Hello, everyone and thank you for joining today's Sun Nuclear webinar Recommendations for Quality Assurance of a 1.5 TMRGRT System. Before we begin today's presentation, Please note that all attendees are muted. We do encourage your questions and you can enter those at any time in the questions box on the left side of the screen and we will address them at the conclusion of today's presentation. A recording of today's presentation will be sent out after the broadcast to everyone who has registered. Today's presentation will be delivered by Victor Yakovenko, an assistant professor and associate director of the treatment delivery of Treatment Delivery in the Department of Radiation Oncology's Division of Medical Physics and Engineering at UT Southwestern Medical Center. Doctor Jacobenko received his bachelor's in physics and his master's in nuclear physics and particle physics and particle physics at Terra Sherenko National University in Kiev, Ukraine. He subsequently received his PhD in particle physics at Paris Said University in or say, France. Doctor Jacobenko then completed his postdoctoral fellowship at the Princess Margaret Cancer Center within the University Health Network in Toronto, Canada. And Prior to joining UT Southwestern, he completed a radiation oncology physics residency within the Department of Radiation Oncology at the University of Toronto in Canada, where and worked as a medical physics physicist, sorry at Sunnybrook Health Sciences Center. His research interests include Mr. guided adaptive radiation therapy, development of novel detector systems for radiotherapy and dosimetry for emerging treatment modalities and stereotactic radiotherapy. And now, without further ado, Doctor Yakubenko, the floor is yours. Thank you, Andrew for introduction and hello everyone and thank you for opportunity to share some of the experience that has been gained over the years. So I'll be talking about QA in 1.5 Tesla emergency systems. So I don't have any disclosures to disclose. So this is the outline of the talk. So we're talking about some motivation why we why we do a majority and what are the challenges and physics behind it because that will drive the selection of the QAI would I would say right up from the MRQA probably will not be that different from regular immersion, but the Leanak part will be quite different from the conventional 1. So then we'll talk about integrated systems and then some QA considerations as well as some view on the quality assurance program as a process and then some summary outlooks and short conclusions. So let's go into a motivation part. So the first thing is that MRI is providing superior soft tissue contrast. So this is one of the main motivations to do a majority. This will also provide opportunity for real time MRI based tracking and gating. Recently this has been deployed for the Unity system. There's also a line adaptation of relation treatment and that's and that's a very convoluted term. So online adaptation right now is happening on the anatomy alone. But there is Mr. provides a lot of opportunities to go deeper, much deeper into what is meant by adaptation. So we can talk about Mr. biologic marker, something that can help us understand how tumor responds and react on that and adapt treatment based on that. So that's a Holy Grail and this is where we're going. Also, all this is coming with no imaging dose. So you can image your patient quite, quite a lot. It's just that like some, some heating of the patient that can happen. So we need to be mindful of that. And also we're adapting to anatomy. So we don't need any fiducial markers. If you're doing like GI or Gu like you don't need that. You're adapting just to make your anatomy. Obviously it comes with some challenges and the first one of them is Mr. Safety. So Mr. Safety means this week will be following considerations for like typical like ACR recommendations from as Mr. SIM. So there will be 4 zones and some patients will not be able to to use this because of the considerations for Mr. Safety. There's also magnetic field introduces some requirements on the quality assurance devices that will be used in the bunker and the immobilization devices to keep patient in place. So all this need to be a more compatible especially if you would like to scan those. Also there is immobilization that is kind of like the clearance is about 70 CM, but also it's kind of constrained even more because there is Mr. coil. So you cannot put whole lot in the bore even though it's you know quite wide relative to the regular Mr. SIM or Mr. diagnostic diagnostic Mr. there's also inherent issues with dramatic distortion. So this is something that all Mr. have also in the in the workflow Mr. does not provide any electron density. So we need to bind that up with CT SIM or perhaps some synthetic CT in the more only planning workflows. Also the workflow is pretty complex and this also demands a lot of resources and resources like human wise. So they will there's a therapist that need to be there. You need the MRO physicist, physicist that will be covering the console physicians need to be available. So all this is quite, quite demanding and the the thing that they would like to focus more about is how magnetic field influences the the dose distribution. So and let me show this. So when the electrons are moving in magnetic field, there will be a force that will be impacting them which is Lawrence force. And you can see this is a cross product. So how this can be visualized is that when after us are traversing in the magnetic field, they will be curving and this curvature will be different depending on the different