Reproductive healthLecture
Artificial intelligence, femtech and automation in IVF
From at-home hormonal screening and ultrasound classifiers to CD138 detection in chronic endometritis and endometrium-on-chip — a tour of where artificial intelligence already adds value across the IVF cycle, and where its real limits sit today.
Audio: MultiSubtitles: EN · UA · RU29 min
Mykola Gryshchenko
Professor, MD, PhD
Professor, MD, PhD, and Head of Sona Academy at Sona Group; an expert in reproductive medicine focusing on medical education and scientific strategy. He initiates and leads the group's educational and scientific programmes, including the Winter School of Reproductology and satellite scientific events.
UkraineLinkedIn →
What was shown
A tour of where artificial intelligence already adds clinical value across the IVF cycle, with an explicit focus on the endometrium. The talk covers deep learning on imaging, femtech ecosystems for at-home hormonal screening, automation in routine embryology, microfluidics, and the emerging endometrium-on-chip model.
Key findings
Hertility, a UK accredited at-home femtech ecosystem, currently screens 18 reproductive conditions including 5 with direct endometrial relevance, combining patient questionnaire, home hormonal tests and AI-assisted teleconsultation. A convolutional model classifying endometrial receptivity from ultrasound reaches 95% accuracy on 402 images. AI-based CD138 counting for chronic endometritis matches expert pathologists in sensitivity, specificity and accuracy. AI-assisted hysteroscopy flags septum margins, ostia and high-perforation areas. Endometrial single-cell atlases (interactive, QR-linked) are now public. An endometrium-on-chip model with stroma, epithelium and vasculature responds to E2/progesterone dosing in a dose-dependent way.
