Email updates

Keep up to date with the latest news and content from BMC Veterinary Research and BioMed Central.

Open Access Methodology article

Pregnancy detection and monitoring in cattle via combined foetus electrocardiogram and phonocardiogram signal processing

Gaetano D Gargiulo1234*, Richard W Shephard1, Jonathan Tapson12, Alistair L McEwan3, Paolo Bifulco4, Mario Cesarelli4, Craig Jin3, Ahmed Al-Ani5, Ning Wang6 and André van Schaik2

Author affiliations

1 HEARD Systems, 502/143 York St., Sydney, NSW, 2000, Australia

2 University of Western Sydney, Penrith, NSW, 2751, Australia

3 The University of Sydney, Sydney, NSW, 2006, Australia

4 “Federico II” The university of Naples, Naples, 80100, Italy

5 University of Technology, Ultimo, Sydney, NSW, 2007, Australia

6 University of New South Wales, Sydney, NSW, 2052, Australia

For all author emails, please log on.

Citation and License

BMC Veterinary Research 2012, 8:164  doi:10.1186/1746-6148-8-164

Published: 17 September 2012

Abstract

Background

Pregnancy testing in cattle is commonly invasive requiring manual rectal palpation of the reproductive tract that presents risks to the operator and pregnancy. Alternative non-invasive tests have been developed but have not gained popularity due to poor specificity, sensitivity and the inconvenience of sample handling. Our aim is to present the pilot study and proof of concept of a new non invasive technique to sense the presence and age (limited to the closest trimester of pregnancy) of the foetus by recording the electrical and audio signals produced by the foetus heartbeat using an array of specialized sensors embedded in a stand alone handheld prototype device. The device was applied to the right flank (approximately at the intercept of a horizontal line drawn through the right mid femur region of the cow and a vertical line drawn anywhere between lumbar vertebrae 3 to 5) of more than 2000 cattle from 13 different farms, including pregnant and not pregnant, a diversity of breeds, and both dairy and beef herds. Pregnancy status response is given “on the spot” from an optimized machine learning algorithm running on the device within seconds after data collection.

Results

Using combined electrical and audio foetal signals we detected pregnancy with a sensitivity of 87.6% and a specificity of 74.6% for all recorded data. Those values increase to 91% and 81% respectively by removing files with excessive noise (19%).

Foetus ageing was achieved by comparing the detected foetus heart-rate with published tables. However, given the challenging farm environment of a restless cow, correct foetus ageing was achieved for only 21% of the correctly diagnosed pregnant cows.

Conclusions

In conclusion we have found that combining ECG and PCG measurements on the right flank of cattle provides a reliable and rapid method of pregnancy testing. The device has potential to be applied by unskilled operators. This will generate more efficient and productive management of farms. There is potential for the device to be applied to large endangered quadrupeds in captive breeding programs where early, safe and reliable pregnancy diagnosis can be imperative but currently difficult to achieve.