Transcript for:
Ultrasound Compression and Dynamic Range

hi learners it's em from sono nerds and this video is on unit 16 compression and dynamic range so unit 16 compression dynamic range now we've already talked about the idea of dynamic range and compression when we discussed receiver functions and contrast resolution so this unit's mostly going to be a review but another opportunity to understand the concepts just a little bit further so compression and dynamic range are used interchangeably but to understand the concept it might be helpful to think of dynamic range as the result of compression section 16.1 compression recall that we have two stages in the compression process we have first compression and second compression during either step of compression there are three rules that must happen first the largest signal must remain the largest if you have really really strong signals they still have to be coded as the strongest signals out of the group second the smallest signals must remain the smallest if you have a very weak signal coming back it also has to be recorded as the weakest signal you can't start flip-flopping things if you do you're going to start introducing errors and the third is that the range or the choices of decibels that are being displayed of the signal in between is reduced and this statement probably is where it starts to get a little bit confusing the idea behind the range of the signal in between being reduced is that we are reducing the vast amount of choices into a smaller group we'll go over some examples later in the unit but the biggest part of these three rules is that when this occurs and compression is done correctly no errors will be introduced we won't all of a sudden have weak signals as strong signals strong signals as weak signals or all of a sudden have things that should be anechoic being displayed as echogenic and vice versa so the whole idea is that the range is reduced so the machine can either process the signals correctly and that's going to be between components within the machine during our first compression or we can improve our contrast resolution during second compression so speaking of first compression let's go over that again first now remember first compression is an automatic function of the machine it's going to be performed automatically so this is not controlled by the sonographer and it has to occur so each component of the machine can accurately process the signal as it gets to it so in general the dynamic range is really going to decrease the more the information is processed the transducer is going to have the highest dynamic range where the archival system is going to have the lowest and we see that in this chart here the transducer has a dynamic range of 120 decibels where the archive has a dynamic range of 10 to 30 decibels and so that means that the transducer is capable of processing a trillion times larger signals than the weakest signal that it can process that's the 120 decibels remember for every 10 it's multiplied by 10 so it's 10 to the 12th it's how strong that strongest signal can be in the transducer compared to its weakest if we flip that down to the archive though at 10 that means the strongest signal can only be 10 times larger or at 30 decibels a thousand times larger than its weakest signal so we're talking by the time that we are printing pictures out on that little thermal printer connected to the machine we're talking about taking signals that can reach up to a trillion in differences down to like tens or hundreds or thousands of differences so this is why dynamic range has to be reduced this is why compression occurs really as those signals are processed we don't want to lose the information about that strongest or weakest signal so we need to compress them together make their differences less drastic without losing the hierarchy of strength and again through compression we make sure that each system component does not introduce any errors into the image and remember that dynamic range is measured in decibels it's going to be a comparative number let's look at some examples that occur in the current textbooks that are out in the market right now for ultrasound physics dr sydney edelman has a green understanding physics book it is his fourth edition and in this chapter that he discussed dynamic range he uses the bathroom scale as an example for compression in this example we first discussed the idea that people have weight and their weight can really be anything 0 5 10 200 250 350 pounds there is an actual weight to that actual person that is their real life weight however when they step onto a digital scale in their bathroom that digital scale might only be able to display weights from 30 pounds to 300 pounds and so what we see is a compression of the information of that person's actual weight so on the lighter side of things if a person doesn't even meet the threshold of 30 pounds the scale might not even know that they're on it or just weigh them at 30 pounds so it is compressed everything from 0 to 30 as just a flat 30. on the other side on the heavier side of things if the scale can only go up to 300 pounds anybody who weighs over 300 pounds and steps on that scale is going to be weighed at 300 pounds based on that scale now the important part about this example of compression is that if you're 15 pounds you're not all of a sudden going to weigh 300 on this bathroom scale you're just going to be lumped into the 30 pound range same idea on the other side if you weigh 340 pounds you're not all of a sudden going to weigh 10 pounds on the bathroom scale you're gonna weigh 300 pounds lumped with everybody from 300 and above so the real life weights have been