Dognosis uses AI and dogs to detect seven types of cancer

Dognosis, a startup based in Bengaluru, India, is developing a non-invasive cancer screening method that combines canine olfactory abilities with artificial intelligence. The project utilizes 15 trained dogs, including a Labrador mix named Chloe, to detect seven types of cancer by sniffing human breath samples.

To capture precise data, the dogs wear 3D-printed helmets and harnesses equipped with sensors. The AI system analyzes the dogs' brain activity, respiration, and body language to convert their scent reactions into actionable data. This approach leverages the fact that dogs possess approximately 300 million olfactory receptors, significantly more than the five million found in humans.

Preliminary research involving over 1,500 subjects across six hospitals indicates an accuracy rate exceeding 90% for detecting head and neck, breast, lung, gynecological, upper digestive tract, lower digestive tract, and urinary system cancers. The method involves patients breathing into a cotton mask for approximately 10 minutes. Dognosis, backed by venture capital firms such as Accel India, aims for a commercial launch in 2027, positioning itself alongside similar emerging firms in Germany and Israel.