New imaging modality: Ultrasound!

[Andy Perrin]

Several times on this board, people have mentioned the possibility of a "sound camera" for making images from sound waves. Of necessity, the waves would probably be ultrasound because the diffraction limit really makes things blurry if the wavelength is much beyond the millimeter range. Conventional "B-mode" ultrasound machines do not make passive images the way a camera does, using available sound waves from the environment. Instead they work more like RADAR/SONAR/LiDAR and send out a pulse of ultrasound and use the time-of-flight of the echoes to form an image. While I would eventually be interested in a passive ultrasound camera, I noticed on eBay that the price of conventional ultrasound machines has fallen enormously, and veterinary models (for pregnancy testing of dogs and cats and other farm animals) were in the ballpark of $600 for the most stripped-down models. I decided paying $550 (for used/refurbished) was a reasonable amount to splurge on as a toy. So I bought the DAWEI S0 Portable Veterinary Ultrasound. You may read info about this particular device here, although as I said, I bought mine on ebay. 

 

 

One of many drawbacks of ultrasound as a modality is that, to generate an image, it assumes the speed of sound is constant in all substances, and for a medical ultrasound, not only constant but roughly equal to the speed of sound in water (1500 m/s). So you are restricted to imaging the following things:

1) things similar in density and speed of sound to water

2) there is no (2). See number 1.

 

That leaves you with body parts, stuff made of plastic, most other living or formerly-living things including flowers and fruits/vegetables, and...that's it. You won't be looking inside anything with a hard outer case.

 

On top of that, ultrasound does NOT travel though air (this is why they slap ultrasound jelly all over you at the doctor's office) so you need to either coat the probe in something (water, dish soap, and glycerin all worked pretty well), OR submerge your object in a tank of water. The probe, but not the electronics, is waterproof to 1 meter. 

 

So far I have had fun peering at my organs in cross-section — that's another thing, you really have to get used to imagining things in cross-section — and I have bought a little turntable that I can computer-control which should allow me to take multiple images at different angles in a systematic way. The turntable hasn't arrived yet from China. Then I plan to try to do some tomography and make 3D ultrasounds. The frequency is adjustable from 2-5MHz, so it should be possible to do "tri-color" images also.

 

Meanwhile, here we have my middle finger:

 

And inside my finger:

 

[Stefano]

...guess we need a new tag now!

 

Interesting concept. Air is very bad for sound propagation, it seems.

 

If it works underwater, you have an imaging environment. One thing you could try is building sound lenses using materials with various sound propagation speeds. Not for imaging, just to see if it works, to see if you can distort the image like optical lenses do.

 

TriColour imaging sounds very interesting.

[lukaszgryglicki]

My father has a device which is used in railways to scan for metal internal fractures - I don't remember the ultrasound frequency but I *think* it is in similar range - a few Mhz.

 

[Andy Perrin]

Lukas, even if it is, it will have its own assumptions about what the constant speed of sound in, say, concrete is, which means it won’t work for anything but stuff with concrete’s sound speed. So wouldn’t work in the body, for example. The nature of the method is to calculate distances from time of flight, so you measure time and convert to distance using the assumed speed. If the actual speed is different than assumed, you would get garbage answers if the rest of the electronics worked, but since many things in the device have to be synchronized for it to work properly, you probably wouldn’t see anything. 

[lukaszgryglicki]

Yes, it is specially designed for metals (steel?) and is quite bulky.

 

[dabateman]

So now you can make a summer student have an engineering project.  Create a tunable ultrasound device calibrated for water, steel, iron , concrete,  ete.

I have seen a railroad device once. It can detect cracks, minor air gaps in iron to avoid potential points for breaking and derailments. 

I have also seen many times the ones in hospitals for intra human imaging.

Mixing the two could be interesting.  

Sound moves faster in iron if I remember correctly.  