compressed into a range between 30 and 300 pounds on the bathroom scale and we can kind of see that in our graphs here we have if we could just take everybody's weight plot it on their actual weight plot it on a graph we could get something like this we would get exactly what they are on our graph if we had a scale that could display exact weights but remember our scale only has a range of 30 to 300 so our light people now become all 30 and our heavier people all become 300. so we have compressed the actual weights into a range some don't meet the threshold some are over the threshold but they are not flip-flopped mixed around they stay on the light side they stay on the heavy side and then you have your ranges in between another common book that is used by many programs is frederick krumkaw's book i believe that book is up to a 10th edition but in his book he talks about compression in the idea of age and height and in this example he talks about a brother and sister the sister is five years old and the brother is 15 years old they have 10 years of age in between them so their range is 10 years but now we are looking at their actual height where the 5 year old sister might only measure 40 inches the 15 year old brother probably measures somewhere around 65 inches if we were to add 15 years to their age now the brother and sister are 20 and 30. the sister has grown a lot so maybe she's more now towards the 64 range and the brother has probably only grown a little so maybe he's more around the 68 inch range so when they were 5 to 15 10 years apart still they had a large dynamic range there was a big difference between them their heights were compressed they are still 10 years apart but the difference has reduced so the options of heights between them are reduced and we see that again on our graph we have the five-year-old sister compared to the 15-year-old brother in blue we have a large range of heights in between their two graft heights compress that and what we end up getting is a very small range in between their heights the sister is still younger the brother is still older they're still 10 years apart we've just changed the differences in their heights to be reduced so again weakest or smallest is still true strongest or tallest is still true we've basically just reduced our choices or made it less drastic the changes in between the two so we can apply these two examples to our transducer going to the receiver so we're really going to see a similar occurrence with the ultrasound system components the transducer might be able to process signals up to 120 decibels but the receiver might only be able to process signals up to 80 decibels so the signals above 80 decibels maximum cannot accurately be used by the receiver so on the stronger side of things compression needs to occur to bring all those signals coming back above 80 compress them all down to the maximum that the receiver can handle that's the compression of the values on the flip side of things maybe the receiver is not going to be able to process anything that's weaker than 10 decibels even though the transducer might be able to so again everything needs to be pre-amplified and lumped into that 10 decibel range so the receiver can further process the signal as it comes back from the transducer again the important part is that the weak signals are still on the weak side of things the strong signals are still on the strong side of things so again just to walk through the graphs here this is our transducer the transducer can take those real life those analog signals coming back from the body and it can handle all of that it's got from 0 to 120 super super sensitive to the signals that are coming back it's going to send all that information over to the receiver the receiver is not as sensitive it has a dynamic range of 70 so 80 to 10. when this happens it needs anything below 10 to be amplified up to that 10. so two three four decibel signals are going to be amplified to 10 they're still going to be on that weak side and anything stronger than 80 is going to be reduced down to that 80 still on the strong side we have compressed the values instead of having a range of 120 we have a range of 70. we've kept weak weak strong strong so no errors are going to be introduced as the signal moves into the receiver so that is first compression anything that happens between the system components is going to be a first compression it has to happen so each component of the system can continue to process the signal that is being received switching over to second compression we need to remember that second compression is controlled by the stenographer we have knobs on our machine that allows us to do that so your knob might be labeled compression as it is in this image or it might be labeled dynamic range and by adjusting this knob you are adjusting the spread of choices or the grays that the system is going to use to display the reflector strings now typically what happens it's going to do more compression on the weaker sides so it's really going to assign more blacks to the weak signals it's going to assign more whites to the strong signals and then we get just a little bit more differentiation of our signal in between so let's take a look at some examples of second compression when you are increasing the dynamic range you are increasing it as if you were going towards a transducer you're trying to make it so you can see all the real life signals that are coming back so increasing your dynamic range means you're going to increase the number of grays that the machine is using across the spread of signals what tends to happen when you increase your dynamic range is that we get an image that uses a lot