Video
library

Dr. Luennon: Thank you very much for an excellent introduction, Carston. So, my name is Yani [inaudible 00:00:16] Luennon. And I'm working as the future chief physicist at the Hospital District of Southwest Finland. I'd like to welcome everyone to this joint Turku University and Brainlab Webinar. If we could get the next slide. In this slide, you can see the agenda of the event. So, today, we have six international speakers from different hospitals in Europe from Denmark, Finland, Germany, and the UK. These presentations will be covering SGRT, IGRT, ExecTrac for SRS, ExecTrac Dynamic workflows, Elements treatment planning system, and patient follow-up using contrast clearance analysis. And finally, there will be another 20 minutes for questions and answers. All those questions that cannot be treated during this session will be collected and treated afterwards by email. First, I would like to introduce our first speaker from our home field. Heikki Minn is a professor and head of the department of oncology and radiotherapy at Turku University Hospital. He also works as a chairman of the board for the Finnish Oncologic Society. So, welcome, [foreign language 00:01:51] Heikki.
Dr. Minn: Thank you, Yani. It's a pleasure to be here joining this symposium. And I hope you can see my screen now or is it...? So, I would like to start little bit about telling and introducing how we became the first certified Novalis Circle Center in the Nordic and Baltic Region and how this process went and what other requirements for the staff and also the Lineax. And this will be a brief introduction to the topic. And first, these are my disclosures. And I would like to a little bit tell you about our home city, which is Turku. And it's located in the southwest coast in Finland. It used to be the old capital of Finland before Helsinki took over. So, we are quite small, only 184,000 inhabitants compared to Helsinki, which has 650,000 inhabitants. And we have this old medieval castle, which dates back to the 13th century. And this is a photograph, which I took on the day of independence, which was the 100th anniversary of the day of independence in Finland. And I would like to just tell how beautiful are Archipelago between Sweden and Finland. I think we have several thousands of islands, which are really nicely located between Turku and Stockholm. So, we have a very nice time when we have a vacation visiting those beautiful islands.
So, in Finland, at the moment, we have 44 Linear accelerators. And they are located in five university hospitals, which are here. If you see the numbers here on the map, they are colored in red. And then, we have also seven regional hospitals, which also have Linear accelerators. So, the total number of 44 Linear accelerators in Finland is 44. And considering the number of the population we have, we have 5.5 million people living in Finland, I think we have such a good number of Linear accelerators. But as you can see, it's densely populated here in the south and very sparsely populated in the northern parts of Finland.
The most common cancers, which is quite typical, inland, we have prostate cancer, and we have 5,000 new cases every year. And the same thing goes for breast cancer. We have approximately the same new number of new cases of breast cancer per year. And that survival rate is very high for both cancers. So, at the moment, the 5-year survival for breast cancer is 91% and for prostate cancer, it is 93%. Colorectal cancer among both genders is the next common cancer. And here, the survival figures are a little bit worse. We have a 64 or 65% 5-year survival rate for men. And for women, we have a 67% or 69% 5-year survival rate. And then, the total number of new cancer cases in Finland in 2019 was 35,357. So, it's the second most common reason for mortality. Only the cardiovascular disease have higher figures. In 2019, we had 13,000 deaths for cancer in Finland.
So, our radiotherapy department is well-staffed. We have, at the moment, approximately 55 people working as staff in our department. We have more than six radiation oncologists. In Finland, actually, all oncologists are trained all for medical and radiation oncology. But when they have become certified, they get additional training for radiation oncology. Then we, of course, have a couple of residents. We have six medical physicists and one physicist in training and 33 therapists already, of course, and then some additional staffing, including nurses and service engineers and technicians. We treat approximately 1,500 to 1,600 patients per year. And we have...every day we treat more than 100 patients. In our population referral RL, which is in southwest coast of Finland, it's 485,000 people, which are living in our area.
So, our fleet is completely Varian. And we have four TrueBeams. And two of these Varian TrueBeams are equipped with the ExacTrac system. And then the new kid in our block is the Halcyon, which we had installed since 2019. And we can use all of the modern techniques to treat these patients. We also are a certified center for giving bone marrow transplants. So, we have a field of doctor radiation specialists here also. Then, we have also one HDR Afterloader, which is mainly used for gynecological cancer. So, this is now we are talking about radiosurgery and still radiotherapy. So, this is not new in Turku or Finland. We started, actually, our program already in 1995. And as you all know, it was very different.
At that time, we didn't have the micro-MLCs. But we had the fixed collimators. And we had two types of collimator sets where the set which was usually molded, custom-made collimators for a fixed space. And then, we had kind of a cylinder one with different apparatuses treating with Dynamic arcs and no on-boarding, which as you see it kinda bulks which we put onto around the head of the patient. And then we just put the isocenter in place. And this patient, actually, he was treated with Dynamic arcs. But even then, we very soon noticed that cure is possible for certain cases, of course, especially for benign lesions. Like, this is an AVM, a very typical case, where you have complete disappearance of the AVM noticed. So, this was very successful already from the beginning. But, of course, the technical realization of the treatment decision was more complicated compared to the current phase.
So, now, as we have started with the more modern techniques and also this Linneaux, [inaudible 00:09:54], which I just showed you using image guidance, and we can treat lots of patients. And this is just figures from our craniologist after radiosurgery or radiotherapy. And we have the very typical indications for cranial radiosurgery and the radiotherapy. So, benign tumors not only AVM but we also treat meningiomas, of course, and pituitary adenomas. For the malignant cases, the brain metastasis is the most common indication. So, as you can see, we have now treating. I think last year was a little bit different because of the COVID outbreak. But I think we approached between 70 and 100 patients annually for the year for cranial radiosurgery and radial therapy.
So, this is a workflow for cranial radiosurgery and therapy visits designed by our physicist, Eleanor Wright [SP]. So, I give credit to her. And this is, I think, a very difficult process. And from the day the patient enters our department, we, of course, do imaging. We do all the segmentation, the image co-registration, and then the use of steps, which take approximately one week, including approval of the plan and the treatment certification by the physicist. So, actually, we use quite a lot of time for this process because we think that because, in many cases, we have benign tumors. And we, also, of course, want to reduce the toxicity of the treatment as much as possible and really not compromise the dose in the tumor.
