LEARN #1 – week 115 of my PhD

Good Morning world! Here is the first edition of LEARN, your favourite  weekly blog series from now on! This is the recap of what I have learnt last week:

Psychedelic image of the Hough Transform for a sine wave

Monday: Hough transform – Despite being extremely surprised that on a Monday I was motivated enough to learn new things, the hough transform was extremely fun to learn. It has a very simple math that allows you to extract linear features of an image. I needed that for extracting some information on sine waves in my pictures. Although for my problem I would need to define my own transform for sine waves decompositions, the simple application of the simple hough transform creates damn-good-looking images!

From left: kymograph of a measurement I am doing (no spoilers for now); extraction of maximum intensities position; fitting of results with various sine wave functions.

Tuesday: I learn that I could spend a week implementing my hough transform, or using the properties of my image to do a simpler analysis in just one afternoon (thanks to the suggestions of a clever friend of mine).

Typical Scottish Chick

Wednesday: Some of my experiments involve chicken embryos, during development. In total, from fertilisation to hatch it takes 20-21 days. In my study, though, I am looking just at the first 2 stages, during the first 6-7h of the embryo. In particular I am interested in the transition between stage 1, when the embryo is formed just from the area pellucida (a thin layer of epithelial cells) and a collar all around called area opace, and stage 2, when it start to develop what is called the primitive streak. What I learn on Wednesday is that the feature that appears in this transition, Koller’s Sickle, last just 30 min (at most). So this would be the timeframe of my experiments!

Ant-Man and his scary army

Thursday: After a disturbing intro about Lotus Birth, and still pondering about chick embryos development, another one of my clever friend pointed out how convenient would it be for humans to develop in elastic eggs. His argument was based on the fact that the humans’ head size is limited by the way we deliver our offspring. In fact, we have merely a brain:body ratio of 1:50, while small ants would go up to 1:7. One could argue that the number of neurons in ants is more than 300000 times smaller than humans, but I would reply by warning them about the secret plan ants must have. In fact, while one might think the humans are the highest source of change on planet earth, they move just about 35 billion tons of earth every year against the about 50 billion ants move around the globe*. And since ants have a very strong social intelligence, it is reasonable to ask “Why are they moving all this stuff?”, “How are they organising it?”, “Are they aiming to rule Earth?” and more importantly “Why isn’t Antman the most powerful superhero of all?”

PCA of a fish

Friday: Finally, just before to head to the pubs for drink, another clever friend of mine (that unfortunately does not have a blog to link to) explained me about Principal Component Analysis. I found it extremely clever, and I suggest you to check this interactive explanation! (It’s 01.21 am- I am quite tired, and still need to shower and I also hope to get some sleep before working: sorry if I don’t explain it myself)

See you all next week! In the meantime, don’t forget to LEARN Every day A Remarkable Notion! 😉

*  This is calculated assuming that ants move 25 metric tons of soil per ha every year. This number is obtained by averaging the only scientific data I found online (64 tons for the P. Badius Ant, 1 ton for very small ants as the Trachymyrmex septentrionalis30 tons as reported here for non specified ants, 10 tons as reported for Australia for generic ants).
The average is not weighed, and I felt satisfied with it just because it turned out to be more conservative than 50 metric tons per year, the most commonly reported number on the subject (but of which I was not able to find the source).
The calculated amount of tons is then multiplied by the surface area of earth that is inhabited by ants (pretty much everywhere except Antartica).