A week of meat grading in Iowa
I’ve been working with the British meat grading industry for 6 years now, and earlier this month, I also had the opportunity to visit Switzerland for the second time to work on ultrasound meat grading for sheep.
This week, I flew to Ames in Iowa, to see how the CUP Lab are running their ultrasound technician training programme, and learn how the Americans are collecting data on intramuscular fat percentage (IMF%) in cattle. There are other data processing laboratories in the USA, but it was Quality Assurance Manager Emily’s willingness to discuss her thoughts on equipment and the training process that persuaded me to choose the Iowa team.
Intramuscular fat is not something that we are currently concerned with in the United Kingdom, but it is important to some of our international members. Fat and muscle depth are obviously measures we already take routinely here in the UK and in many European countries, but the US method is slightly different from ours in terms of transducer positioning and measurement points.
The first day of the course was spent making sure that everybody was introduced to basic anatomy and imaging landmarks, and familiarising the students with ultrasound images – what makes a good one, and what makes a bad one. Ultrasound physics was barely mentioned, which was both a relief (I like to think that I’m already well versed on this topic!) and a surprise. It soon became apparent why this is of so little importance: the CUP Lab requires its technicians to adhere to very specific settings on each of the ultrasound machines it has approved for use. There’s no point in learning what all the buttons on the scanner do if you’ll only be using it for meat grading work, because touching them would be an act of self-sabotage.
In fact, the Lab’s most recent software update will now alert the user if any settings are out of line, and reset them. The software also allows the technician to quickly and easily enter a new ear tag number, and save each respective image using hotkeys. The only downside to the reliance upon external software is that for many models of ultrasound system, the technicians have to transport laptops around with them, loaded up with the Lab’s software. This is not a criticism of the Lab; it’s just a limitation of today’s technology, and one that I’m sure won’t even exist 10 years from now. It’s certainly a vast improvement on the previous Aloka situation. But, it does mean investing in yet another piece of kit, on top of pricey software and extortionate ultrasound machines (again, not a criticism of the CUP Lab – they don’t set the equipment pricing).
Equivalent technology can of course be purchased from other manufacturers for around 20% of the price of an Ibex or ExaGo, but unfortunately for anybody wanting to collect analysable IMF data, it’s only ECM and E I Medical devices that the Lab have invested in validating so far since the Alokas. Mark explained to me that the main reason behind this was the need for continuity, and both of these companies had pledged to continue producing and supporting these models for long enough so as to make the Lab’s investment worthwhile.
Fat depth, muscle depth and muscle area can all be calculated independently on any machine itself through the ‘measure,’ ‘distance’ or ‘trace’ functions, and do not require any additional software. For anyone wanting to send their data into US breed associations, however, the measurements have to be performed through an official UGC-approved Lab, not by the ultrasound technician, so people’s equipment options in the USA are always going to be limited to those which these laboratories are prepared to process data from. IMF% calculation is something that definitely cannot be performed on the machine itself, even for private data collection, as it requires specialised proprietary software developed from years of research and painstaking testing. The CUP Lab claim that their software has over 200 parameters for IMF calculation. I’d be interested to know how this is possible, but unsurprisingly, they are not willing to disclose such commercially sensitive information.
Good news for us
For British and European farmers, the strict requirements of US breed associations are of little or no concern, and interest in IMF% is currently low. This gives those wanting to scan cattle in these countries much more freedom with regard to equipment choices. Given that consistency is no less important, however, cattle and sheep farmers will invariably rely on organisations such as Signet Breeding Services in the UK or Sheep Ireland to collect data and generate Estimated Breeding Values (EBVs). Self-employed independent scan technicians are rare in Europe when it comes to collection of beef and lamb data, either because of the established existence of government-funded projects with nationwide coverage or because meat grading is entirely absent in these countries. Heinz Pluess is one example of an independent technician, and he was actually the person who introduced the concept of ultrasonic meat grading in sheep to Switzerland.
The UGC technique
The USA Ultrasound Guidelines Council has a set protocol with regard to which images are taken, and how. Obtaining a rump image is relatively straightforward, can be learnt in about half an hour, and perfected over the course of the week.
The chosen image for IMF% calculation is taken across the 11th, 12th and 13th ribs, and this image is definitely more challenging. Acorn fat (the hook of the ribeye) and the spinalis dorsi have to be excluded from the image, yet without sinking so low down the animal’s flank that the muscle depth is truncated. One also has to differentiate between lumbar and ribs, being sure not to mistake one for the other. Lumbar are flatter and wider; ribs smaller and rounder, but it is only with proficiency in scanning technique that it becomes possible to manipulate (‘roll’) the transducer and distinguish them easily.
This image is so important that technicians are requested to take five, removing the probe from the animal each time, in order to obtain independent images from which the IMF% can then be averaged from the acceptable ones.
The third and final view is the of the entire ribeye. This requires placing the transducer inside a well-oiled standoff (an image can be taken without a standoff, but at least one standoff image is required for fat depth measurement, as the standoff moves the fat layers from the near field of the ultrasound transducer), between the last two ribs. Signet in the UK will take a similar image further back on the animal, at the last rib and third lumbar vertebra, so the US method differs from ours here.
I found this view by far the most challenging. It can vary a lot between cattle in terms of visibility of the required landmarks, and with great variation in size. In larger bulls and steers, it can be difficult to get the entire ribeye into one shot. It is sometimes difficult to avoid drop-out in some of the highly marbled animals when using the standoff, so this is where taking an image without the standoff comes in handy. The technique I found most challenging of all was bringing the intercostals into view, and not a rib. Either option presents as a bright white line, but intercostals are a double line at a shallower angle that one can only really begin to recognise with confidence after scanning at least 50 cattle. Until then, it’s all a bit of a shot in the dark, or like throwing a netball or basketball behind your head at the height and distance you think the hoop must be, but with no way of really knowing how close you’re getting until you actually start getting them in.
By day five, the students are relatively proficient and able to self-correct in most cases. At this point, a three minute time limit is introduced, in preparation for the UGC certification examination (for those who require it for their work) – and also for the practicalities of the real world. Three minutes is generally more than enough time to obtain the required set of images, and most trainees seemed pleasantly surprised at the ease at which they were able to gather their pictures within this time frame.
Beyond fat, muscle and IMF%
Of course, I didn’t just learn the UGC technique from my week in Iowa. I witnessed first hand some of the fiercest political debates gripping America today – a country starkly divided over peanut butter M&Ms versus Reese’s Pieces – but I also saw how with targeted practice of and adherence to a strict and limited protocol, people with no background in ultrasound technology can be taught to obtain excellent images, whether for diagnostic or measurement purposes.
This is an idea which is very much being explored in human medicine at the moment. With the miniaturisation of technology, cost-effective portable ultrasound scanners can be put into the hands of people in remote locations, to help those who have no access to a hospital. Trained to perform a limited number of views – of the heart using a Vscan, for example – these people can then send their images back to a lab hundreds of miles away (even in another country) for interpretation.
Those who question whether ‘lay’ scanners can be taught to produce high-quality, reproducible images within a limited time frame should come and see the CUP Lab technicians in action. After a week of practicing on just three (very specific) views, its technicians are capable of reproducing these same views over and over again in a range of cattle back in their home States. As well as interpreting these images back at the lab, Mark, Emily, Holly and their team score the image quality as acceptable, marginal or rejected, thus continuing to give feedback to the technicians and ensuring that standards are maintained.
August 10, 2019
August 08, 2018