Regarding the body radiotherapy, it has also become a very routine process in Turku. And we started mainly from lung cancer or lung metastasis. [inaudible 00:12:07] We have treated only 25 patients. And also, we had a program for SBRT and prostate cancer, multi-disciplinary, that I will shortly show you examples of that. And then, a little bit later, we also started to treat liver tumors, including both the [inaudible 00:12:28] carcinoma and the liver metastasis. But enough, of course. And now, we treat spine metastasis using recurrent bone metastasis. And this is the protocol for SBRT. And it's very similar, actually, a little bit some differences but the basic [inaudible 00:12:57] to the brain was produced. So, these are just a couple of papers from the "Prostate SBRT," models of the trial. We were able to show together with these colleagues from different countries that the toxicity is very low. And we have this kind of urethra-sparing techniques where the dose is set 36.25 to [inaudible 00:13:19] and 22.5 Gray placed in urethra in 5 fractions. And I think we will be able in a couple of years, we will also be able to publish the biochemical outcome very soon.
A couple of just slides from our plans for SBRT in lumbar spine, 3 X 8 Gray is quite standard. We have also, of course, other fractionates, depending on the target and the lesion. And this is for the liver, also quite high doses but very focused. And, actually, now it's already more than one year since we received the certification in December 17, 2019. And I would like to acknowledge all staff, which were very committed for this process. It really took a great effort from all the staff, including the radiographers, the physicists, and the physicians. So, I would like to give thanks for Manuel Todorovic, who is one of the speakers today, for completing this and doing the audit for the certificate. So, this is our team who is now present here in this room. And I would like to thank them all. And after these topics, we are ready for questions from the audience. So, thank you very much.
Dr. Luennon: Okay. Thank you very much. Heikki. Our next speaker is Samantha Bennett from GenesisCare in Oxford UK. She is a lead radiographer for clinical delivery and service development of SRS and SABR support. She also works as a project driver for SABR implementation, as well as Brainlab and Elekta integration. You are very welcome, Samantha. Be my guest.
Samantha: Thank you very much. Okay. So, I hope you can all see my screen.
Dr. Luennon: Oh, yeah.
Samantha: Perfect. So, what I'll be going through today is the beginnings of an audit that we started within our center looking at IGRT versus SGRT. And given our advance in technologies, is internal anatomical verification necessary in this day and age with radiotherapy? So, to give you a background as to GenesisCare Oxford, we are a private, independent medical center because we are not attached to a big hospital. And we are the world's leading private medical provider in oncology care. So, we have centers in Australia, Spain, America, the UK, and China. Now, at my particular center in Oxford, we operate a 1 Linac Model. And I have an ElektaVersa HD. Also, on-site, we have a Linac, a chemotherapy suite, also our clinical trial. We have a 3T MRI with DTI capabiility, PET-CT, and outpatient. So, we have a lot going on in a very, very small center.
So, within my center, we also have four radiographer model. So, there's myself and my three senior radiographers. And in 2018, we had VisionRT installed. Now, really importantly, GenesisCare Oxford is deemed as a complex SABR site. So, this means that we can treat spinal metastasis with SABR, complex afterwounds, including those with abdominal compression. Also, we can treat any nodes within the pelvic area and any other SABR sites that our colleagues are not competent to. Now, in 2019, we had Brainlab ExacTrac installed. And since then, we have treated approximately 50 SLS and SRT patients. Moving onto this year, in 2021, we had our 6Mv FFF commissioned for SRS and SRT. And in March this year, we conducted this new project, which allowed us to treat our spine SABRs with ExacTrac and the R6 DOF-catch. Now, since then, we have treated two spine SABR patients. Given that it's COVID our spine SABR referrals have got ever so slightly. But the basis of this whole presentation is an evaluation of these spine SABR patients and a evaluation on IGRT and SGRT.
So, a bit more about GenesisCare Oxford. As I said, we're single in our site, which means we have quite a unique business continuity plan. So, we're the only site to have ExacTrac within the UK. And in the UK, we have 15 other single Linac sites. So, in the case of breakdown, which is always very rare, there is a bigger thought to have as to how we organize our patients, whether we delay, treat at the weekend, treat in the evening, or change to a daily [inaudible 00:18:56]. But most importantly, our SRS and SABR implementation. So, as we all know, ExacTrac allows us to take precise KV imaging very, very quickly. And we can also enable our robotic catch to move within six degrees of freedom, which is essential for anything with curvature, spine, brain. The beauty of ExacTrac, as we all know, there is a zero percent chance of collision risk, unlike when we traditionally use our cone beams when we have any floor to it.
Now, to support or SRS and SRT planning, we use Elements as our treatment planning software. But we also use Pinafore to plan all of our other radiotherapy patients, including our SABR patients, which makes it even more unique. But I'll get onto that at the very end. Now, to tie this all together, we use Mosaiq as our care packaged delivery chemotherapy and radiotherapy.
So, as we've all seen many, many times, this is an example of one of our SRT treatments from a patient we treated a few weeks ago in Elements. But what does this all correlate to within the treatment room? So, we've got quite a busy bunker. And as you can see in the background, we have our Versa HD Linac. Now, on the far right, far left, and at the back in blue, we have our VisionRT surface-guided radiotherapy cameras. Highlighted in pink are our Brainlab ExacTrac panels. And at the very back, we have the Elekta HexaPOD iGUIDE camera that will point reference to a frame and allow the sit/stop couch to move. Now, within the control area, as you'd expect, with all this equipment comes a lot of monitors. So, we are relatively cramped. But on the far right-hand side, we have Vision supported by the middle by Elekta, far left we have got an iGUIDE, and the fourth Brainlab ExacTrac in the box on the left-hand side.
But most importantly our surface-guided radiotherapy is functionality. So, the top picture really demonstrates what SGRT looks like when we have the cameras on in the room. So, it does shine a red infra light pattern on the patient's skin, which enables us to move to the isocenter. There is a functionality for us to have gating with the Linac. And on the bottom of this slide, you can see two pictures. So, on the right is an example of a treatment capture whereby the area in green is the area that we're monitoring during treatment. And back on the left is something called video deformation that we use for a lot of our breast folks so we can visually monitor the patient. And that pink outline is the day's capture. So, that can incorporate anything new that we've done with the patient.
However, what do we think about when we think of surface-guided radiotherapy about in clinical practice? So, it's safe to say, and especially within the UK, SGRT has really picked up in popularity over the past year. And it is absolutely fantastic for mitigating the use of tattoos. So within GenesisCare Oxford, and, of course, the UK, we haven't tattooed our patients since installation, so since 2018. And that does include our SABR patients. So with that is the reduction and psychological impact of having a permanent reference point. And SGRT is very, very easy to use and set up the patient too. But can surface tracking only be safe and effective in the accurate positioning and monitoring with SABR and cranial SRS patients?
