 Good morning. Okay, I will try to speak a little bit about VUMAT, Volumetric Modulated Archaeotherapy. Why a little bit? Because I think that today, the scientific community are discussing a lot about commissioning and quality assurance on VUMAT. Because there are a lot of open problems regarding VUMAT commissioning and quality assurance. So the aim, the purpose of my presentation is to give you some idea of what is VUMAT and some reference about commissioning or quality assurance. Even if, to be honest, in my institution, we are not working today with the reference of TG, the guidelines that are usually used with VUMAT. Because our experience is that it's necessary to work not only with the machine, and usually we are too much interested to the quality assurance of the machine. But it's important to get the line and the water we have as a reference, just to have a point, a starting point. Okay, VUMAT is a dynamic treatment modality. A treatment modality that is usually defined as a complex modality, because during the delivery we have a gantry, the gantry is moving with MLC, and we have even a dose rate modulation. So because of three parameters that are synchronized during the treatment, our idea is it's very complex, even related to the quality assurance. These three parameters, the synchronization of these three parameters are necessary to reach a good plan in terms of conformity index, in terms of homogeneity, and in terms of dose sparing or that risk. In my mind, it's important to know the different steps that were in the road to the VUMAT. So, sorry, it's easier for me. The first idea that we have on VUMAT is the conformal archetype. A dynamic archetype that means that the gantry is moving around the patient during the treatment, and the field match the target thanks to the beam eye view. No modulation is inside this treatment, and the dose rate is constant. This is the characteristic of conformal therapy. But conformal therapy is able to reach a good conformity index during the treatment with only a conformation, thanks only to the MLC and the gantry rotation. In 9823, Chin showed that if we work with the gantry rotation, collimator motion, that means MLC, but even collimator rotation, and we work with dose rate variation that at that time means different weights, it's possible to hardly improve conformal dose distribution to this, from this. The first step was brain, working on the idea that we can change the weight of the field, that means in certain things, we are modulating the dose, rotate paper, the first paper about the Fluence Intensity Modulation Concept. If we are able, but you know I think about IMRT, if we are able to modulate the Fluence inside the field, we are able to improve our homogeneity and conformity index. The first example is the Tomo Therapy. Tomo Therapy is a machine, it was the first machine, we have the first machine in 1993, and it is the first example of the therapy in which we have a gantry that move around the patient. In this case, we have a binary collimator, but thanks to the binary collimator, it's possible to modulate the Fluence. Tomo Therapy has a problem, the treatment had a problem, maybe that today is not, but the treatment time is very long because the coach moved inside the board during the treatment, but at the same time, Tomo Therapy was a good starting point to improve IMRT. IMRT was the first idea that not only we need to conform, to move the gantry around the patient, not only we need to conform the field using the beam eye view, but we need to modulate even the dose during the rotation of the gantry. Using the Lenac machine, the Lenac we can use, oh, sorry, we can use the conbin, and thanks to the conbin, we can have a time of delivery very optimized, very short respect to the Tomo Therapy. I would like to show you the difference, and this is the first example. If we compare 3D CRT with IMAT and IMRT, we see that the idea of IMAT, so the idea that we have many gantry but many gantry, oh, sorry, the gantry rotating around the PTV. Conformed to the PTV, the beam eye view, and with different weight, we see that it's not possible, it was not possible the first time to obtain the same homogeneity that we have with the starting fix that IMRT treatment. Why? We have, we are using, working if we change the sides of the collimator, and we don't use the beam eye view, we see that it's possible to move to a homogeneity distribution, conformed to one, and we see the dose over the target is more, the deviation is given less, and at the same time, the dose to the region at risk, we are sparing the organ at risk. And again, if we move with three hungers, seven hungers, 11 hungers, 33 hungers, so at the same time, we reduce the dis homogeneity inside the target, and it is important to observe that the integrated dose, it's the same. This data, it's very important because in our mind, often we think that because the vomit or the conformal arc, as the gantry that rotates around the patient, the dose, the integrated dose may be higher than for IMRT. This is not true, and this is the demonstration. So starting from this experience, dosimatic experience