Handheld ultrasound scanners

Go To Last Post
32 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Nothing to do with AVRs, but can someone explain to me how the handheld ultrasound scanners detect position/motion? The ones they use on pregnant women, or old farts like me, when checking for Abnormal Aortic Aneurysm.

Do they have inertial sensors? Cameras like optical computer mice? Do they simply paste together images based on matching overlap, like dendrochronology? Or is there some other, blindingly obvious method that I've missed?

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

They more or less work as Asdic.

Send out a HF 'ping'  and process the echos coming back.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks, but that doesn't answer my question. How do they know where they are, relative to my insides?

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

they do not. They just put it at your body and see what picture is there. As all the organs have somewhat of a special response to the ping they can build a picture, that is all they go by.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

So why do they move it about on the surface of my body?

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I guess the question is: are they moving it around to build up a single picture ("scanning"), or are they moving it around just for a different view - as you would move a normal camera to get different views from different angles ... ?

 

EDIT

 

I guess that the typical "sector" shaped picture that you get suggests the latter ... ?

Top Tips:

  1. How to properly post source code - see: https://www.avrfreaks.net/comment... - also how to properly include images/pictures
  2. "Garbage" characters on a serial terminal are (almost?) invariably due to wrong baud rate - see: https://learn.sparkfun.com/tutorials/serial-communication
  3. Wrong baud rate is usually due to not running at the speed you thought; check by blinking a LED to see if you get the speed you expected
  4. Difference between a crystal, and a crystal oscillatorhttps://www.avrfreaks.net/comment...
  5. When your question is resolved, mark the solution: https://www.avrfreaks.net/comment...
  6. Beginner's "Getting Started" tips: https://www.avrfreaks.net/comment...
Last Edited: Wed. Jul 29, 2020 - 09:34 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I don't know. But it's not a camera, is it? It's an ultrasound device. A camera has a whole matrix of sensors for each pixel.

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The scanner can only see a small field..so they have to move it around to see more.

But they can use some image processing smarts to line up things in 3D

 

If you took a 3D cad drawing of a toaster & sliced it up & threw the slices on the table , you could probably figure out how to put the slices back together.

Especially if you had another pile with more slices taken from a slightly different angle ...together they hold all the clues needed for reassembly.

 

I just realized I created an example of using slices on a toaster....interesting

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So like I said in my original post, " Do they simply paste together images based on matching overlap, like dendrochronology? "

Thanks.

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0


John_A_Brown wrote:
But it's not a camera, is it?

I don't know - my point is that, perhaps, it is like a camera.

 

It's an ultrasound device. A camera has a whole matrix of sensors for each pixel.

Who's to say it doesn't also have an array/matrix of sensors ... ?

 

EDIT

 

Wikipedia seems to thing that it does:

 

Wikipedia wrote:
Ultrasonography (sonography) uses a probe containing multiple acoustic transducers to send pulses of sound into a material

 

https://en.wikipedia.org/wiki/Medical_ultrasound#Sound_in_the_body

 

again, that would explain why you get that typica "sector" view ...

 

Top Tips:

  1. How to properly post source code - see: https://www.avrfreaks.net/comment... - also how to properly include images/pictures
  2. "Garbage" characters on a serial terminal are (almost?) invariably due to wrong baud rate - see: https://learn.sparkfun.com/tutorials/serial-communication
  3. Wrong baud rate is usually due to not running at the speed you thought; check by blinking a LED to see if you get the speed you expected
  4. Difference between a crystal, and a crystal oscillatorhttps://www.avrfreaks.net/comment...
  5. When your question is resolved, mark the solution: https://www.avrfreaks.net/comment...
  6. Beginner's "Getting Started" tips: https://www.avrfreaks.net/comment...
Last Edited: Wed. Jul 29, 2020 - 10:38 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

OK. Thanks. Looks like the multiple sensor thing may be more for phased arrays than camera analogy, but interesting stuff, anyway.

 

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 2

If only someone could device an algorithm to detect duplicates...

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

We could have a whole thread about that.

 

And then duplicate it all 4 months later ...