of gray tones sometimes they kind of look really washed out they're almost all gray there are still a few whites with the strong strings and there's a few blacks with the anechoic areas but really we're going to see kind of an overly gray picture so using our graphs to kind of help us remember that most of our systems are capable of displaying 256 shades of gray and if we are using a dynamic range of 70 decibels the machine is going to assign the weakest weakest as zero as a true black for every change of a 0.2 decibel we're gonna see a new gray but what ends up happening is that most of our usable echoes that are coming back are going to fall in this mid-range area and that's why our picture looks really really great because almost everything is being mapped into these light grays again we do have some whites as we can see in the picture here we have some blacks as we can see in the picture here but for the most part everything is just kind of those mid-level grays so when you have a wide or high dynamic range the machine is going to spread those grays out along all of the signal and we end up with a really low contrast image so wide high dynamic range means we're using more grades for the signal and we end up with a really low contrast image so let's compare the sun to a low or narrow dynamic range image so in this image we are at 70 for our dynamic range now we've dropped it down to 30. you can see that here in the corner here so what first has happened is that the machine is going to reduce the dynamic range so remember we were zero to 70 now we're 40 to 70. so we've reduced our dynamic range to 30 70 minus 40 is 30. and what it's doing when it does this is basically saying any signal that is coming back under 40 decibels just to sign it black that's where we're going to start so 40 is our new zero anything under that assign it to black from there then we have 30 decibels to spread out over the 256 shades of gray so now instead of for every basically 0.3 decibel change we got a new color of gray now for every 0.1 decibel of change we are getting a new shade of gray so really what we're doing is taking all of the signals above 40 the one that's making up most of our picture anyway we're now saying put the 40s in the dark grays keep the 60s and 70s up towards the white but use all those shades of gray in between compare that to our 70 decibel range 40s probably started right about here at those mid-level grays that's all we were using for most of our picture were these really bright grays kind of into the mid-range that's what most of our picture was made out of that's why it looks a lot grayer compared to our low dynamic range because now we're using all of the shades of gray to display where most of those echoes are returning from so the biggest thing high wide dynamic range is going to use more grays and low contrast where we see narrow low dynamic range creating a picture that basically uses less grays and therefore it's a high contrast image and we can see that in our image down here the ivc was black before it's still black now just looking at the pictures you can see that this is very washed out kind of very gray the ligamentum venosum is this white part here it just jumps out at you so much more in the low dynamic range image it's just a lot more contrast in this image our whites are whiter our blacks are blacker and we just are using a lot more of those shades of gray to express the parenchyma of the abdomen now your compression map is going to change depending on your machine and depending on your settings this is a pretty common one where just more get assigned to black sometimes you're going to have more assigned to a white kind of like in those graphs that we saw earlier when we were talking about the receiver and about the bathroom scale so it really depends on the compression map that your machine is using but it's mostly common for again for those low level echoes just to get assigned to black and then using the rest of the grays to display those other decibel signals that are coming back so here's yet another example of low contrast versus high contrast high dynamic range versus low dynamic range so look at this top example here 90 decibel dynamic range that means that for every just about 0.4 decibel change we are going to be using a new shade of gray but remember most of our echoes are coming back in that 40 to 60 decibel range and so all we're really using in this picture are those mid-level grays we're going to have very few blacks very few whites we're really just hanging out in those mid-level grays because so much of it has been mapped to those compare that to the 60 decibel range now we're starting to get some true whites we're starting to see some true blacks we are using all of the grays to display a smaller range of signals and then compare that lastly to again to our 30 decibel range bright bright whites dark dark blacks we're getting a lot more contrast in this image so high dynamic range very gray kind of washed out low contrast low dynamic range lots of blacks and whites high contrast those are the main key things you got to remember about dynamic range and contrast to finish up this section then i just kind of want to cover more of a clinical discussion about why we want low dynamic range in certain situations versus high dynamic range and other situations in older systems the limiting factor of dynamic range was really more closely tied to the memory of the scan converter remember if we started with a one bit memory we could only display black and white while that was super high contrast it didn't do much for our contrast detail or spatial resolution so when we think of super high contrast that's going to be a