Tissue Elasticity Evaluation – Clinical Applications - SD

Introduction

Hello, my name is Brian Garra, and I'm from the Washington DC Veterans Affairs Medical Center. I'm going to be talking to you today about elasticity imaging and clinical applications, and how you can use it in your practice. Today I'm going to be speaking to you about tissue elasticity imaging and some of the clinical applications that are currently available using this new technology, which is actually pretty exciting. The goals of this talk will be to summarize elastic and elasticity imaging methods to review some of the major diagnostic applications, both oncologic and non oncologic, to review the role of elastography and monitoring of treatment of tumors. And finally, to discuss some of the future potential of elasticity imaging and elastography.

Elasticity Imaging Methods

To begin with, I'm not going to discuss in great deal detail the methods by which grams are created, but I will briefly overview them. There are two major methods for doing elasticity imaging. One is called static methods, and the second group of methods are called dynamic. To begin with, static elastography is traditional elastography, the method by which the term elastography was created and invented for. It's basically the imaging equivalent of palpation, and it represents an image of tissue strain, which is produced by measuring tissue displacement as a function of distance from the transducer, applying the compression force to the tissue. When you apply compression to tissues, stiff tissues give low strain values, and soft tissues give higher strain values. You can think of a strain value as the amount of displacement of the tissue related to the distance of the tissue and soft tissues, the tissue near the transducer or the compressor displace more than the tissues further away, whereas with stiff tissues, the tissues near the transducer and tissues further away from the transducers displace the same amount. The final image produced in a static gram is one of relative stiffness. It's a relative image, very similar to an MRI, and thus, it's very difficult to derive specific fixed values from within that image.

How is static elastography performed? Typically, it's performed by doing a slow compression of the tissue, often with the ultrasound transducer, with tracking of tissue displacement. This is usually done from the raw ultrasound signal instead of the image, or detected ultrasound signal, because that gives higher resolution. The relative stiffness images are good focal disease, but not so good for diffuse disease. It's the most mature of the elastography or elasticity imaging technologies, and it can result in very high quality imaging. As I mentioned earlier, elastography is a, is actually an image of tissue strain, and this is, this slide is meant to illustrate how a tissue strain is computed. So you have a transducer on a block of tissue, for instance, as shown on the left hand drawing. You have a soft lesion marked s and a hard lesion marked h the com. The transducer then compresses the tissue, and the amount of compression is magnified on this for purposes of illustration, but it's actually a very small amount. It's usually like a half a millimeter, and it's barely perceptible to the person experiencing the pressure. And if you look at the two images, you, it's hard to tell the difference between the pre compression and the post compression image, but if you compress the tissue as shown on this slide, you have the soft material which deforms and sort of flattens out, and you have the hard lesion, which moves as a unit. So the hard unit doesn't change its shape, it just moves as a solid unit. And then if you measure the distance or the displacement of the front edge of the lesion s compared to the back edge of the lesion s you get two displacement. And you see the amount of displacement occurring at the back edge is less than the amount that occurs in the front. And that difference between the change in displacement between A and B is the strain. So you can see soft lesions produce a larger strain value than do hard lesions, because a hard lesion, the back edge displaces the same amount as the front edge pretty much, and so you have virtually no difference in the displacement. Thus the strain is almost zero.

So how's this displayed? Larger values of strain are displayed as wider areas on a gray scale image, and areas of lower strain are displayed as blacker areas on an image. And here's a classical breast cancer shown on gray scale ELAs agram. And here you can see the breast cancer on the sonogram produced from the same radio raw data as the ELAs agram, and then corresponding ELAs showing some of the characteristic features. In other words, the cancer is much harder than the surrounding tissue. It has lower strain values and thus appears dark, and it also appears larger than the lesion does on the gray scale image.

