Infertility is the failure to achieve a clinical pregnancy after 12 months or more of regular unprotected intercourse, while subfertility means a delay in achieving pregnancy. Several factors can contribute to these conditions, such as autoimmune diseases like systemic thyroid and celiac diseases, rheumatoid arthritis, systemic lupus erythematosus, and endometriosis. Some of them may overlap, sharing a common immunological milieu.

In this scenario, medically ART can help women carry on pregnancies if they are in stable disease remission and receive adequate treatment. However, the number of oocytes that can be ovulated under stimulation is highly variable, so the outcome of any IVF is linked to many factors, not only biological but also connected to the mean success rate of the fertility center.

For these reasons, the need to find methods for identifying the best gametes and embryos to guarantee a successful pregnancy is becoming pressing. AI can be essential to develop prediction models for pregnancy outcomes for patients undergoing ART treatments.

AI currently represents a modern way to obtain algorithms from a set of observable data that could help clinical decisions

Inclusion Criteria:

• Women with age below 43 years old
• IVF cycle with fresh semen sample that could be used
• progressive motile spermatozoa expected count of minimum 1 million following density gradient purification on the day of oocyte pick up

Exclusion Criteria:

• Women with uterine abnormalities
• Women with chromosomal and genetic abnormalities
• Positivity to serological test (HCV, HBV, HIV, VDRL-TPHA)

Women undergoing ART procedures (n= 200, ) at SSD Ostetricia e Ginecologia 2 - PMA of Fondazione IRCCS Policlinico San Matteo, Pavia (Italy)