CRYSTAL Study
Defining Data and Structural Requirements for Digital and AI-Enabled IVF Cycle Monitoring
Anusch Yazdani for the CRYSTAL Study Group
Follicle size rated essential
Oestradiol rated essential
Expert clinicians
National Delphi consensus
Study Question
What data elements and structural features are required in IVF cycle monitoring records and what is required to enable digital analytics and AI-supported clinical decision-making?
Summary Answer
The CRYSTAL Delphi consensus identified high-priority monitoring parameters and preferred data presentations to create a cycle monitoring record that supports clinical decision-making and digital analytics.
What is Known Already
"No agreed standard defining which parameters to monitor or how this information should be presented to support safe decision-making."
Clinical decisions during IVF stimulation rely on the interpretation of complex, rapidly evolving ultrasound and hormonal data, yet there is no agreed standard defining which parameters to monitor or how this information should be presented to support safe decision-making.
Digital health tools and artificial intelligence are increasingly proposed to support complex clinical decision-making. In IVF, cycle monitoring relies on multivariate, time-dependent hormonal and ultrasound data, yet documentation practices are heterogeneous, frequently unstructured and poorly standardised.
This variability limits reproducibility, clinical auditability and readiness for advanced analytics. Currently, there is no consensus framework defining the data content and structural requirements necessary to support safe, guideline-aligned digital or AI-enabled decision support in IVF cycle monitoring.
Study Design & Methods
Study Design
Participant Disciplines
- Reproductive endocrinologists
- Fertility specialists
- Trainees
- Nurses
Key Findings
Parameter importance ratings — % of respondents rating essential
used purpose-designed software
relied on handwritten records
prefer direct scanner import
An evidence-based foundation for AI-enabled reproductive care.
By defining both what should be monitored and how data should be presented, the CRYSTAL consensus provides an evidence-based foundation to inform the standardisation of data content and the structural prerequisites for digital and AI-enabled monitoring.
The consensus supports future guideline development and the safe implementation of AI-supported reproductive care — tools that augment rather than override clinician judgment.
Respondents explicitly identified interest in advanced analytics, including pattern recognition, prediction of response, and risk estimation, provided these tools complement clinical expertise.
Ethics Approval
Approved by The University of Queensland 2025/HE000908.
Keywords