Dynamic Methods

Now, moving on to dynamic methods, which are also often called elastography. In a dynamic method, a vibration is induced in a tissue rather than produc, than just a single compression. This vibration causes waves to trans transmit across the tissue, and you either image the waves traversing the tissue, show the tissue movement, or you measure the velocity of the waves traversing the tissue. All of these can be done, and all of these can can be used to produce elasticity images. You may get a relative stiffness image, or you actually may get quantitative estimates of tissue stiffness. This method, since it is more quantitative, can be used for both focal and diffuse disease. And a couple of examples of systems that do this are the supersonic imagine machine, which creates shear wave images using a dynamic method and magnetic residency elastography, which also measures shear wave velocity. The first dynamic method, however, just tracked the vibrations as they traverse tissues, and this method was called sono elasticity imaging and was developed at the University of Rochester. Here's an example given to Mees courtesy of the University of Rochester, showing a prostate gram, where areas of tissue movement as detected by a color doppler are shown in green, and areas of decreased tissue vibration in response to the vibrator placed up against the tissue are shown as black. So you can see the prostate cancer outlined with the red arrows shows up as a dark area against the softer prostate tissues, which vibrate more during the imaging procedure. This method of imaging depends not only on the stiffness of the material, but also on the viscosity of the material. So what we call the viscoelastic properties of the material are being tested here, but it is a good method for detecting stiffer materials alone. Magnetic residency elastography is produced by generating, by placing a vibrating device up against the patient, and watching the sheer waves or the vibrations as they traverse the tissue inside the patient. This is a head, a MRI ELAs gram showing the transmission of the shear waves inside the brain and in the video and in the muscle as static images and in a phantom as static images. And by measuring the wavelengths between the vibra, the ShearWave vibration waves, the velocity can be inferred and the stiffness of the material can be assessed. This ShearWave ELAs gram is produced by an ultrasound device, which generates shear waves and tissues by sending a specially formed ultrasound pulse that generates shear waves or vibrations in the tissue. The velocity of the shear waves that are produced are estimated, and in the region of interest marked by the box, the color denotes the relative stiffness of the tissue inside in kilo past gals, although this is an estimate that is derived from an equation which may not always apply in normal tissues, this feature is not yet approved in the United States for the quantitative color bar, but is approved for just imaging without the quantification.

Clinical Applications

Let's move on to elastography applications. I've divided them into two main categories, oncologic applications and other applications.

Oncologic Applications

Talking about the on oncology applications, breast cancer is the first one that gained popularity. It was the first one to be described in a clinical study in 1997. Other applications that appear promising are lymph node evaluation, thyroid cancer, prostate cancer, and other cancers. Almost every tumor in the body has been studied using elasticity imaging. Treatment monitoring is also a promising area of investigation. What are some of the imaging features of tumors? All tumors appear to have increased stiffness relative to their benign surrounding tissue or benign tumors in the same organ. This increased sif stiffness could be related to increased interstitial pressure caused by leaky capillaries, or it could be due to traction on the surrounding connective tissues by the tumor cells themselves. But for whatever reason, tumors tend to be stiffer than surrounding normal tissues or benign tumors. Another feature that is seen on elastography of tumors is that they have increased size often relative to their sonographic image. This is particularly true in the breast, but is true to a lesser extent in some other organs. The final feature is malignant tumors often attach themselves to the surrounding tissues, and there are certain techniques that can be used to image this degree of attachment. This is a classic appearance of a breast cancer, an invasive ductal carcinoma in a patient studied with elastography. Now, in this image, the sonogram is here on the left hand side and the ELAs gram on the right hand side, and the tumor is right here. As you can see, the ELAs shows the tumor to represent a very dark or hard lesion or stiff lesion, and the lesion appears somewhat larger than its counterpart on the sonogram. These are both characteristic features of malignancy in the breast using elastography. I mentioned the possibility of detecting the attachment of a tumor to surrounding tissues. Typically, benign tumors do not attach themselves to surrounding tissues very well and are called mobile are, and they move and rotate and slide against the surrounding tissues, whereas cancers extend themselves into the surrounding tissues and attach themselves firmly. This can be picked up by a new method developed called axial shear strain elastography. In this method, the shear strain is shown in the colored blobs, which appear at the edges of the lesion. So the lesion on the axial shear strain, ELAs gram is here, and these blobs occur at the borders. Now, when there's a slip border, in other words tissue, the nodule is movable relative to the surrounding tissue. These areas of shear strain are very intense, but very thin. When the tissue is attached to the surrounding tissues, they're less intense, but also quite a bit larger. So this appearance is quite characteristic and is very helpful in distinguishing benign from malignant lesions. As you can see, the difference between a benign tumor with its thin dog ears and the malignant lesion with its thick areas of a actual shear strain is quite pronounced.