So, to give an example of what this looks like, this is what we would import as the diacon data for a T9 spine SABR patient. So, you can see on the image we have a very big band, a quiet area, going across the patient's upper breast and thorax. Now, the team has placed this region of interest as it is in location to the isocenter. But how does this correlate to any form of internal imaging? So, in this case, we set the patient up for this picture. And then we took our ExacTrac images, which showed us it. And that's how we're using the two systems together at the moment. But most importantly, after we've taken that ExacTrac image, matched and moved the bed, we want to update this SGRT system as to where we have now moved to. And what I present before you is something called a reference capture. So, you can see that ROI has some reduction in its quality. You can see there's a lot of extra structures around the patient, especially inferiorly. But it's also picking up the HexaPOD frame. So, these are all things to consider when we are using multisystems in one patient's pathway.
So, what was our audit about? Our audit was about comparing the isocenter positioning and the accuracy of two systems on my one Linac. So, ExacTrac proceeds to the SGRT component. And VisionRT SGRT was installed to use and compare shifts. And the table position generates it by the ExacTrac system. But what does that imaging flow really look like? Now, as I stressed, we only have two radiographers on set when we treat any SABR patient. So, there's a lot going on in a very short space of time. We also only schedule our SABR patients for 20 to 30 minutes. So, we set up to SGRT. And when we treat any spine SABR patient for anything C and L, we use something called the windboard.
So, we do not use [iandubile 00:24:16] at all. To help us with that if we have any patients that are C spine or lower C, upper T, we can immobile them in a head and neck [inaudible 00:24:25] with an open portion of the face marked to allow our VisionRT cameras to detect the patient. However, we ensure the RSGRT camera is on. And then, we apply our first set of ExacTrac images, unmatched, as you can see. We perform an auto-registration. And when they're happy and if it's intolerant, we push those moves to the sit/stop couch, reimage to ensure the correct place, and we take a treatment capture to update our SGRT system.
Now, to ensure that we are definitely in the right place, my team then acquires a [inaudible 00:25:03] CT. We are purely checking for gross error at this stage. We are not applying any moves or any shifts from this image. The reason we're doing that is because spine SABR with ExacTrac SGRT is still a relatively new workflow for the team. And this is part of our first 20-patient audit process. When we're happy, and due to the length of time that it takes to acquire this cone beam, we then repeat our ExacTrac images to make sure that we are very much happy that the patient has not moved. And this will be used in accordance with SGRT. We record a new treatment capture. And then, we begin to deliver treatment. So, the majority of our spine patients are treated with two-up. We do not use triple X for these patients.
So, how do I know if my patient's moving throughout treatment? You would hope that SGRT is able to detect these millimeter movements of the patient. However, at present, that just doesn't quite seem to be the case. And I'll provide you with an example. One of our L-5 patients had an arc going from 180 to zero. My team will always take ExacTrac images at either zero or 180 to ensure the patient hasn't moved. ExacTrac came up to say that we had moved over at that time. However, this didn't appear on the SGRT system. So, already, we have some points to consider when we're looking at tolerances between SGRT and Exectrac. But if we are happy that the patient hasn't moved, given our ExacTrac images, and we have either not moved or moved and reacquired, we then complete the fraction.
But what do these tolerances really look like? So, to put it in comparison, we have our cranial ExacTrac tolerances on the far left-hand side of the screen. So, our initial ExacTrac image in time was 6 mills, 3 degrees. But then, the subsequent imaging will be half the amount, half a degree with our cone beam tolerances being 1 mill, 1 degree. But I stress we do not move from this image. Now, when we look at spine, we start in the same fashion with our ExacTrac images. So, we have a 6 mill and 3-degree tolerance with our subsequent ExacTrac images being 1 mil and 1 degree, our cone being 1 mil and 2 degrees. But when we have a look at SGRT, it is unable to detect anything smaller than 1 mill when you're looking at any body area to monitor. So, when I say that our tolerances are 2 mill if we start at zero but +2 mil, -2 mil, all of it is essentially a continually moving target because the patient's breathing, as you would hope anyway. It's also important to consider that your [inaudible 00:27:43] is always set at 3 degrees for the same reason.
But what does this look like and what has the data really showed us? So, what you're seeing on your screen at the moment is at the very top we have our ExacTrac data. And at the bottom, we have our VisionRT SGRT data aligned in the same way with pitch, roll. You're then looking at our translations of that long [inaudible 00:28:05]. And what you're looking at is the averages distances at the isocenter. Now, the standard deviation value that's been calculated do represent that ExacTrac. And these fractions delivered is accurately able to position the patient to the isocenter with significantly smaller standard deviation than VisionRT. Now, this data really screams out to myself and my team. With this discrepancy that can be seen with my SABR patients, it makes me think twice about using this for SRS, particularly with multiple brain metastasis based on a one isocenter position setup.
Another point to consider is at the moment in GenesisCare, we plan our SABR patients for the 3-millimeter tolerance. So, if we were looking at ExacTrac, we would be able to safely deliver with this 3-millimeter side. However, if we used VisionRT alone, we would need to increase this PTV volume by at least 1.2 millimeters to account for this gross change in the standard deviation. So, as it stands, this can be an option to use for some SABR patients. It does not look at present like an option to use solely for SRS, considering our accuracy [inaudible 00:29:21] to a half-degree. Another part to consider as well is when we are defining the region of interest, we have to be aware if there's any motion, so breathing, especially in the abdomen. And could we improve this with a DIVH technique? So, a breathe hold technique or an expiration hold technique for these patients?
But, ultimately, we need more data. We've only had two patients come through. The only good thing is we have had two very different types of lumbar spine, which gives us a good overall comparison to start this audit with. But with all audits, that has been a bit of a challenge. So one of the main challenges of the audit and with the SGRT camera position is the gantry blocking the view of the patients with the cameras. So, what this means is that as the gentry moves around and the camera is blocked, the SGRT system will struggle to detect the patient motion. There is a functionality for gaiting, which will then stop the machine from delivering the treatment. However, in these spine SABR cases, we do not enable gaiting because we are relying on the ExacTrac system more. Our new software does allow for a video review from an unobstructed area. So, we will reflect back to the beginning of the presentation where that pink outline was showing around the patient. But from the very end camera at the foot of the bed, which allows us to view that patient throughout the whole treatment but it doesn't give the overall ROI picture.
Now, I stress this, and I will continue to stress this throughout this audit. The position of the ROI is essential when we treat any patient with SGRT. And whilst we are given some instructions from the manufacturer as to where to place this ROI, it very much relies on clinical judgment, especially when you're looking at very small areas within the thorax. You can something called a accessor or a centroid whereby you place your region of interest at an area that is more stable but, ultimately, away from the isocenter. But, as I said, this relies on clinical judgment to use that functionality and to then be aware of how this may impact on your isocenter position.
