A new approach to disease detection being explored in India is bringing together two very different technologies: the extraordinary sense of smell of dogs and the analytical power of artificial intelligence.
The initiative, described as “Dognosis,” is investigating whether trained dogs can help identify biological signals associated with cancer from a patient’s breath, while AI translates the animals’ responses into measurable data.
The concept could eventually offer a less invasive way of investigating certain diseases, although the approach remains an emerging area of research and should not be regarded as a replacement for established medical tests.
How the Dognosis Method Works
The process is designed to be relatively simple for the patient.
A person breathes into a cotton mask for approximately 10 minutes. The mask is then sealed and transported to a laboratory, where trained dogs are exposed to the collected breath sample.
Researchers use artificial intelligence to analyse the dogs’ reactions, including changes in brain activity, respiration and behaviour.
The objective is to convert the dogs’ biological responses into data that researchers can study for possible disease-related patterns.
Why Dogs Are Being Used
Dogs possess an extremely sensitive sense of smell, allowing them to detect chemical compounds that humans cannot easily perceive.
Scientists around the world have investigated whether trained dogs can identify volatile organic compounds associated with certain diseases.
Research has explored canine detection in conditions ranging from cancer to neurological diseases and infectious illnesses. However, the ability of dogs to detect disease in controlled research settings does not automatically mean that the method is accurate enough for routine clinical diagnosis.
That is where AI could become important.
AI Could Help Standardise Canine Detection
One of the challenges with using animals for medical detection is that their responses can be difficult to quantify consistently.
Artificial intelligence could potentially help address that problem by analysing multiple signals at once and identifying patterns that may be difficult for humans to recognise.
Instead of relying solely on whether a dog reacts positively or negatively to a sample, researchers can examine behavioural and physiological information and potentially build computational models around those responses.
If successfully validated, such an approach could open new possibilities for non-invasive diagnostic research.
What Nigeria Could Learn
The development also highlights opportunities for Nigeria’s growing technology and healthcare sectors.
Nigeria has universities, medical research institutions, technology startups and an expanding artificial intelligence ecosystem that could collaborate on unconventional approaches to healthcare.
However, biomedical AI research requires more than technology alone.
It would require properly designed clinical studies, qualified medical and scientific teams, ethical oversight, reliable data, appropriate regulation and rigorous testing before any such system could be considered for widespread medical use.
Partnerships between universities, hospitals, technology companies and research institutions could provide a foundation for investigating locally relevant applications of AI in healthcare.
A Promising Idea That Still Needs Evidence
The combination of canine biology and artificial intelligence is certainly unusual, but its potential should be separated from what has already been scientifically established.
Further research would be necessary to determine how accurately such systems can detect cancer, whether results can be reproduced across different populations and whether they can perform reliably outside controlled research environments.
For Nigeria, the broader lesson may be just as important as the technology itself: investing in interdisciplinary research could help local scientists explore new ways of using AI and biological data to address major healthcare challenges.