Breast Elastography

Well, what's the imaging procedure used for breast elastography? Typically, you use very light compression or just breathing motion. You place the probe on the patient's chest and just let breathing motion create the ELAs grams, and you often create a short video and several still images to show it, because the ELAs agram will show the lesion during certain phases of the breathing cycle and not during other phases. So to assess lesions whether they're present or not, and how stiff they are and how large they are, you really need to watch at various stages. During the compression process, typically both radial and anti radial imaging are performed, and as I mentioned, you use a mixture of cine loops and static images, and it's important to check the quality of the ELAs gram before using it for diagnosis, because it's very possible to create inferior quality ELAs grams that look like there's no tumor there, whereas a higher quality one will show the lesion and be more useful for evaluation of benign versus malignant. Now, it's not obvious to the eye which lesions, which ELA grams necessarily are high quality versus low quality. So most manufacturers have provided quality indicators to help distinguish a high quality ELA versus a low quality one. And most of these are based on the mathematical agro mathematical algorithms used, and a quality, numerical quality assessment that is performed by the imaging system. For example, the semen systems gives you a numerical readout of quality factor, and for quality factors above 70, it produces an image and allows you to assess it for quality factors. Below 70, the image is grayed out so that it can't be used for diagnosis. The ultrasonic system on the right uses a pie chart, and the goal is to have as many of the green pie shapes lit up in green as possible. The more pie shapes are the pie sections that are lit up, the higher quality the ELAs gram is. So it's important to have this available any system so that you can use that to assess which one, which of the images you should be looking at. So what are some of the imaging features of cancer in breast markedly? Increased stiffness, as I mentioned, increased size relative to the sonogram. Now the increased stiffness comes in several different flavors. You can have the uniformly stiff or hard lesion, which is uniformly dark, or you can have a hard center with a soft periphery, or you can have a soft center with a hard periphery in the case of a necrotic tumor. Or you can have multifocal areas of stiffness in a stiff but not terribly stiff lesion. Here's an example of a classic breast cancer, which is fairly uniformly stiff. You can see the lesion here larger than what it appears like on the sonogram and also relatively hard compared to the surrounding tissue. Here's an example on the left hand side of a hard centered lesion with a softer periphery. Again, the lesion is a whole, is larger than the sonogram, but the hard center is a relatively small component of a larger stiff, relatively stiff lesion. And on the right hand side is the opposite. Now, here we have a stiff lesion that's actually the same. It's the same as this one. You have a stiff lesion with a stiff center and a soft periphery shown in color rather than gray scale imaging. Here are a couple of other examples. The large lesion on the left hand side is a fairly stiff lesion, but with multiple focal areas of increased stiffness within it that may correspond to these microcalcifications that are present within the tumor. Again, this is an invasive ductal carcinoma. On the right hand side is a lesion where there is a stiff periphery shown in red on this ELAs gram where stiff corresponds to red rather than blue. And the blue center represents a softer center. So this is the stiff periphery soft center variant. So what about benign lesions on elastography? The characteristic features could be that it's not visualized, its stiffness often is very similar to the surrounding tissue, especially for fibrocystic nodules. Oftentimes, the lesion can be softer than the surrounding tissue, and this is the most reliably reliable indicator of a benign lesion. Some the size may be smaller on the ELAs agram than the sonogram. If it's a pronounced decrease in size, that's useful. If it's just about the same, then it's per perhaps not so useful. And for cystic lesions, there's a characteristic correlation pattern that occurs on ELAs grams. Here's an example, a couple of examples of fibroadenomas, a benign lesion of the breast. On the upper left hand side, you can see the lesion on the sonogram and the ELAs agram, and it looks very similar to the surrounding tissue. And although you can see the lesion, it'd be difficult to pick this out as a specific lesion if you didn't