Welcome to nano at tech seminar series here at the Institute for Electronics and nanotechnology. Before I introduce today's speaker, just a word about our next seminar, which will be in a couple of weeks on November 8th. And it will be, the speaker will be Professor Minoan, who's from the Georgia Tech School of Mechanical Engineering. So we have another mechanical engineering speaker today, professor Zhi Hou in the School of Mechanical Engineering here at Georgia Tech. Professors sheet got his bachelor's and master's degrees in mechanical engineering from Shanghai Jiao Tong University. Before getting his PhD at the University of California, Berkeley, where he was also in mechanical engineering and he was also working in the Materials Science Division at Lawrence Berkeley Laboratory. He joined Georgia Tech in 2018, where he is currently an assistant professor in Mechanical Engineering and also Program faculty in bioengineering, the Institute for bioengineering and biosciences and Biomedical Engineering. He's won a number of awards and been recognized by including an NSF Career Award and the Office of Naval Research Young Investigator Program Award. And with that, I will turn it over to Professor Shi. Hope this works. Okay. So everyone can hear me. Good. Thank you, David for having me here. And I'm transitioning from mechanical engineering as David just introduce, and I'm an assistant professor here. So today it's my honor to share with you our research on the ultrasound for brain imaging and therapy. So first of all, when talking about AutoZone is a type of acoustic wave. So the first thing we need to do before we actually get into the research is to define what's acoustic wave. And actually Wikipedia, I'll give you a very nice summary of the definition of acoustics. Even though they are people talking about, don't trust Wikipedia. This definition here is correct. So it is the branch of the physics that they always the study of all mechanical waves in gas, liquid, and solid. In terms of type of acoustic wave, it can be classified into propagation of vibrations, salt, ultrasound, infrasound. So before I get into the details, first, when talking about putting wave, does anyone have some kind of idea? What would you think about as acoustics? Anyone? Yes. Salt. So what kind of song? Music? Yes. Yeah, the sound of the room, the solder there, I actually pronounced from my voice. And then you here, there's one of the songs right here. That's the pressure wave we hear, we sent from our ear. One thing. Any other things you can think about other than the sound wave that we hear? Yes. Jet engine is part of the sound wave that we can heal. Anything else? Yes. Echolocation use the marine animals. There's some kind of like a echoes the animals use, which is part of the infrasound that we will cover later. Anything else? Yes. Yeah. Waves travel in water. That's true. So I'm not sure how many of you here are married. Okay. 12. So do you have kids? Good. For your spouse? Did you actually accompanying her to the OB test in the hospital? What do they do there? Exactly neutral zone to view the infant inside the body of your spouse, right? So that's another type of acoustic wave. The people studying. Anything else you can think of. Of course, there's infrasound the animal use the echolocation to after he tried to find the direction of moving and also like to try to detect their prey, right? So, yes, sonar is another type of acoustic waves. They're forced into the range of our audible. Sound wave, right? And we have some study of the sonar as well in my lab. Today. I'm not going to talk about that. But I'm talking about infrasounds other than echolocation. I'm not sure if you offer many ways something there's going to be very, very dangerous to our humankind's sales TV introduce. I got my degree from Berkeley. And if in a state area for five years was the biggest written off the Bay Area earthquake. Seismic earthquake. That's right. And this is actually the picture of the eye 88 freeway in 1989 when the earthquake hit Oakland in the Bay Area. And it was a tragic death. It was actually happening during the rush hour when people just try to get back home in the evening and this double decker freeway actually crash. A lot of people were trapped. They result in the high casualty during that moment. So that's something else that we tried to actually study. And there are a lot of study, active study trying to see how people can try to read the rock and the propagation of the assessment wave. Trying to reduce the damage to the infrastructure or the buildings to save lives. But in terms of saving lives, can of course a wave do something other than trying to study how to avoid assessment kidney stones. Kidney stones, there's something called lithotripsy that's using the acoustic energy to try to break the kidney stones into small fragments. And then people can actually try doing. The patient can actually try to repel the storm outside the body. There's also something very hard nowadays in terms of research is called a high-intensity focused on. So what they do is they try to focus the sound wave onto the target. And then which resulted in the high-intensity ultrasounds and a focal spot. Then it will heat up the first part and also result in something called cavitation that I will cover later. And then this will ablate the focal spot. And assuming the focus mode is a tumor, for this case, is a liver tumor that's actually putting in place in the clinics nowadays. Actually, you can look for this kind of treatment. If you have a liver tumor, if you have some kind of tumor, pancreas tumor, you don't want to actually have the surgeon cost from your body. You can choose this method. The benefit of using this is non-invasive. And