Artificial intelligence techniques for embryo and oocyte classification

Abstract

One of the most relevant aspects in assisted reproduction technology is the possibility of characterizing and identifying the most viable oocytes or embryos. In most cases, embryologists select them by visual examination and their evaluation is totally subjective. Recently, due to the rapid growth in the capacity to extract texture descriptors from a given image, a growing interest has been shown in the use of artificial intelligence methods for embryo or oocyte scoring/selection in IVF programmes. This work concentrates the efforts on the possible prediction of the quality of embryos and oocytes in order to improve the performance of assisted reproduction technology, starting from their images. The artificial intelligence system proposed in this work is based on a set of Levenberg-Marquardt neural networks trained using textural descriptors (the local binary patterns). The proposed system was tested on two data sets of 269 oocytes and 269 corresponding embryos from 104 women and compared with other machine learning methods already proposed in the past for similar classification problems. Although the results are only preliminary, they show an interesting classification performance. This technique may be of particular interest in those countries where legislation restricts embryo selection. One of the most relevant aspects in assisted reproduction technology is the possibility of characterizing and identifying the most viable oocytes or embryos. In most cases, embryologists select them by visual examination and their evaluation is totally subjective. Recently, due to the rapid growth in our capacity to extract texture descriptors from a given image, a growing interest has been shown in the use of artificial intelligence methods for embryo or oocyte scoring/selection in IVF programmes. In this work, we concentrate our efforts on the possible prediction of the quality of embryos and oocytes in order to improve the performance of assisted reproduction technology, starting from their images. The artificial intelligence system proposed in this work is based on a set of Levenberg-Marquardt neural networks trained using textural descriptors (the 'local binary patterns'). The proposed system is tested on two data sets, of 269 oocytes and 269 corresponding embryos from 104 women, and compared with other machine learning methods already proposed in the past for similar classification problems. Although the results are only preliminary, they showed an interesting classification performance. This technique may be of particular interest in those countries where legislation restricts embryo selection.

Cytoplasmic movement profiles of mouse surrounding nucleolus and not-surrounding nucleolus antral oocytes during meiotic resumption

“…This analysis was performed ten times (tenfold cross‐validation of the FANN), and allowed to determine the fairness of the recognition as “True‐SN recognition rate” (specificity), “True‐NSN recognition rate” (sensitivity), and “SN‐NSN overall recognition rate” (accuracy). FANN specificity, sensitivity, and accuracy reflect the classification performance—that is a parameter ≤100% and ≥90% equals excellent performance; <90% and ≥80% is good; <80% and ≥70% is fair; <70% and ≥60% is poor; <60% and ≥50% is a failure (Manna, Nanni, Lumini, & Pappalardo, ).…”

Charting new frontiers in In Vitro Fertilization (IVF): The Role of Bioengineering

“…However, the attributes of 'optimal embryos' and the methods to identify embryos with a high developmental potential are still unknown entities (Cruz et al, 2012). In most cases, embryologists select embryos by visual examination and their evaluation is totally subjective and has several limitations (Manna et al, 2013;Montag et al, 2013;Santos Filho et al;, Freour et al, 2012.…”

“…Several studies exist that utilize novel imaging techniques to visualize or quantify intracellular components of gametes and embryos with the intent of correlating localization of organelles or molecular constitution with quality or outcome (Manna et al, 2013;Jasensky et al, 2011).…”

Gray level Co‐occurrence Matrices (GLCM) to assess microstructural and textural changes in pre‐implantation embryos

“…These features provide a depth of analysis that is not otherwise readily detectable; five features of particular interest are the degree of both local and global homogeneity, heterogeneity, smoothness, and randomness (Ou et al, ). A number of machine learning techniques are starting to be applied to the analysis of phase‐contrast images of embryos for noninvasive embryo quality analysis (Manna et al, ; Hernandez‐Gonzalez et al, ), but have not yet been used to survey fluorescence images of embryos.…”