Week 7: The Challenges Of Imaging And Annotating
Hello Everyone! Tis’ week 7!
This week, I wanted to address some of the challenges of imaging and annotating. (pretty self-explanatory from the title). While most types of exploratory research gradually accrue findings and evidence toward a particular hypothesis, my research is different. Rather than conducting assays, a fancy word for tests, and running computational models, I’ll only find results once the data sets I’ve been establishing are analyzed. The final analysis of the probability maps that will be generated from my annotations shouldn’t take too long (hopefully). Even so, generating good probability maps is challenging. So what are the challenges?
I won’t pretend to know all the technicalities of imaging and I won’t speak at length about the challenges of imaging because I didn’t personally take the images. However, there are some factors that one needs to consider when analyzing images, especially those taken via confocal laser scanning microscopes. The point spread function and chromatic aberration are two characteristics of imaging that I’ve experienced in my research.
- Point Spread Function
PSF is a phenomenon in light imaging where an image blurs in response to light coming off of the object interfering with itself. The PSF is a way of determining the quality of images because it affects the resolution of the actual image. PSF also causes different wavelengths of light to lead to different resolutions. Shorter wavelengths of light will result in a smaller PSF and better resolution while longer wavelengths of light will do the opposite.
- Chromatic Aberration
CA is a phenomenon in which lenses can’t focus on all wavelengths of light. This is caused by different wavelengths of light focusing at different distances from a lens. Practically, CA leads to shifts in different wavelengths of light in either the X, Y, or Z direction, leading to the downstream correction being needed.
These shifts and blurs lead to challenges in annotating because they need to be taken into account. Shifts are generally corrected pre-annotation; however, the PSF blurring is more difficult to accommodate. Seeing images for what they are takes time and I’ve experienced a steep learning curve while analyzing these images.
I’m hopeful that within the week, I’ll finish up a probability map and move on to analysis. Thank you for reading and have a great week!