Another challenge is the HexaPOD frame location. So, anyone who uses HexaPOD or has that integration, you cannot have your HexaPOD frame installed or placed any further away than 65 centimeters from the isocenter. And it must always be indexed inferiorly. So, if we're looking at anything L or lower or more inferior in a patient's spine, there is a challenge with the way we immobilize our patient. So, [inaudible 00:31:51] and a knee rack to place these patients in. The HexaPOD frame can abut either the patient or the [inaudible 00:32:00], which makes it quite hard to settle into position. But most importantly, at present, it's the synchronization of all of these extra devices, all of these extra facilities that we have, that is the hardest part to make sure that we're all getting right. The patient files must be opened in a certain order, as well as the Linac, Brainlab, and our SGRT system, all being able to pull the correct patient data and all not be fighting with each other for the main use of our time and system.
So, what are we really understanding at the moment? So, at present, surface-guided radiotherapy and spine SABR is showing a reduced accuracy compared to our IGRT imaging with our main challenge being the obstruction of our patient or of our view of the ROI with the gantry angle and our reliance on video deformation. But most importantly, it's my team's and my responsibility to make sure that we are placing this region of interest or [inaudible 00:32:56] in the most appropriate area possible.
So, looking forward with IGRT, SGRT, and SRS, what are the considerations that we have? So, at the moment at GenesisCare Oxford, we use ExacTrac with the full three-layer mask for SRS and SGRT patients, which allows us to enable full pics for these multiple isocenters and take additional imaging at [inaudible 00:33:25] account for the sit/stop movement. That's just not something that SGRT can currently do well. And it's not something that we use for our SRS patients. We use SGRT with couch kicks for our GBN patients. From the moment you have a couch rotation, the SGRT definition of the patient's space does reduce in quality.
As we've said throughout this presentation, intrafractional motion movement. So, SGRT, because the tolerance isn't so different to ExacTrac, we are always going to be fighting against two different systems and two different priorities. So, it's another thing for my clinical team to be aware of. But most importantly, using many systems at once with little synchronization does rely on the radiotherapy, radiography, clinical judgment, and prioritization. Ultimately, it encourages a greater risk of error.
But moving forward, what can we do to further improve the service? So, to start us off, we are looking at our physics and machine advancements. So, at the moment, as I said, our PTV expansion is currently 3 millimeters on spine. But this is because it is defined in Pinnacle. We do have the option to plan these patients in Elements. And that's something that our new physicists who we recently onboarded are looking at doing. This will enable us to have a reduction in those margins because we can correct to the sit/stop couch. But an important thing to consider, is at present, we will not be able to treat any of our spine SABR patients with FFF because we are the only site within the UK network who has that clinically available and commissioned.
But is there a revolution on the horizon? And we think we've got it. So, from June 2021, GenesisCare is looking to have Dynamic installed, which we will be using for all of our SRS/SRC, and our spine SABR patients. And it's really essential. We've all acknowledged that we can no longer rely on one solution to set up and image and track the patient. And Dynamic really seems to have this ability to incorporate all those areas of concern into one user-friendly, easy solution. So, I'm very excited for that to come along. Thank you very much for your time and listening. And, as before, if you have any questions, please save them until the end and we'll be happy to answer. And a big thank you to my team here at GenesisCare Oxford.
Dr. Luennon: Okay. Thank you very much, Samantha. And then, I would like to introduce our third speaker, Dr. Lucie Calmel, a Ph.D. from Herlev & Gentofte Hospital Denmark. Lucie did her master's studies in radiology and medical imaging at the University of Toulouse and her Ph.D. studies at Paris Sud University in France. [foreign language 00:36:38], Lucie.
Dr. Calmels: Thank you very much for the introduction. Yeah. So, I will share my screen. Okay. Can you see my screen?
Dr. Luennon: Yes.
Dr. Calmels: Yes? Super. So, thank you very much for the introduction. So, I'm a medical physicist at Herlev Hospital in Denmark. And I will talk with you today about ExacTrac and intra-fractional motion for SRS brain metastases treatment. So, first an introduction of who we are. So, we are quite a large center. We have 10 matching units with 4 Varian TrueBeam, 2 Halcyon, 1 Ethos system, 1 MRI Linac. We have also 2 Siemens [inaudible 00:37:30], one dedicated MRI, and 3 Brainlab ExacTrac system. And we treat around 3,000 patient per year.
And I would like to introduce briefly the implementation for the SRS treatment in our hospital. So, before June 2018, we treat no SRS treatment in our hospital. They were all referred to a Naval hospital in Copenhagen. But in June 2018, an original directive asked us to do SRS treatments. And from summer to autumn 2018, we have implemented the HyperArc solution on a TrueBeam with a 5-millimeter central leaf called MLC called [inaudible 00:38:38] 120-MLC and with the ExacTrac system. And we have treated the first patients in October. In February 2019, we have transferred all the HyperArc treatment on the new TrueBeam with a 2.5-millimeter central leaf width called HD-MLC. And, again, we use the ExacTrac system to verify the position of the patient. And today in May 2021, we have treated about 300 patients with the HyperArc technique within three years.
I would like to briefly explain a few important steps for the implementation of the SRS technique. So, first, for the patient, immobilization, the HyperArc system is delivered with a setup device called Qfix encompass and with dedicated mask. Then, it is important for the radiotherapy team to learn how to use the system to achieve good positioning of the patient during the treatment and to minimize movement of the patient. Then, we have also modified our IGRT workflow. And for this purpose, we use the CBRT imaging and the ExacTrac system for the initial setup of the patient. And then, for each calibration, we verify the position of the patient with the ExacTrac system. And we correct the position of the patient with a 6- degree of freedom of the couch. And we have done that for the first year of treatment. And I will explain later in the presentation how we have divided our IGRT workflow now. But we can say that the CBCT system and the ExacTrac System were well correlated with the test within 0.75 millimeter. And the total runs for the patient setup for both CBCT and ExacTrac systems the first four months. So, with 120 MLC it was 1 millimeter 1 degree. And then, we decrease the [inaudible 00:40:17] two-day runs .5 millimeters 0.5 degree with the HD MLC.
So, the quality of the runs for the SRS treatment consists of the machine performance check, the Winston Lutz Test, and the cure rate of the patients. I will explain here how we have performed the Winston Lutz Test for the treatment of SRS brain metastasis. So, the Winston Lutz Test is used to verify the correspondence between the mechanical isocenter of the field and the radiation center and was performed by scanning the Brainlab flat phantom that you can see here on the image and this Brainlab phantom incorporate [inaudible 00:41:01] in the center. And the aim of the Winston Lutz Test is to map the isocenter of the imaging with the center of the sphere. So, the phantom was placed randomly on the couch of the treatment.
