Chick embryo growth modeling using near-infrared sensor and non-linear least square fitting of egg opacity values

Abstract

Non-destructive monitoring of chick embryonic growth can provide vital management insights for poultry farmers and other stakeholders. Although non-destructive studies on fertility, hatching time and gender have been conducted recently, there has been no available method for embryonic growth observation, especially during the second half of incubation. Therefore, this work investigated the feasibility of using near-infrared (NIR) sensor-based egg opacity values—the amount of light lost when passing through the egg—for indirectly observing embryo growth during incubation. ROSS 308 eggs were selected based on size, mass and shell color for this experiment. To estimate the embryo size precisely, we fit various mathematical growth functions during incubation, based on the opacity value of incubated eggs. Although all the growth models tested performed similarly in fitting the data, the exponential and power functions had better performances in terms of co-efficient of determination (0.991 and 0.994 respectively) and RMSE to explain embryo growth during incubation. From these results, we conclude that the modeling paradigm adopted provides a simple tool to non-invasively investigate embryo growth. These models could be applied to resolving developmental biology, embryonic pathology, industrial and animal welfare issues in the near future.