iDAScore & embryologist experience: Reducing variability in embryo evaluation

A newly published study in F&S Reports explores a critical question in modern IVF: how does embryologist experience influence agreement with AI-driven embryo ranking, and what does this mean for clinical outcomes?

Bridging embryologist expertise and artificial intelligence

Embryo selection remains one of the most important, and complex steps in IVF. While AI-based tools are increasingly used to support decision-making, understanding how they align with embryologists across different experience levels is essential.

This study evaluated over 1,300 PGT-A cycles and compared manual embryo selection with iDAScore v2-based ranking. 

Key findings:

1. Concordance matters:

Overall, agreement between embryologists and iDAScore was observed in ~58.6% of cases, increasing with experience level.

2. Experience drives agreement: 

Junior embryologists: ~51% concordance 

Senior embryologists: ~57% 

Highly experienced (Senior+): ~63%

3. Clinical outcomes trend higher with agreement: 

Fetal heartbeat rates (FHB) were numerically higher when AI and embryologists agreed, highlighting the potential value of consistent decision-making. 

4. AI shows greatest value in complex cases: 

The benefit of AI support appears strongest when selecting from a moderate number of embryos, where decision-making is less straightforward. 

"Our latest research shows that adjunctive use of the iDAScore algorithm in embryo selection improves evaluation consistency and objectivity, offering the greatest benefit to less experienced embryologists. In a rapidly changing field with limited embryologist resources and many new professionals entering the workforce, integrating iDAScore supports efficient training, reduces subjectivity, and helps ensure consistently high-quality embryo selection and ultimately improving patient outcomes."

- Jørgen Berntsen, Senior Product Specialist, Vitrolife Technologies, Denmark.

Note: iDAScore is currently unavailable in selected markets and is not available In the United States.

 

What this means for IVF laboratories

The findings reinforce an important message: AI is not a replacement for expertise but a complement. For experienced embryologists, AI can enhance consistency and confidence. For less experienced embryologists, AI can serve as a valuable training and decision-support tool. However, across all levels, AI contributes to standardisation of embryo assessment. A previous RCT demonstrated that iDAScore evaluation was 10 times faster than standard evaluation without compromising clinical outcomes.

Supporting better workflows and outcomes

By reducing subjectivity and variability, AI-assisted embryo evaluation has the potential to:

• Improve workflow efficiency 

• Support more consistent decision-making

• Enhance overall laboratory performance

"These findings reinforce the importance of consistency in embryo assessment and highlight how decision-support tools can help reduce variability between embryologists. Standardisation in evaluation is an important step toward improving overall laboratory performance and supporting better clinical outcomes."

- Mark Larman, VP Medical Affairs, Vitrolife Group, North America

 

At Vitrolife Group, our goal is to support embryologists with tools that support performance, regardless of experience level, while maintaining full control in the hands of clinicians.

Advancing embryo selection through AI and expertise

As this study highlights, enabling more consistent and objective embryo selection, by leveraging both embryologist expertise and AI, clinics can work toward improved consistency, optimised workflows, and better patient outcomes. 

Explore the full study to learn more about how iDAScore and embryologist expertise together are shaping the next generation of fertility care.

Want to learn more about how EmbryoScope and IDAScore can support your IVF laboratory?

You may also be interested in:

Illingsworth et al., Nature Medicine volume 30, pages3114–3120 (2024)
Bori et al. Fert. Steril. 2024 Articles in Press October 14, 2024
Ueno et al. J Assist Reprod Genet. 2022; 39(9): 2089-2099