strength of the magnetic field and the energy of electrons. So especially with when electrons are going from like a lower density like from lung for example to an increased density of the material, they can loop back and they can deliver more dose. So this was this is the work of the one of the inventors of the 1.5 Tesla emergency system when they did the Monte Carlo study. And you can see like point spread kernels of a secondary electrons and they so this is for the low energy. But you can see as the we increase the strength of magnetic field, they start, I'm going a little to the to the left. And this is the direction of the Lawrence force itself. Another thing is that so this is the simulation of the different slabs. So this is density of one. This is density of point to five, which is like lung and and another one of the density of water. And you can see as they penetrate. So there is some curving that is happening. And I can maybe write this up. So it will be like this and they will come back and they will deliver those again and something similar on the exit. So and come back and then we'll deliver those here. So that is something that we need to, you know, keep in mind. And this is something that will help us select how we do the QA in the in the system. Another thing, what's what, what else is happening to the dose distribution? So electrons are deflected professionally towards the average direction of the Lawrence force. And you can see there's no Lawrence force here and there's Lawrence force that is moving everything to the right in the direction of the Lawrence force. So the entire dose distribution shifts laterally and then the maximum of the dose is not anymore and the central axis. So this introduces the the asymmetry of the profiles in the lateral direction, in the direction where the Lawrence force is as well is that there is this. This is especially pronounced to the small fields. You can imagine if you have two by two field. So this can shift so large is that it actually. So imagine this is central axis like the gradient can be directed on the central axis. And this obviously can be a big issue, a big challenge for proper DOS imagery. So also there is an increase and a symmetric electrical penumbra as well because electrons are being deflected in this side, but there's also divergency. So this penumbra will be little sharper than this one. And all this is because of the Lawrence force. So another perception on this is when electrons so electron turn effect when they return back and they deliver the dose as more as more significant on the heterogeneities. So this is this is water. So this is water and then goes into the lung and it comes back and delivers the dose and you can see the speak there. So all this exactly is due to the Lawrence force again. So there is a bunch of physicists and engineers and they came together and they decided to do something about that because that was something to the thing to be taken into account. So Monte Carlo treatment planning system to account for this effect. So that is relatively logical solution. If you would like to have an integrated system, you need to somehow make sure that the ARAF noise from Leanak will not interfere with the imaging NMR. And then you need to somehow create a low, low magnetic field where the Leanak is so that magnetic field will not interact with the electrons in the waveguide. So this is where the new metal comes in, high permeability metal that basically creates a local low field of low magnetic field where the LYAC is placed. There's also a solution, how about we use just cobalt and then we don't need to any electrons in the waveguide. So all this was in works at the dawn of the in majority. There was some engineering things that need to be solved and then there was an alternative solution. How about we be simple and be the first one to create a majority facility and we'll have a linear accelerator and Mr. scanner separate. And it actually worked all almost worked pretty well for Linux part. It does work for for a bracket part. So this is the Emerus Amar on rail system. This is Amar in this room in the central room, and it's on rails. And you can travel to the bracket suite or it can travel to the Linux suite. And so this was built in Princess Margaret Cancer Center. And right now there is a bracket under image Mr. image guidance is done in the bracket suite. So you can, you can appreciate right away that there is challenges to it. So it's not fully integrated system. So there is times when MRR doesn't see what's happening to the patient, especially when it goes back for treatment on the lean neck. And then you need to have a couple of settings, well specifically 3 settings for shimming because magnets can move into three different directions. So that introduces a very, you know, complexity to the workflows on the on the QA side, it makes QA simpler because you will need to be just having separate QA for Leanag and then for MRI. So how these were overcome. So again, this is different picture when magnet is in the room and then there's a special ARAF, ARAF doors is this one. So they would close this and they would park the Leanag behind these RAF doors and then patient will be imaged here and then the leaner will go back into the room. There's another door that is closing then part over here, they would open it and they would deliver the treatment for the patient. So it's, it looks pretty complex and it was complex, but then there was advancements done in how the system can be integrated. And right now we have this integrated systems. I should make a remark that now you're not able to use the .35 