you can see the patient actually back to work the next day of the surgery. While in a traditional surgery, you will see the patient lying in a bed in the hospital for weeks just to recover from the wound. And also moving up to the frequency here, you'll see some missing slot here. So that is something called acoustic microscopy. When we increase the frequency to above gigahertz. And the wavelength is very, very small, such that even though you're still diffraction limited. But people can actually use that to actually see the molecular structures using those very high frequency ultrasonic. So this is the range that covers everything of the acoustics. In my seminar today, I'm going to focus on these two, imaging and therapy. So talking about imaging, we all know that in hospital, if you have kid, you're probably already like a commonly with your spouse or your already go into the OB test and then you know, that ultrasound can help you to see the baby inside your body. But have you ever heard about ultrasound imaging of the brain? And guess? Yes. Yeah. And the neural dust which is using the I guess like what you're talking about. It's the photo acoustics. So the neuron thus we are actually absorbed the optical wave or optical energy. And that results in the emission of an ultrasound and you detect the ultrasound wave to try to reconstruct the image, right? But in that study, what happens? That's a very critical point in all those studies. I just give you one example here. In this Nature Medicine paper, looking at what happened here. They tried to do the ultrasound imaging of serverless car off the mice. But look at this portion here. Just right beneath the ultrasound probe. What did they do? They remove the skull. So this was a funny story there when orbitals, I'll start to actually be in practice. The frontier area or the pioneer of AutoZone actually tried to tell us if they can actually use the ultrasound to see the brand. And what ends up is what they see is just a Very noisy random signal, and that ends up proven to be the scattering of the skull. So let's see what happens if we don't remove the skull. Let's just do look at this wave simulation here. If we just send the ultrasound towards this goal and the brainiest inside, normally in the clinical setting is not a submerged in water, is actually this layer is the ultrasound jail. But yes, the same mechanical property of the waters are we just regular replace it with water just for convenience in our study. So what happens is that tried to see how sound propagates through the skull. And there were some experimental study is trying to study the acoustic parameters of the skull. In the 1980s, you'll sing the ex vivo humans go. So we actually use these parameters from that paper, the density and bulk modulus, sorry, Does the aerosol speed after you can use the bulk modulus, but we tend to like to use the sound speed. So it doesn't matter to mimic the effect of median of the skull here. And then try to do the full simulation to see if we send in a 1.5 mhz ultrasound will happen. If we look at the transmission line. Along this line here, we should expect some transmission and refraction, right? Let's typical when you send a sound wave or optical wave onto some kind of interface like the grass, when you actually see through the glass, you will see some kind of refraction. You'll see some kind of transmission through, right? So we are trying to measure what's the transmission and what's the refraction. See here. The blue curve here is actually the pressure field of the ultrasound wave. And the red curve here is the amplitude of the pressure wave. So what happens here is you can see that you will see a strong interference between the incident wave and the reflection from Moscow and the acoustic transmission through if the scalp is very low here. Okay, so that's typically what happens and that's why currently there's no grant imaging using ultrasound for adults. Infant is a different story because the skull haven't crossed yet. Okay. So yes. Oh, Scholar is a good reflective of some because it's too stiff and also density is much higher. So in terms of the so-called acoustic impedance, is much higher than the soft tissue. So that's the impedance mismatch there results in the strong reflection. Right? So I just talked about the impedance mismatch. So how to solve this problem? They will, people are talking about, okay, maybe we can develop some kind of something called a complimentary metal material trying to match the impedance. So this is what their desires. So what happens is they try to actually design a metamaterials such that the density and so speed, or the negative value of the density and sound speed of the skull. Then you will result in something with negative density and negatives on speed. That sounds weird. But people can actually achieve them using something called a monopolar and dipolar reference in acoustics to achieve effective negative density n, sound speed is not a static, is a dynamic parameter. So what it means is for ultrasound, what it sees is a negative density. But if you actually measure it without the ultrasound, is still give you a positive density. Okay? So if you're interested in, I can actually give another lecture. Then negative density stuff, how you can actually achieve that. But today we'll just focus on how to actually use their negative density and memorials to design this complimentary metal material. So what happens is they try to use this density to be negative. And it's asked me to the negative of the scalp parameters. Then that's just see what happens here. If we send in a SEM ultrasound towards this bi-layer and then look at the same transmission line. This is what happened. It will still see a strong