straight black and white when we got up to 256 shades of gray though now we're almost to the point where our eyes can't even manage all of the grays that are being used because we can really only see some text say 30 some say up to 64 different shades of gray so the compression that occurs actually is more beneficial for our own eyes and that's really more where we're at with how compression works so typically what we want to do is reduce the dynamic range so we have a higher contrast image and therefore our eyes can pick up on those subtle changes that are being displayed now remember when you decrease your dynamic range what typically happens is that those really weak echoes are going to be assigned to a black shade of gray it's just going to kind of compress them all down into those lower level echoes sign them all as black and when we are doing vascular work or echo we typically want our pictures a little bit more black and white because we normally want anywhere that there's blood like in the heart chambers or in a vessel we want that to be mostly anechoic but if there's thrombus in there thrombus typically takes on a low level echo and if your dynamic range is too low you might actually end up hiding it within that kind of compression that occurs with those low level echoes so i kind of have an example of that here say this is our real tissue we're looking at a vessel and it has some thrombus in it so if we decrease our dynamic range take a low dynamic range those weak echoes that are coming back from the thrombus those are going to be assigned black or nearly black and it might get kind of hidden in the vessel the really nice part is though we're going to have a nice black lumen we're going to have a bright white echogenic wall and kind of darker tissue around it and that's going to look nice for our eyes but we might miss pathology because we have reduced our dynamic range too far so when we are looking for pathology that typically returns low level echoes we might actually want to use a little bit more of a high dynamic range the higher dynamic range is going to make everything look a little bit more gray we have a better chance of seeing those low level echoes because it's getting assigned its own gray instead of being compressed into the black shades so this makes dynamic range actually very important to the sonographer not only for being able to pick things out because you don't want to hide things but you also want to use to really help elevate the diagnostic value of your images as well so it's really a balance act between hide things or make things look really pretty with those bright white walls and anechoic blood within compare that though then to say like a liver lesion if we have real tissue this is our liver lesion with some mets in it maybe when we have high dynamic range the high dynamic range might hide those echoes that are going to be very similar to one another say the liverpanchama was assigned like a grayshade number 213 and the pathology that's in it is coming back at a great shade level of 215. are our eyes going to be able to see that probably not there's got to be a larger gap in between the signals that are coming back to change the grays enough for us to see them so if we compress the signal down we change the gap to be in a smaller gap and now a wider difference of grays are going to be seen on that liver parenchyma so low dynamic range when we're looking at tissue looking for things that might be very subtle within the tissue is going to be important where a high dynamic range could possibly hide them so in the end we see that low dynamic ranges are going to improve your contrast resolution however that contrast might be a little too contrasty and might end up hiding some pathology as well so as a sonographer not only do you have to be aware of these pitfalls of the changes that you're making on your machine but you also have to use them it is not a set it and forget it you want to change different settings if you think you see something change your dynamic range see if you can get something to pop out a little bit more it's very important that we understand how to use our machine to create the best diagnostic pictures possible and that is it for our review of compression and dynamic range the biggest takeaway from this is to remember those three rules strongest say strongest weakest stay weakest and we're going to reduce the range of signals being spread out over those 256 shades of gray high wide dynamic range produces grayer pictures which offer low contrast images where our narrow or low dynamic range is going to produce high contrast images because they are using less grays technically what they're doing are using all of the shades of gray for a smaller range of signals remember too then that there is that first and second compression first compression automatically has to happen so that the signal can be processed by each level of the system components transducer has the widest dynamic range where the archive has the narrowest dynamic range so i really highly encourage you the next time that you are scanning change your dynamic range see what happens to the picture notice how when you go to low dynamic range your picture becomes more black and white notice when you change your dynamic range to the high end it's going to become more gray kind of more washed out play around with those settings so when you're asked questions about it on your tests or on your boards you have a very working hands-on knowledge that experimental knowledge that you gained from practicing it in the clinical setting again you do have some activities in your workbook and then those open-ended nerd check questions to go through as well