Predicting Success in the Embryology Lab: The Use of Algorithmic Technologies in Knowledge Production

Area of Science:

• Sociology of Scientific Knowledge (STS)
• Reproductive Medicine
• Health Informatics

Background:

• Time-lapse (TL) imaging is increasingly used in fertility treatments for embryo selection.
• TL technologies promise enhanced embryo knowledge and standardized selection, but evidence for improved pregnancy rates is inconclusive.

Purpose of the Study:

• To analyze local algorithmic practices in embryology labs using TL imaging.
• To investigate knowledge creation, standardization, and the impact of algorithms on professional expertise.

Main Methods:

• Ethnographic data collection across five UK clinical sites.
• Analysis of human-machine interactions within TL imaging workflows.
• Examination of standardization efforts and knowledge production.

Main Results:

• TL imaging alters laboratory routines and training requirements.
• Human input into TL algorithms creates situated uncertainties, challenging algorithmic authority.
• Algorithmic knowledge production is embedded within local practices and expertise.

Conclusions:

• Algorithms in TL imaging do not simply add knowledge but actively reshape professional practice.
• The locality of algorithms and AI requires further exploration within STS scholarship.
• Understanding the situated nature of algorithmic decision-making is crucial in clinical settings.