EUVIMEDEUVIMED entdecken
Crossref

L26/P-058 A Comparative Study of Artificial intelligence (AI)-Assisted and Embryologist-Selected Sperm Using Sibling Oocytes: Embryological Outcomes Following Intracytoplasmic Sperm Injection- A Pilot Study

A. Atared

Human Reproduction · 01.07.2026 · 41 (Supplement_1)

DOI: 10.1093/humrep/deag083.394 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von A. Atared.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Study question Does AI-assisted sperm selection incorporating motility, morphology, and non-invasive DNA fragmentation assessmentimprove embryological outcomes in ICSI cycles compared withconventional manual selection? Summary answer AI-assisted sperm selection represents a promising, objective approach that yields embryological outcomes comparable to manual selection, with potential advantages in cases of severe male-factor infertility. What is known already Sperm selection for ICSI has traditionally relied on the embryologist’s subjective assessment of sperm motility and morphology. Recently, AI-based sperm selection systems have been introduced to enhance objectivity and reproducibility.However, most currently available models focus primarily on conventional parameters, such as motility and morphology, and their impact on embryological and clinical outcomes remains uncertain. A newly developed AI system (GC-ViT; VitruvianMD) integrates analysis of motility and morphology with a validated, non-invasive assessment of sperm DNA fragmentation, providing a more comprehensive sperm selection approach that may help identify sperm with higher genomic integrity without compromising viability. Study design, size, duration This single-center, observational pilot study using a sibling-oocyte design was conducted in Saudi Arabia between March and May 2025. Two sperm selection approaches for ICSI were compared: conventional selection by senior embryologists based on motility and morphology (Group 1, control) and AI-assisted sperm selection incorporating motility, morphology, and DNA fragmentation assessment using the GC-ViT system (Group 2). Normal fertilization and usable blastocyst rates were evaluated. Participants/materials, setting, methods A total of 32 treatment cycles were included, involving patients with at least eight mature oocytes (mean female age 35.4 years). Sibling metaphase-II oocytes were randomly allocated to conventional embryologist-selected ICSI (Group 1; n = 276 oocytes) or AI-assisted sperm selection (Group 2; n = 279 oocytes). The GC-ViT system utilized a two-step process: initial screening based on motility/morphology at × 200 magnification, followed by DNA fragmentation assessment at × 400 magnification using a validated, non-damaging method. Main results and the role of chance In Group 1 (manual selection), the normal fertilization rate per mature oocyte was 63.3%, and the blastocyst utilization rate per 2PN embryo was 49%. Comparable outcomes were observed in Group 2 (AI-assisted selection), with fertilization and blastocyst utilization rates of 65.8% and 56.0%, respectively. In a subgroup analysis of cycles involving severe oligoasthenoteratozoospermia or testicular sperm, AI-assisted sperm selection demonstrated a trend toward higher fertilization rates compared with manual selection (72.2% vs. 56.5%), while blastocyst utilization rates remained comparable (56% with AI vs. 48% with manual selection). To date, embryos derived from AI-selected sperm have resulted in nine live births or ongoing pregnancies. Given the pilot nature and limited sample size, no formal statistical significance testing was performed, and results are presented descriptively. Limitations, reasons for caution The principal limitation of this study is the small sample size, which limits statistical power and precludes definitive conclusions. As a pilot investigation, these findings should be considered exploratory and require confirmation in larger, well-powered prospective studies. Wider implications of the findings AI-assisted sperm selection incorporating non-invasive DNA fragmentation assessment may help standardize ICSI practices and reduce operator-dependent variability. This approach appears particularly promising in severe male-factor infertility, offering an objective method for selecting sperm with favorable genomic integrity without adversely affecting laboratory workflow. Trial registration number Yes

Zitieren

APA

Atared, A. (2026). L26/P-058 A Comparative Study of Artificial intelligence (AI)-Assisted and Embryologist-Selected Sperm Using Sibling Oocytes: Embryological Outcomes Following Intracytoplasmic Sperm Injection- A Pilot Study. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.394

Vancouver

Atared A. L26/P-058 A Comparative Study of Artificial intelligence (AI)-Assisted and Embryologist-Selected Sperm Using Sibling Oocytes: Embryological Outcomes Following Intracytoplasmic Sperm Injection- A Pilot Study. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.394

BibTeX
@article{Atared2026,
  author = {Atared, A.},
  title = {L26/P-058 A Comparative Study of Artificial intelligence (AI)-Assisted and Embryologist-Selected Sperm Using Sibling Oocytes: Embryological Outcomes Following Intracytoplasmic Sperm Injection- A Pilot Study},
  journal = {Human Reproduction},
  year = {2026},
  volume = {41},
  number = {Supplement_1},
  doi = {10.1093/humrep/deag083.394}
}
Kostenlos starten →