have the sonogram to compare it with. So this is a characteristic appearance of a fiber adenoma. Here's another example, another similar example of a fiber adenoma with a different machine. Again, the lesion on the sonogram and the lesion looks very similar to the surrounding tissue on the ELAs gram, although it is possible to detect it. And the third example, the lobulated mass at the bottom with the ELAs gram showing bands of stiffness and bands of softer material interspersed with one another, which can be a characteristic appearance of a fibroadenoma. Fibrocystic disease may appear as a nodule that you can detect, but oftentimes they're hard to detect or they may not be detectable at all. Here's an example of a fibrocystic nodule on sonogram and on the ELAs agram, it's detectable by this sort of ghost appearance at its borders, but otherwise would be very similar in stiffness to the surrounding tissue. This small fibrocystic nodule on the right hand side looks relatively soft on the ELAs agram and would not be classed as a suspicious lesion from malignancy. Breast cysts, as I mentioned, may have a characteristic pattern within them, and this is caused by de correlation of the echoes inside. Most elastography systems where they estimate strain use a cross correlation function to determine how much tissue displacement there is. This presumes that the material that is being compressed is solid. Well, since the material in assist is not solid, but rather fluid, the algorithm breaks down and various types of software handle this breakdown in the algorithm differently. On the Siemens machine, you get a characteristic bright, bright center with a darker periphery as a characteristic appearance for a debris fill cyst on the ELAs gram. The Hitachi, on the other hand, when an images cyst may produce this tricolor appearance, red, green and blue in a layered appearance within a cyst. Both are characteristic appearances and both are suggested of cyst, but the rules that apply to one machine do not apply to the other 'cause. It depends on how the programmer programmed the software to handle this de correlation and how it's going to appear on the ELAs gram image. So for looking at cyst, you'll have to learn once you have a system how a cyst appears on that system, and those rules that apply to that system will not be generalizable to other systems with different elastography software. So what about performance for of breast elastography for detection of cancer, typically the areas under the RC curve range between 0.9 to 0.96. It works best as an adjunct with sonography because it's not a reliable detection method, as you could see from the previous ELAs grams that I show, there are many dark areas on an ELAs agram that could represent cancers but don't, 'cause you don't see a lesion corresponding to that area on the sonogram. So it's much more useful as an adjunct where you detect the lesion using sonography and then try to evaluate whether a lesion is suspicious for benign or malignant using elastography. One thing that has been found is that considerable training can be required for new users, although a user may be able to create an elastic gram fairly quickly. Learning how to interpret the various nuances, which images are best for interpretation and the various appearances of lesions on elastography can require some amount of time. And a consistent training program has not really been developed, but I expect over the next few years, such training programs will appear and performance will improve at this point in time. I think some users have found breast elastography to be extremely useful and others have not found it to be useful. And I think it's partly due to the ease of use of software and partly due to the amount of training that has been supplied with the instrument. Well, how would you use breast elastography for patient management? The first use is to confirm that a suspected benign lesion is truly benign. This allows you to look at a lesion that you already think is probably benign and move it from say, a birads three category to a birads two category, thus reduce reducing the number of benign nodule biopsies. In my personal experience with our experimental software, we were able to reduce the number of bi benign biopsies by about 15% without missing a significant number of cancers. Finally, it can be used to identify debris containing cysts. As breast ultrasound systems have become higher frequency and higher quality, the old rule of seeing no echoes inside a cyst no longer applies. Most cysts have some internal echoes making it harder to identify cysts than it used to be. Elastography can help with this by demonstrating the characteristic correlation pattern within the cyst that will help you make that diagnosis.

