At an IVF clinic today, a developing human embryo sits inside an incubator equipped with a built-in camera, photographed automatically every five to twenty minutes around the clock, generating a continuous, unbroken visual record of its entire development — never once removed from its stable environment to be checked. An AI model then analyzes that whole morphokinetic sequence and outputs an objective viability score, no embryologist’s eye required.
At a captive breeding facility for the endangered crested ibis, one of the rarest birds on Earth, a 2026 study describes essentially the opposite approach: researchers candled 98 eggs — periodically holding them up to a bright light to see the developing embryo inside — capturing an average of just 14.5 images per egg across the entire incubation period. The study itself flags specific windows, days 7 to 15 and days 23 to 29, as critical periods when most embryo deaths occurred and that “require particular attention.” Under the current monitoring regime, those critical windows are being watched roughly twice a week, not continuously.
Scientific Foundation
Time-lapse embryo monitoring has become a mature, if still debated, technology in human IVF. Continuous imaging systems like EmbryoScope pair automated, uninterrupted photography with AI scoring algorithms — one well-cited 2019 study using such a system analyzed clinical data and roughly 50,000 images from 2,182 embryos, achieving an area under the curve above 0.98 in predicting blastocyst quality. A 2025 review is candid that time-lapse imaging’s benefit on live birth rates overall remains debated in large trials, but notes it shows clear, specific value in contexts like predicting genetic normality to reduce unnecessary biopsies, and researchers are actively working on next-generation models that adjust developmental thresholds for individual patient factors like age and hormone levels, in one case improving live birth rates by 14 percent in poor responders.
Captive breeding programs for endangered birds, meanwhile, largely still rely on candling — a technique essentially unchanged in its basic principle for centuries — performed at intervals during incubation rather than continuously. The crested ibis study is valuable precisely because it’s rare: a systematic, published reference describing normal egg development patterns and identifying exactly when mortality risk peaks, information that required manually compiling 1,422 candling images across dozens of eggs to establish. Other conservation egg-monitoring approaches reviewed here take entirely different technical routes, like using GPS and accelerometer tags on wild Cinereous Vultures to infer incubation behavior indirectly, rather than any kind of direct visual embryo monitoring, continuous or otherwise.
Cross-Domain Connection
Both fields are trying to answer the same question — is this developing embryo on a healthy track, and precisely when might it be at risk — from image data captured through a shell or a culture dish. IVF solved the frequency problem by automating continuous, non-disruptive imaging entirely, removing the need for anyone to physically intervene to check on the embryo. Endangered bird conservation, working with eggs that are typically far more limited in number and often irreplaceable in a way individual IVF embryos usually aren’t, is still relying on periodic manual candling — precisely the kind of intermittent-observation approach IVF moved away from because it can miss anomalies developing between checks and requires physically disturbing the embryo’s environment each time.
Given that the crested ibis study itself identifies specific narrow mortality windows that current biweekly-ish candling can only catch after the fact, adapting continuous, non-invasive time-lapse imaging technology — already mature, commercially available, and validated at scale in IVF clinics — into endangered species incubation could let conservation programs catch developing problems inside those critical windows in near real time, rather than discovering a failure only at the next scheduled candling session.
What Remains Undemonstrated
No research reviewed here describes deploying continuous IVF-style time-lapse imaging systems on endangered bird eggs; every conservation approach found here still relies on periodic candling, remote tracking, or other indirect methods. There are real technical differences that complicate a direct transfer: avian eggs are opaque shells rather than the optically accessible culture dishes IVF embryos develop in, meaning continuous imaging would likely require either specialized shell-penetrating imaging techniques or accepting a much lower-resolution signal than the direct microscopic view IVF systems rely on. It’s also unproven whether the AI morphokinetic models trained on human embryo development would have any transferability to bird embryo development, given the two are governed by fundamentally different developmental biology, or whether entirely new species-specific models — the kind of painstaking, image-by-image groundwork the crested ibis study represents just one species’ worth of — would need to be built essentially from scratch for every endangered species a program hoped to monitor this way.
Why It Matters
Every egg lost in a captive breeding program for a critically endangered species represents a meaningfully larger fraction of that species’ entire future than a single lost IVF embryo represents for human reproduction broadly — the stakes per individual egg are, in a real sense, higher. A field that has already built and validated continuous, non-invasive, AI-scored embryo monitoring at clinical scale has technical infrastructure that conservation programs working with far scarcer, far more consequential eggs currently don’t have access to.
The Human Dimension
There’s something worth noticing in the fact that a technology built to give hopeful parents peace of mind about one embryo at a time has, quietly, gotten far more sophisticated at watching over developing life than the tools protecting some of the rarest birds left on the planet. Fewer than 30 crested ibis eggs hatched in that one study’s breeding season. Each one mattered enormously. They deserved, if anything, more continuous attention than a single IVF embryo gets — not less.
Sources:
1. “EmbryoScope time-lapse system,” Vitrolife: https://www.vitrolife.com/why-vitrolife/the-patient-ivf-journey/embryoscope-time-lapse-system/
2. “Time-Lapse Imaging in IVF: Bridging the Gap Between Promises and Clinical Realities,” International Journal of Molecular Sciences (MDPI): https://www.mdpi.com/1422-0067/26/19/9609
3. “Artificial Intelligence in Routine IVF Practice,” PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC12784763/
4. “Use of time-lapse technology and artificial intelligence in the embryology laboratory: an updated review,” PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC12225204/
5. “Candling Analysis of Egg Development in an Endangered Bird Species Crested Ibis (Nipponia nippon),” Ecology and Evolution: https://onlinelibrary.wiley.com/doi/full/10.1002/ece3.73797
6. “Annual Review 2024/25,” Institute of Zoology, Zoological Society of London: https://cms.zsl.org/sites/default/files/2025-12/IOZ%20Annual%20Review%202024-2025.pdf
7. “Using GPS and accelerometer data to precisely record egg laying, incubation and chick hatching of Cinereous Vultures in-situ,” PMC: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12246797/
8. “Breaking the Egg Barrier: Synthetic Eggs and the Future of De-extinction,” Earth.com: https://science-technology.news-articles.net/content/2026/05/19/breaking-the-egg-barrier-synthetic-eggs-and-the-future-of-de-extinction.html
Idea originated at artificialideas.org. Article researched and written by Claude Sonnet 5. Published at artificialideas.org
Artificial Ideas 