Human and animal health in mutual interaction
Better picture of sick animals
Sick animals behave differently from healthy ones. But how can you monitor animals’ behaviour during infectious disease trials without having to keep watch over them around the clock? WUR researchers are working on methods for tracking such behaviour. This will make animal trials more reliable, help refine and cut back on animal testing and improve animal welfare.
Infectious diseases such as bird flu are a threat to the health of humans and animals alike, and have a major impact on animal welfare. Trials with animals help researchers to learn more about infectious diseases and develop vaccines. “In these trials, the animals are infected and their symptoms monitored,” explains Harmen Doekes, a researcher and lecturer at Animal Breeding and Genomics. Aspects such as the temperature, blood values and behaviour are all recorded to determine how a pathogen such as a virus affects the animal. “The disadvantage is that you can’t spend all day next to the enclosure observing the animals,” says Doekes. “That means our measurements are only snapshots.”
Additionally, the presence of humans can also influence the animals, not just in their behaviour but also in their body temperature, for example.
Being handled can be stressful for them, which makes the body temperature rise. Doekes: “If we can carry out continuous measurements without anybody having to be present, the measurements will be more objective and the animals will experience less stress. This will also reduce the number of animal trials as we will be able to collect much more information per animal. When they are sick, animals can become less active and withdraw from the others, but they can also become more restless. So their behaviour doesn’t just tell us about the progression of the disease in that animal but also about contact with other animals in their group and therefore the risk of them infecting others.” By monitoring behaviour, the researchers get a better understanding of the effect and spread of the disease, and they may therefore need fewer animals to achieve the same research results.
‘If we can use sensors for continuous, automated measurements, the measurements will be more objective and less stressful for the animals’
Doekes and his colleagues investigated three methods for continuous monitoring of animal activity with the aid of various sensors and other systems. For instance, they studied pixel changes in cat videos to measure group activity before and after infection with COVID. “If there is no movement in the video, the intensity of the various pixels remains the same. The more movement there is, the more pixels change per unit of time. So the pixel changes let us estimate the overall movement in the video. The disadvantage is that we don’t get any information about the behaviour of individual cats. And that is crucial in research on infectious diseases. For example, if you want to know how a disease is passed from one animal to another, you need to know which animal is which.”
In a second experiment, the researchers used ultrawide-band (UWB) sensors to measure the activity of sheep and pigs. They placed the sensor around the animal’s neck or ear. The sensor then communicates with what are termed ‘anchors’ fitted to the barn walls. “In contrast to the pixel method, this gives you information about individual animals,” says Doekes. “You can see where the animal is in the area at different times and that lets you monitor its movement. But all you know is whether or not the animal is moving. If it does move, you still don’t know exactly what it is doing.” That is why the researchers combined the sensors with camera images. That produced relatively accurate data on individual animals.
‘Information about the behaviour of individual animals is important in efforts to reduce animal testing’
The UWB method has downsides too, explains Doekes. “The sensor weighs 26 grams, so it is too heavy for chicks, for example. And you need to fit each animal with a sensor separately.” Another disadvantage is that the anchors need to be suspended at least 20 centimetres from the wall. That means if an animal stands next to the wall in the enclosure, it will be in the system’s blind spot. According to Doekes, it also takes a lot of effort to install the equipment. “We usually run trials of two to three weeks. Because we work with infectious diseases, we have to sterilise everything. That means between each trial, everything has to be removed and cleaned thoroughly, which is hugely labour-intensive and time-consuming.”
Doekes therefore prefers to work with as little equipment as possible in the barn. “Ideally we want a system that identifies the animals and combines this with video images.” Facial recognition doesn’t exist for animals, so in their following experiment the researchers used QR codes. Each animal was given its own QR code on a mini-rucksack to let the computer recognise individuals in the camera images.
In autumn 2022, the scientists started an experiment with sheep, using both accelerometers and these QR codes. An accelerometer is a small rectangular sensor weighing about 11 grams. “While the UWB sensor measures the animal’s position, the accelerometer records the movement in three directions but can’t tell the animal’s location. The accelerometer doesn’t need anchors in the barn, which means it requires less effort as a method,” says Doekes. The researchers attached two accelerometers to each sheep: one to an ear and one on the sheep’s back. “As expected, the sensor on the ear recorded much more movement than the one on the back. But the activity patterns over time were comparable for the ear and the back.” This is an important finding for the researchers because it is easier to attach a sensor to the ear in some animal species. Doekes: “Ear movements are also interesting in their own right because animals use them to express their emotions. Hopefully we can learn more about that.”
‘By monitoring behaviour, researchers get a better understanding of the effect of diseases’
The results so far from the experiment with sheep show the benefits of continuous automated measurements. “During the period three to ten days after the infection with Toxoplasma, the sheep were less active during the day (the usual active period for sheep) and exhibited increased activity at night (which is usually when they rest) – whereas the people looking after the animals had only scored the sheep as slow/lethargic during the daily measurement moments from days five to seven.”
At present, the scientists are analysing the video images with the QR codes to get a picture of specific kinds of behaviour, such as eating and drinking. The researchers want to develop the method further in follow-up studies so that it can be used for other species such as chickens and cows.
According to Doekes, the above methods are not just useful for getting a better understanding of the relationship between disease and activity and behaviour. “They also let us detect subclinical symptoms in an early stage, or even
phenomena that we would have missed altogether, such as subtle changes in the animal-specific behavioural patterns.” Another important goal is improving animal welfare, says Doekes. “In animal testing, we have what are termed humane end points, in which you decide to put the animal down as otherwise it would be suffering unbearably. The sensors allow us to monitor the animals continuously and if things are not going well, we can spot that sooner.” In future, Doekes hopes it will also be possible to process the data in real time. “Then the animal carer will get a text message, for example, to check up on an animal as it hasn’t moved much in the past few hours.”
“My hope is that we can use the lessons learned over the next few years to implement a real-time monitoring and warning system for a range of animal species that will let us optimise the trials with infectious diseases and possible reduce the number of trials required.”
Share this article
WHO Harmen Doekes, Researcher and Professor Animal Breeding & Genomics
RESEARCH Activity tracking in infectious disease trials
TEAM Norbert Stockhofe, Rineke Klaassen-de Jong, Ronald Petie
The animals portrayed in these pictures are not the ones that were part of the research project, since a photoshoot would disrupt the research environment too much.
MORE INFORMATION This project is part of the Next Level Animal Sciences (NLAS) innovation programme.
Participating researchers of Wageningen University & Research collaborate with various partners to develop new research methods and technologies within the field of animal sciences. NLAS consists of three research directions, namely sensor technology, complex cell systems and data and models.
