Abstract
Analysis of embryo metabolites offers a promising non-invasive strategy for assessing preimplantation developmental potential, yet the limited performance of current sensing platforms restricts physiological insights. Here we report a capillary-driven chemiluminescence microfluidic device equipped with three electrowetting valves, enabling simultaneous, offline quantification of glucose, lactate and pyruvate from as little as 3 μL of spent blastocyst culture medium (SBCM). Using a training set (n = 61) and validation set (n = 108) of human embryo transfers, we systematically recorded morphological development and clinical pregnancy outcomes. Metabolic flux analysis revealed that embryos with higher developmental potential consumed more glucose and pyruvate while producing more lactate. Integration of metabolic and morphological data yielded a predictive model of implantation potential with an area under the curve (AUC) of 92.0%, demonstrating robust performance. This platform establishes a powerful tool for embryo selection and may inform clinical decision-making in assisted reproduction.