Top Tips:

  1. How to properly post source code - see: https://www.avrfreaks.net/comment... - also how to properly include images/pictures
  2. "Garbage" characters on a serial terminal are (almost?) invariably due to wrong baud rate - see: https://learn.sparkfun.com/tutorials/serial-communication
  3. Wrong baud rate is usually due to not running at the speed you thought; check by blinking a LED to see if you get the speed you expected
  4. Difference between a crystal, and a crystal oscillatorhttps://www.avrfreaks.net/comment...
  5. When your question is resolved, mark the solution: https://www.avrfreaks.net/comment...
  6. Beginner's "Getting Started" tips: https://www.avrfreaks.net/comment...
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have a sneaking suspicion JC is going to know more about this than most of us ;-)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Image-based sensing approach extracts the relative positions by analyzing the image feature, for example, speckles, instead of depending on position sensors [24]. According to the phenomenon of speckle decorrelation, the speckle pattern should be the same if two images are acquired at the same position, which results in nondecorrelation. However, the decorrelation is proportional to the distance between two images. To obtain the relative translation and rotation between two images, the acquired images are divided into small subregions. Calculated decorrelation values can be used to analyze the relative position and orientation of adjacent 2D images. Using this scanning protocol, operators are supposed to move the transducer at a constant velocity in linear or rotational manners to guarantee appropriate intervals. 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Interesting topic.

 

Just this month at our ER Doc staff meeting we had a vendor bring in their newest, latest and greatest, ultrasound machine for a demo. 

It costs about $50,000 base price, but I don't recall which probes that included.

 

Disclaimer: As an old timer, I rarely perform ultrasound myself.  I often order a formal ultrasound study, which is then performed by an ultrasound tech, and is read (interpreted) by a radiologist.  examples might be to examine a patient's heart, wall thickness and motion, and heart valve function; or look at an aorta, (the big blood vessel coming off the heart and supplying blood to the rest of the body), to see if it has an aneurysm, (weak bulging spot in the wall), or a dissection, (a tear within the layers that make up the wall); or an OB scan to see the fetus, (size, development, fetal heart activity,...), or a patient's leg to see if there is a blood clot, (DVT), in the leg; or their eye, (to see if the retina is detaching, etc), or to locate large blood vessels for putting in a "central line", (a really big IV catheter), or to scan the patient's abdomen for blood swishing around inside from a leak somewhere from trauma; or rather frequently to check their gall bladder for gall stones; etc.  The list goes on and on. 

 

My younger partners learned how to perform many of these ultrasound scans, and interpret them, during their training as an intern and as a resident.  I learned a little of it over the years, but am clearly nowhere near as proficient at it as "the kids".  I'll use the ultrasound for trauma exams when I work on the helicopter, but in the ER I usually rely on the ultrasonographer.

 

When I think of US in electronics I'm usually thinking of 20 or 40 KHz frequency transducers for ranging or motion detection.  Things like robots tracking parallel to a wall or avoiding an obstacle in front of them, or measuring how full a container is as it is filled, or burglar alarms, etc., all come to mind.

 

Marine usage for depth finders and submarine locating also use sonar, which is essentially ultrasound in a water medium.

 

Medical ultrasound is typically performed at 1 - 20 MHz, as an old school engineer I tend to think of that as R.F., not ultrasound...

 

A single transducer, pencil like probe, can be used to listen to a fetus's heart rate by listening to the doppler shift of the returning ping signals due to the motion of the baby's heart.  The "envelope" of the frequency shift is played through a speaker and one can hear a whoosh, whoosh, ... with each heart beat.  That is perhaps the simplest medical ultrasound device and usage.

One can purchase a fetal heart rate doppler ultrasound these days in stores for ~ $20 USD.

 

The "probes" come in different frequencies, and as single transducers and as arrays.  How the pings are send, received, and decoded is based upon what one is imaging.  In the old days there was B Mode and M Mode, now days there are too many "Modes" for me to keep up with.

 

Pretend that the US "beam" was a single, narrow, line/ray/laser beam.

Aim it through one's chest at the patient's heart, for example.

Send a single ping, (for example).

As the US wavefront travels through the body some of the energy is reflected back at every tissue interface, (different tissue density).  So the beam goes through the skin, and hits a chest wall muscle.  Some energy keeps going through the muscle, some is reflected back.  Far side of the muscle it hits some more "soft tissue".  Again, some goes through it, some is reflected back at the interface. It hits the surface of your heart, some is reflected back, some keeps going forward.  It hits the inner chamber of the heart, full of blood, again some is reflected back.  It hits a thin heart valve leaflet, some is reflected back, it hits the back wall of the heart muscle, then more soft tissue, then your back muscles, then your back's skin...

 

Over time, following the ping, one receives the various reflected pulses.  For a robot looking for an obstacle it is a binary decision, either there is a returned signal within X meters in the front of the robot, or there isn't.   For our patient, above, there are a bunch of reflected signals, all of them important.

 

If one aims the beam at the patient's heart valve, and plots the return signals as a function of time, across the screen, (penetration depth on the vertical axis, time on the horizontal axis), one sees a bunch of lines across the screen.  Each line is a reflection from the US energy hitting a different tissue.  The vertical axis is time or depth or distance the ping traveled.  So I see nice stationary lines at different depths, and can see how thick your skin is, how thick your muscle layer is, how think your heart wall is...

