Start at Home

pupdate 28102023

/5/ pupdate 28102023

While Florin sniffs a urine same from a prostate cancer patient, a fully non-invasive brain-computer interface powered by EEG records her neural activity. From studies in humans and mice, we know data from non-invasive brain recordings can encode meaningful properties, such as a specific identity of an odor from a finite subset of odors. Therefore, it seems plausible that the "scent signature of cancer or anything else that a dog is trained to look for" (Guest, 2021) can be decoded through technology, and such technology will not cause any harm or discomfort to the partner dog, while also relying on relatively affordable and tested cognitive neuroscience methods.

We now have the ability to do this at Dognosis. We expect, and indeed hope, many others will develop and refine ways to decipher the canine mind. Our current prototype (DogSense v0) is still relative crude and clunky, with plenty of low-hanging design and engineering fruit remaining to be plucked. In 3 years, we expect to be able to boost data-collection affordances by at least 2 orders of magnitude. At the turn of the decade, expect soft and machine-washable electro-composites, that wrap around your dog's head like a sports bandana, to be available for the price of an Apple smartphone. The advance of HD-DOT technology (e.g Gowerlabs), with their promise of fMRI spatial resolution from a portable headset, is also cause for deep optimism.

Far more challenging is the work required to parse meaningful signal from what is a low-res electro-optic imprint of a complex and dynamic cognitive process. The noise of EEG, the lag of HD-DOT, the chaos of your average canine, all these will have to be overcome, and even then, we are tasked with the astounding task of understanding the mind of another species.

But as they say, what better place to start than Home?