Back

Assessing the Accuracy of a Smart Collar for Dogs: Predictive Performance for Heart and Breathing Rates on a Large Scale Dataset

Jarkoff, H.; Lorre, G.; Humbert, E.

2023-06-11 physiology
10.1101/2023.06.09.544347 bioRxiv
Show abstract

We present an evaluation of a smart collar for dogs that measures resting heart and breathing rates. The study involved 40 dogs of various breeds, ages, and sizes. The collars accuracy in estimating heart rates, breathing rates and detecting heart pulses was evaluated by comparing the measurements from the collar with those from a portable ECG device, completed by labels annotated by trained humans. The results show that the collar provides accurate and reliable measurements of heart and breathing rates, with a (symmetric) mean absolute percentage error of 0.38% and 1.42% respectively, and a F1 score of 98.04% for the detection of heart beats at 50ms precision. Our study demonstrates the potential of Invoxias Smart Dog Collar as a medical-grade tool for continuous, noninvasive, remote monitoring of vital signs in dogs, helping to improve the quality of life for dogs, detect early signs of illness, and strengthen the bond between dogs and their owners.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.