Though I am not sure why you would want a tunable one. Maybe for dams, to detect air and what penetration to prevent failure? 

[Alaun]

The "classic" US-Scope uses the A-mode, the amplitude mode.
It is rather simple.

 

Your probe sends a short US-puls and then the probe listens
to the respond signal.

 

To visualize this, something very close to an oscilloscope is used.

At the y-axis, the amplitude is shown
and along the x-axis, the signal is shown according to the time scale.

 

The start puls is used to trigger the signal.

 

As with aconventional scope, you can adjust the time frame on the
x-axis and the amplification of the signal on the y-axis.

You need to control the time on the x-axis firstly, to adjust
for the thicknes of your sample as you want the full resolution
on x up to the first back wall echo.

 

And with the same adjustment on the time, you cabn easily account
for a wide range different speeds, usually you use a sample with
defined thickness to adjust for speed changes. (e.g. even within
different aluminium alloys you can have quite different speeds).
You need the calibration, to determine the position of faults
inside your sample.

 

An by the way, concrete is a bad example for a sample, you will
probaly only see a lot of "grass" (sic!) because there are so
many internal surfaces due to all the pebbels it is made from ;-)   

[Andy Perrin]

23 hours ago, Alaun said:

The "classic" US-Scope uses the A-mode, the amplitude mode.
It is rather simple.

Huh? I do not know where you are getting your information but medical ultrasound typically is B-mode, which is 2D imagery+time. If A-mode is still used anywhere, I haven’t seen it sold online. Of course there are more uses for ultrasound than medicine as Lukas pointed out, but I do think B-mode and “higher” are standard across the board these days. 

[Alaun]

I was referring to the instrument for searching cracks in steel (therefor also the reference to aluminium alloys)

Nowadays these are often combined with e.g. xy-drives, so you can scan the a surface with the probe,

With these data collected, you can even creat 3d-images.

Also nowadays often you do not have just one probe but up to hundreds in one "probe holder".

But even the instruments only with the A-mode are still quite common, when it comes to material inspection.

 

You can also find US-microscopes, which use frequencies much higher (into GHz), they also work as

scanning US-devices collecting the A-mode signal plus the xy-position information.

[Andy Perrin]

I see!

[dabateman]

On 3/13/2024 at 6:39 PM, Andy Perrin said:

I see!

Or do you hear?

[Andy Perrin]

No, not even dogs can hear at those frequencies. 

[dr_narval]

Andy, please show us some fruits with this magical device!

[Andy Perrin]

I am waiting (impatiently) for my computer-controlled turntable to arrive from China. Once that arrives, I will be able to do some tomography, I hope!

[dabateman]

51 minutes ago, Andy Perrin said:

I am waiting (impatiently) for my computer-controlled turntable to arrive from China. Once that arrives, I will be able to do some tomography, I hope!

What and from whom did you order?

 

Is like a stackshot rotation base with controller? 

 

I am currently researching macro rails and considering building an automatic one. But I keep getting pulled into the WeMacro rail. I like the software. 

[colinbm]

@dabateman What magnification are you planning to use & what is the smallest steps do you plan to use ?

[dabateman]

I will most likely default to 2:1. But I would like to try out my 80x, 60x, 40x UV,  20x, 15x UV, 10x, 4x and 2x uv objectives.

I need to do a shoot out at 1x, 2x and 4x with my various lenses.

 

I have deep UV capabilities at 1x, 2x, 9x, 15x and 40x.

 

As for steps, I would need to calculate that over the imaging depth, with 30% overlap.  It would be great if cameras did that which support focus bracketing. 

 

CrispyBee is encouraging me to build a rail using a THK KR2602B. But I haven't seen one on sale cheap yet. There are large ones for cheap. But not at 1mm or 2mm per turn.

