Can a miniature device based on artificial intelligence replace a qualified veterinarian and become a real find for modern sheep farmers? This idea comes to the fore in the story of a new AI stethoscope that researchers at Olds College in Canada are testing in partnership with the scientific team at Vibrome Canada. The technology promises to enable the detection of parasitic diseases and pregnancy in sheep in a farmer-friendly format, raising questions about the future of diagnostics in animal husbandry.

Modern farmers are increasingly facing a shortage of skilled professionals who can quickly and effectively diagnose diseases and pregnancy in animals. Experts estimate that the prevalence of parasitic infections in sheep in Canada results in significant economic losses each year. Many farmers cannot afford regular ultrasounds, which limits their ability to identify problems promptly and plan farm workflows. Meanwhile, there is a growing goal to optimize costs and ensure herd health using affordable technologies.

How imPulse Una works – technology and meaning

The breakthrough came with the development of the compact AI stethoscope imPulse Una, which combines sensors and machine learning methods.

The device is capable of detecting a wide range of vibrations, from infrasound (low-frequency vibrations that humans cannot hear) to ultrasound (used in medical diagnostics). The device analyzes various acoustic signals, including those that are not perceptible to the human ear. All data is sent to the Vibrome cloud system and processed by algorithms.

According to researcher Yaogeng Lei, the collected signals can reveal a lot about an animal’s health, physiology, and even age. However, for the training algorithm to produce reliable results, a large amount of data must first be collected — this is the stage at which the researchers are currently working.

AI is becoming a handy assistant

But despite the amount of work ahead, the technology looks promising. After all, it is based on AI, which has been proving itself for several years in a wide variety of areas, from content creation to customer service. Every month, the quality of various AI algorithms continues to improve. One of our authors realized this while preparing this article, using the example of customer support. After a two-year break, he decided to return to virtual gambling. To achieve this, he analyzed search results and selected several top online casinos, including Parimatch, 1win, and Mostbet. In the end, he decided to register here, in the 1win app, as it offers the widest selection of games. During registration, he had a few questions, which he addressed to customer support. As it turned out, customer service was provided by an AI-based chatbot, which quickly and efficiently solved all his problems. What’s more, it acted more professionally than a live operator.

This discovery only reinforced our author’s conviction that AI has become significantly more sophisticated over the past few years. And that means it is more helpful in solving a wide variety of tasks.

What the experiments showed

The choice of the Haemonchus parasite, also known as the barber’s pole worm, to test the new method was due to the scale of damage this organism causes to sheep flocks.

Adult worms live in the stomachs of animals and feed on their blood, which in mass infections leads to anemia and even death. A recent study by the University of Calgary found that the parasite is resistant to the main drugs used on 20 Canadian farms, making rapid diagnostic methods particularly valuable.

To test the effectiveness of imPulse Una, scientists divided the animals into four groups: two control groups and two infected with parasite larvae, one of which received antiparasitic treatment after 28 days. Data from the device was collected from five anatomical points: the heart, lungs, neck, stomach, and intestines. As the researchers note, after just 28 days, the device was able to distinguish between infected and healthy sheep, and objective verification using blood markers confirmed the accuracy of the new method.

Pregnancy testing

Further experiments raised the question: if an AI stethoscope is capable of accurately detecting patterns associated with infection, why not attempt to detect pregnancy in sheep? Following this logic, the team conducted a series of tests comparing signals from different parts of the body, including the area where the fetus develops, with the results of traditional ultrasound examinations.

Data was collected at different stages, and then experts analyzed the dates of lambing and the number of lambs born. As Yaogeng Lei emphasizes, one of the long-term goals is to provide farmers with a tool for independent and straightforward “pregnancy checks,” which will increase the autonomy of small farms. The authors of the study note that the algorithm still requires refinement, but the initial results are comparable in accuracy to those of classic ultrasound.

Prospects and challenges for implementation

It is clear that to create a fully functional system, it is necessary to collect even more data on the condition of animals and to test the tool’s stability under various conditions.

According to experts, the difference between an AI stethoscope and a traditional ultrasound device lies not only in its simplicity of use but also in its ability to detect subtle patterns in acoustic signals. However, questions about reliability and versatility remain unanswered. In European countries, similar projects are being implemented primarily in scientific laboratories, which highlights the novelty of the approach.

The range of potential applications for this technology goes far beyond animal husbandry. Developers are exploring the possible use of AI stethoscopes for the early detection of complications during pregnancy in women. In particular, conditions such as preeclampsia and gestational diabetes. However, a serious clinical basis and additional testing are needed to confirm the reliability of the analysis for humans.