Tesla emergency system. If you have one, you know that you may be able to service it but not use it. So there's Electa Unity system and there's Aurora RT. So this is the slightly different design because it uses the parallel of beam and magnetic field. So they're partly in that direction, which minimizes the Lawrence force. But that is .5 Tesla. We'll be focusing on one where we're focusing on 1.5 Tesla machine, which is elect Unity and model Linux system. And this is like a schematic of the layout. So you have a split magnet here, so MRI from Philips. And then there is the bore, the the gantry that is outside where you have a linear accelerator with different parts of it. And I will have to say that the linac itself is relatively straightforward like so there's no bended magnet, there's no steering magnets. The complexity of the whole system is comes from the fact that it's integrated with MRI. So let's look at the parameters of the of the beam. So the beam is FFF. This is the strength 1.5 SCD is pretty long, which means that we're delivering those slower than on the conventional FFF. Moreover, when the beam is penetrating to cryostat here on the way it hardens a little bit. So the shape of the beam is slightly different. What you would, what you would think is FFF, it's a little hardened. So in terms of field size, so lateral is 57 and 22 is a soup in. So if there's like very long fields that you would like to treat for the patient as going to be a challenge. MLC is 7.12 mils at the ISO center. This is damage from the bore. There's no climatic rotation. Couch can be moved only up and down and inside outside the bore. So there is no couch motion during the adaptive workflow. It is realized by moving the segments of the of the MLC and there is currently no VMAT, but it's in works. So all this gives us so the the challenges that we saw about the shift of the profile. So this led to a like a series of research that was done by people about how the reference, the symmetry formal is need to be modified. And the consensus was to use a special factor KBQ that will be accounting for the effects coming from magnetic field. And there is a working group, the G351 that is working currently on it. And their recommendation is to use a quality beam specifier TPR 2010. It's a, it's a group that is working also with Europe, the European colleagues. So it has to be homogeneous. And this is the parameter that is very representative. It doesn't have the sensitivity as a percentage depth goes at 10X and then how exactly you would do it? So there is currently a couple of ways how you can set up you, you obviously set up at the ISO center and you still go at SAD set up depth and 10 by 10 field size. But you can calibrate your machine at 0° gantry or at 90° gantry. And initially this is something that people considered because there was a helium level that could impact the output. But if your feel of the magnet is under 91%, healing level is not a problem at that point anymore. But still a lot of people have their machine calibrated at 90°. So absolutely avoiding any possibility of that. Some machines calibrate is still at 0. There's different issues that you can, you know, account for how you would set it up and you need to have water equivalent length to the chamber. So there's some other things to consider. If you calibrated 90 then for the for the chamber. So waterproof reference glass iron chambers is something that should be considered. Typically the orientation is done anti parallel to the beam to the magnetic field. So magnetic field coming out of the bore you set up the chamber facing like in the bore here and then the chamber central should be isocentric because any shifts that for effective point to measurement is actually encompassed in the KBQ correction factor. So you can either build the in house made tank. This was in house made in Sunnybrook or you can have a commercially available 1D water tank as well. So this is something that was presented at WPM this year. This recommendations from TG351 group about what correction factors should be used for different type of chambers and you can see there is a spectrum of chambers available. I think it's also available as a poster and the at the at the context side. So you'll be able to look that if you want. OK, so this is that relates to the reference, the symmetry and calibration of the machine. Now if we go to a daily quality control task, so again at the dawn of the emergency and the recommendation of TG group to do any QA underwater. So you could in house build a cylindrical jig that is filled with water, filled with water over here. And I believe this is in the cell chamber. So you'll be able to be very confident that you're the symmetrical measurements. Alternatively, you can have a commercially available daily daily QA devices that you can set up at the at the couch itself and it's pretty reproducible, pretty convenient as well. The only thing that you will need to have electronics outside of five gauss line. So just keep that in mind. But this is something you'll be advised by the vendor. Another thing is that all these electronic devices should not be scanned by Mr. because that can damage them with the Eddy currents. So just keep this in mind. I think this one you can scan though OK for daily QA control test for MRI, something that is also performed on the phantom that has been provided by the vendor. And this is Phillips imaging quality, quality tests. And that gives us an opportunity to measure the noise of the uniformity, a linearity resolution and central frequency and also the transmission gains. This is a daily task that is