 So dear, good morning dear participants of this course. It's nine o'clock, or past nine o'clock. And I would like to start with this session which is on imaging in radiotherapy and patient data acquisition. Let's start with the first picture which I think is quite nice. It shows an idealistic picture of treatment. And what I would like to say is by far not reality. Because of what? You see the beam. And I think this is something which I was told some years ago and I think this is a very nice understanding of the situation. The problem is we do not see the tumor or the target. We do not see the beam. But we want to match exactly the beam onto the target. So this is something, this is a challenge. And of course to get information on the tumor or the target, the clinical target volume, you need pictures imaging. And this is now the topic of this course. I have used again material from this book and I have also used some material from a book which is issued from in our institute. It is 3D conformal radiation therapy and it is a disk-based book so you can get the information out from a disk. So this is what I want to deal with. The need for and the types of patient data, segmentation methods, image registration, display of retches that images sequence as image fusion. I also want to touch some words on patient treatment position immobilization devices. It is not directly imaging but it has to do with the problem, to match the beam to the tumor or otherwise to match the target to the beam. I will spend words on conventional treatment simulation, computed tomography-based simulation, conventional simulator versus CT simulator and also want to say some words on magnetic resonance imaging for treatment planning but not so much. Okay, let's start with the discussion on the need for patient data. Within treatment simulation and calculation process, the patient anatomy and tumor targets have to be represented by a model. Again, we don't have the patient itself. If we prepare, if we do the calculation, if we do treatment planning, we need a model for the patient and nowadays such a model is normally a three-dimensional model. But I may ask because here are many people coming from countries where two-dimensional treatment planning is still, as far as I know, having seen the situation in Indonesia in one hospital, it's quite a mixture. It's sometimes, it's a transition from two-dimensional to three-dimensional treatment planning, which, a situation which is not easy. So this is an example, take out a book. This is an example of a 3D model. We have here the clinical target volume in violet. It's quite within the two, this is a long situation. Then we have the both lungs here as the organ at risk. And we also have the spinal cord here. And this is the outer contour of the patient. So such things are quite nice to see and can be nicely used for treatment planning. But I want to go now on some more general considerations on patient data. Patient dimensions are always required for treatment in order to get the monitor units or especially to get the monitor units. And the amount of required data depends on the treatment planning method which is used, two-dimensional, three-dimensional, the dose calculation method. May I ask you, how many of you are using now modern treatment planning systems? I think it is a majority now, yeah, I think so, yeah. So it is not so easy anymore to understand what's done in this calculation and those calculation methods. I will go in the next lecture, I will touch some of these ideas because in any case, the calculation of those is always an approximation. There is some advantages, disadvantages and I think it is useful to know a little bit about what may go wrong or what are the difficulties if one is doing treatment planning. In my eyes, and I think you agree with that, there's really a risk now we can do so many things on a computer, it's like playing a game. We can rotate the patient, we can rotate, we can do everything. We see pictures, but it's not by far the reality. It has nothing to do with that but I think it's a quite nice thing. I got some pain in my back and I got an MR image of the spinal cord and what is seen, the very strange thing that there was this, how you call it, the inner part of the spinal cord, yeah. It shows some structure like it was split in a wing, very strange. I never know that. The thing is if I do the picture, I could be what's going wrong with me. But I feel nothing, I know nothing and for me it's now very obvious. Picture is one thing and the reality and truth is the other thing. I think this is a very general consideration of imaging. Nevertheless, here I was discussing the amount of required patient data and we need as patient data, not directly for treatment planning but for the treatment positioning, also other data such as landmarks, anatomical or artificial landmarks or other items. Nowadays it's quite modern style to take into account the breathing of the patient and to do a tracking or other things, a gating, is someone doing this here? The gating of breathing, no. You do it now. Again, it may work, it may work but the problem with that, let me just tell you some things about that, is that it takes some time from the signal to the response of the accelerator that and this time maybe typical, say, 0.2 seconds or something like that. It is fast but in that time the moving goes on. The problem is it should be prospective, it should say what is in a few seconds, in a few tenth of seconds, what is the position? Which works well