Non-invasive Detection of Chick Embryo Gender Based on Body Motility and a Near-infrared Sensor

Abstract

Culling of male day-old chicks in layer production is a big global ethical issue. As gender differences exist in the activity of human fetuses, such a physiological difference may also be true for chick embryos. Hence, this study investigated embryo (ROSS 308) gender differences based on this body motility. A near-infrared (NIR) sensor was used to measure embryo motility and separated it in frequency domain. Principal component (PC) scores from body motility strength (i.e., signal power in frequency domain) were used for gender classification using machine learning approach. The formation of sex organ and hormonal differences to be the reason for the male to be significantly more active (p ＜0.05). The findings could contribute to resolving animal welfare issue.