Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Randomized Controlled Trial
. 2024 Nov;30(11):3114-3120.
doi: 10.1038/s41591-024-03166-5. Epub 2024 Aug 9.

Deep learning versus manual morphology-based embryo selection in IVF: a randomized, double-blind noninferiority trial

Peter J Illingworth  1 , Christos Venetis  2   3   4 , David K Gardner  5   6 , Scott M Nelson  7   8 , Jørgen Berntsen  9 , Mark G Larman  10 , Franca Agresta  11 , Saran Ahitan  12 , Aisling Ahlström  13 , Fleur Cattrall  5 , Simon Cooke  2 , Kristy Demmers  14 , Anette Gabrielsen  15 , Johnny Hindkjær  16 , Rebecca L Kelley  5 , Charlotte Knight  2 , Lisa Lee  5 , Robert Lahoud  2 , Manveen Mangat  2 , Hannah Park  17 , Anthony Price  18 , Geoffrey Trew  8   19 , Bettina Troest  20 , Anna Vincent  8 , Susanne Wennerström  13 , Lyndsey Zujovic  12 , Thorir Hardarson  10
Affiliations
Randomized Controlled Trial

Deep learning versus manual morphology-based embryo selection in IVF: a randomized, double-blind noninferiority trial

Peter J Illingworth et al. Nat Med. 2024 Nov.

Abstract

To assess the value of deep learning in selecting the optimal embryo for in vitro fertilization, a multicenter, randomized, double-blind, noninferiority parallel-group trial was conducted across 14 in vitro fertilization clinics in Australia and Europe. Women under 42 years of age with at least two early-stage blastocysts on day 5 were randomized to either the control arm, using standard morphological assessment, or the study arm, employing a deep learning algorithm, intelligent Data Analysis Score (iDAScore), for embryo selection. The primary endpoint was a clinical pregnancy rate with a noninferiority margin of 5%. The trial included 1,066 patients (533 in the iDAScore group and 533 in the morphology group). The iDAScore group exhibited a clinical pregnancy rate of 46.5% (248 of 533 patients), compared to 48.2% (257 of 533 patients) in the morphology arm (risk difference -1.7%; 95% confidence interval -7.7, 4.3; P = 0.62). This study was not able to demonstrate noninferiority of deep learning for clinical pregnancy rate when compared to standard morphology and a predefined prioritization scheme. Australian New Zealand Clinical Trials Registry (ANZCTR) registration: 379161 .

PubMed Disclaimer

Conflict of interest statement

Competing interests P.J.I. has provided paid advice to Western Australia government on medical cases and has previously (until June 2022) held shares in Virtus Health, a commercial IVF company. C.V. has received consulting fees, lecture payments and travel expenses from pharmaceutical firms with a commercial portfolio in IVF, and has previously (until June 2022) held shares in Virtus Health, a commercial IVF company. D.K.G. has received a research grant from Vitrolife to develop culture medium. S.M.N. has received consulting fees, lecture payments and travel expenses from pharmaceutical firms with a commercial portfolio in IVF. J.B. and M.G.L. are employees of Vitrolife and own shares in Vitrolife, who have a pending iDAScore patent. R.L.K. has received a grant from a pharmaceutical firm with a commercial portfolio in IVF, has received speaking fees and travel expenses from Vitrolife, who have a pending iDAScore patent. R.L. has received lecture payments and travel expenses from pharmaceutical firms with a commercial portfolio in IVF. M.M. has received lecture payments and travel expenses from pharmaceutical firms with a commercial portfolio in IVF and has previously (until June 2022) held shares in Virtus Health, a commercial IVF company. H.P. and B.T. are employed by a company where Vitrolife funded research. T.H. was previously employed by Vitrolife, who have a pending iDAScore patent. The other authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Patient flow diagram.
The number of patients randomized in the electronic case report form (eCRF) to the study group and the control group then, following exclusions for discontinued intervention, the number analyzed per protocol.
Fig. 2
Fig. 2. Non-inferiority analysis and risk ratio per center.
a, Crude and adjusted risk difference (red circle represents mean risk difference and blue error bars represent 95% CI of the risk difference) in clinical pregnancy rates between the iDAScore and the control groups (a negative risk difference indicates lower clinical pregnancy rate in the iDAScore group) per intention-to-treat (ITT) analysis (n = 1,066) and PP analysis (n = 1,002). b, Risk ratio (red circle represents mean risk ratio and blue error bars represent 95% CI of the risk ratio) of clinical pregnancy rate for each participating center (a ratio >1 indicates better results in the iDAScore group).
Fig. 3
Fig. 3. Time use for day 5 evaluation in the iDAScore group and control group.
Mean and standard deviation for all patients (n = 38) and for embryo cohort sizes from 2–5 (n = 9), 6–9 (n = 16) or 10 or more (n = 13) in the treatment.
See this image and copyright information in PMC

References

    1. Topol, E. J. High-performance medicine: the convergence of human and artificial intelligence. Nat. Med.25, 44–56 (2019). - PubMed
    1. Lang, K. et al. Artificial intelligence-supported screen reading versus standard double reading in the mammography screening with Artificial Intelligence trial (MASAI): a clinical safety analysis of a randomized, controlled, non-inferiority, single-blinded, screening accuracy study. Lancet Oncol.24, 936–944 (2023). - PubMed
    1. Alpha Scientists in Reproductive Medicine and ESHRE Special Interest Group of Embryology. The Istanbul consensus workshop on embryo assessment: proceedings of an expert meeting. Hum. Reprod.26, 1270–1283 (2011). - PubMed
    1. Storr, A., Venetis, C. A., Cooke, S., Kilani, S. & Ledger, W. Inter-observer and intra-observer agreement between embryologists during selection of a single day 5 embryo for transfer: a multicentre study. Hum. Reprod.32, 307–314 (2017). - PubMed
    1. Steptoe, P. C. & Edwards, R. G. Birth after reimplantation of a human embryo. Lancet2, 366 (1978). - PubMed

Publication types

LinkOut - more resources