for healthy people but it is always a risk for people who have a real lung disease, it may go wrong, something like that may happen and then suddenly it's very different. So, all these techniques who are trying to follow the breathing circle are working very well with healthy patients but may have problems. So, we still have to introduce and develop techniques who are very, very fast to really respond immediately to everything what's happening. So, patient information required for treatment thing varies from rudimentary to very complex data question. So, one can use just distance red on the skin, one can use manual determination of contours, acquisition of CT information over the large volume, you can introduce image fusion which is all the use core registration using various imaging with all the data such as CT, MR, and PET. And also other advanced method of IGRT and my friend Pavel Kokolovic will tell you a little more on the use of IJD technique which has an increasing role in this treatment of cancer. I personally think that the progress in radiotherapy, one should say before we had continuously a very good progress in radiotherapy. If one asks my neighbors or my family and the reputation of radiotherapy is normally not so good. It's seen as a palliative method very often but we know if we have to figure that radiotherapy is really successful in many things, in many situations. So, it becomes more and more successful but in order to have more success, more progress, I think the Q will be the better imaging. And I think IGRT and even what is now coming on the market in combination with imaging on directly within the accelerator system. This is a code B and CT what we have but even now the MR systems are developed directly together with the accelerator. I think, I believe this will be a major step forward but then you for the first time you can really see much more better the target, the humor because due to the soft tissue contract which is offered by MR systems. So, this is a typical old fashioned way of getting contours which is I think not done anymore. Yes, I think in the last 10, 20 years the development in any country has dramatically changed the duration but maybe if you look in the textbook of the agency such pictures are still to be seen there. So now it's over. This is still done imaging using a simulator, using a kilovoltage imaging. This can be taken for comparison with port films during treatment but I think probably, could look at yesterday made the point which is quite interesting. Use of a simulator was quite distributed everywhere in a time where it was too expensive to have a special CT for simulation of a regulatory program. And it has as he told us a small advantage that you have to position your patient a second time. And there may be something ever introduced and it's nice to see. I've never heard this argument but I think it's very good argument. If you use the same data which you use for treatment planning and for simulation it is there is no difference between. Of course, there's again a difference if you really then position the patient but only one influence of mistake. Radiographs are particularly important for a regular field. This is an example where this is a prostate. I think it's a prostate there which should be buried with a block field so you can directly draw the field side and the blocks onto such a radiograph and then help to construct these blocking systems. Again, I would like to ask block field is maybe also old fashioned now. Who's still doing blocked field? Yeah. Oh, not so old fashioned. Yeah, it's still done. So you don't have a multi-leaf collimator. Okay. Yeah, it is. Yeah. Yeah, it's still good. There may be even a risk sometimes. I have seen the risk which is a multi-leaf collimator because it cannot exactly fit it. So the block feed can be shaped much more accurate to what you want to have. And especially if you have some multi-leaf slices which is a 0.5 centimeter or even one centimeter and you have problems in the lower end and the higher end. And I have seen that people are doing mistakes with that. They think it's quite nice but it was too much blocked. So I think it needs a lot of practical training to really understand how to work with the multi-leaf collimator. You can buy all this equipment now if you are buying new equipment from Varian or Electra. There's only two remaining companies. They will offer you of course a system with multi-leaf collimator. Everything is fine. You will offer you the treatment planning system. Everything's fine. And I've seen that people who are not so well trained do not really understand the problems involved. And it has all to do with the margins. One has to be cautious with the margins. Using a multi-leaf collimator. So suitable slice spacing is of course then the problem. If you construct a model of the patient you may have different slice spacing. This is a recommendation to use 0.5 to one centimeter torax, 0.5 for the pelvis and 0.3 for head or neck. The point is now that structures relevant for radiation treatment can now be identified on the CT slices. The following image processing procedures applied to anatomical structures are typical CT-based procedures. The one is which is called segmentation. You can say that's only drawing the contours but the name, the scientific name segmentation. The other name is the process of matching images obtained from different imaging devices called the registration. Or if it's all the used core registration. Segmentation process in particular