interference between the instance and refraction. Transmission is still very low. And also talking about be more imaging of ultrasound, especially like ultrasound, imaging is relying on the scattering off the scatter to actually reconstruct the image. So we should not only focus on the transmission through this direction, but also the backward direction. So what happens in a backward look at the backward eruption, still strong reflection, low transmission. So what's wrong here? What's wrong here? Something is wrong here because they ignore this imaginary part of the density and solved speed. So what does it mean by the measuring part of the density and Salisbury in acoustics. Actually what happens here is the imaginary part of this two parameters corresponds to the energy loss of the ultrasound propagating through the layer. Because skull is a porous medium. So it's like a salt absorbing sponge. Now you put a run the speaker to actually absorb the sound wave from those boundary. I Lego a porous, same thing for ultrasound. It will absorb the sound wave. One results in is it will heat up and skull. The energy is just used to heat up the skull without transmission. And also the problem here is what we are talking about, the impedance mismatch. The immediacy is actually given by the product of the density and the sound speed. And what you'll see here is because of the imaginary part will contribute to the real part of the impedance. If you ignore this imaginary part, of course, you won't be able to get a perfect matching of the impedance. So that's why we wouldn't have had an interesting if we consider the imaginary part. Let me see what happened. We include this immigrant partly into the study and then just send an ultrasound through the bi-layer of this so-called non Hermitian complimentary metal material. And look at the transmission line here. Now you'll start to see the total transmission. Same thing for the backward direction. It will start to see the total transmission. So now we have this total transmission through the scar layer. Can we do the PMO imaging? Let's try to test their originally tried to actually see if we can use the B-mode image to see this sphere, which we assume is a brand human in the model. And we send in a plane wave e to use the prime with ultrasound imaging to construct the VMO imaging. And then we look at this because the field in this region, because that's the region where the ultrasound probe can actually scan through. And then that's where you can actually read the information from the ultrasound. Also, we slept one line that's typically when you have a stationary ultrasound probe. That's why you will read from this RF signal and now reconstruct the image. So let's just look at the scattering of this line. Here. You'll see the shadow of the scattering. That's the shadow that you use to reconstruct the B-mode imaging in the clinical ultrasound. Now with the scar, you will see here seven view or field. And the seven lines of reconstruction, while you will see, is just some kind of noise as expected, because there's this direct scattering from the skull. And this image is not a brain tumor but after this call. And then with the previous design of the complimentary manner material, what happens is it's traveling improves. It will start to see some kind of shattering effect. But the signal to noise ratio is too small. The Johari get the way to reconstruct a B-mode image of the brain tumor, shift the conversation of these complimentary metamaterials. So instead of using the land commission complimentary metal material, what we get is you will still see now, you'll start to see this shattering effect, which is similar to the case without the skull. That's what you will use to reconstruct a B-mode image of the autosome. So some of you may argue the most car is not fluid. Heel, if you, if you pay attention to the slides here, we show here is we have the pressure field inside the skull. What does it mean by having pressure field inside of the skull? Will only measure the pressure field in fluid. The solid, you will measure the stress field or strain field, right? So someone may argue that you are trying to simplify the model to be too simplified than the scar becomes fluid. But we all know that's called solid. So we will hand the model, the scar with solid using the business model and then fill in the porosity and permeability of the skull that we measured. And then try to do this study again. And now you'll see total transmission, even with the solid model. For the bad word disruption. It's a little bit complicated. You see a near total transmission. But also you'll see this kind of a beating effect of the amplitude. And we hypothesise that this building is that is because of the mole conversion between a longitudinal wave and share with Ian silo solid model. But we hope this bleeding event on the noise is not too high, such that we can still reconstruct the B-mode image of the brain tumor through if the scalp. So let's try to see what happened if we use the sorting model here. So when we use this model for the pills go, of course you just see the scattering of the skull, right? So the image you reconstruct, it's basically just the scope for the complimentary mixing complimentary metal material. What happens is even with the solid model, you'll still see the shadowing. In fact, the noise. You can see the noise being more severe, but the signal to noise ratio is still high enough such that we can still reconstruct a B-mode image of the brand human. The question lies here. This is all theory. To make it a practice to omega, to be practical, we need to do any experiment. And now we need to achieve those legged negative density, native sound speed, and also the narrative