Prostate Elastography

Now moving on to prostate elastography. It was actually the very first proposed application for elastography. And so elastography, when they were both developed around 1990 prostate cancer detection is increased with elastography, with sensitivities reported in the range of 75 to 85%. But quality images are much more difficult to obtain than they are with breast elastography. One of the reasons for this is the tight curvature Of the endo rectal probe, which makes it difficult to apply consistent pressure against the tissue of interest, the prostate gland. Thus, it's very difficult sometimes to interpret an ELA gram because the amount of displacement produced in this direction, for instance, is much larger than the amount that would be produced in this direction. This various methods have been used to produce compression on the prostate gland. This instrument used a balloon for compression. So a balloon was inflated to apply fairly even compression in all directions rather than a handheld in rectal probe. And in this example, you can see the prostate malignancy shown as a stiff lesion on the left hand side of the image and corresponding to this large area on histology. Note, however, that the small cancer on the right hand side is not really seen, and that's points up. One of the deficiencies of the method is that it may be insensitive to small lesions. Here's another example using the Hitachi system, which is one of the first clinical systems that was out there detecting a large prostate cancer, showing it as a stiff lesion on the ELAs gram corresponding to the cancer shown on the histology slide here. So the challenges of prostate elastography are the cancers are often not visible on the sonogram. So the task is now a detection task rather than just a characterization task. And as we have seen, ELA ELAs grams can be a problem 'cause they can create many false positive areas if they're used alone for detection. Compression with the endo cavitary probe is difficult, as I mentioned, because of problems with inconsistent amounts of compression pro vided in each direction because of the tight curvature of the probe face, there's a lot of strain decay because of the tight curvature and lateral motion occurs, which causes de correlation and decreases the quality of the ELAs agram. Finally, we already talked about the small size of many prostate cancer foci will make it very difficult to detect them. However, the clinical sniff of small cancer foci is questioned in much of the literature these days, and this may not be as big an issue as we might think. So near term applications for prostate elastography, the main one is biopsy guidance. If you see a stiff area on an ELAs agram and you're planning to biopsy anyway, then it would make sense to biopsy that stiff area and make sure that you get a few extra cores in that region to confirm that there's not a cancer lurking in that area. With sheer wave imaging using one of the dynamic methods where vibration is applied to the prostate, the issues of compression by the endorectal probe may disappear and it may allow for detection of smaller cancer foci, but this still has yet to be determined.

