IVF bovine oocyte classification and selection

The current technology for in vitro fertilization (IVF) has a long history of merely 40 years. The addition of all live births from either fresh or frozen oocytes transferred from a single oocyte collection is known as the IVF productivity rate. The ability to select the best quality eggs is the key to minimizing the number of oocytes generated, procedural cycles for potential failed attempts, and lowering cost and maximize efficiencies. The current standard for egg selection is based on defined morphological criteria. Technicians usually look for an identifiable pattern such as cytoplasm, perivitelline space, zona pellucida, size, etc. Concomitantly oocyte is also screened using genetic analysis that enhances selectivity. However, the same selection procedure is not possible for oocytes. It also has been shown that eggs with similar morphology performance might have drastically different metabolism. Here we proposed a solution where we focus on developing advanced non-invasive imaging methods to map the energy metabolism of oocytes using fluorescent lifetime imaging (FLIM). Combining multiphoton excitation, hyperspectral imaging, and machine learning approach to minimize instrumenting costs, minimize laser exposure to the bovine oocyte, and maximize predictably results for oocyte selection whether fresh or frozen. Hyperspectral imaging (HI) offers the capability of collecting intensity of autofluorescence signals from the oocytes with less laser exposure minimizing the risk of photo-toxicity. Using the phasor approach to HI enables the use of multiparametric analysis of the emission in the range of 350-500 nm range (in the range of emission of metabolic cofactors in oocytes). Complete profiling of the entire egg in three-dimensional space shows its morphological and metabolic features is then generated, showing significant differences for super-stimulation with follicle stimulating hormone, in-vitro maturation, IVF, and in-vitro culture produced oocytes.