refers to the well-known ICU volume that have been defined as principle volumes for treatment planning. This is of course tumor gross volume, the clinical target volume, the planning target volume. The gross tumor volume is important because for the purpose of diagnosis of the stage, the GTV is the most important indicator for measuring tumor remission and therefore for measuring therapy success. Again, it suddenly comes to me when I was visiting this hospital. It's so important to have an information of the success that requires patient follow-up. And I know that I can really easily imagine that that is something which is again not easy to organize. If it's a huge country and people are coming from one corner of the country for treatment planning and then they get the treatment and then they go back far away and they never will come back. So how treatment follow-up can be done with that and how the information on success or failure can be get. This is something which I think is very, very important for the self-education for oncologists. And I have seen that in the hospital that it has not been organized because it's so difficult to do. I think it is one of the key things to really to get across with this point. You must see what if we've done something wrong or not. The GTV presents that volume which has to be irradiated. The clinical target volume and the planning target volume of course are also important for segmentation. There is one thing which may not so clear and even in my home institute in Heidelberg it's not so clear. Of course, these two volumes, the gross tumor volume and the clinical target volume are poor anatomical volumes which has nothing to do with the technique. It's a patient, it's a disease, nothing else. So it has to be done by the doctors who knows the disease, who knows the spread. Though they know the gross volume and they have an idea of what may be involved. Very often the doctors came up and they, I will give you the planning target volume. And they tell you, this is a planning target volume and not the clinical target volume. We have debates on that and I know there are courses in Europe and Estor to carefully educate people, the doctors especially, they should focus on the clinical target volume and they should leave the discussion on what is the planning target volume to the people who are known, who knows very well the advantage, disadvantage of the accelerator. Another point is, you can make a ratio of the volume, the clinical target or planning target volume, clinical target volume and by getting this number you can get information how well the planning is done over several patients. So if the doctors are giving you the planning target volume, you are missing this information. This is one example of segmentation with a brain tumor, this is the, and here it's already the planning target volume. I'm not sure whether this is the planning target volume. I would assume that the doctors have given here the clinical target volume. This is the brain stem, the head contour, the eyes, the optic nerves, the chiasm and so on. So a typical result of contouring of segmentation. And this again is what can be done with such segmentation if you use the series of slices to get a 3D model of the patient. All segmentation algorithms can be divided into groups. One is region-based approaches. Region-based approaches try to find an area of people with similar properties, similar gray values, similar Hounsfield units and the border between the volume of interest and background is thus defined by a cut-off value. This cut-off value may be determined by an algorithm or maybe introduced by the user itself. And again, I want to show this picture in more detail where we have here. And now I call this clinical target volume. And what you can see is that obviously such an automatic process will not run at all because here to see this as a clinical target volume, this requires a knowledge of the doctor, nothing else. We as medical physicists, we can never do that. We should not do that. However, and I think this is taken from John and I forgot the name before I found. This is a situation if different people are drawing the clinical target volume. Here are two cases, the brain. This is a result of eight radiation oncologists, two rated diagnostics and two new researches. So they get quite different shapes of the clinical target volume. And you can tell, you can use this as saying, oh, not well educated, but this is not true. It is extremely difficult to find out or to know what the clinical target volume. And again, it needs a lot of experience to identify the clinical target volume. There are another method for segmentation measures. This is the edge detection algorithm. Yesterday, Pablo Kogoloch was saying that our eyes is especially sensitive for seeing edges. And I want to tell him already, it is a nervous cell, the cell itself which behind our retina, this cells always has synapses or going to other nerves. They are reducing the signal or the increases of these nerves. And if you have a homogeneous picture, these are canceled out. So the two neighboring cells or neighboring nervous cells are reduced and at the same time it's coming back. It is increased so they are balancing out. If you have an edge, then there is really an increase of signal in our cells which is a nervous cell which goes in the brain. So this happens already just in the cells here. This is edge detection, which is, you can quite a huge story on that. What we really see is calculated. So it's