IVF Embryo Aneuploidy Predictors from Morphology and Biomarkers

124 embryos from a single fertility center collected between December 2021 and December 2022 were analyzed by Achmad Kemal Harzif. This dataset contains measurements linking embryo morphology, morphokinetic timings, cell-free DNA, and microRNA expression to chromosomal ploidy status. It was used to build predictive models for aneuploidy, achieving an AUC of 0.745 for miRNA-191.

Use Cases

• Predict embryo aneuploidy status using miRNA-191 expression levels and expansion category.
• Model the relationship between pronuclear fading time (e.g., 21.9 vs. 19.9 hours) and ploidy outcome.
• Analyze the correlation between cell-free DNA metrics and chromosomal status for non-invasive PGT-A.
• Build a classifier using third cell cycle synchrony duration (13.4 vs. 17.5 hours) and miRNA-372 fold change.
• Validate predictive performance via ROC analysis metrics like AUC, sensitivity, and specificity reported for miRNA-191 and miRNA-372.

Strengths

• Contains 124 embryo samples with confirmed ploidy status (50.8% aneuploid).
• Reports specific, statistically significant fold changes for biomarkers like miRNA-191 (8.1 fold) and miRNA-372 (6.5 fold).
• Includes model performance metrics (AUC, sensitivity, specificity) for key predictors.

Limitations

• Small sample size of 124 embryos limits statistical power and generalizability.
• Data is from a single center over a one-year period, introducing potential geographic and temporal bias.
• Specific column names and raw feature values are not provided in the description.

Provenance

Source

Achmad Kemal Harzif via figshare.

Collection Method

Cross-sectional study using non-probability consecutive sampling of day-5 blastocysts.

Time Range

December 28, 2021 to December 31, 2022.

Freshness

Data covers collection up to December 2022; last platform update was April 2026.

Geography

Single, unspecified fertility center.