integrated into the workflow. Another one that I saw this being done not every day, but I believe that this is something to consider for the daily one just because it's a like imaging and radiation field coincidence test. So this is MRI isocentric coincidence to radiation isocentric MV and this is something that is acquired during the commissioning of the machine. And then this information is also used in the proven planning system while adaptations are done on the on the console. When images arrived from the imaging system for planning, they already inherited shift. So it's important to track it and if there any upgrades, any interventions done on the Linux itself or the MRI, it's a good idea also to check it right away. The phantom is also provided by the vendor and it's visible on Mr. with a special copper sulfide liquid and also ball bearing that there's zirconium. So they're not ferromagnetic, but they're visible on Mr. as well. OK. Also something that I mean, safety is, is an important aspect of it. And then you would imagine it's going to be very similar to what we typically measure at all conventional machines. But here the, the something different is the healing level. So this is something that we need to check and the healing level is fine. It's not dropping. There is no metal in the in the bore. But this is part of the daily QA routine for the weekly task. So geometrical distortion is something that is done on the weekly basis and you can see that this is a phantom that has seven slabs is also provided by the vendor. The central slab, the 4th 1 is located directly at the at the ISO center and it has almost 2000 of different and more visible fiducials with a known location. So once we image this phantom, we can then compare with what we acquired and what was known as as a model of the phantom to come up with the with the with a correction for for distortion. Another task that can be done alternatively to the vendors provided with the ACR phantom. A lot of clinics have this implemented in their routine, so you can measure some similar but just with different phantom, different software, but similar parameters, but also some additional ones as well. OK for the monthly pass. So again, TG351 recommends to use TPR 2010 as a beam quality specifier and if you're within that range, that is, that is recommended by TG, then you're able to use the correction factors KBQ. So energy check makes sense to do and really the way you can do it is in that 1D water tank. So you can do, you know, level 20 and 10 and then you come up with TPR, which is the, the, the very straightforward way of doing this. Alternatively, what you can do is make some sort of surrogate, a surrogate device and you can see this is something that was done. So there's again a development in Sunnybrook and you can create slabs that are attaching to that cylindrically cylindrical phantom with field and water and a cell chamber. And then you can basically clip it on and measure at 20 and 10. Obviously the the commissioning of this device should be done with the water tank so that you can sign what like any correction factors that are needed for you know, lack of maybe scattering conditions. So that's also an option. OK, well let's move on for the next 1. So for the next 1 is a picket fence. So again, picket fence performance of MLC etcetera, team ties that need to be done for deliveries where the IMRT is used. So in, in in Unity system, the APID or MVAC device, the MV image controller device only can see a portion of MLCS, so about 22 by 8 1/2 field size, while this whole this whole length is 57. So you're basically measuring all the central part and you're not able to see what's happening on the periphery. So for the, you know, weekly tasks or daily tasks, you can do the the picket fans for the centrally located MLCS and this can be done with the vendor provided software Aqua. But also if you would like a more rigorous 1 so you can go to a film. So that's as far as I know this is the only way to measure the whole 80 pairs of MLC leaves. And there is a trick to how this is measured because if you can remember there is there is a doesn't matter go. There's a dose shift due to the Lawrence force to another to mitigate that. The film is sandwiched between the copper plates and the copper plates are about at least 2 mils. I think United States is this 3 because of the interest system. So you would slab it in between this copper plates and then whatever electrons that are there about to come back and deliver those again and smear your dose. You absorb them right away with the copper. So that is a way to mitigate that. Another thing is the so for the annual, you can also do this with with film and you can do this with with the copper as well. So this is for the starshot measurements to measure the radiation that the damage of radiation isocenter and then in order to attenuate those secondary electrons and prevent the dose shift, you use the copper plates, but you also remove the center of it. So you become a copper ring. And you can see that over here, it's much more similar to the conventional star shots, but over here you can see that it's smearing this part. So it definitely makes the analysis more robust and and the result of your analysis are much representative. OK, for the annual like profile and PDD and output factor measurements. So this can be done with a 3D water tank and this is PDW beam scan in Mars. So we position directly on the on the couch. This is just a recent image from our annual and you can do relative measurements with semi flex 3D or micro diamond for the smaller fields and here the motion itself. The motors