IDEA Bio-Medical learned of this challenge from our clients, so we created a novel deep-learning AI (artificial intelligence) algorithm to automatically analyze their images for them. The AI training process is complex and error-prone, requiring substantial time and effort that our biology-focused clients were not able to do themselves. The AI we have developed is now able to robustly identify objects like the zebrafish contour and its  internal anatomy that are clearly visible to us humans.  We now want to make this tool available to all zebrafish researchers.

Related Stories

Many processes investigated using zebrafish take place or begin in specific organs or regions within the body.  They can be marked with fluorescence labeling to image the tissue or cells of interest. But these must be distinguished from undesired signals and autofluorescence.

For example, in the zebrafish larva, hematopoietic stem cells (HSC) originate in a specific area within the tail termed the caudal hematopoietic tissue (CHT).  It is analogous to the mammalian fetal liver and relevant to leukemia in humans.  Studies looking at proliferation of GFP-labeled HSC need to focus specifically on the tail and ignore those fluorescent signals present elsewhere, such as circulating cells and yolk sac autofluorescence.  

Athena is currently the only software enabling high-content screening in this context.  It permits automated cell counts exclusively within specific parts of the fish to maximize both throughput and data extracted.  It stands in contrast to common approaches such as simple intensity thresholding, which ignores the visible anatomy.  Instead, all fluorescence is gathered together, irrespective of its origin, producing noisy data.  Manual selection of the tail would improve quality, but severely limit achievable throughput.  

Athena Zebrafish is a stand-alone software developed and offered by IDEA Bio-Medical to empower all zebrafish researchers.  It enables true high-content imaging in zebrafish for the first time.

Athena permits parameter-free zebrafish analysis using simple bright–field images.  It automatically detects zebrafish embryos and larvae up to 5 days old (dpf), extracting the fish contour and much of its internal anatomy (yolk sac, eye, notocord and more), along with body regions of the head, trunk and tail.  For each of these objects, the software measures the morphology (area, length, and shape) and can detect fluorescence in associated color channels. Both fluorescence intensity and spot/structure detection within specific anatomy are supported.  The software is suited for a broad range of researchers and accepts multiple image format types output from nearly all microscope manufacturers.  

A free 14-day trial version of Athena Zebrafish is available for download by new users from the IDEA Bio-Medical website.