Thyroid Elastography

So moving on to another application, thyroid elastography, this is a logical one. The gland is readily accessible. A linear probe can be used to apply the compression and the ultrasound and nuclear medicine criteria for diagnosis of cancer are known to be unreliable. More nodules are being detected that require classification and large numbers of biopsies are being performed. And the goal of the ELA gram would be similar to breast cancer, where you would try to use the ELA gram to try to determine which nodules to biopsy and also which ones not to biopsy. So for cancer imaging, thyroid cancers are stiffer than benign lesions just as they are in the breast. No, currently no size difference criteria have been reported, so you're depending alone on the stiffness of the material. The areas under the RRC curve values range between 0.86 to 0.95 for distinguishing cancer from benign lesions. It appears to be useful for reducing benign nodule biopsies. One study reported a reduction of 60% in the number of benign nodule biopsies when Elastography was employed. So that's a pretty promising start. Here's a few examples of thyroid ELAs grams. This is from one of the first papers published on the subject by Dr. Leic and shows a meine lesion on the ELAs, agram and sonogram. Here's the sonogram showing the lesion outlined in the dots. The carotid artery is marked with a C and on the ELAs agram, the lesion is barely perceptible it being essentially the same stiffness as the adjacent thyroid tissue. A thi a thyroid cancer on the other hand, shows up as clearly stiffer than the adjacent thyroid tissue on the ELAs gram, even though it's very difficult to see reliably on the sonogram. This is another method and employs color encoding of the ELAs gram here and a right-sided thyroid nodule located here on the sonogram with the carotid artery. Here, the ELAs gram shows a lesion that has relatively stiff areas with low strain values close to zero within it, and the relatively softer thyroid gland around it. So one should not think that thyroid elastography is necessarily very easy. Some of the problems with thyroid elastography are that the sloping neck contour produces lateral and outer plane motions, which degrade the quality of the ELAs agram. The pulsations of the adjacent common carotid artery can actually be used to create an ELAs agram and have been in some published articles. But if you're using classical elastography where you apply pressure with a probe, the pulsations will degrade the quality of the ELAs. Now, there also have been reported studies showing very lower observer agreement on ELAs grams, but in reviewing all these studies, you have to realize that all ELAs are not created equal. Some algorithms are more effective than others because we're at an early stage of development of the technology. Finally, the effect of increased thyroid stiffness on nodule visibility and strain index is uncertain. Strain index is a method by which all oftentimes thyroid nodules are evaluated. Instead of just evaluating the strain within the nodule directly, you take the ratio of the strain in the nodule to the strain in an adjacent tissue such as thyroid or neck muscle, and use that to help normalize or control for differences in compression caused by different operators. The of course, if you're using thyroid tissue as to calculate your strain index and the thyroid itself is abnormal, you may end up reducing nodule visibility and get abnormal strain values, strain index values that are not reliable for detection of cancer. So this all needs to be evaluated still. This is an example of a study where papillary cancers were evaluated and there was very poor observer agreement. But I think by looking at the ELAs grams, you can, you would probably agree that perhaps the algorithm is not as advanced and that the lesions are relatively hard to identify on these images and you would expect poor observer agreement. Thus, all ELAs grams are not necessarily the same and different manufacturers are progressing at different rates with their image quality. So what about quantification? I already briefly mentioned strain index and I'll mention, I'll cover that again at this point. If you have another static gram, how can you quantify or semi quantify the amount of stiffness there is in a nodule? Rather than just saying, I think it's hard or soft or in between? One way to do it is to put pressure sensors on the ultrasound probe and that allows you to determine how much pressure is being applied with the ultrasound probe, and thus you can compute the actual stiffness of the material. Another way is to do a relative strain computation. In other words, a strain index where you calculate the strain from a lesion by drawing a region of interest in the lesion. Take the mean strain from a nearby normal structure such as fat tissue or the surrounding organ and computer strain ratio. And finally, the third method would be to use a calibrated standoff pad, where you put a standoff pad of known stiffness between the ultrasound probe and the tissue being examined and use that to compute strain ratio. Here's an example of a strain ratio implementation. You can see that on this example, here's a breast lesion and they're computing a strain ratio by taking the region of interest in the region and comparing it to what they believe is more normal tissue out here. And you get a compu computated computed ratio here. Now, the each manufacturer currently is using different tissues for normalization and obviously the tissue selected is up to the operator. So the values are gonna depend on the operator and the machine being used to a considerable extent, but they may be more reproducible in the end than just looking at the image without doing the strain ratio. Well, what about current usefulness of thyroid elastography? As I mentioned, it's currently potential usefulness is to detect of all the candidate nodules. Say if a patient has five to 10 nodules, which ones are you gonna biopsy the largest ones, which is what we do now, which is known to be unreliable, or are you going to use the stiffer nodules, which is what elastography could help you with? In addition, you can evaluate lymph nodes in the neck for potential biopsy at the same time you evaluate the thyroid.

