Ultrasound for beef meat grading in the UK and USA
Ultrasound has been used to measure fat and muscle depth in sheep and cattle in the United Kingdom for many years. Signet, part of EBLEX, have been measuring these two parameters as part of their estimated breeding value (EBV) calculations for over 20 years, as well as muscle area for their Breedplan clients. Signet use cutting edge ultrasound equipment and technology, including their own custom-made linear meat grading transducer. Yet, there has been no need to change or add to the standard measurements they take on farm: fat and muscle depth and muscle area are well-validated components of their EBVs. After all, in our market here in the UK, consumers demand lean meat. The combination of a larger muscle depth/area and a lower fat depth is the most desirable outcome, and these parameters can be confidently measured with current technology and protocols.
In the US market, however, demand for fattier ‘marbled’ meats is high. This market demand has resulted in ultrasound being used to estimate not only ribeye muscle area, but also intramuscular fat percentage (IMF%), as a far more important parameter for genetic improvement in their herds than just subcutaneous fat depth. In the United States, quality control is ensured by farmers through using an Ultrasound Guidelines Council (UGC) certified technician, in a similar way to how dog owners in the United Kingdom can ensure quality by using an Animal Ultrasound Association (AUA) certified scanner.
The images the technicians acquire need to be analysed by specialised software, and so the images are sent through to and processed by a UGC certified laboratory after acquisition. The technician will take several acquisitions in the same location, and the results will later be averaged. This is important because, more so than other imaging modalities, ultrasound suffers from issues of inter and intraoperator reproducibility, unless adequate training and controls are put in place.
Nevertheless, there is no perfect correlation between IMF% and marbling grade. One of the leading data analysis and training centres in the States, CUP Labs, claim this is because IMF% is a quantitative, objective measurement, whereas marbling is qualitative. These facts are true, but since the price commanded at market is based upon a marbling score, there is a slight disconnect between the ultrasound-derived meat score and the score it may receive at slaughter. There’s also the highly subjective issue of taste quality – is this more closely correlated with ‘true’ IMF%, or visual marbling score?
A final issue is that CUP Labs imply that their IMF% software has been tested against chemical extraction derived IMF%, but there is no published data to support this. This is not to say that it doesn’t exist; merely that it is not readily available at time of writing. The software estimates an IMF% based on the amplitude of the returning echoes, which will differ between fat and muscle given their different acoustic properties. However, it does this from the resulting image that the technician captures, presumably through a reading of pixel brightness coding. An adjustment of frequency can completely change the ways in which the fat and muscle cells reflect ultrasound (objects the size of or smaller than the ultrasound wavelength will scatter ultrasound. Since adjusting frequency changes the wavelength, it will increase or decrease the number of scatterers and change the on screen appearance of the insonated tissue). An increase or decrease in output power or gain (level of amplification of the returning signal) can similarly cause bright regions of intramuscular fat to ‘bloom’ (increase in displayed area) or even be completely obscured.
The resolution of the transducer is also important. A transducer with poor spatial resolution will not be able to resolve two reflectors as separate if they are spaced closer than half a spatial pulse length apart in the axial direction, or closer than half a beam width apart – at best – in the lateral direction. As many technicians in the US still swear by their 10 year+ old Aloka systems, spatial resolution may be a limiting factor and may also mean that baseline data used to develop IMF% software will not be calibrated to data collected with more modern machines. Having spoken with CUP Labs, there does seem to be an attempt to make IMF% software compatible with more modern systems which does reassuringly imply a rigorous scientific approach. Training courses also include learning to calibrate one’s system and ensure reproducibility.
The downside, however, is that the ultrasound technician is currently very limited in equipment choice, and must pick from a handful of very expensive options that may not necessarily deliver the best value for money.
Moving things forward
The differences in our two markets are interesting, and there is certainly the potential to learn from one another and make new advances in a field that has not seen much change in over a decade.
Over the next few months we will be trialing meat grading equipment on cows and sheep, not only using standard routine UK measurements but also incorportating popular US measurements as well as some new experimental ideas. Our results will be shared with the AUA community.
February 17, 2022
March 13, 2020