are ultrasonic, so they perform pretty good in the Mr. environment. But if you utilize them a lot, they can be some heating. So just keep in mind on that. Another annual task for the geometric distortion that can be done on the annual basis with this phantom from Modus. And this is again, in addition to dramatical distortion that is done weekly. You can also use this one, you know, that just has the same parameters but different vendor phantom and software, but also measure how the machine, how the MRI is performing with active and passive streaming, for example, and some noise reduction as well. So this phantom is pretty heavy. So I saw instances where QA person wasn't able to lift it. So just keep that in mind as well for the patient specific QA. So there's multiple options currently on the market that you can do patient specific QA with. And there's the arc check Mr. there's Octavius Mr. and Delta Mr. as well. So these are pretty good for pre chat, pre treatment, quality assurance as well for the annual, you can utilize some of them to do the output versus gantry, which is important parameter for making sure everything's fine across that. For example, if we're going to some more specific things like small field of symmetry. So the micro diamond is a, you know, pretty logical detector to be used, but then again, like it has some effect of magnetic field on its response that need to be, you know, corrected or quantified for alternatively, what can be done is the use of scintillator detectors that are not affected by magnetic field and doesn't have doesn't need the correction for going to the smaller fields and you can have. So there's some vendors that are available as well. If we're talking about motion management QA, again this the comprehension management or motion monitoring is currently being deployed at the size that have 1.5 Tesla emergency system. And there is there's a ways, there's a way that it needs to be commissioned, accepted and commissioned. We're talking about end to end tests. So this devices are actually designed to be scanned with MRI and you absolutely need to scan them in order to check the performance of the gating system for example. But also for end to end they would work fine. The only thing to note is that Quasar system is actually better integrated into the workflow of the latency measurements. While these use system has a moralistic tissue equivalent anatomical structures and can simulate some sophisticated respiratory motion. OK, some special considerations for QA. So I'm becoming more into the summary one I know that time is pressing. So something to note is that most devices for Alina QA aren't designed to be imaged by MRI. So keep this in mind so that you don't damage your device except maybe for end to end and some gating performance analysis that have a designated devices to be scanned on it. Electronics for the some equipment need to be outside of five gauss line. There is there's still some equipment that need to be developed in house like that copper plate solution. Something to note is that recommendation of the G351 group is to measure your output in water, any small air bubble on the chamber or if you perhaps thinking about using solid water for measurements, you need to make sure you don't have air bubble there or air gap because that can significantly impact how you measure your output. And then the recent deployment of this comprehension motion management will require some amendment to the existing UA programs and developing some, you know, most of forward devices because measuring the performance of gating system with the phantom everyday is going to be pretty challenging. So maybe it's going to be something different, some pneumatic system, something that will just show that gating actually works and integrated in the morning routine perhaps. So something to keep in mind and some final recommendations that I would like to bring to your attention. And a lot of you probably have this already realized that create program is a process. And I would, you know, this methodology of this little pyramid you can basically apply to anything like anything that you are bringing to clinic would probably follow this general steps. So you would you would have a scope, you would have a multidisciplinary team brainstorming together. What are we going to do if we're going to treat? Because then if if it's like a BRT or this is like a small field or some even smaller field, what are the equipment to be used? So you need to have then that decision to go to the equipment phase and then you can go to the personal phase. What, what personal need to know in order to run the equipment? Is it the equipment like emergency system itself? Or is it just something simulator base that you need to have the training for that? And then you at the end end up with a program that is pretty robust, but also the program need to be able to change and improve. And at the end you will have this feedback loop because every time the scope will be changing, you're probably adding something. And in general, create program, ask the questions how can I get better and why we do this? So those are important questions which makes this whole create program business process and then as a summary and outlook as well. So everything that was said in this talk, in this or that way are represented in this list of did you report some guidelines? So some of them were published in little in 2021, some of them were published just this year like in April, this one. And you can always go back and you can and you can look of what are the things that will be helpful for your create program. There is