vix.ing · top · new · best · stats · spec

High spatially sensitive quantitative phase imaging assisted with deep\n neural network for classification of human spermatozoa under stressed\n condition

2020/02/18 by Ankit Butola, Butola, Ankit, Daria Popova +19
Physics and Astronomy · Agricultural and Biological Sciences · Medicine · #Digital Holography and Microscopy #Seed Germination and Physiology #Sperm and Testicular Function

paper · pdf · doi:10.48550/arxiv.2002.07377

Abstract

Sperm cell motility and morphology observed under the bright field microscopy\nare the only criteria for selecting particular sperm cell during\nIntracytoplasmic Sperm Injection (ICSI) procedure of Assisted Reproductive\nTechnology (ART). Several factors such as, oxidative stress, cryopreservation,\nheat, smoking and alcohol consumption, are negatively associated with the\nquality of sperm cell and fertilization potential due to the changing of\nsub-cellular structures and functions which are overlooked. A bright field\nimaging contrast is insufficient to distinguish tiniest morphological cell\nfeatures that might influence the fertilizing ability of sperm cell. We\ndeveloped a partially spatially coherent digital holographic microscope\n(PSC-DHM) for quantitative phase imaging (QPI) in order to distinguish normal\nsperm cells from sperm cells under different stress conditions such as\ncryopreservation, exposure to hydrogen peroxide and ethanol without any\nlabeling. Phase maps of 10,163 sperm cells (2,400 control cells, 2,750\nspermatozoa after cryopreservation, 2,515 and 2,498 cells under hydrogen\nperoxide and ethanol respectively) are reconstructed using the data acquired\nfrom PSC-DHM system. Total of seven feedforward deep neural networks (DNN) were\nemployed for the classification of the phase maps for normal and stress\naffected sperm cells. When validated against the test dataset, the DNN provided\nan average sensitivity, specificity and accuracy of 84.88%, 95.03% and 85%,\nrespectively. The current approach DNN and QPI techniques of quantitative\ninformation can be applied for further improving ICSI procedure and the\ndiagnostic efficiency for the classification of semen quality in regards to\ntheir fertilization potential and other biomedical applications in general.\n

Related