Using computer vision to improve care for desert jerboas with hair loss

Jim Crocker
13th November, 2025

Using computer vision to improve care for desert jerboas with hair loss

An adult jerboa from the study.

Photo adapted from: Boulanger et al. / CC BY (Source)

Key Findings

  • This study, conducted on Lesser Egyptian Jerboas, used computer vision to accurately track behaviors and assess welfare
  • Increased grooming behavior was strongly linked to alopecia (hair loss) in the jerboas, especially in smaller enclosures without visual barriers
  • Computer vision proved as accurate as human observers in identifying jerboa behaviors, particularly short actions lasting under one second, offering a more efficient analysis method
Understanding the wellbeing of animals in captivity is a significant challenge, particularly for species with fast, complex behaviours that are difficult for humans to accurately observe. The Lesser Egyptian Jerboa, a small bipedal rodent, presents this challenge. These animals are naturally active and their movements are quick, making detailed behavioural analysis difficult – a problem exacerbated by the fact they are nocturnal. Recently, researchers from the University of Michigan and University Centre Sparsholt[1] investigated whether computer vision – using artificial intelligence to ‘watch’ and categorise behaviour – could provide a more accurate and efficient way to assess the welfare of captive jerboas, specifically in relation to a skin condition called alopecia (hair loss). The study focused on validating the use of an open-source computer vision toolkit to analyse jerboa behaviour. Traditionally, assessing animal behaviour relies on human observers watching videos and manually categorising what the animals are doing. This method is time-consuming and prone to human error, especially when dealing with rapid movements. The researchers compared the accuracy of trained human observers to that of computer vision algorithms, assessing how well each could classify different behaviours. A key metric used was ‘accuracy’ – how often the correct behaviour was identified – and the ‘intraclass correlation coefficient’ (ICC), which measures the consistency of observations. The results showed that the computer vision system performed remarkably well, achieving accuracy and ICC scores comparable to those of the human observers. Importantly, the computer vision system was less affected by the duration of the behaviour, proving particularly useful for identifying very short actions – those lasting less than a second. This is significant because the study found that 34% of behaviours observed in the jerboas lasted less than half a second, highlighting a limitation of human observation. Having validated the computer vision system, the researchers used it to investigate a specific welfare concern: alopecia observed in some of the captive jerboas. They analysed activity budgets – a breakdown of how the jerboas spent their time – before and after introducing different types of environmental enrichment (items designed to improve their environment and wellbeing). They found a strong link between alopecia and increased grooming behaviour. Interestingly, grooming was more common in jerboas housed in shorter terrariums and without opaque dividers between enclosures. Conventional rodent enrichment, such as wheels and tunnels, did not have a significant impact on behaviour. Further investigation ruled out inflammatory causes for the alopecia, leading the researchers to suggest it may be ‘psychogenic’ – meaning it originates from psychological stress. This study builds upon previous research demonstrating the unique biomechanical adaptations of jerboas[2]. Jerboas are bipedal, meaning they walk on two legs, a trait they share with humans but evolved independently. Earlier work has shown that the fusion of bones in their feet[2][3] is crucial for withstanding the high impact forces experienced during bipedal locomotion, and that their escape movements are unpredictable, likely aiding predator evasion[4]. Understanding these fundamental aspects of jerboa biology and behaviour is essential for providing appropriate care in captivity. The use of computer vision in this study represents a significant advancement in animal welfare science. It allows for the rapid and accurate analysis of large amounts of behavioural data, something that would be impossible with manual observation alone. This technology can be used to tailor husbandry practices – the ways in which animals are cared for – to the specific needs of each species, ultimately improving their wellbeing. The findings regarding the link between alopecia, grooming, and enclosure characteristics highlight the importance of providing jerboas with spacious, private environments.

HealthBiotechAnimal Science

References

Main Study

1) Refining animal care through technology: Addressing alopecia in Jaculus jaculus with validated computer vision analysis

Published 11th November, 2025

https://doi.org/10.1371/journal.pone.0330143

Related Studies

2) Convergent metatarsal fusion in jerboas and chickens is mediated by similarities and differences in the patterns of osteoblast and osteoclast activities.

https://doi.org/10.1111/ede.12320

3) Metatarsal fusion resisted bending as jerboas (Dipodidae) transitioned from quadrupedal to bipedal.

https://doi.org/10.1098/rspb.2022.1322

4) Unpredictability of escape trajectory explains predator evasion ability and microhabitat preference of desert rodents.

https://doi.org/10.1038/s41467-017-00373-2


Related Articles

An unhandled error has occurred. Reload 🗙