And two KV-image at and 0 and 270 degree were used to align the sphere with the imaging at the center. And an automatic couch shift were used to move the phantom to the imaging isocenter. And a control was done with the ExacTrac system. Then, a field of 1.5 by 1.5 square centimeter were required for full gantry rotation, 5-couch rotation, and 8-collimator rotation. The data were analyzed with the [inaudible 00:41:54] software as you can see here from the Figure 2. And the Winston Lutz Test results demonstrate the mean total deviation of 0.3 millimeter plus, minus a standard deviation of 0.07 millimeter.
At this time of the study, we have treated 32 patient with the 120 MSC. So, the MLC was 2.5-millimeter difference. And 170 patient with an HT MSC. We have treated more than 60 of the plan with only 1 brain metastasis. And we have applied a 2-millimeter GTV PTV margin for all the brain metastasis. More than 60% of the brain metastasis received one fraction of 18 Gray. If the brain metastasis is larger than 3 centimeter, we have used 3 fraction of 9 Gray. And if the brain metastasis is close to one organic [inaudible 00:42:54], 1 fraction of 13 Gray were applied. The GTV volume were compressed between 0.01 cubic centimeter to 90.6 cubic centimeter.
So, before each treatment, as mentioned before, the bony structure registration were performed between the planning CT and the coming CT the ExacTrac system. And prior to each treated arch, the infra-fractional motion of the patient was corrected with the ExacTrac system and the 6 degree of freedom of the couch for the 132 first patient. The results have shown that the IGRT treatment was delivered in a reasonable time slot. So, an overall treatment time of 30 minutes was observed the first four months of treatment, as you can see here on the orange box plots on Figure 1. And it was a treatment with 120 MRC. And then, the overall treatment time is decreased to around 20 minutes for the treatment with HD MRC as you can see here on the blue box plots. So, the overall treatment time is very similar to a non-SRS treatment. And it included ExacTrac image acquisition or registration of the image with the planning CT and the couch adjustment if needed for couch rotation. So, this is a very high-precision treatment with high-dose for fraction.
So, I will explain an introduction. We have tried to divide our workflow in two workflow. And I will explain here how we are doing that. So, first, we have extracted the ExacTrac data acquired and recorded for the first 100 patient treated with the HT-MCL. Then, we have created and analyzed the data of two workflows. So, one workflow with ExacTrac and one workflow without ExacTrac. So, the first workflow with ExacTrac for this patient, we have corrected at each couch rotation for the positional error of the patient if there were outside [inaudible 00:45:10], so 0.5 degree and 0.05 millimeters. And for the workflow without ExacTrac, we have used the ExacTrac only for the initial setup. And after at each couch rotation, the position of the patient was not corrected. And we have recorded the residual setup error for the last ExacTrac image required before the beam on for the workflow with ExacTrac. And we have recorded the residual setup error up until and including the first setup error outside [inaudible 00:45:45], so 0.5 degree and 0.5 millimeter. And the recorded value for the two workflow were used to calculate the magnitude of the brain metastasis center displacement from the ideal position. And this maximum displacement is called E.
So, on the Figure 2 here, we represent the maximum deviation of the center of the brain metastasis here as a function of R in millimeter. And R is the distance from the center of the gray metastasis to the isocenter of the plan. And this solid line represents the workflow with ExacTrac. Y is a dotted line represent the workflow without ExacTrac. And we have plot the maximum deviation E for the means plus one standard deviation. So, this is a circular marker and the mean plus-minus 2 standard deviation. This is the three other markers. The green line here denotes a 1-millimeter GTV-PTV margin while the red line here denotes the 2-millimeter GTV-PTV margin.
So, according to this figure, we can conclude that the single target and center of the brain metastasis within 2 centimeters from the planned isocenter. The workflow without ExacTrac is sufficient to cover 95% of the brain metastasis. But for the multi-brain metastasis with a center further away than 2 centimeter from the planned isocenter, the workflow with ExacTrac is needed and must be used to cover 95% of the brain metastasis. And yet, as you can see on the graph here, the multi-brain metastasis in 90% of all cases were situated further away than 2 centimeter from the planned isocenter. We can also conclude that GTV-PTV margin could be adjusted if we use the workflow with ExacTrac. So, if the center of the brain metastasis is further away than 3 centimeter, so here we have to use 2 millimeter [inaudible 00:48:00] margin. But if the center of the brain metastasis is situated within the [inaudible 00:48:05] from the planned isocenter, the 1-millimeter GTV-PTV margin could be used.
So to conclude, our experience has shown that the workflow with ExacTrac is essential for the SRS treatment with a TrueBeam Linac for single isocenter brain metastasis treatment with a VMAT plan. We have used the workflow with ExacTrac for the first 130 patients. Then, for the brain metastasis situated further away than 2 centimeter from the planned isocenter, our studies suggest that the patient positioning errors must be verified at each couch rotation with the ExacTrac system to avoid compromising of the brain metastasis coverage. At our hospital, the workflow using ExacTrac was used for more than 80% of our multi-brain metastasis treatment with a GTV-PTV margin of 2 millimeter. For brain metastasis situated within 2 centimeter from the planned isocenter of a single brain metastasis, our studies suggest that the workflow without ExacTrac is sufficient to ensure brain metastasis coverage.
And we used the workflow without ExacTrac for a single target and multi-brain metastases within 2 centimeter of the planned isocenter. However, if a brain metastasis is close to one organ at risk, we use a workflow with ExacTrac to ensure that we don't increase the dose delivered to the organ at risk. And now, we try to investigate if we can reduce the margin so for the GTV-PTV margin to 1 millimeter. And we think that we can use 1 millimeter if the brain metastasis is situated at less than 3 centimeter from the planned isocenter. and if we use the ExacTrac system to verify the position of the patient at each couch rotation. And I would like to thank all of the team who have been involved in this project. And thank you for your attention. And if you have question, I will be happy to answer during Q&A round table. Thank you.
Dr. Luennon: [foreign language 00:50:29], Lucie. Our fourth speaker comes from Rigshospitalet in Copenhagen, Denmark. Dr. Nikolaj Kylling Gyldenlove Jensen. Nikolaj did his Ph.D. in medical physics at the University of Western Ontario, Canada. He's been collecting his clinical experience in many different hospitals. And he is a board-certified RT medical physicist in Denmark. Welcome. [foreign language 00:51:05], Nikolaj.
