The review assesses AI’s ability to predict embryo chromosome conditions using image analysis for non-invasive IVF screening. AI algorithms have shown potential in accurately predicting embryo ploidy, although further research is needed to improve reliability. A study evaluated AI models for this purpose, finding a pooled diagnostic performance of 0.67 sensitivity, 0.58 specificity, and an AUC of 0.67. Factors influencing model outcomes include algorithm type, DSS category, external validation, and sample size. While AI shows promise in supporting embryo assessments, current models lack the accuracy to replace invasive testing methods and should be used as supplementary tools.
Source link
AI shows promise for predicting embryonic health without invasive testing