by Shepard, said that Q observed that IMRT means N field, so what, M intensity, intensity level. Probably the idea is that the quality of the plan is related to the multiplication of N and M. So we can reduce N and enlarge M, or we can reduce M and enlarge N, so the idea is that increasing the number of gantry angle, we can reduce the number of intensity level. Why we needed to move on this direction? Because in vomit, we have no intensity level differences. So the idea is that if with stepping shoots, seven field IMRT give us a good plan quality that means that we are able to reach our dose constraints, PTV coverage, we are able to reduce the door to organize the risk of rectum, parothid, and so on. So the same would be possible with 68 gantry angle, with no modulation, because the factor is equal. It's only an idea, but this is the idea of intensity modulated archetype, and it's easy to understand with this picture. Static gantry IMRT means that we have different entry angles with a field that is, it's modulated. So we have a different fluency level. In vomit, we have a lot of gantry angle. It's a different form system to system. And for every gantry angle, we have a field. Every field is not modulated, but it's different weighted in which way, changing the dose rate of the system. And the superposition of every field give us, if you look, for example, to this gantry angle, you see that in that point, we have a three level of modulation. But just behind, we have only two level of modulation. So we are able just with small field, very near, but not modulated to obtain the same modulation that we have with a fixed gantry angle. Just one, no, sorry. The question could be, is it true? Are we able to obtain with this method the same quality of plan that we are able to obtain with the IMRT? Probably up to now, we are not able to answer to this question. There are plenty of paper and the way we do it. We will see after trying to compare IMRT, Vimat with only one arc, Vimat with two arcs, with more arcs, how many arcs we needed to reach a good plan. But to have an answer is very difficult. And this is the question, it is the reason that make it difficult even to choose a right, we will see method for quality assurance. We don't have up to now a metrics to measure, to compare plan. In my center, we use the coverage of the PTV and some data, contact data, for example, on organ at risk, but in other center, they use another metrics. So when we look at all these people, we have no answer. But probably we must be confident that the plan must be good enough to guarantee a good coverage, to guarantee low dose at organ at risk. And probably with Vimat, we are optimizing something else and so could be important move from IMRT to Vimat. And what are we improving? Okay, IMRT, for example, if we look at this paper, IMRT is largely better, is superior for better PTV coverage. Only, I would like only point out that probably tomotherapy is better that Vimat. Up to now, we can say that probably this is true. And a lot of article supports this idea. The same is not for IMRT. So it's better, IMRT seems to be a little better, but here Vimat is a little bit superior to IMRT. What is clearly true is that if we look at the MU per fraction, we have the MU reduced to a third, respect to the MU of IMRT. So if we look at the time of delivery, we have the time reduced from five minutes to one minute, we say for one hour, two minutes for a treatment, yes. No, not so true. My experience, yes, Vimat treatment is around two minutes. Being on time, it's around two minutes. So if you have two arcs, it's one minute for arc with two gray per fraction. But the, yes, but the wall time, the beam on time, yes, only beam on time. The beam on time is around two minutes, not less. Because even if we work with the maximum dose rate set in your machine, the system, because you need to guarantee the accuracy of the MLC, but we will see later, works usually with less dose rate. And so the time is longer than one minute. But it's one of the reasons that it's not easy to understand if Vimat is better than here. Regarding time, it's easier, but about the distribution, we are not able to decide. But it's two. I can spend only experience with Varian machine. And with Varian machine, if you have a two gray per fraction or 2.2 gray per fraction, if your treatment is a simultaneous integrated boost, is not too far from two millimeters. Even in head and neck, it's around two millimeter, two minutes, in my experience. Eugenia? You are seeing that at the neck, three opportunities, it's natural, so maybe there's a different between the modulation of the angular speed and the dose rate. With Varian machine, it's not too easy to stress the modulation. So the time is quite fixed, in my opinion. Varian, the dose rate is changing, but with electric power, it's just the steps. Like we start with 600 and then three. Not now, not now. Now it was, but with the new machine, it's not a two. They are able with the same, yes. Yes, it's not a two. It was a two, but now it is not. But in any case, it's a two that, okay, we are discussing about two minutes, three