parameters of the management part. Remember the neck, remember the management part of those parameters are the lost a part of the skull. So when we flip that to be negative, it means we need some acoustic gain. In a nature, there's no acoustic and it's not like objects. You have non-linear material to actually achieve those kind of like optical gain in acoustics. We just couldn't find it in Malaysia. Instead of trying to get help on nature. We tried to look for help from electrical engineers. So let's try to actually use this type of elements that weighs two pieces of piezoelectric material. One being used for sensing and the other being used for the emission of the ultrasound. And then we connect them with a feedback control circuit. That feedback control circuit can actually control both the phase and amplitude of the feedback. So this was one of the paper published in the earliest days of this active acoustic metamaterials. What happens is the total reflection and transmission will be a result of the sum of the passive refraction positive transformation and also the active component of the refraction and transmission. And those active components either controlled by this given parameter of the feedback circuit. So we can actually play around with the game parameter to modulate this active component of the reflection and transmission to try to achieve that your family parameters we want. For the ultrasound. This is our actual design. For the metal material. We have to use this three-layer structure. So the purple layer here is the hydrogel. And the reason we use hydrogel is because it's mechanical property is very close to water. So we have a very good coupling between water and hydrogen. And then we embed those piezoelectric material inside, one being a transducer, one being the sensor, and connect them with this feedback control circuit. And the reason we put an L gap here, yes, because in this case, we isolate the ultrasound from getting through. So the passive component of the transmission is zero, which simplified a little bit in a design. For the circuit control. What happens is we will have to face component using this all pass filter. And also we will have this amplifying circuit here. So what happens is we can actually modulate the resistance of this R1 to control the face and modulate the R4 such that we can have power input form this op M to drive this unit cell such that we can have energy input to achieve again we want and we just met, as I just mentioned, this. Total refraction and transmission is actually a result of the sum of the passive part and active part. We know the total. The positive part of the transmission is zero, but we still need to go ahead and measure the refraction part, right? And then we can actually, we also calculate through the feedback circuit, what's the active component of the refraction and transmission? By adjusting the resistors R1 and awful, we can adjust the given value G1 and G2 to try to achieve the active component of the refraction and transmission such that we can actually achieve the effective parameter for the ultrasound that we want. So this is one of my student actually trying to adjusting the R1 and offal component of the circuit and then try to see if we can actually achieve the parameter we want. Those density, this cava is bulk modulus are all normalized to the value of the skull. Okay? So you can see here the real part 1.5 mhz test our target frequency, or g of negative one. So it's actually minus of the density of the skull. Same thing for the bulk modulus. Same thing for the imagery part, which means there is an active again. And the measure of hot Florida parsimonious as well. And then my student went ahead to actually fabricate this guy. So this is the sample we have. You can see the piers of materials inside. One serving as the transducer at the outer surface is the sensor loop, loop through this leg of Wilders feedback control circuit. And then this is the embedded in a hydrogel. And you can see in the middle there's an air gap there. Isolate ultrasound to actually transmits roof. And my student needs after testing the fact that parameter of this metal material in the lab and the water now. And also we have this ex vivo humans color here. So we can naturally try to do the sentencing by measuring the transmission and reflection throughout the scope to retrieve the effective parameter of the skull such that the West, we know that effective hammer into of the skull. We can then turn the control circuit to achieve a negative value of those and then the mantle material will be put in effective, right? And this just shows you the underwater measurement here. You can see with other skull, you'll do your spec high transmission because there's nothing blocking the sound wave was the skull. You will see a strong chance. You'll see a strong reflection from the skull. And the transmission is just very low, almost noisy, almost nothing here. So where's this parameters measured? We can actually try to retrieve the faculty family that, that's what it's being done in that paper that was in the eighties for the ex vivo human score. And we can actually try to put this parameter into effective to try to see if we can achieve a total transmission through the skull. And then the next step will be testing this for PMO imaging. That's the story of the imaging. Next story I'm going to share with you is the cerebellum. And particularly, we're interested in two types of disease. One is called cerebral venous sinus thrombosis. What stays? Sometimes people just showed, therefore CBSD. This may sound very strange name. But if I say one word, you probably are all familiar with it. This is a type of stroke. Okay. So normally when we refer to stroke, what it