Hi, my intern and I were working on imaging V. corymbosum buds in a confocal microscope (LSM900). Our plan was to have the same settings for all the images, but we changed (by mistake) the gain and the laser intensity on each image. Is there a way to normalize all the images so they are comparable? We are targeting two different states of pectins (LM19 and LM20), and the plan was to see how their relative quantity changed across three sampling dates. Sadly we don't have time to stain the sections again and obtain new images, so we are searching for a way to use the already-acquired images.
Thank you!
Yes, it is possible to normalize the images so that they are comparable despite the variation in gain and laser intensity settings. Normalization is a common technique used to remove variations caused by different acquisition settings and bring the images to a common scale for meaningful comparisons.
Here's a step-by-step guide on how to normalize your confocal microscope images:
1. Identify Reference Regions: Choose regions in the images that should remain constant across all the samples, ideally areas that are not affected by the gain and laser intensity changes. These regions will act as reference points for normalization.
2. Extract Pixel Value: Extract the pixel values of the selected reference regions from each image. This will create a dataset of reference values for all images.
3. Calculate Normalization Factors: For each image, calculate the normalization factors by comparing the reference values to the corresponding values in the other images. The normalization factor can be calculated as the ratio of the reference value in each image to the mean or median reference value across all images.
4. Apply Normalization: Divide each image by its corresponding normalization factor. This will effectively normalize the pixel intensities across all images, making them comparable.
Here's a simplified formula for the normalization process:
Normalized_image = (Original_image / Reference_value_of_image) * Mean_or_Median_of_Reference_values
Remember to perform the above steps separately for the LM19 and LM20 images if they are stained differently or require separate normalization.
After normalization, you can proceed with your original plan to compare the relative quantity of LM19 and LM20 pectins across the three sampling dates. Keep in mind that normalization helps to make the images comparable but does not correct for potential changes or degradation in the samples over time.
Always validate the results and carefully interpret the data, considering the limitations of the normalization process and other factors that may influence your analysis.
Camilo Villouta My recommendation to you is that if you will be planning an intensity profiling for your experiments you should always take the imaging at the same intensity.
You may consider comparing changes in the area of signal (after background signal subtraction) or count puncta of signal per area. Since gain and laser intensity are different -> big SD for intensities -> difficult to compare or even not possible. I do not think that it is good practice to compare intensities in that case.
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Extraction