Dr. Jensen: Okay. Thank you. So, I will take you through the workflow of the ExacTrac Dynamic as we perform here at Rigshospitalet in Copenhagen. At Rigshospitalet, we've been working with ExacTrac for more than 15 years. We installed ExacTrac version 4 back in 2005. We've been working with ExacTrac Dynamic since early 2019 when we had a prototype of the system installed. We've done user testing with the Varian integration and we were also involved with the CE marking process. We currently have six ExacTrac Dynamic systems installed in the clinic. We only have six TrueBeams. So, we will not be getting any more systems any time soon. As of yesterday, we had treated more than 200 patients with ExacTrac Dynamic since the first system went live in October 2020. More than half of the patients we've treated have been cranial SRS patients. But we've also treated quite a lot of glioma and other brain cancers. We have treated a few bone SBRT. And we've done it now using it in pediatric patients as well for the dose reduction compared to [inaudible 00:52:28].
The example workflow I will show you is a female patient with astrocytoma. It's a post-operative patient treated 60 grams of [inaudible 00:52:40]. The tolerances that we used are .7 millimeters and 1 degree for X-ray and 1 millimeter 1 degree for surface. The mean tolerances are it's a fractionated treatment. And we had the option of moving this patient to an older clinic with slightly poorer mechanical accuracy. The patient will be treated in the green mask, which is a full-face covering mask, which does introduce a little bit of problems with the monitoring, but you'll see that as we go.
So, I'll show you a video recording of one of the fractions for this patient. So, I'm gonna show you ExacTrac will be waiting for the patient plan to be opened on the Linac. Once the patient plan is opened up on the Linac, ExacTrac will then attempt to open the plan. And if the plan has been exported to ExacTrac, ExacTrac will give you this review screen where you can choose to reject and auto-authorize, in which case ExacTrac will not interfere with the treatment, or you can confirm the plan indicating that you wish to set up this patient with ExacTrac. If the plan is not in ExacTrac, then here you can send... It will simply reject them and auto-authorize. So, once you confirm the plan, it will go into the pre-positioning workflow. You have to mode up on the Linac. Otherwise, ExacTrac cannot properly communicate movement. You could position this based truly on surface information.
So, using the light dots that Samantha showed earlier, the thermal camera's not in use here. The patient is currently within field view of the camera. But we just saw the shadow of a nurse putting the immobilization mask onto the patient. And in a minute now, the patient will appear in view. So, here the patient is now in view. ExacTrac has identified that the patient's there, has calculated the shift. This shift is then automatically sent to the Linac. And the Linac couch will then perform this motion to move the patient to isocenter. Once in this position, you have the option of correcting the patient if you see any large rotations. In this case, that was not necessary. So, we could move straight through.
The next task to perform is the area of interest for the surface camera tracking. Because this is a later fraction for this patient, there was already defined an area of interest over the face of the patient. And it's the brighter area in the center of the image. I have to pause it here because to verify this took only a few seconds at treatment. Once this image is done, we will go into the stereotactic X-ray direction. For this patient, we wanted to start the gantry at 180. So, I'll spare you kind of to watch the gantry move all the way down. And then, we'll do stereoscopic X-ray imaging. As you know from all previous versions of ExacTrac, yes it will calculate an automatic [inaudible 00:55:56]. That's great to verify with the spyglass. We allow RTTs to verify the automatch with any of the reviewing tools at hand. This shift is sent to the Linac to perform the couch motion and a set of verification images are then acquired. We have a surface policy [inaudible 00:56:20] whenever we perform a couch motion to acquire verification images, and you'll see that in a minute.
Here, you can see the longitudinal is still outside of the .7 millimeter tolerance. That's something we've seen happen when we perform large couch pictures especially. So, it's not particularly a surprise to us. So, we just drew an extra couch correction, go back and acquire another set of verification images. And in this case, yes, we're within .2 millimeters of where we want to be. A quick verification and we can proceed to treatment. Once we're in treatment, the surface and thermal camera will take over monitoring the patient. It should be said that the values are initialized by whatever accepted deviation from zero was accepted at X-ray. As you can see, we do not have any imaging in the view. We don't have any triggered images. The amount of review time here is less than 30 seconds. And with the automatch taking about 5 to 8 seconds to perform, it would leave us with so little beam on time that we've simply decided that for fractionated treatment, this is not going to give us much benefit. But it would have been impossible here to define X-ray images taken during the beam either at a 45-degree angle [inaudible 00:57:54]. And you can see in the bottom left which of the X-ray tubes can actually successfully acquire an image. [inaudible 00:58:05] to see both of them we would be able to.
Obviously, with a full-face mask, this is just for a second, some limitations in terms of the surface tracking as we're tracking the mask and not the patient. We have found that if we make these masks sufficiently tight-fitting, the patient has a very difficult time moving under the mask without also disturbing the mask enough that the system will detect it. We then move to our third and final treatment beam, which is at a couch rotation of 85 degrees. And here, you'll just have to sit through the rotation. So, when you perform a planned couch rotation with ExacTrac, it will automatically try to perform the surface area of interest that was defined at the beginning of treatment.
If the system thinks it's succeeded, it will just move on. If it is unsure of the result, it will pop up into this review mode. You can then change it as necessary. In this case, it has done a perfectly acceptable job. And you would just confirm to move on. We have a policy to always review it even when the system is informing us that it has done a sufficient job. And that's just based on our early experience with the system, that sometimes it would include volumes at the boundary of the patient, which may then spill over onto the gantry itself, which would cause a lot of deviation. And here, we just confirm that ExacTrac had done a good job and move through to X-ray collection for the post-calibration field.
And here, you'll see that we're actually within tolerance of .7 millimeters after the couch rotation. So, there's no need for an additional set of images. We'll just go straight to treatment. And, again, we have another 30-second beam so we don't have any triggered images for this patient. And here, you can obviously see that since the X-rays for this beam had a slightly higher deviation from isocenter, that also means that the surface tracking was initialized with a greater offset. But as the patient here is very static, it doesn't change direct treatment. An option could be set here for ExacTrac to have a gaining function where it would turn off the beam if the patient moves. For these fractionated treatments, we generally don't allow that. But it is something that we're using for our FBITs and pediatric patients. And in a second here, you should see the mask start to be removed from the patient. And there we go.