Lymph Node Elastography

This moves us on to lymph node elastography, which is another logical application. Many superficial lymph nodes are available, are accessible to compression and to acoustic radiation force imaging where ultrasound beam is actually used to produce displacement or vibrations. The increased stiffness noted on elastography seen in other cancers is also true for malignant lymph nodes. So nodes involved with cancer are stiffer than benign nodes. This has been shown to have relatively high specificity for tumor involvement with areas, areas under the RC curve of between 0.87 and 0.90 and up to 0.97 for the combination of sonographic features. Andto graphic features. There have even been a few papers published showing endoscopic ultrasound as being useful for evaluation of deep abdominal lymph nodes and mediastinal lymph nodes. Here's a few examples of lymph nodes. Here's an example on the left of a lymph node meta metastatic disease where you see the lymph node and it has a rather plump non benign appearing appearance and on the ELAs shows up as being very stiff consistent with possible malignancy. A reactive node often has a more normal architecture with an echogenic hilos. In this case you could worry about the possibility of an eccentric metastasis in this location. But you can see on this reactive lymph node, which is a benign proliferation of lymphoid tissue caused by an infection or an other inflammatory process, the lymph node is relatively soft, making it far less likely than it's that represents malignancy. Some other organs that can be evaluated include the liver looking at hepatocellular carcinoma and evaluating for degree of liver fibrosis for evaluation of liver fibrosis. Grading of liver fibrosis has been relatively successful using elastography and elasticity imaging methods. Most of these are produced by the non-imaging device called the fibro scan for hepatocellular carcinoma. Most of them are stiffer than normal livers and preliminary work has been promising with areas under the RC curve as high as 0.94. However, more recent results have shown that in some cases there's considerable overlap between hepatocellular carcinoma and other lesions, and that metastatic lesions may be more detectable than HCC in the liver. It's clearly more work needs to be done in this area, in the testicle. There are very few reports out there, but those reports are promising for pancreatic cancer. The initial results using endoscopic ultrasound have been very promising, and there's been one report at least of successful cancer imaging using a transcutaneous approach in the skin. Elastography has been used to detect benign processes such as scleroderma skin abscesses, decubitus ulcers, and it has been even shown to be somewhat promising for melanoma. The one area where elastography has shown no success so far has been in the single report of a salivary gland tumor where elastography was unable to reliably deter, differentiate the lesion from benign.

