RSNA, COUEism, and the Eye Q Test

Hi, Folks, I hope you are all enjoying a marvelous Thanksgiving.

We are all obsessed with the allure, secrecy, and efficiency of acronyms. Medical education makes abundant use of mnemonics, and who doesn’t dream of inventing a meme that propels itself through society? I’ll get to the foundation for this quirky intro later.

I’m just back from attending the globally international Medica, where I presented material about dedicated use of ultraportable hi-def ultrasound in the NICU. There were three massive fieldhouse sized halls filled with a hodgepodge of every kind of medical electronics and optics. There were more different and varied diagnostic imaging units than I have ever seen anywhere, and, encouragingly enough for the future of ultrasound, a large number of gel and aseptic wipes’ manufacturers and distributors. I thought that an appropriate topic for this RSNA edition of the Practice of Ultrasound would be a personal guide to evaluating the potential of new equipment, seen for the first time – like walking a path of objectivity before entering the forest of sales promotions and advertising.

A Few Basics
Everyone involved with clinical ultrasound will have preferences for commercial equipment. You’ve trained and practiced, and you need to be comfortable with equipment rely on it completely. Very few people in the ‘outside’ world understand that there is a life and death component in working with patients. Obviously, this does not typically take the form of an ER- or OR-centered TV drama, but, missing a lesion or condition that could have been caught early has grief-years of consequence for the patient. There are major limits to what we can diagnose, or exclude, we just try to do the best we can. But, we cannot allow our work to be compromised by suboptimal equipment or by inexpert operation of the equipment we have. Equipment choice remains personal, because it involves technical performance compounded with the patients we see and what we need to get out of the exams we do.

Brand loyalty needs to be reassessed from time to time. Ultrasound manufacturers are not professional sports teams, because their definitions of winning are not the same. The enthusiasm of new development and the willingness of the young company to respond to the needs of its relatively few supporters may give way to the exigencies of business when it is large, when the priority for the company may shift away from new developments. And, even more sadly, small companies that do not gain traction, even with superb clinical devices, may not be able to sustain their initial progress. This is a particular problem for accessories for exams and procedures. So, start or end  your face, but booth hop and see as much as you can with an open mind.

Coueism
I started with a title: checking out ultrasound equipment which contracts to COUE. Something popped into my memory from about 60 of my years ago: Every day, in every way, I get better and better, and an attribution to Emile Coue. A quick check, thank you Google, informed me that he was a French pharmacist who came up with a panacea self-help program precursor of psychotherapy just before 1900, and that this had become a popular, and very inexpensive, fad in the U.S. in the 1920s as the ever optimistic, Coueism.  We all want to do better and better, we hope that equipment gets better and better, and how better to reconcile those than by checking out everything new at RSNA?

Better and best are slippery notions. It used to be canon that ‘pictures speak for themselves’, and we all still look at pictures that vendors display. Presumably, they are shining examples of what the equipment can do under ideal circumstances. Be very wary when images are of uneven quality, no matter what the diagnostic delivery might have been. If some of the images are noisy or low contrast, then we cannot expect a constant level of performance in our routine use of the unit. Another major issue occurs with gray scale images with a tinted background, especially any shade of blue, except for special circumstances like neon green night vision for needle placement and candle light or skin-toned fetal 3D portraits. That is a deal breaker for me, because it implies a basic misunderstanding of image perception and makes me wonder what other basic errors are also present. Look in any radiology journal: X-rays are always in black and white, there is a reason for that, you should all know it, and it applies to ultrasound as well.

Noise is ever the enemy. I look at fluid spaces near a reflective object, are they clean or obscured? A related issue is screen size. Images can seem crisp on a very small screen, because it’s harder to see noise. Look at magnified prints or large screen monitor displays. Most of the handheld devices fail this test. It’s not that they cannot be used for some simple tasks, but that is all they can be used for. One of the hallmarks of radiology is that we are called on to do a variety of tasks every day, repetitive single use forms of ultrasound don’t work for us.

There is a related issue, which has to do with adjusting equipment on the fly. A lot of vendors, intent on increasing their sales to nonradiologists, provide units that are not adjustable; they may rely on arbitrary presets. I dislike presets, because they are loosely based on anatomy and have nothing to do with pathology, which is what we are trying to detect. Beyond that, patients are not standardized, so we always need a way to optimize our equipment for individual patients and for different image fields in an exam. There are relatively few controls that are necessary at this point in time, since there are full field focusing and adaptive gain algorithms on the high-end systems. Tissue contrast remains an essential adjustment and not all manufacturers address this directly. I hate the notion of handheld units, because they cannot be adjusted while the exam is in progress.

The larger companies at Medica had live scanning done by their full-time, expert demonstrators. I lingered in front of a repetitive scanning of a slender woman about 28 weeks pregnant. Image quality was pretty impressive, even though there was low contrast and not much real gray scale, plainly the operator didn’t know better. But what struck me was that he spent most of the time showing off simple and ordinary things. The heart was done pretty well, the brain not at all. When the probe was flying from the head, after doing a circumference, to the abdomen to do another perimeter, I saw the thyroid zip by. It really showed well, as did the larynx and other things elsewhere in that female fetus, none of which merited any stopping or inspecting. What I want to see at a live demo, or in still images or videos, are things that are not the everyday things that we all have been doing for the last 30 years. I want to see something new or that I may have trouble with. There will be live scanning at this RSNA, look for something novel as an indication that the unit can provide something new for you to use clinically.

Two Classes of Equipment
Let’s diagram an ultrasound system as three components: The transducer, the display, and the beam former. Displays and transducers have been improving continuously for years, and the best in these areas are pretty much available to all manufacturers. All equipment is digital. It’s the beam former that is the secret, algorithm-packed black box, part of the unit that makes all the difference. The high-end equipment uses real time, dynamic beam forming in which all the factors that go into image formation are optimized on the fly as the images are formed and updated. There are a few hundred factors to be optimized, and they are interactive in generally unpredictable ways. It’s like a small scale version of the genome with its repressors and de-repressors, polymorphisms, and on and on. There are feedback loops that can make a system unstable or generate strange and bizarre artifacts. But when it works, clean modern day ultrasound images are a treat to behold and work with. Dynamic beam formers require a lot of computational power. They tend to be housed in larger units both for their CPUs and GPUs but also for power requirements and heat dissipation.

Physically small units are of two sorts. The most familiar has the display and beam former integrated. One of the nicest I saw at Medica looked like an iPhone, but much thicker and heavier, because of the chip it used. The other variety had the beam former housed in a chassis between the transducer and the display unit, which was a laptop, tablet, or smart phone analog.  I prefer the modular design. There’s enough switching and pre-amp circuitry in probes already and adding more throws off the balance and ease of use to no real advantage. Modern probe cables are very thin and light.

The beam formers of these units are static not dynamic, noise reduction and image adjustment are limited, and there is usually one or, at most two, transducers, so that applications are limited technically and clinically -  good for training, but not necessarily what one would use when it really counts. It’s interesting that the big companies with high-end equipment at Medica, did not display any small units, I know, because I asked specifically. A Lithuanian manufacturer of a small modular unit told me that they made the beam formers for the laptop and platform units of most of the big companies. A low-end unit uses what I think of as a ‘static’ beam former. A dynamic beam former in a tiny unit is still high end, a static beam former in a big unit is low end. This is a case where size does not count.

Call on The Phantom
I always try to get an idea about performance, before I talk to anyone from any of the manufacturers, and I rely a lot on what I hear from colleagues who use equipment themselves. But I sometimes avail myself of a small, portable and pretty uniform phantom that we all carry around: an eye. It’s a demanding target technically for a unit for both detail and contrast resolution, and it’s a sand trap for noise potential.  I tend to like high frequency applications, and I hate to get gel on nice clothes or ruin a silk tie. This is one of several identical images that I made myself at Medica with a U-Lite EXP with a linear array and a frequency setting in its mid-range:[[{"type":"media","view_mode":"media_crop","fid":"54380","attributes":{"alt":"","class":"media-image media-image-right","id":"media_crop_5284914977715","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"6806","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"height: 250px; width: 250px; border-width: 0px; border-style: solid; margin: 1px; float: right;","title":"I placed the probe on my left lower eyelid and aimed up. If you look closely you will see my tiny forming cataract, hopefully it won’t get any bigger or denser. Pupil diameter was 3.6mm by electronic caliper. ","typeof":"foaf:Image"}}]]

As I am writing this, I decided to call this the EYE-Q test. I have never had an opportunity to name a specific test before, so I don’t want to pass up this opportunity. Q for quality: this is high-end, dynamic beam former performance, pretty good for a 600-gram ultraportable tablet. Try it out, if you know how to scan, it’s easy, and it keeps you in the driver’s seat in your preliminary equipment testing.

Happy and enjoyable booth hopping! If you see something that you really like at the RSNA, add it to this blog and share your thoughts with your fellow ultrasounders.