minutes, but when we speak about three minutes with Vumat, probably we have seven, eight minutes be more on with Armourty. So in any case, we are sparing time. And the sparing time doesn't mean that we can treat, yes, means that we can treat more patients. But the important thing is that we can treat a patient in a very short time so we can be more confident about movement of the patient during the treatment. So probably we are reducing residual errors that we know even after a verification of the patient's top with can be an important vision or brain lab system as you want. We have all the time residual errors, but if the time of treatment is very short, probably these residual errors is smaller. Okay, so this is very important. Which system we have to plan with Vumat? Different TPS, Variant Eclipse, the system is called Rapid Arc, Phillips Pinnacle, Smart Arc, Electa, with Monaco or Oncenta, Massaplan that is working, yes. And the right search with Vumat module. All the system works in a different way. The same as with IRMRT, because even with IRMRT, the optimization model, we have a different optimization module, a module that use a leaf sequence inside the optimization and so during the optimization, you take into account all the mechanical limitation of your delivery system, or you can have software optimizer then works with two steps. The first one, optimizing the fluence, that in that case means something else, we will see. And after the optimization of the fluence, we will introduce the mechanical limitation of the accelerator. Mechanical limitation that with Vumat are very strong because with IRMRT, we need to set the maximum leaf speed. With Vumat, we need to set the maximum leaf speed that must be, must work together with the maximum acceleration of the gantry, the maximum speed of the gantry and at the same time, the maximum dose rate or minimum dose rate that we can use. Usually, every system try to work with the gantry rotation working at the same speed. Why? Because it's very difficult to accelerate or decelerate the gantry because of its weight. And this could be introduced an error on the gantry position. And this error can create inaccuracy on the delivery dose. So even variant and electa and at the same time every TPS that we have seen before in the list must work trying to set, to fix the speed of the gantry. So we can work on, we will work at the first time on the M and C speed and dose rate modulation. Pinnacle, okay, probably we can go on. This is just, I would like to show, to compare, but it's not important, the two modalities, Pinnacle, master plan and the right search that try to work in two steps. As we said before, at the first time we say they fix the arc, fix the limitation of the gantry and then they optimize a fluence for a fixed number of the gantry. And from this they split act, splitting the fluence in a different gantry angle, but in the second step. At the second step they introduced the limitation of the delivery machine. Eclipse doesn't work in this way. It's not possible to set in Eclipse a rapid arc. The number of the gantry angle, it's set by the vendor equal around two degree, but there is at least two papers that shows that is important to work with a small gantry, because when you would like, you must, in a commissioning, in a quality assurance, so you have to compare the calculated dose with the delivered dose. If you calculate with the gantry that is not so fine in degree, but for example, you use four degree or eight degree as it's possible with some TPS. At the end, we will see a picture later. We will have a difference between the delivered dose at the calculated one, because the calculated is very different from the delivered one. We don't take into account all the gantry delivered, okay? So we can say that Eclipse has fixed the gantry angle to the minimum up to now possible. The optimization at the first time takes into account the limitation of the MLC and of the speed of the gantry. To conclude, if we needed to move to Vumat, as I told before, we can't say if Vumat is better than IMRT from the dosimatic point of view, but we are sure that it's better in term of time. So what we would like to know now, if it's better or it's equal to IMRT in term of the accuracy of the dose delivered to the patient. And we need to move to the commissioning. Commissioning, what does it mean? To be sure that the calculated dose matches with the delivered one. Problem, dose rate, variable dose rate, variable gantry speed, dynamic MLC movement. We are very confident with dynamic MLC movement. We have a lot of experience about the accuracy of the MLC position with IMRT, but we don't know nothing, nothing, but we are knowing something about the synchronization of the gantry and the MLC movement at the same time, variable dose rate. So we need to think some test able to give us the information if this synchronization is accurate, is enough accurate to guarantee the dosage of the patient. But at the same time, we don't need to work a lot. To do a quality assurance is very expensive in term of time. So we need to understand what it is necessary, probably