 

Down near the bottom of the screen I see a saw tooth waveform across the screen.   That means that the distance from the probe was changing, nearer-farther, back and forth, over time.  That thin line is your heart valve's leaflet, moving back and forth as the valve opens and closes with each heart beat.

 

Still with me?  This is a lot easier to explain to someone with a probe in hand, and an image in front of you...

 

Now let's step it up a notch.  Instead of one transducer element, let's have an array, so I can see the reflections along a beam from multiple "parallel" beams, simultaneously.  Now instead of a 1-D image vs Time, display I can have a 2-D image vs time display, and I can see a lot more information in the image.  (2-D, I see depth into tissue and tissue reflections along one axis, (each beam), and I see the second axis as the stack up of parallel beams, now giving me a 2_D image, or a 'Plane", i.e. a cross sectional cut.  Time is an additional dimension.)

 

Now I use my wrist to orient the linear array's beams so that they slice through the patient's body at just the right orientation in 3-D space to visualize what I want to see, (your gall bladder, your appendix, your heart, ...).  Recall that the beams are not narrow like a laser beam, but they are still pretty narrow, and I am essentially visualizing "planes", slices through the patient's body.

So cool, especially to see it in action, rather than listen to my description of it.

 

Now add color to the Doppler shift and I can see the velocity and direction of the blood flowing through your heart value.  Does the valve work, or is it leaking backflow of blood when it is supposed to be closed?

 

If I am looking at your forearm, trying to find a vein in which to start an iv, I can see the round cross sections of arteries and veins, as their walls, and the blood within them, are different acoustic densities than the soft tissue in which they lie.  Both look like round circles in cross section.  How do I know which is the artery, which I (usually) want to avoid, and the vein, which I'm looking for?  If I now gently press the transducer against the skin then the low pressure, soft walled, vein collapses and flattens out.  The high pressure, muscularly walled artery maintains its circular cross section.  Easy.

 

Now I can use my wrist to rotate the transducer array 90-ish degrees an instead of seeing the vein in cross section, I can see the pipe running lengthwise through your arm.

 

There are now many different scanning "Modes", and image reconstruction techniques, all designed to optimize obtaining certain information.  An ultrasonographer doing a formal heart US might take an hour of waving the probe over your chest, aiming the beam at certain parts of your heart, in certain angles, and with different scanning modes, to obtain all of the images and data needed for the formal echocardiogram report.  When I do a trauma "FAST" exam, looking at about 4 items, it takes < 1 minute to do the procedure.

 

The coolest development in US, in my opinion, is 3-D obstetric ultrasounds.  One can generate a 3-D appearing image of the fetus.  Totally useless for me in the ER, and our new $50K+ machines can't even generate one of these images, but still so cool to look at.  Google 3D fetal ultrasound, then click on images, to see a bunch of baby ultrasound images, all generated with "sound" waves and echo data analysis.  This makes "The Hunt for Red October" sonar look like the dark ages of crystal radios.

 

Wiki Medical Ultrasound has a very cool photo of an US video image of a heart showing the four chambers of the heart and two of the four valves, with their leaflets opening and closing.

I should have just referred to this page image and not tried to write my own comments!

 

Oh well, hopefully that helps a little bit.

 

JC, ER Doc, but definitely NOT an ultrasound expert! 

 

Edit: typo

 

 

 

Last Edited: Wed. Jul 29, 2020 - 07:17 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Funny I asked this when I went for an ultrasound of my kidney.  I asked the operator if she knew what teh jelly is for and she did and also explained what goes on...most of what is posted above.

 

In the 'wand' there are several crystals that are resonating at different frequencies.  The various internals of us respond differently to the frequencies so the tuned receivers send back different signals to the machine.  (she did not know what goes on inside the machine and thats ok).  The image on the screen is the field of view the scanner has so they have to keep moving the 'wand' until they get a full image/partial image of what they are looking for.  Experience is how the operator can decipher the images.  The jelly is to both lubricate teh 'wnd' so it does not bind against the skin, and it also is there as a matching agent for the sonar waves coming off of the 'wand'  Muck like we use matching gel in fiber optic connectors.

 

 

Whether or not she was wrong I do not know, but she has been doing this work for 20+ years so I am gonna think she knows what she is talking about.

 

JIm

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Give me a pingVasiliOne ping only, please."

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The wand is an array of piezo transducers.the frequencies involved are up to the likes of 2MHz. Rather than using the electromagnetic wave, they use the physical wave and the goo is used to improve the coupling.
Since an array of transducers are used, techniques like phased array are implemented. There is a lot of processing that is performed.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
The wand is an array of piezo transducers.the frequencies involved are up to the likes of 2MHz. Rather than using the electromagnetic wave, they use the physical wave and the goo is used to improve the coupling. Since an array of transducers are used, techniques like phased array are implemented. There is a lot of processing that is performed.

