84 Chapter 3 DISCUSSION AND CONCLUSION This paper introduces an extension to the OrganelX e-Science website that automatically detects and segments organoids upon uploading a CZI file containing different stack images. It is also offers functionalities such as selecting organoids of interest, computing their area, and counting the number of organoids in each stack image. The performance of the website was evaluated using two scenarios, in which OrganelX has shown the ability to accurately segment organoids. Regardless of the small variability in individual sizes, OrganelX is able to track down in time the size or multiple organoids with similar outcome as manual measurements. Moreover, OrganelX offers the possibility to easily increase the sample size (e.i. from 25 organoids up to hundreds) in order to produce a more descriptive result. 10 Segmented CZI files with more than 30,000 detected structures were made publicly available. Most organoids of up to 10000 µm2 of area can be easily detected by OrganelX. As depicted on Figure 2B, OrganelX is also able to measure organoids between 150000 µm2 and 200000 µm2. The upper limit of detection lies somewhere 200000 µm2 and 25000 µm2. It is to be noticed that for this quantification, each individual stack image has been regarded as an independent image. OrganelX has only been tested on murine liver organoids, which present a common cystic morphology. However, it is possible to also measure any other types of organoids. As long as these organoids have an approximately spherical morphology they should be detected. The method has not yet been tested on non-cystic morphologies such as murine intestinal organoids. For the future, OrganelX will be trained to work with different morphologies. DECLARATION OF COMPETING INTEREST The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
RkJQdWJsaXNoZXIy MTk4NDMw