it's necessary, a good method of commissioning, but we must be relaxed about equality assurance. It's very old and it's an article by Ling. I don't remember, but 2008, it's true, 2008. But it's probably the first and the only paper that we can take as an example to do the commissioning and even to catch test for the quality assurance. Ling has a big story with IMRT and so know that we need to check the MLC position. We are able to do with the fancy test strip where you check the position, the accuracy of the position of the accuracy of the gap of the MLC. We are able to do this with IMRT static field, but we need to do the same with the Gantt movement. Look that these tests are done with film. It's important this aspect. We can do the same test with portal vision. It's easier to do with portal vision, but we must be sure that our portal vision doesn't move because this test not only guarantee us that the position of the MLC is correct, is accurate, but is an example of synchronization of MLC movement and Gantt position. If during the movement the PV is not stable, we will have a wrong information. So please, before you need to check the stability of your portal vision, then we need to check the ability to vary does rate again to speed at the same time. Again, we make a copy of the same test that we thought with IMRT. We need to be sure that in that case it was the MLC, in that case is the Gantt speed. We will change the Gantt speed and the does rate for every strip with the idea that the does delivery for every strip must be the same. And to check, we will go to compare with a uniform field. And if the superposition is good enough, it means that our system is working well, good enough. What doesn't mean good enough? We need to decide a matrix. Matrix for commissioning. TG 119 suggests 3% of three millimeter for quality assurance for patient. I don't remember if suggest some criteria on a gamma matrix for commissioning. Now it is clear that 3% of three millimeter as gamma matrix, you know gamma matrix. Okay. The three millimeters, three centimeters is now clear that is not good to solve some errors. But even, but this is a very open problem. Is not good when you are sure that you are using a method to measure the error that is able to see less than two millimeter, two millimeter, two percent. What I mean, probably I will not use gaff comic. I will not able to solve 2% dosimatic differences. I will be able with portal vision. So 3% of three millimeters is probably too large, but is accurate for a large number of center that are using different system to check the accuracy of our system. So what I would like to point out is that when you select a criteria for commissioning, the first question is which system I am using to test my delivery system or my TPS. This is the first question. Going on, it's possible to have this test for rapid art is possible to catch all the tests for the delivery machine from my variant website. And these tests are the same that are implemented in Epica that is a student very useful for quality assurance of the machine and for commissioning. In my mind, it's not useful for patient query. Another article, but I have no time. I can't spend more time on this paper even if I think that there is a good paper to understand a method for commissioning. And the important aspect of the paper by Anvanesh is that Anvanesh has clearly in mind that we need to test delivery machine, TPS, and patient. Delivery machine link does a lot of tests so we can catch tests from link paper or from variant, if we have a variant from my variant website. But TPS is the important aspect. We need to check to be sure that the TPS is rightly commissioned for the Vomit system. Why? Sorry, because this is an example. Probably you are not able to see the field. This is the open field. So we are giving those only in this area. So we have small field, a large coordinator opening, small field of access, and a lot of my MSC, very small field, very near. Usually when we make commissioning for the TPS we don't check this kind of field. So it's important to move on this kind of test. And Anvanesh has done a big offer to depict some tests, small field, on axis of axis, tongue and groove, to be sure that our system is able to depict tongue and groove effect. Gantry test in which we test the delivered dose from a small sweeping gap with gantry rotation on axis and off axis. So in my opinion, this is the important point and all of us must repeat this kind of test. We have done in our institution and I know it's not a good thing, but we had to repeat a lot of measurement just to rightly commission our TPS for Vomit. This is the picture I told before you that if we plan with gantry angle, two degree gantry angle, four degree gantry angle or eight degree gantry angle, there is paper by Masi. We have a difference from the calculated and the delivered one. And this is because we are calculated in different way that the delivery and system. At the end, we must check the plan, plan patient plan, plan it with the delivered one. Thanks to a test plan verification, good for commissioning. And again, we, we must check the calculated dose with the measured dose. How many system we have? A lot, 2D system, circular system, 3D system, maybe? 