 

So the lady that answered my question was pretty spot on I'd say.

 

Jim

I would rather attempt something great and fail, than attempt nothing and succeed - Fortune Cookie

 

"The critical shortage here is not stuff, but time." - Johan Ekdahl

 

"Step N is required before you can do step N+1!" - ka7ehk

 

"If you want a career with a known path - become an undertaker. Dead people don't sue!" - Kartman

"Why is there a "Highway to Hell" and only a "Stairway to Heaven"? A prediction of the expected traffic load?"  - Lee "theusch"

 

Speak sweetly. It makes your words easier to digest when at a later date you have to eat them ;-)  - Source Unknown

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB, RSLogix user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

it's not a camera, is it?

 More like RADAR...

 

These came up recently:

ultrasound probe tear-down:  https://www.youtube.com/watch?v=...

Machine/display tear-down:  https://www.youtube.com/watch?v=...

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks Bill, I got my daily Dave injection. One can never stop there, so that's a couple of hours I'm never getting back :)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

obdevel wrote:

"Give me a pingVasiliOne ping only, please."

cheeky

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Give me a pingVashiliOne ping only, pleashe."

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

barnacle wrote:

"Give me a pingVashiliOne ping only, pleashe."

 

The name'sh Bond, ... :)

 

Vaguely back on topic, I have seen my liver on a live ultrasound scan. You really do not want to catch Hepatitis; fortunately is was B not C, and self-limiting. One of the joys of working in further flung parts of the world.

 

I use one of those cheap ultrasound modules to sense the water level in my allotment water butt. It's accurate enough for measuring 1/4, 1/2 and 3/4 full.

 

Does anybody remember ultrasonic TV remotes ? The type where you could randomly change channels by jangling the coins in your pocket.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I doubt you could produce an accurate scan of a fetus by jangling coins in your pocket, but I suppose anything's possible, with enough post-processing.

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Wow, I'd not seen the ultrasound machine tear down video before.

Quite impressive, both for its overall design, and to actually construct it.

 

I was a little bit surprised at the number of jumper wires / modifications done on the commercially released units.

 

The wide spacing for the isolation seen on one board had some really expensive burr brown chips, haven't seen those in a long time.

 

When imaging the heart one typically has the EKG displayed across the bottom of the ultrasound, so one can correlate the electrical activity and the mechanical activity of the patient's heart.

 

To do that one has EKG electrodes on the patient, and an EKG (analog) front end.

 

Given that patient's sometimes have a cardiac arrest, and might get defibrillated (shocked), and the ultrasound electrodes are on the patient, one doesn't want to fry the ultrasound machine with a several KV volt signal.  Hence the attention to the isolation.

 

Looking at all of that electronics is incredibly impressive.

These days one can fit the entire probe and processing within the probe, for a hand held unit, at least for the 2 MHz ish units, (not the 20 MHz units).

 

The new model our group just looked at looked like a laptop on top of the cart, with a small box under the laptop part.

 

Thanks for posting the link, very cool.

 

JC

  

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:

I was a little bit surprised at the number of jumper wires / modifications done on the commercially released units.

 

 

It's interesting how expectations have changed. For something of that complexity, cost and production quantity, having field mods is expected. Most of the 'fixes' these days are software updates. These could incorporate new fpga bitstreams or workarounds for silicon errors along with bug fixes/enhancements for the code itself. Thus the 'fixes' are invisible.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm certainly familiar with SW upgrades, (patches, fixes, new features, etc.).

 

It just caught my eye as to how many wired fixes there were, on both sides of some boards, on multiple boards.

 

I've seen the inside of our CAT Scanner, and the inside of our monitor/defib/pacer units, and I don't recall seeing any "post production" modifications.

 

That was also a 1990's vintage machine, so perhaps mods were a little bit more frequent back then?

 

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That's a staggering amount of electronics at work--very impressive. Hard to imagine all of it thrown in a dumpster, like an old pizza box.

Looks like a fun project to debug!

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I don't recall seeing any "post production" modifications.

For things of that vintage, that complexity, and that small of a target audience, a certain number of jumpers isn't "post production", it's PART of production.

You're talking about thousands of dollars to get each board made, you can certainly afford to have a tech install a "reasonable" number of jumpers ($hundreds/board or less), especially since they probably already have to be doing pretty extensive probing of all those test points and such...

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

There's also the point that while a new PCB would certainly require recertifying, modifications to an existing design 'fundamentally the same' probably don't...

 

Neil