Hemaxi Narotamo

There are errors in the Funding section. The correct Funding statement is: CC and MS were funded by FCT—Foundation for Science and Technology, Portugal, under project UIDB/50009/2020. https://www.fct.pt/ H.N was funded by Fundação para a Ciência e a Tecnologia (FCT), under Doctoral Grant 2020.04511.BD. https://www.fct.pt/ AM was supported by the National Research Foundation, Singapore, Investigatorship award NRF-NRFI05-2019-0006. https://www.nrf.gov.sg/. Reference 1.

Abstract Background: Analysis of vascular networks is an essential step to unravel the mechanisms regulating the physiological and pathological organization of blood vessels. So far, most of the analyses are performed using 2D projections of 3D networks, a strategy that has several obvious shortcomings. For instance, it does not capture the true geometry of the vasculature, and generates artifacts on vessel connectivity.

Eyespot measurements To train the U-Net to perform automatic eyespot measurements we randomly divided dataset2 into 80% images for training (101 eyespots) and 20% images for testing (24 eyespots). From the center X,Y coordinates and total area annotations in dataset2 we were able to create square eyespot cropped images, which were resized to 128x128 pixels. The results presented below were obtained by applying the U-Net trained models on the test set.