Thermal Imaging Detection

Thermal imaging using infra-red cameras has shown promise overseas as a non-invasive means of detecting temperature anomalies that may indicate animal distress or disease. Agriculture and Agri-Food Canada has had a research programme running for some time looking at detecting health problems in feedlot beef cattle using cameras mounted around water troughs.

AgResearch has also been working in this area. Rural Delivery has previously featured the use of IR imaging to detect pain and emotional stress in animals, and AgResearch is also looking at other uses such as determining “contentment” as well as general unwellness and specific diseases, according to Dr Jim Webster, AgResearch’s team leader, Farm Systems North.

“Some years ago we established that infrared cameras can pick up subtle changes in heat distribution and we know that many health and welfare factors change the heat emitted by an animal so it is a very useful spectrum to be observing,” he says.

“Recently the cost of IR cameras has fallen dramatically. EID is now widely used and farmers are becoming more comfortable with routine use of electronic technology, so from an animal welfare and health point of view, IR cameras now have the potential to contribute to the monitoring of animals in various environments.”

Dr Webster points out that in many farming situations the strategic placement of cameras by feed or water troughs or in dairy sheds could help compensate for the difficulties that farm staff have inspecting each animal closely, especially as the ratio of animals to people gets larger or where farm staff may not be experienced enough to recognise problems. Canadian work has shown additional benefits from feedlots cameras that can often pick up problems some days before they become obvious even to an experienced stockperson.

Calf health and mastitis detection in cows are two areas that AgResearch is currently looking at. Other uses could include any ailment that alters the heat loss pattern of an animal eg. lameness and facial eczema. Less obvious but also a possibility is looking at metabolic efficiency in animals, says Dr Webster.

“There is potentially a huge economic benefit if you can select animals that are more efficiently converting feed into protein, and the hypothesis that we have been testing is that the animals that are less efficient in turning feed into milk or growth give off more heat, like a less efficient light bulb,” he says.

“However, we know that farmers don’t want to spend any more time than they have to in monitoring stock so we want to avoid producing an alert when there isn’t a problem. It means that a lot of research needs to go into how we monitor animals taking into account natural variations and then removing ‘noise’ from the results so that when we flag the animals that are deviating from the norm that means there is a real benefit.”

At present AgResearch staff at Ruakura are remotely monitoring calves in Canterbury with cameras mounted around automated feeding stations.

“Calves are prone to diarrhoea, BVD, rotavirus and other viral diseases, and we are combining infrared technology with automatic calf feeding systems. The ability to pick up problems early is one of the key benefits of infrared because as soon as animals become infected there is a cascade of cellular reactions that cause heat, so very early in the process you can detect local heat production way before you can pick it up through rectal temperature or by observing the animal,” says Dr Webster.

“We are looking for different sorts of patterns depending on whether the problem is, say, a respiratory disease or a viral one. The aims are a reduction in deaths and improvement in welfare of sick animals and also in others that they are likely to infect if they are not promptly treated.”

IR technology lends itself well to non-invasive monitoring which offers advantages of measuring animals in an undisturbed state. Handling an animal to take conventional clinical measurements such as temperature or blood samples can start a cascade of responses that may interfere with the diagnosis itself and alter the animals’ behaviour for some time.

LIC is working with AgResearch to realise the potential for IR to detect mastitis. Cameron Marshall, LIC’s Analytics Programme Manager, says that there are two main aspects to the testing regime, which has just started. The first is to collect thermal images of cattle while they are in the farm dairy and to see how they correlate with information from other mastitis detection systems and from herd testing.

“The data we collect will help us understand the heat images and what they correlate to on an animal, specifically relating to mastitis and other health events. We also need to learn about protecting the camera in the dairy shed, ensuring that even though it can get wet and dirty, it can still get good images,” he says.

“We need to learn whether we can get good images from a single fixed position on a rotary platform when cows are moving past, so there are a lot of mechanical engineering aspects that we want to try out. Complicating factors could include diurnal temperature variations in individual cows, ambient temperature variations, differences in normal udder temperatures, and we need to understand a lot of those environmental aspects and the impacts they have. For example, is the morning sun at the dairy where we are working at present going to be a problem?”

Cameron points out the difficulties in producing a system that will detect real differences. Will a cow with subclinical mastitis go clinical, and can IR detect any difference? Some cows have a high somatic cell count but don’t develop clinical mastitis, and some are self curing.

“We don’t want to be drafting off animals and expect farmers to treat them if it might be unnecessary, so that’s the sort of thing we are trying to understand by analysing the data” he says.

“I think at the end of the day it is not going to be a silver bullet for treating mastitis, but will give you an indication when a cow is at risk and how severe her infection might be. Then you would do an RMT test or just keep an eye on her.”

By combining the IR reading with a cow’s computer records it may be possible to flag a potential problem early. If a cow has a history of getting mastitis it is more likely that a hot udder or quarter indicates infection that may soon become clinical and therefore should be treated. Ultimately it should mean more milk in the vat.

Currently an LIC scientist is programming an IR camera to work with Protrack so photos of an animal can be automatically linked to its EID and form part of its records.

“We see IR technology as having a lot of potential to tie in with existing farm automation systems and acting as another pair of eyes for farmers looking for mastitis, and there is the possibility that it could extend to other areas as well,” says Cameron.

“This is still an area of research in it’s infancy for us, until we start taking a lot of images and marrying them up with actual results from animals we won’t know how capable the system is.”