "role" does this coefficient have ? In whichfurther calculation can it be involved ? Thank you very much@@CraigDaly

    • @CraigDaly
      @CraigDaly  9 месяцев назад

      Hi @@lounablc8885 The coefficient is a single value that represents noise in both the red and green channels. In many cases there will be more noise in one channel than the other. Therefore the value itself can't be applied to the original images to be used as a noise-reduction filter. The value is only of use inside a co-localisation calculation and is correcting a calculated correlation value. Noise reduction in a single image would use a different process. Hope that makes sense. Regards. Craig.