also there is also actually this was published and in April this is published earlier. So APM is is catching up. So there is important reports that are coming out. They've gone to a certain level of approval, so 351 for the reference to symmetry. So this is coming soon. This will be very useful. This one is specifically talking about quality assurance to G352 and that's something that is coming up soon as well. So just keep in mind on that one, there is a a lot of help coming from the AWPM. And then as a, as a conclusion, so I would like to again stay that emerging ecosystems are very complex. However, they have not been realized to full extent. There's still biomarkers that we need to integrate into our workflows, understand them that will be super helpful. But there's still a topic of research mainly then Mr. guided adaptive relation therapy has become a new relation therapy paradigm that enabling a high quality personalized patient care. So right now adaptive therapy is slowly focusing on anatomy, but already it has becoming a big thing how we can treat patients better. Early adopters of emergency systems, they were dependent on in house developed devices for the leaner quality control test initially. But now there's a spectrum of devices that can establish a very robust quality assurance program for your emergency system and they are well represented by multiple vendors. So it's not a problem of finding the one for you. Something that I mentioned is that APM is actively catching up on the guidelines with QA and mortgage reports are going to be coming in the next years and emergency is still relatively new. So it's a great opportunity for research and innovation. Personally, I enjoy working with that system. And also I'd like to acknowledge that the results presented in this work has been obtained in work with some extraordinarily multidisciplinary clinical research teams. And I'm really thankful for their support and inspiration over the last decade. And I thank you for your attention as well. Thank you Victor for your presentation today there. Would you have a couple of minutes here? So if there are any questions, I'd go ahead and encourage you to go ahead and and place those questions in the lower left hand side of the application and we will answer those as they come in. There are just a couple of quick, quick questions here. 1 You mentioned that the upcoming APM task groups and you know possible recommendations that may come from that. Do you foresee in any ways that you might alter your existing QA based on on recommendations there? Or is it too early to say at this point? Thank you for the question. I think the main charges of the G groups is actually summarize all the work that has been done. So there is there is quite a bit of you know early adopters that were working on it and you have multiple. Paper saying that we did approach this way and then that way and that way so it's kind of all over and obviously before TG is there this is what we're doing we're going to different sources and see how we can learn from other experiences. So I would say that those DJ reports will be a good summary and some good recommendations as well significantly alter I'm sure. I think you can update the quality or the accuracy of your symmetry by utilizing like correction factors in the KBQ. So this is something that I saw this published recently and that's like a conference obstruct, it's in a 2G report, but it's not like as a 2G report report, it's an abstract. So those I think can be already being integrated. All right. Another question, you'd mentioned adaptive radiotherapy and and the role that Mr. guided radiation therapy plays in that when it comes to things like gating and kind of queuing that process. Are there any tips and tricks that that you've developed at UT Southwestern for that? For gating, so that's a good question. And I think this is something that is a hot topic right now because there is a, there is a handful of centers that have gating already. We don't have getting educational faster right now, but we're getting actively ready for it. And you can always reach out. The community people are very helpful if you would like to know more. So the G352 is actually coming up with the list of QA tasks that you need to do in order to assure the adequate performance of the gating system. In the meanwhile, we know that the latency is 1. So that's going to be a well integrated process with one of the vendors Motion Phantom. So that should be a pretty, pretty straightforward process. Then also you can do some symmetrical measurements with film or some end to end. So those are still I think people are still accumulating a lot of you know, checklist what do we do? And but I think it is going to be like a booming topic in terms of paper, well, quantity of papers as well. Probably the next six months as more and more systems centers will be will be having the gating on site. Well, it looks like we've reached the end of our questions here. Doctor Yekovenko, thank you so much again for your presentation. For those of you attending that would be interested in future Sun Nuclear webinars, you can find those on sunnuclear.com under the Resources page. There is the opportunity to sign up for upcoming webinars that I want to thank everyone for attending. And again, thank you to you, Doctor Yokovenko. Thank you very much for having me, all the best. _1728289408419

This webinar will cover specific features of MR-guided radiation therapy systems in the presence of strong magnetic field, which drives considerations for equipment selection and quality assurance program development._1728289408575