3D system, maybe? We are not sure that is a 3D system, but could be. So, we check, we compare, calculated with measure. We need to do this. How many patient, how many test patient, five, 10, 11, but the important things is the criteria. Gamma, gamma index, two millimeters. I repeat the same thing, two millimeters, two percent, three millimeters, two percent, 90 percent, 95 percent experience. Our check, we are using our check. Our check, look. In a TG 142, we need to check the laser localization with a tolerance of 1.5 millimeters. Our check with our check because your dosimeter is in a circular surface is not possible to correct for translation. So, what we see with our check is only the gain rotation, an error in rotation. We are not able to see an error in translation, X translation, lateral translation. But this means that one gain to your rotation, one rotational error is equal to millimeter on surface phantom. So, two millimeters, two percent is not a good enough for our check because we will see our setup error at not the delivery error is our check around the dosimeter, I don't know. At the same time, we needed to know that, for example, the system in the last version, not now, the new version, they correct this problem, but the first version of the our check has bigger error with field bigger than 15 centimeter. We haven't 15 centimeter with IMAT or VUMAT, but we have a lot of field outside of axis, so in a range of 15 centimeter. So, we need to pay attention about our dosimeter. And just the last thing, so on the dosimeter. This data measured from different dosimeter of error. Hussein did a very interesting work, intentionally introduced the MSE error from 0.5 to two millimeters. This is measurement results of the error. As you can see, different dosimeter have different mean value of gamma and minimum because we have from 0.5 to two millimeters, but when we use the software with the dosimeter, we are not able to find a statistical differences from this dosimeter. So, the software implemented by the vendor is seems able to correct for dosimeter resolution, characteristic of the dosimeter, so on. It seems. I would like only to say this one. Is it true that we need to spend a lot of time on patient quality assurance at the machine? Is it true that we are not confident on the delivery system? Many papers now are telling us that we must be confident on the Linux delivery. We are not confident on the TPS. So, we need quality assurance and we need to make measurements on patient to test the TPS, not the Linux, because rapid dark is not more critical that the IMRT, and we are confident with the IMRT. We know the error in IMRT. You see, we have no differences with post-8 and a little bit with add-and-neck. More with IMRT if we introduce error, shift error in position of the MSC. And at the same time, the system is not sensible to random errors and we have a lot of random errors, but we haven't systematic errors. We needed to check our system. We needed to be sure that is as a long term stability, but we must relax about some error. Two millimeters random errors doesn't matter. There are a lot of articles that tell us this. And at the end, because I have no time, this has the data from the Varian Linux. What we see here is that we have no errors about Gantry during the rotation. This has the data related to 2,0600 delivery. And you see that the error in Gantry angle is quite randomly distributed around zero, not for all the arcs. We have a little bit of shift, but 0.2 degree doesn't matter. In the same way, we have no errors. Okay, we have errors up to one millimeters in gap width, but it's randomly distributed. So, Santa, yes, it's not possible in it. Because when you check one millimeter, when you check differences in those around one percent, it's not possible to use any kind of instrument. So log file, these are data from the log file. You must be confident on the log file. Yes, but about this, I think a lot of paper. The only problem is that you check the system only 50 millisecond. And so you are not able with 600 in our machine, the MU, the maximum doser rate is 600. So you don't have many point of measurements during the treatment. You have many points if you work with 100 MU per unit. But in any case, the long-sterm stability, here you can see only the error on the MSC position is quite good. And this is more than one year of check on our machine. So just to conclude, I think that VU-MAT is an important technique that we need to implement in our institution. I don't think that is need a big offer more than we have done for the MRT. So if we are confident with MRT, we can be confident with VU-MAT. We need to do quality assurance on the patient treatment, not to check the machine. But to check our TPS so properly, we need to reduce measurement of the machine because we have no dosimeter able to tell us nothing but we need to work more about TPS commissioning and about long-term stability to check this with a good frequency. This is my personal opinion. Thank you.