[colinbm]

@dabateman for my 100x objective I would love this rail setup.....
https://flamicroz.com/en/-macro-focusing-rail-/167-3192-ultra-rail-mini-v2-kit-incl-usb-controller.html#/93-qooli_reil_shutter_release_cable-panasonic_cl_l1

Some of the others I have see are pretty wonky at higher magnification them 20x.
With the 100x objective, I can not play music while photographing !

[Andy Perrin]

David, “ComXim MT200RUWL20 Remote Control,WIFI,USB,Rotating Electric Turntable for Photography,Display,Support Secondary Development” is the device. I bought it on AliExpress. It can be controlled directly using serial commands from MATLAB (or anything else that can send serial commands) and they have published their API online. Those were the deciding factors for me. They claim 0.5 degree resolution. 

[dabateman]

3 hours ago, colinbm said:

@dabateman for my 100x objective I would love this rail setup.....
https://flamicroz.com/en/-macro-focusing-rail-/167-3192-ultra-rail-mini-v2-kit-incl-usb-controller.html#/93-qooli_reil_shutter_release_cable-panasonic_cl_l1

Some of the others I have see are pretty wonky at higher magnification them 20x.
With the 100x objective, I can not play music while photographing !

They are ripping you off.

This one would be better and cheaper: 

https://www.mjkzz.com/product-page/thk2001a-based-rail-system-usb

 

That looks like the Qool-rail-250 rail which you can buy directly here:

https://www.mjkzz.com/product-page/qool-rail-250

[dabateman]

Andy those are cheap and I even see them on Amazon.  Interesting, product and will have to think about that. Thank you. 

[colinbm]

@dabateman Thanks for these links Dave.
I will keep an eye on them, but none have a release cable for Sigma fp :-(
 

[dabateman]

1 hour ago, colinbm said:

@dabateman Thanks for these links Dave.
I will keep an eye on them, but none have a release cable for Sigma fp :-(
 

Based off this you can make one.

https://www.reddit.com/r/sigmafp/comments/qpj18t/so_dudes_i_reverse_engineered_sigma_fp_cr41_cable/

 

Looks like only 3 and 4 from the tip are tripped.  These controlers use 2.5mm plugs so you can buy a release from ebay or make one using a 2.5 mm to 3.5 mm conversion cord.

Select the Canon RS-c1 cable. That will be 2.5 mm to 2.5mm cord. Then get a female 2.5mm to 4 pole 3.5 mm cord and splice it so that 3 and 4 get tripped.

 

If your are just going to use your Fp, than the Qool 250 rail is all you would need.

I am thinking of using my Fujifilm Gfx cameras.  So the weight is quite a bit more.

 

This looks like all you would need.

https://www.ebay.com/itm/203353803963?mkcid=16&mkevt=1&mkrid=711-127632-2357-0&ssspo=Bp-z_5qhSIG&sssrc=4429486&ssuid=RxqKLNPYQr-&var=&widget_ver=artemis&media=COPY

[colinbm]

@dabatemanThanks for this.
I'll have a look at it & see what I can do.

[Andy Perrin]

8 hours ago, dabateman said:

Andy those are cheap and I even see them on Amazon.  Interesting, product and will have to think about that. Thank you. 

The company makes TONS of similar models, but not all have equivalent features and functionality. Be careful to buy the right one. I could not find the USB controllable one on Amazon at the end of February, it was out of stock, but I see it there now. It has many good reviews on Amazon from technically-oriented people.

 

https://www.amazon.com/ComXim-Photography-Turntable-Shooting-Direction/dp/B079L1L5G8/ref=mp_s_a_1_1?crid=1MLJJ62TYRVTJ&dib=eyJ2IjoiMSJ9.GhGNsNGy1UCdKWMnNJKirQ.D01ZWVJxys_OX_QSC5bj6i870p29NFeO0EKQqFDKwHw&dib_tag=se&keywords=ComXim+MT200RUWL20&qid=1710736000&sprefix=comxim+mt200ruwl20%2Caps%2C199&sr=8-1