Non-Oncologic Applications

So talking about malignancies and tumor ablation, most ablation techniques including cryotherapy and radiofrequency ablation cause coagulation of tissue and result in increased stiff tissue stiffness. CT and ultrasound are not able to reliably distinguish ablated tissue from injured or unaffected tissue unless a contrast material is used. In the case of ultrasound, while our f ablation occurs, for example, microbubbles appear in the tissue, but they disappear soon after the treatment making it impossible to determine where the treatment was applied a few minutes after the treatment has been concluded. Elastography, on the other hand, because it image is tissue stiffness rather than back scatter, shows a good correlation between the size of the lesion on the elastography and the ablation zone. This has been shown both for RF ablation and high intensity focused ultrasound studies with other ablation techniques are needed. And there is even one chemotherapy treatment study underway to look at the appearance of chemo treated tumors using elasticity imaging. This is an example of anela gram of a prostate following a creation of a lesion in the prostate using high intensity focused ultrasound. As you can see on the gray scale image, there's a slight increase in echogenicity here of uncertain significance, but on the ELAs agram, the area of treatment shows up as a markedly stiffer area that is very well defined and easily seen. Some other applications of elastography include hepatic cirrhosis, arterial wall and plaque characterization, venous thrombosis, graft rejection, musculoskeletal applications, and the evaluation of lymphedema starting with hepatic fibrosis and cirrhosis except for breast. Most of the work done using elasticity imaging and elasticity evaluation has been done on hepatic fibrosis. This is a result of a non-imaging device called the fibro scan, which can give you a numerical readout of relative stiffness of the liver after just simply holding the device up against the side of the patient for a few seconds using this device. A number of papers have shown areas under the ROC curve between 0.84 and 0.89 for distinguishing grades one through two fibrosis versus grades three and four of fibrosis and cirrhosis. As expected, since it's not an imaging device, the accuracy degrades when there's thick overlying tissue or when there's little space between the ribs, which interfere with the ultrasound beam. Newer imaging systems such as the Siemens S 2000 and the supersonic Imagine machine, which can estimate stiffness of liver quantitatively or semi quantitatively, but also image where the, where the stiffness estimate is being made are showing potential areas into the RRC curve of greater than 0.9 for some very initial work. And further work in this area is underway For arterial plaque and arterial wall stiffness. Detection of vulnerable plaque has been shown using intravascular ultrasound strain imaging. However, because intravascular ultrasound imaging is relatively invasive, transcutaneous methods are under development. Recent work has shown that arterial wall stiffness estimation is reliable for detecting change, increased wall stiffness prior to development of interval thickening and development of plaque, thus helping to assess patients at risk for developing, developing arteriosclerotic disease. This is work that's underway and is very promising. This is an example from several years ago of strain imaging of an atheros plaque showing that there is a relatively soft area in the plaque right in the area of vulnerable plaque seen on the microscope. Um, slide Venous thrombosis imaging using stiffness is a logical outgrowth of vascular imaging. The stiffness of venous thrombi is known to increase with the age of thrombus, so acute thrombi are softer and pose a higher risk of pulmonary embolism than to older thrombi. Some initial work has shown the ability of strain imaging to distinguish acute from more chronic thrombi. And it's known that gray scale imaging can help with this distinction but is not reliable by itself. So perhaps combining the gray scale imaging and elasticity imaging for evaluation of thrombus will be more promising. Here's an example of venous thrombosis with normalized strain being calculated on the lesion of acute thrombus. On the left hand side, you can see the calculated strain is 5.23, whereas in chronic thrombus shown on the right hand image, the normalized strain is only 0.14. Showing stiffer thrombus on the right hand side consistent with chronic thrombosis. This Several types of allografts including kidney and pancreatic transplants have been assessed in early work using stiffness. Imaging. Transplant rejection involves both inflammation and fibrosis, so it's expected that the plant transplant undergoing rejection would increase in stiffness. Liver transplant fibrosis has been known to be detectable using the FibroScan, so it's expected that using the FibroScan one could detect a change in stiffness of, for instance, a superficial organ such as a renal transplant. And in fact, at least one paper has shown a significant difference in stiffness in kidneys with the GFR less than 50 using the fiber scan versus those with the GFR greater than 50. These imaging case reports that was with a non-imaging device, but imaging case reports are just appearing and we'll have to wait and see what appears. But this is a promising area of development. In musculoskeletal ultrasound extensive work has been performed on using magnetic resonance elasticity imaging, and that work has shown that muscle stiffness increases as expected with contraction of the muscle and with tension placed on the muscle. So elasticity estimation may be useful for quantifying the degree of contraction that occurs during therapeutic maneuvers. This is always a big issue, especially in the physical therapy community where they want to be sure that the patient receives feedback that the muscles they're hoping to contract during the therapeutic sessions are actually contracting Preliminary work on tendons, ligaments and hylan cartilage showing stiffness changes in relation to pathology are underway. Finally, lymphedema has been evaluated using elasticity imaging. A special technique called poor elastography is used to evaluate lymphoedema. Poor elastography is a technique that uses the poissons ratio, which is simply the strain in the axial direction divided by the strain in the lateral direction. Normal tissue has a poissons ratio of 0.5, whereas emus tissue shows a poissons ratio that initially may be near 0.5, but declines rapidly over time as fluid leaves the tissue being compressed and the tissue starts to move back towards the center line. The goal is to quantify the amount of fluid versus fibrosis for staging of lymphedema and the rate at which the fluid leaves the compartment and the rate at which the poissant ratio changes may be related to the quantity of fluid present versus the amount of fibrosis. Here's an example of a poor ELA gram of a normal arm in the upper row of images and a lymphedema arm on the lower row. Red represents a poissons ratio of 0.5 and blue represents a poissons ratio of zero. As you can see, the lymphedema arm has a poissons ratio much lower than that of normal tissue. And as time progresses over from four seconds to 10 seconds, the amount of tissue showing lower poissons ratios increases and it's possible to plot a curve showing the rate of increase or the rate of decrease in the poissons ratio, which may be helpful in quantification of the amount of edema present.

Conclusion

So in conclusion, elasticity imaging both qualitative and quantitative, is very promising for distinguishing benign from malignant In most cases of tumors for most for established applications, the ones that are most useful currently I think are breast, thyroid, and lymph node evaluation where you're using a linear transducer in a pretty well controlled environment. And there's considerable data out there showing the criteria that you can use to distinguish benign from malignant. But there's still quite a bit of variability and you can expect initially that your results are going to be variable. It's important that standard protocols be developed for creating the ELAs and the proper training be supplied to new users to avoid people getting discouraged with the new technology. Other applications, including evaluation of liver fibrosis, fibrosis in transplanted organs are quickly emerging and will be hopefully useful in the very near future.