So, just to highlight a few features that ExacTrac Dynamic brings that are very different from previous versions of ExacTrac, so, the first one, you saw the repositioning workflow is completely surface-based, meaning that you really don't need tattoos. You don't need markers. You don't have to place reflective markers on the patient or on the couch. You can base this purely on the patient's setup of the day. It will also allow you, and especially for brain cases, to perform rotational correction prior to mobilizing the patient before applying the mask. We've actually found during the rotation of the patient this is something we use for difficult patients. We have found that with our mask system, it's typically not necessary. But for patients who are very cognitively impaired, this is something that we have to do every so often.
Here, I've highlighted just the difference between ExacTrac 6.5 and ExacTrac Dynamic. This on the left is ExacTrac 6.5. And this is actually the same patient treated with the same isocenter at the same time, just two different days, once on an older ExacTrac 6.5 and once on ExacTrac Dynamic. And really, I think the big takeaway here is the very significantly larger field of view. So, with ExacTrac Dynamic, you can always see the facial bones. So, you always get very good image information that's particularly connected to the skull whereas with ExacTrac 6.5 you can get in a situation where most of what you see is spine and mandible, which can both move relative to the skull and potentially and negatively impact your registration and accuracy. And obviously, then, with the monitoring, we don't have to place reflective markers. We don't have to handle that very large horseshoe-shaped array. We don't have to tape anything onto the patient or their fixation. Now, we just put them in the mask, define where we want to monitor, and we move along.
We've done a few studies on isocenter accuracy. This is a software solution that was developed in-house. So, based on [inaudible 01:03:20] images with the last marker, an optimal isocenter position is determined. That's then compared with ExacTrac's identification of where that last marker is. And what we've found with two clincal systems over about six months with this testing performed every month is that, on average, we're actually seeing deviations from ExacTrac to MV isocenter of less than .1 millimeters with a standard deviation of about millimeters. So, the systems are quite stable and they're very mechanically accurate when calibrated properly.
So, there's one thing, as I said, trigger imaging is something that we do very rarely here at Rigshospitalet. You can set your trigger imaging based on monitor units. So for [inaudible 01:04:09] fields or standard MLC fields with gantry entry, you can base it on gantry angle. So, you can do monoscopic imaging at all the 45-degree gantry positions. And you can do stereoscopic imaging for the gantry at 180, 92 seconds, and zero. You can also trigger on-surface deviation. So, if the surface ever goes outside of tolerance for a set amount of time, you can define a delay in that. You can then ask the assistant to trigger imaging if it can. And if it can, it will potentially hold the beam and ask you to move the gantry to a position where you then can perform your trigger imaging and verify if the patient has moved or not. Again, we use full-face masks for cranial patients. This is typically not something we do. And really when looking at data of patients being treated and then having couch sit beams where we will re-image them, we have found that they tend to be very stable. So, that was it for me. Thank you. If you have any questions, I'll be happy to stick around and answer them during the Q&A session.
Dr. Luennon: [foreign language 01:05:28] Nikolaj.
Dr. Jensen: [foreign language 01:05:29]
Dr. Luennon: Even we can see that we are a little bit behind of the schedule. I would like to remind that please send your questions to the chat. I can see that there is no questions at the moment. And next, I'd like to introduce our fifth speaker from the University Hospital Hamburg, Eppendorf, Germany. Dr. Manuel Todorovic is the head of clinical medical physics and auditor for the Novalis Certified Program. [foreign language 01:06:09], Manuel.
Dr. Todorovic: Yeah. Hello. Thank you for introducing me. Good to see you again. Can you see my presentation? Yes, right. It's there, hopefully.
Dr. Luennon: Yes, it is there, and it's good to see you again, as well.
Dr. Todorovic: Okay. Perfect. So yeah, I want to talk about multiple brain metastasis SRS and cranial SRS, so, about the two software solutions from Brainlab and then, a little bit in terms of complex case planning. So, yeah, here's my little disclosure. We will skip that. So, the next around 15 minutes, I will talk about what is needed to get theoretically a good dose distribution. And then, I will speak a little bit about how can treatment planning be done. I will show you different treatment options for multiple lesions. So, we saw one. We saw HyperArc. I will talk about Elements. I will compare these both a little bit to give you an idea what's up with each solution, what are some features maybe that come up with these. I will talk a little bit about cranial SRS, not too much. And then, I will show you a case report just to give you an idea how we use these in our clinics.
So, first of all, I would like to motivate a little bit why in the end we are talking about SRS for multiple lesions. So, in the past, we all did whole-brain radiotherapy for patients that got multiple lesions within the brain. But we all do know that whole-brain radiation is a toxic treatment. As you can see here on the pre-RT image, you see it here. And you see another image of the same patient 18 months after RT. And obviously, we see we do a little bit to the brain. We harm the patient. We destroyed a lot of material there. And the question is do the benefits of the whole-brain radiation therapy justify the risks, obviously, coming with them when you look at the post-RT image?
The good point is we all thought okay, that might not be the best solution. So, there was a little bit modification coming out for whole-brain radiotherapy. So, nowadays, a lot of people are doing whole-brain radiotherapy in combination with hippocampal sparing. But here, what happens in this case if you do that? So, we see here a brain of a patient. Let's see. He got multiple lesions and here we've got another lesion over here, which is quite large, which we would like to integrate maybe with the boost within the whole brain radiotherapy. And on the left, you can see what would happen if we do an SRS boost or just an SRS for the big lesion in this case. And on the right, you can see the dose distribution when we do the hippocampal sparing whole-brain therapy with an integrated boost.
And the good point is you can see the blue areas here. We really did a good job in hippocampal sparing. But if you look at all these green zones here, you see it's still 99 or more than 99% of the whole brain received the prescribed dose. And prescribed dose in this case, when we're talking about whole-brain radiotherapy might be 30 Gray or 33 Gray, depending on your fractionating scheme. So, still everything else got 30 Gray. So, this will happen as well. So, we will spare the hippocampal area maybe. But this will happen, for sure as well. So, maybe that's not the last call or the best idea in this case. It might be for some situations we do that. But most of the time, we do SRS treatment for multiple lesions.
So, what are there for different treatment options in this case or planning options combined with treatment options? So, the first thing up here is we could use multiple isocenters and arrange Dynamic or HyperArc or beams or whatever we did. Like we did in the past when we're treating just one single lesion or so, we can do that for multiple lesions, as well. So, we use 1 iso per lesion, 3 to 5 Dynamic HyperArcs, table angles per lesion or beams. But what you can see here up in this little image is as soon as you use...if you try to use these for lesion numbers larger than three or four, you will, obviously, see that you got a little bit of a space problem up here. So, every beam here for this isocenter will also influence situations on the other isocenters. So, this is completely manual-driven. This is foreplanned. So, you will just arrange everything around one lesion. Then, you go to the next lesion, put the iso there, arrange the beams, and just see what happens. So, this is really, yeah, a time-consuming approach. And in the end, you will not come out with a good solution for that.