means is the blood clot formation in the artery in the brain. And you can also have this type of stroke, which is the broccoli crown formation inside this venous sinus. But it's a real case. I mean, it's a real case before the COVID. What happens is for this CBSD, you will have this black rock formation inside this venous sinus. And the reason for this blood clot formation is because some kind of mechanism that effect the platelet in the blood releasing those five marine structure, the harvest those Brazil and then block the venous sinus here. Normally you would just see in the artery because the diameter of the artery is usually two millimeter. And talking about this venous sinus, this is the anatomic structure of the human brain. Waves. The venous sinus, the diameter is usually once in the middle. So we're talking about five times difference. Okay? So for the treatment of the stroke in artery, nowadays we have a lot of already very developed that technology is to treat them. But those technology cannot be used for this kind of treatment because we are talking about five times difference of the diameter. And also why this is very important is because the frequency of this disease is now much higher because of the COVID, because of the virus that actually will stimulate the platelet is that the sinus and then release those fiber in structure result in Blackhawk inside assignments. And you haven't mostly for young people. For the traditional stroke. It happens mostly for old people. But for this, it happens in the young people. That's why you see those kind of like a side effect of the COVID vaccines. There are people talking about like they have had that or sometimes joke. That's what happened. Okay. So that's no effective treatment for this. Was the increase in frequency and is life threatening? Why is life threatening? Because once you have the blockage of the signers, what happens is the pressure of the blood. Well, the brow is trying to enter flow back to the heart. But once it has been blocked, the pressure is building up. And then you would heal upon the vessel wall, result in hemorrhage which is going to be fatal. That's why this is a deadly disease. The other this is we're trying to focus on is called a pulmonary embolism. So what happens is normally you will have plaque formation in those femoral veins, which is inside a lag. And then those crowd dislodge from those veins. And then just clumped together forming a big crowd going through the inferior vena cava, going through if this right ventricle and then getting into the pulmonary arteries and then embolize over there, which is why it's called a pulmonary embolism. And what happens when you have this kind of disease? It will stop the blood circulation of your lung and you will stop the gas exchange of the lung. So you won't be able to brace if you have this disease. And also what was a pressure will build up because this is artery. So the pressure inside the artery is much higher. So usually what happens is you will very, very rapidly build up the pressure inside the right ventricle result in failure of the right ventricle. And this is the third most common cardiovascular disease in Award. And about 400,000 patients die every year, just here in the US because of this disease. Because of the size of the Claude Steele, that's no effective treatment. So what should we do? Someone tried to invent them, this Castile rise the device trying to pull this card out. So what happens is this is a 24 French catheter. If you're not familiar with, was 24 French is actually eight millimeter diameter. And then when we talk about these vessels here, the Palmer arteries, the diameter is usually 2 cm. The reason people use 24 French Cassidy is because that's the largest cluster that's proven to be safe to be used in the cardiovascular structures. So what happens is once this cast of gas into the palmar artery, they will just deformed into this shape. I'm trying to launch on the Cloud and then pull it back out. Is there any problem with this treatment? Yes. Exactly. That's very dangerous because you have to enter drug. They are there all the way, right? But also the proton is, those structures are in contact with the vessel wall. So it's very easy, if you are Kelly's is very easy to tear apart those vessel wall and then resulted in fatal hemorrhage. So what happens is this is not a good way to treat it. Nowadays. I had only include a video that show by my collaborator in Grady Hospital. In Grady, what they do is they still open your chest, pull the crop off. It's very bloody and high-risk. So is there a way to have better treatment? Nowadays, people are trying to see if ultrasound can help. And there's something called a sauce and releases. The principle is based on the acoustic cavitation. As I just mentioned, there was something called acoustic cavitation and the focus ball of the hive as we focus on ourselves. So what's acoustic cavitation? You may not familiar with this acoustic cavitation, but when talking about boiling water, I guess that everyone has the experience. Right? When you boil the water, what happens? You'll see the bubbles. So what happens is you usually like the increase of temperature into this vapor phase and then you'll start to see those bubble formation. But remember, ultrasound is a pressure wave. So there's a way we can actually have the intensity of the ultrasound to be high enough such that the pressure drops under this phase transition points such that we can still get into the vapor phase without increasing the temperature. Instead of going this direction, we're going in this direction. What happens when you have bubble formation? This is usually what happens. Usually happens near the interface of solid and liquid. When