Design: Multi-centre prospective observational cohort study

Setting: Multiple fertility centers in three different international clinical settings.

Patients or Participants: 64 IVF/ICSI patients undergoing fresh embryo transfer from participating centres.

Interventions: Each patient underwent a four-minute B-mode transvaginal ultrasound (TVUS) scans performed one hour before embryo transfer (ET). 23 features related to frequency, amplitude, power, velocity, and coordination were extracted using strain analysis from TVUS speckle tracking results. Three probabilistic classifiers, i.e., support vector machine (SVM), K-nearest neighbors (KNN), and adaptive boosting (AdaBoost), were employed to discriminate uterine activity as either favourable or adverse to ongoing pregnancy rate (positive foetal heartbeat at 11 weeks gestation). Prior to machine learning, feature selection was performed by correlation filtering, least absolute shrinkage and selection operator regression, and sequential forward selection. The proposed method was evaluated by a nested 8-fold cross validation.

Measurements and Main Results: Age, BMI, type of treatment, stimulation protocol, duration of infertility, number of previous IVF cycles, gravidity and parity did not differ statistically significantly between pregnant and non-pregnant groups (p>0.05).Our results suggest that features related to coordination, velocity and frequency of the uterine peristalsis are strongly associated with clinical pregnancy, and SVM demonstrates the best classification performance between successful and unsuccessful pregnancies, with an average area under the receiver operating characteristic curve of 0.81.

Conclusion: We developed a machine learning framework to improve the prediction of IVF outcome based on TVUS recordings of patients from multiple centers. Our SVM model identified significant
uterine motion features and demonstrated reliable and generalisable classification performance. This work can provide useful means to support clinicians for clinical decision-making prior to ET and possibly
enhance IVF success rates.