Before doing Integrated density measurement in image J software is iit necessary to invert the Western blot image (from black protein band and white background to white protein band and black background). What could be accurate measurement ?
Thanks in advance for your expert opinions.
Hello Pranav,
The integrated density is defined as "the product of Area and Mean Gray Value" (http://rsb.info.nih.gov/ij/docs/menus/analyze.html). So, you basically want to measure the overall intensities of your bands. Think of it like that: The more photons are detected, the higher the pixel values in your image will become (assumed there is no clipping or saturation). The emitted photons are proportional to the loaded amount of whatever in your band is. The intensity values in an image vary between 0 (for no light detected) and some maximal value. This maximal value is defined by the bit-depth of your image - a 8-bit gray scale image for example has values 0 to 255 - basically 2^8 possible intensities. Imagine an image as a big matrix (image width x image height) of intensity values. How these values are displayed, depend on the LUT (look-up table). 0 is usually displayed as black and the maximal value is displayed as white and the values in between follow a gradient. Now envision that you could display the pixels in any other way. For example, the other way around - 0 becomes now white and the maximal value is now black. This would be a so-called inverted LUT. people somehow got used to bands being black rather than white and that's why gels are often displayed with an inverted LUT (e.g. on .gel-Files). So in that sense, I guess you're asking the wrong question. It's not so much about the bands being black or white but rather their underlying intensity values. If the intensity values of a pronounced band are higher than the intensity values of a background region, you don't need to invert the image. However, if the values are inverted, then you need to invert your image prior to measuring. What the inversion does, is to swap the intensity values. What you want and what you need for a proper measurement are bands that have higher intensity values than the background. If you're not sure whether the image is inverse, just select a region from the background, measure it and do the same with a band region. Compare the average intensity values and decide whether it's necessary to invert the image.
I hope my long explanation helped you
Martin
Dear Pranav
Martin gave an excellent explanation. I just have one thing to add regarding accurate measurement. When dealing with 'bright' bands and 'black' backgrounds, it will be important to keep in mind that the 'bright' bands you measured within your chosen ROI could also include a certain degree of background intensity values (i.e. the background intensity of the gel is not absolutely zero). Therefore background subtraction would be necessary to achieve accuracy.
HTH.
Chenshu
Is it, generally, beneficial to convert the image to binary and than measure the integrated density?
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing