NDT Discussion Forums

Hi,
I'm developing an A-mode ultrasound system to measure bone depth (typically 10-30mm). I've been using a 10MHz 25mm focal length spherically focused immersion transducer. I need to use a standoff to buffer the distance to the target bone so it is in the nice range of my focal length. Right now I fill a glove finger with water (an effective technique shown to be my an echographer) and use it in between the transducer and the skin surface. The problem is that this is cumbersome and not a clinically acceptable solution. I've tried using Aquaflex gel pads which are clinically used but seem to disintegrate if pressed too hard and are also cumbersome as they require using an extra hand (one to hold the transducer and to hold the pad). I've seen bubblers, but again, these seem impractical as they require a water source. I have seen pictures in other publications of a plastic-looking cone attached to the end of the transducer but no details on what that is. Does anyone know of a simple standoff solution that attaches to the end of the transducer that can solve this issue? Alternatively, does anyone know of an immersion transducer that can be used directly on the skin that would have a focal length in the range I need (~10mm)?

Thanks,
Alon

I have a couple suggestions. Although Olympus NDT is a manufacturer of industrial rather than medical ultrasonic equipment, a number of biomedical researchers over the years have successfully used our products for tissue thickness measurement. You describe your application as measuring soft tissue over bone, in the thickness range 10 mm to 30 mm. Based on our previous work, I would recommend trying a 5 MHz, 13 mm diameter contact transducer with an epoxy face, such as a Panametrics-NDT V609-RB. We also have a 10 MHz, 13 mm diameter V611-RB, although my experience in related tests is that 10 MHz is a higher frequency than customers have typically used for soft tissue measurement in that thickness range.

You write that you're working with focused transducers. The contact transducers above are unfocused, and unfocused beams have been shown to work in applications that may be similar to yours. If you need focusing, it is certainly possible to use one of our focused immersion tranducers in a direct contact mode, using coupling gel between the face of the transducer and the skin to eliminate the need for a water path. However due to the practical limitations of acoustic lens design, the shortest focus that we could offer in a standard 10 MHz immersion-type transducer would be approximately 12 mm, a bit longer than the 10 mm that you have requested.

We also make various bubblers that can potentially be used for this type of test, but as you say they require a water supply and usually some way of collecting the water stream runoff as well.

Finally, I am aware of the solid cone-shaped plastic waveguides that you mention -- not to be confused with cone-shaped water bubblers -- but our (industrial) experience is that they seldom work well because of the large amount of noise that is reflected from within the cone, obscuring echoes of interest. Thus we do not offer them.

Thanks for the response.

I am trying to measure relatively shallow bone so I chose 10MHz to maximize the axial resolution since I didn't need to penetrate deeply. My reason for using a focused transducer is to maximize lateral resolution. My assumption was that the unfocused contact transducers will have relatively poor resolution, determined by the element size, which you mention is about 10mm. I believe the focused 10MHz transducer should provide an order of magnitude better resolution in both dimensions, but please correct me if I'm mistaken on this. I picked the transducer based on theory and have only used the one transducer since. My concern with using the 12mm focal length transducer without a standoff is that if the bone is too close to the transducer source, my measurement may not be accurate (or worst case, may not be measurable at all due to near-field noise). Again, this is all theoretical, so I'm open to more practical advice.

Any idea where I might be able to get a plastic wave guide to test?

Thanks again.
-Alon

You are correct that a focused transducer will theoretically have better axial resolution than an unfocused tranducer, but lateral resolution is independent of focusing. Whether the improvement in axial resolution will be functionally significant in this case is certainly your call based on your test needs, however our experience with industrial applications suggests that in this type of test it will be nowhere close to an order of magnitude, probably more like a factor of two or three, because even an unfocused beam will show significant sensitivity to axial position when aimed at a cylindrical reflector like a bone. With respect to depth measurement, focusing can typically affect amplitude but not measured transit time.

Since you will be measuring pulse transit time from the skin surface to the bone echo, near field effects will not introduce timing errors -- near field pressure differentials are normally a consideration only in amplitude-based tests. You will see some echoes from the various soft tissue boundaries between the skin and bone, but they should be smaller than the reflection from the soft tissue/bone boundary.

As for conical plastic waveguides, I'm afraid that as I wrote earlier we don't offer them, but maybe somebody else will be able to suggest a source.

Here's a pointer to an article that demonstrates my simple understanding of resolution:

http://www.animallab.com/articles.asp?pid=209

Based on that, frequency is closely tied to axial resolution. Lateral resolution at the target location, however, seems related to the "narrowness" of the beam. From what I understand, a focused beam will be more narrow than an unfocused beam. Please correct me if I'm misunderstanding.

I mention near-field effects (and could be confusing that for "ring-down") due to noise I'm seeing in areas in the signal close to the transducer face. They seem to linger for several millimeters and that seems longer than what I should expect from ring-down. I'm definitely seeing noisy results which I'm guessing are due to sensitivity of axial alignment. So while I understand I should theoretically see a clear spike for the bone interface, that's not usually the case. And with a standoff, I'm seeing two peaks (on an ideal signal) - one for the standoff/skin interface and another for the skin/bone interface.

Thanks,
Alon

May I recommend that you contact Tom Nelligan at Olympus NDT Waltham who solves most of our conventional ultrasonic problems. However, you are correct about focused beams; they should give better lateral resolution. I'm not surprised that you are seeing signals from both the standoff/interface and skin/bone. if the acoustic impedances are different, you would expect to see signals. As for the noise etc., best to describe your set-up to Tom at "tom.nelligan@olympusndt.com".

Yours, Michael Moles

That article discusses phased array imaging rather than conventional single-element transducers, but the basic physics is all the same. Beam spot size for an unfocused transducer at the end of the near field is related solely to element diameter; in fact the -6dB spot size will be .257D where D is the element diamter. In the far field, beam spreading for a transducer of a given element diameter with increase as frequency decreases, thus a higher frequency transducer of a given size will have a narrower beam at a given distance in the far field. In the case of focused transducers (whether an electronically focused phased array or a lens focused conventional transducer), the focal spot size is related to frequency and focal length, getting smaller with higher frequencies and shorter focuses. You can find the equations for calculating spot sizes in this document from this web site: http://www.olympusndt.com/data/File/UT-technotes.pdf

So basically, you're correct. A focused beam can potentially be much sharper than an unfocused beam, and with all other things equal higher frequencies focus more sharply than lower frequencies.

I notice that article talks about working with small animals... I was assuming that you were looking at human bones that are fairly large diameter when I suggested unfocused transducers, but I can certainly see the advantage of using a sharp focus if the test target is a tiny mouse bone.

I'm afraid that without seeing your actual setup I can't visualize the specifics of the close-in noise that you mention, but if it's transducer ringdown it would be present whether you're coupled to the target or not. If you're using gel pads for coupling, then the source is reflections within the gel layer. If you're coupling the concave face of the transducer directly to the skin, it's probably reflections in the couplant layer trapped between the lens and the skin.

The article was just something I found that provided a clear description. I am in fact working with human bones with relatively large diameters. The trick is that I'm trying to localize highly accurate points on the bone. But now I'm wondering if there's a tradeoff between processing and accuracy. I mean that with my current setup, I think I'm maximizing theoretical accuracy but in practice, I need to juggle using a homemade standoff, gel coupling, and a finicky transducer that gives me noisy signals if my axis isn't perpendicular to the target surface. As a result, I think my theoretical accuracy goes out the window if most of my signals are too noisy for my processing to guarantee a good data point. I'm debating trying out the unfocused transducer with an epoxy face that you suggest if it will make the overall result better (less theoretical accuracy, but if I can more easily capture clean signals I might be better off).

I have images that show the occasional ideal signal and how I process it as well as the more typical noisy signals that I'm getting. If there's a way to post them, I'd be happy to show you what I'm seeing and how I deal with it.

The hardware setup is standard - pulser/receiver, capture card, transducer, one of various standoffs I've tried (with the water-filled glove finger being by far the best), coupling gel on every surface (some viscous off-the-shelf gel/lotion), and it's applied to the leg bones (femur, fibula, tibia). As I'd expect, thicker tissue areas give me significantly more noise, but areas with the most shallow bone (like on the shin) often provide the best peaks for bone detection. If you can think of a transducer solution that can accurately find the bone in deeper tissue as well, that would also be a big benefit for me.

Looking through my signals, I think you're right that the noise is always there and is probably ring-down. The problem is that it affects the first 3-5mm of my signal. Bone may only be 7-10mm below the surface of the skin so if I don't use a standoff, it may affect my measurements. This is related to why I was asking about a transducer with smaller focal length. I was assuming a short focal length transducer must have better ring-down properties and allow me to accurately and directly measure bone that's only a few mm's below the surface.

If you'd like to discuss this further, my suggestion would be to e-mail some typical waveforms to me at <tom.nelligan@olympusndt.com> . I'm afraid this forum does not support image posting.

Again, we've had other customers over the years who have successfully used our industrial instruments and transducers for this sort of biomedical research application, so we should be able to supply what you need.