you try to boil the water, you can see the bubbles forming at the bottom of the bottle, right? So this is usually will happen. And then once the bubble forms near the interface, what happens is because of the asymmetry of the pressure. It will deform this bubble. No entry, you will form something called a micro Jadi because of the asymmetry of the pressure. Once you have a jetting onto a solid surface will happen. It's like you have a waterjet. If you are familiar with like the machine Hall in mechanical engineering, you will see like if people try to use the water jet to edge through the manuals. And this is what happened. This is something that the Navy first discovered. They see after a long use of the propeller for the summer ring, there are a lot of this damages at the beginning and they didn't understand what's happening. But then they tried to study, and they found that because of a propeller actually in the region created some kind of negative pressure result in the cavitation. And the agile way, this metal surface. Of the propeller. There's something that navy was trying to avoid it. That's why they have now new designs, especially on the Virginia crossed some marine. You don't see this program again. But for biomedical engineering, we're not trying to avoid is we are trying to use it. Okay? So what happens is people tried to send the AutoZone. So if the cardiovascular structural see if the cavitation will etch how the Blackboard. Yes. So we are trying to use the cavitation to get rid of the clot. But look at this, even though it's a non-invasive method, you need to go through if the ribcage and going through a lot of tissues, it's not very efficient. That's why people invented is something called an endovascular ultrasound. Trying to put the transducer on top of the tip of the catheter and then going through the vessels to reach the blood clot and then directly shine ultrasound on top of the backlog. And there was a commercialize the clinical device called echoes, which is now the gold standard of the cardiovascular ultrasound. What happens for echos is give you some statistics for echoes. Usually in this 15 h to Azure, wait a whole cloth for lymphoma. Emily Lynn, talking about just using the normal drug like TPA, usually cost 20 h. So you'll pretty much safe only 5 h. And we're talking about patients lying on the surgical table die in 15, 0 is not going to save the life. So we need something. There can be much faster, efficient, and safe to etch away the clot in a very short term. So let's see what's the program for this intravascular ultrasound. What's the problem of this endovascular ultrasound is because remember here we just send a sound wave. And what's the sound wave? The sound wave is a longitudinal wave. So what happens is once you send a longitudinal wave onto the black cloth, you interact with the blood clot, result in some kind of normal stress into the blood clot. Remember those fibrin structure of the blood clot. They actually resistant to the normal stress. One thing they don't resist them against is the shear stress. So the question is, how can we use the ultrasound to induce shear stress inside of BlackRock? Where's the endovascular autosome? So the answer lies in something called a vertex AutoZone that we start to use. So what happens here? What's the vertex ultrasound? Ultrasound is some type of ultrasounds that has this helical boyfriend when propagates. And what happens is when it actually propagates in the fluid, it will induce this shear flow instead of fluid. So we did a simulation for the blood clot for the traditional endovascular ultrasound, you just see the notice how it's pushing the clawed back and forth without doing anything before the voters order. So you can see the shear off the blood clot inside. And we actually measure the shear stress instead of crop is after four for higher than the shear stress, you will see where's the normal and the vascular ultrasound. And we went ahead and see how are we going to make it okay to make this relatives endovascular ultrasound? First, we need to integrate it into the Canada. And then we need to actually create this Critical wafer and how to create acetyl-CoA friend. We tried to actually have a mechanical phase delay of the four adjacent transducers such that we have this proper phase delay. As you can see here, we have the proper phase delay to generate this helical wave front. And for comparison, we also like made the traditional ultrasound transducer for comparison. And then we measure the AutoSum field using hydrophones. And you can see this spiral pattern or the vertex pattern of the Wakefield for the vertex ultrasound, while for the traditional and the vascular ultrasound, you just see this Gaussian painter. And we tried to see if we can actually induce the shear flow using our categorize, the world has auto saw. When the power is off, we actually put the aluminum powder into the liquid and see nothing just like a random Brownian motion of the Ottoman policies. When apologize on, we see this shear flow. Then. Went ahead and test if this is more efficient than the traditional AutoZone. So we put the, this test in vitro on for scale trying to make sure these two treatment methods are comparable. The reason we focus on the push through force is because we don't want to have a too high pushing force such that it may tear apart the vessel wall. So our standard of comparison is when those pushing force are similar. And we try to actually measure over 30 min of the treatment for LAM vortex and voltage drops out. And you can see with this 30 min, you already edge through. You already recognized the BlackRock and restore the blood flow through the vessel for the traditional non-voters social. So what happens is you