The good point is, at least right now, there are two dedicated solutions for multiple lesions. The first one is the Varian HyperArc solution, and there's the Brainlab Elements solution. We heard already a little bit about Varian HyperArc. Nevertheless, I would just come up with some additional information. This uses VMAT, a VMAT approach. So, you will use one virtual isocenter positioned in the center of multiple lesions. And it will allow you to use three to five VMAT arcs. So, different table angles in total. You can see here how the table angles will be arranged in this situation. And then, again, VMAT to cover all the lesions.
The Brainlab Elements multiple brain metastasis SRS is a little bit different. It also uses one virtual isocenter positioned in the center of mass. But then, it will use Dynamic conformal arcs. So, the same we used in the past here, the same we used to use for SRS treatments. But in the case that you use Dynamic conformal arcs, you need a little bit more table angles in some situations so you can add up to three to seven Dynamic conformal arcs with table angles in total, depending on what you like. So, this is the template-based approach.
Just to give you a short idea why the first approach, modular isocenters, might not be the best solution, here's just a little assumption. TIme and minutes here, number of isocenters, or number of table angles, depending on, yeah, number of lesions or number of isocenters. So, if you just use one...if you only have to treat one lesion, then you see it would take roughly 30 minutes planning time, 20 minutes treatment time on the machine. But if you add up with two lesions, you have to use two isos. So, you have two times the planning time, two times the treatment time. And this easily moves up and ramps up to unbelievable numbers of time you have to take.
So, in our department, we used...we treat... In the mean, we have six lesions per patient or four per treatment when we do multiple lesions. So, if you would like to do that with multiple isocenters, you will end up with 180 minutes treatment time, planning time, and 120 minutes table-on-beam time for a patient. But, obviously, that's not a good approach in this situation. That's not very comfortable for the patient. And it's also very time-consuming, especially when you look at the time that we take in our department. It doesn't matter how much lesions because it's an automatic approach for multiple lesions. We roughly take about 20 to 30 minutes for planning and then another 20 to 30 minutes for irradiation, no matter how much lesions you have as you treat all with a single isocenter in this case. So, this might not be the best approach. We just skip that. But just keep in mind it was out there. and for single lesions, this might still be a good approach in some situations.
So, here's a little comparison between HyperArc and multiple brain metastasis 3.0, which is the newest version in this case. The vendor, obviously, vary on HyperArc, the Brainlab, we heard for multipole brain metastasis, which is already, which is the biggest difference might be that we use VMAT for HyperArch, so an IMRT approach, and Dynamic conformal arcs for multiple brain metastases. And what might be also very interesting is the way the geometry is handled. So in HyperArc, it's fixed. You see here a little screenshot of the HyperArch process. And you see the beam arrangement or the table arrangement. Table angle arrangement is fixed. You can only decide to skip table angles. But you cannot change the position of the table angles. So, you got fixed table angle position. On the other side, you have the Arc optimization within the Brainlab solution So, you can see here what happens. The system starts with an Arc setup. And then, in the moment you start the optimization process, the first approach will be to optimize the Arc arrangements to come out with the best solution, best number of arcs, and best orientation of all the table angles to get you the best possible dose distribution to the lesions and to the OR sparing, in this case.
The degree of automation is very high in the multiple brain metastases. So, it's pretty forward. You push the button. And then everything's happening automatically more or less. In HyperArc, it's a little bit limited because you've got the IMRT optimizer, there again, the [inaudible 01:16:04] optimizer behind that., the VMAT optimizer. And you have to fool around with the numbers there with the constraints and so on. And you see, this is also not very automatically. So, it's quite limited.
Time for planning is quite high for HyperArc, especially the beam calculations, especially when you do not use GPU calculation in this situation, which I would really recommend to keep this at a nice time. For multiple brain metastases, the time is quite low, especially for calculations a dose distribution then because if you use Dynamic conformal arcs, you do need as much control points doing the calculation as you need for IMRT because the leads are just moving on the linear way. So, it's good to know where the leads are in this situation and where the leads are in this situation. And in between there, they are just doing a linear trajectory. So, you do not need to have control points every five degrees or so.
This is just a little comparison between that and then maybe some other differences, which might be interesting. And this is the first one, just for our physicists out there. Multiple brain metastases or all of the Elements come up with a Smartbrush solution, which allows you to contour the lesion quite quickly. We are seeing that already here. So, it's easy enough. You just have to draw the contour in the actual view and maybe in the coronal view or the sagittal view. And you can see that in the movie already running here. And the system will automatically create the 3D model and a 3D contour of your lesion, which is quite nice if you deal with patients with 10 lesions or so. And you do not have to do that 10 times in millimeter distances, slice thicknesses, or every time again, a 2 contour or so. This helps you to speed up the process a lot. So, this is quite nice.
Another thing, which I really want to point out and focus on a little bit, is that within the Elements solution, you've got an automatic distortion correction for MR images. So, MR images are distorted always. So, it's just because the magnetic field is only optimal when there's no patient within. But, obviously, there is a patient. And he distorts the magnetic field and so the image. All scanners, available scanners, come with an automatic distortion correction within the system. But this is just for the distortion correction of the scanner itself. It doesn't take care of the distortion happening due to the patient within the field. And if you look at this here, you can see the distorted and the undistorted image. And if you look at this lesion here, you see that there is quite some movement in this situation.
There's another movement here. There's movement over there. So, you see, there's quite a lot of movement compared to the CT here that's also shown there. And when you think about treating multiple lesions and you position in the isocenter in the middle here, and you do not do a correction for the distortion, you might end up with a position of the lesion that's not where it exactly is within your planning process. So, you have to correct these in some way. One option might be adding a larger margin but a larger margin will end up in the larger PT volume, will end up in a larger V12, as well. And the V12 is something that we want to keep as low as possible just to the risk of radionecrosis. So, this is a nice feature. This can be used for cranial SRS as well.
So, now, just some numbers for a comparison between HyperArc, multiple brain metastases 3.0 and the 2.0 version. Maybe some of you knew that. Some of you have that maybe installed. Just to give you an idea of what comes up if you move to the new version here. So, the first slide shows you just 15 patients picked randomly out of our cohort of patients that we treated since the beginning. And we did plans for HyperArc. We did plans for MBM 3.0 and for 2.0. And all these patients are treated with 3.0 just to keep that clear. So, this is not just some [inaudible 01:20:36] numbers. These are really actual patients. You can see the number of lesions as well here, just to