This week’s blog post was written by guest blogger, Liz Ashforth. For this post, Liz discovered research that shows how computer vision and signal processing algorithms can make health exams safer for both non-domestic animals and their caretakers.

A stethoscope is the tool of choice when a doctor listens to your heart rate and respiration. But when a lion gets a physical exam, the veterinarian can’t just say, “OK, take a deep breath.”

Obtaining vital signs of animals is crucial for assessing their health. But for many animals, it can be quite challenging. Typically, the animals are captured and tranquilized, their vital signs recorded, and then the animals are monitored as the tranquilizer wears off. This process causes the animal stress and can lead to injury. It can be dangerous for their caretakers as well. A new computer vision-based system enables veterinarians and zookeepers to safely and accurately take an animal’s vital signs from a distance. 

Image credit: Kevin Pluck, via Wikipedia

New non-invasive technique safely measures animal vital signs 

As reported by Scientific American, Javaan Chahl, a remote sensing engineer at the University of South Australia, built a device that can capture and analyze images providing insight into animals’ vital signs. The device consists of a high-resolution digital video camera, a telephoto lens, and a signal processing program. MATLAB analyzes the images remotely to determine an animal’s breathing and heart rates.

And it works for more than just lions. According to the journal Sensors, subject animals included the following species: Giant panda, African lions, Sumatran tiger, koala, red kangaroo, alpaca, little blue penguin, Sumatran orangutan, and a Hamadryas baboon. 

Figure 1 Data collection from 10 zoo animals, (a) giant panda, (b) African lion, (c) African lioness, (d) Sumatran tiger, (e) koala, (f) red kangaroo, (g) alpaca, (h) penguin, (i) Sumatran orangutan, and (j) Hamadryas baboon. Image credit: Sensors (Basel). 2019 Dec; 19(24): 5445.

“We’ve done quite a range of animals,” Chahl says. He believes the method could be used on almost any species, including very small ones. “The nice thing about light-based imagery, as opposed to radar or thermal cameras, is that you can go from looking at the moon to looking at bacteria just by changing lenses,” he says. Another benefit of light-based imagery is its high resolution, which allowed Chahl and his colleagues to monitor animals from 10 to 130 feet away. 

Chahl’s experiments extracted signals from the abdominal-thoracic region. In this region, cardiopulmonary activity, such as expansion in the chest, is observable via the images captured by the digital camera. MATLAB analyzes the images and filters out “noise” from the animal’s movement enabling scientists and veterinarians to successfully analyze the animal’s heart rate and breathing rate.

The graphical user interface (GUI) main panel of the experimental proposed image analysing system. Image credit: Sensors (Basel). 2019 Dec; 19(24): 5445.

Temporal signal analysis of the giant Panda, (a) One minute observed time series, (b) the frequency spectrum, (c) the denoised signal using wavelet, (d) the smoothed signal using moving average filter with span equal to 5, (e) the cardiac signal after applying a band-pass filter of 1.1667 to 2 Hz, and (f) the breathing signal after applying a band-pass filter. Image credit: Sensors (Basel). 2019 Dec; 19(24): 5445.

While the program does not provide meaningful results when an animal is running, it is a promising technology for further adoption for both animal and human use. 

Future applications are diverse 

The remote imaging technique to assess vital signs could have many beneficial uses in veterinary practice, conservation and game management, animal welfare, and zoological studies. There could also be human applications for this technology. For example, remote sensing could help with mass health screenings, which is very applicable to today’s COVID-19 screenings at airports and other public spaces. It could also be used in a neonatal unit as a substitute for electrodes that are taped to the newborn.

Ultimately, Chahl’s method can simplify the collection of health data, helping many animals and humans minimize and avoid stressful and potentially damaging physical contact. Now, who wouldn’t want that?