couldn't even reach halfway through what happened. And we did a comparison of the engine speed and also the reduction of the mass of the clock. We found in the vertex AutoZone has about 65 per cent increase in speed and also the reduction of mass. Also remember the voters, non-voters ultrasound transducer here we use is the state of the art which hasn't been used in the clinic yet. The 15th, our ageing of the clinical setup is the echoes, which is much less efficient than this number that has autism. So this is already the state of the art which hasn't been used yet. And we also tried to increase the duty cycle. And we can say like, of course, when you do increase the duty cycle, you increase the power input of the ultrasound and then the actual speed is going to be higher for sure, right? And then remember that the solar system lysis, one of the basic principle is based on the cavitation. So the question is with the vertex and ultrasound and the non-voters options on, well, this vote is almost on a tree, make it less effective in terms of cavitation there will after sacrifice part of the engine speed. So we went ahead and then test the cavitation effect by using our hydrophones to measure the scaffold signal. And then we try to test two parameters. One is the stable cavitation and the other is the inner cavitation. So what's stable cavitation? Stable cavitation is just the vaporization I mentioned. So you will have bubbles and then the bubble are non-linear to ultrasound, so you induce this higher harmonic signals. And then for non stable or inner cavitation, what happens is those pebble, remember it deforms and that explores those erosion of the bubble will result in a bravo signal. So we didn't actually use the broadband noise floor racing and harmonic signal to distinguish the stables cavitation and the inner cavitation. And then we measure if for both the vertex and that vertex option. So what's interesting is for boasts in luck, reputation and stable cavitation, we see an increased dosage unit cavitation comparing with the number of attacks one. Earlier we expand that should be lower because part of the energy was used for the shear stress, interestingly, sector higher. Then we also went ahead and then testing the input power, input voltage duty cycle, and also the pulse repetition rate. We choose to actually use the 80 volt for the input. Because even though we can still increase the speed and also the mass reduction, but at a higher voltage, it will result in damage of the transducer because of the heating. So that's why we chose 80 volt and 7.5 per cent as the ultrasound parameter. And also here you can see some pigs of the licensed speed for different for different pulse repetition rate. One is at the 10 hz, the other end at 10 khz. But and tell her, we just observed some kind of damage because of heating. And that's why we chose 10 khz as the pulse repetition. And also remember, we need to also make sure the push through force is small so that we don't tear apart the blood vessel, right? So this is actually related to the feeding speed. Of course, if you have a higher feeding speed and then you will ask through faster. But what happens is the mass reduction rate is going to be smaller. The reason is once you move too fast, the AutoZone is just touching the center part of the crowd, result in a very narrow channel. So you'll just like breaking apart only the very tiny part of the crop and then just remaining those Lego circumferential part of the record there. So because of those reasons, we choose to use the 3.33 millimeter per minute video speed. And then we went ahead for the in vitro test. First. Safety is very important. So we need to make sure this ultrasound is not going to damage the Vessel wall. And that's why we use the ex vivo bone by injection of n. So for those who are not in bioengineering common meals, like a call, that trigonal plane is just here. So we use the bone mandibular vein, which has a similar diameter of the pulmonary artery that we have in our human cardiovascular structure. And you can see there was the H&E staining. You won't be able to see much difference between vertex options on that vertex and the control which we didn't do anything to it. Which means that ultrasound is not matching the blood vessel. The other important safety issue is we are blocking the clock. So we're blocking, we are actually breaking the Cloud. So we're breaking it into small segments are small fragments. Those more fragments will move. If it's palmar artery, it will move into the digital offerings. And we don't want those crops who actually embolize in the distal arteries to cause programs in the distal region. So we need to make sure that most of the fragments are smarter or like a around 100 microns. And that's what we see for this vertex. So we actually have the fragment to be smarter or around 100 micro, mostly around ten to 20 micron size. So this is safe for the distal region as well. Another safety check is the hemoglobin because this is the endovascular treatment. For endovascular treatment, one of the typical program is usually not just one, not just attaching the BlackRock. But what happens is you will also break apart the red cells. And even for those red cells, they are not like a cluster then from the crowd. So what happens is instead of like a instead of treating the blackboard, you also create a problem of killing the resamples. So we need to actually make sure that hemoglobin or after this treatment is low enough. So we went ahead and do the testing. Now you can see after the sodium ion and of treatment that hemoglobin is about 20, is actually below 25 milligram per deal. In clinics, the haemolysis, which is like a yield start to kill the cells. And hemoglobin is about 40. So we are definitely safe in this criteria. And comparing with the existing endovascular treatments. All those hemoglobin measurements after treatment is way above their criteria. So our treatment method is actually much better than all those existing methods. Okay? So with this, we actually went ahead and then do this in vitro testing with the in vitro 3D model of the cerebral venous sinus crop. You can see how big it is. So this is about a centimeter and then this is about 20 centimeter long clot blocking this super circuit or sinus here. And then we tried to put the vertex order in which the pastor, and then try to actually push through of ways the 3.3 millimeter per minute in speed, trying to see if we can analyze this crowd in a faster manner. And you can see that the video speed is actually eight times faster than I can tell you. This video is only 1 min, which means the edge through the Cloud ways in 8 min, comparing with the one that's already in practice, the echos is 15 h. This is much faster. So let's just wait for the final moment when you'll see the restoration of the blood flow. In. Now you can see the blood flow and we have to add pressure here to simulate the actual circumstance in the brand. Because once you have blockage of the broadcloth, you will have pressure buildup. So we have to have the pressure here. So that's why once we recognize the blood clot, you can see the pressure is just pushing the resales art. And as a conclusion, here I just introduce you some of our research on the medical imaging and therapy using the non-homogeneous accompaniment or metal material for transcranial ultrasound imaging, for the brand imaging, especially if operating or imaging. And we also use the vertex authors are trying to develop a fast and safe treatment method for those disease like cerebral venous sinus thrombosis and also pulmonary embolism that we don't have an effective treatment carpentry sent you and I would like to also send my support from NSF and over now for the support of my research. We have time for a couple of questions. If anybody has got one. What's the distance? What's the distance that the end of the vortex into school or sorry, vortex that ultrasound has to be from the clot to start ablating. It is about 1 mm away. And also just this might be a dumb question. Is there a reason that you need to use vortex or can you just have like, you know, instead of doing the vortex, Could you just have the different transducers like out of phase with each other, like to. Instead. The reason we use vertex is because it's the most efficient way in inducing the shear stress. And it's most often more efficient than just having like alternating, right? Okay, Got you. Thank you. Also, I want to add on top there the traditional intravascular ultrasound is also 1 mm wave on the crime when they started. I would just want one wondering about your pressure fields. 2d graph, graph, graphical ones. How do you map your ultrasound pressure fields like that in 2D. Okay, so what happens is my student loan would actually have those hydrophobic being fixed on a 3D positional. And then once we measure one point and move to the next one, once we measure the AutoSum field and move to the next one, and then maybe odd. Does that answer your question? So what happens is you scan through. So there's no way you can directly measure the length, but you can measure each point. And then just reconstruct measurements. Because we use hydrophones. So the readout is pressure for the vortex shearing. Is that something that requires special instrumentation or is that a straight forward transducer to implement? There are two ways to actually achieve this vertex. The way we show here is using the mechanical phase delay. But more traditional way of doing AS actually you can actually have the socket controller to achieve those face delays roof the circuit delay. And the reason we use this mechanical, just a joke, is because we have mechanical engineering. That's a joke. And the actual reason is we don't have an electrical engineer for this project. So that's why we'll use the mechanical delay. Yeah, I guess mine is a related question. So for the mechanical delay transducer that you showed, is it only capable of doing, let's say, a 11 chi reality, so right-handed only. Yes, that's right. And we hope that we've had naturally have some electrical engineer to actually be on our team wants we get an Azure support. We are still applying for it. Now we didn't actually. One. Another approach is to actually have this alternating high rarity such that it will be more efficient in adjuncts roof. And we expect that to be even faster. So instead of 8 min, we could actually maybe just 4 min and it's done. That's why in the bath. One last question. Anybody that's why in the bath. Sorry. I just had a question about the safety. So basically it seems like it would be relatively safe method to use. I just had a question about using ultrasound which they believe in the beginning you mentioned that I guess if there's enough for surrounded, it can actually ablate certain things. And so I was wondering, it has ever been an issue with using the ultrasound in the blood vessels, is it like is the amount of ultrasound that sees is lower than what would be needed to ablate, let's say e.g. but a black vest like that. That's a very good question. Actually, this FTA guidance on that, that there is something called mechanical index, which related to the power input of the ultrasound. We can know that actually use a very high power input of the AutoZone to cause damage to the surrounding tissue. And we're following the FDA guidance for the development of this treatment here. Alright, let's think first sheet one more time. Thank you.