one thing. What we see is a calculation in our brain. We don't see that the rules. We have calculated in a way and what is quite difficult to understand, each individual do this. Obviously it does the same calculation which each computer is working with the same system which is very strange. This is an example for segmentation for the edge detection algorithm. This is the original picture. This is an edge method. And by defining a cut of value of height, of the height of the parameter change the number of edges found is increased or decreased. So advantage and disadvantage of segmentation method or of course, menu segmentation, the speed is very low, but it's easy to do. Same automatic segmentation may be good in every aspect and fully automatic segmentation seems to be very good in the speed or the probability but it's of course, it's not everywhere available and from my point of view, one should be cautious with this automatic process. Always it needs a control of the well-educated radiocoologist or medical physicist. Now I come to image registration. Modern three-dimensional treatment planning is based on tomographic images of different modalities. So we have x-ray computer macro-free. We have MR images, display soft tissue with considerably better contrast with which allow a more precise differentiation of tissue and we also have now introduced more and more the positron emission tomography. Again, I would like to ask, is it introduced? It is expensive, it's expensive. The pet, no, the pet scanner. It can be used together as a CT pet is one device where you directly can compare and I will show one picture of that. How many institutions have introduced that already? You have access, yeah? My question is, are you using the information of the pet for treatment planning? For fusion, yeah, yeah. But do you put the decision on what to do really on the information on the pet? Sometimes. So with a pet you can do much more things, you can function imaging, you can see because this is biological-based imaging. It's not the imaging of structures. That can be used, yeah? But again, the point is, what you really see with a pet? You see things, like, yeah, you see nice pictures and you can see this is more bright, yeah, and you can contour it, but what does it mean? And again, I refer to my backbone where I see something which I don't know what it is and obviously it's nothing to do with my healthy status. You can see a lot of things with pet, like perfusion, blood supply, or you can, a lot of things. And I know from some projects to introduce this knowledge more and more really as a decisive for the decision of what to do with it and where is the target volume. But still I think it needs even more animal studies for that, what we should have, we should have a comparison of the picture of images with the real anatomical or functional process. And this is a story which still needs a lot of development and research. Again, I personally think I expect a lot of progress in the treatment of tumors by having more fundamental understanding what does the image say to be a pet image. So to be able to use several image modalities simultaneously it is necessary to establish a quantitative relation between the pictures, elements of different images. So the mathematical methods are able to calculate and establish these relations, these are called registration. And this is one example though we have here, theoretical example of course, we have a box which should be going to this, it's two images and we have to do a lot of translation, a rotation, another rotation, again a rotation then we can match these two images together. So it's a typical mathematical process to do that. This is another example though we have here, one, this is a surface picture and this is got from a surface sensitive imaging and they should be matched and it shows how well this can be matched, this is the result of such a matching. Imaging registration can be considered as sort of ticket to the calculation only of the transformation necessary to superimpose information from one image to another one. But you also want to see how well this, so we need a method to display the result of our registration and this is normally called the fusion process so we have here two pictures and these are now the fusion process, we can see the difference between an MR image and a CT image. Now I will leave this and then come to treatment position and immobilization devices. So patient may require external immobilization device for the treatment of pentagon pump patient through the position or the precision required for beam delivery and this is one example on radio surgery which of course needs a very careful positioning and I have taken this picture because this picture is made in 1984. In our institute it shows our radio surgery approach which we have introduced in 1984. It was one of the first radio surgery treatments using an accelerator and from that time radio surgery was found to be quite successful and in many institutions I think also interest now. Radio surgery is a small part but it's a part of the radiotherapy program. In that time there was only one existing unit that was the gamma unit which is very precise and when we started using accelerator we have been always told you cannot do that and accelerator is not, cannot done at all. Nowadays it is, we know that it is not so difficult and what I also think which is that I am attaching that if you introduce a radio surgery program it is a very good occasion to organize a very good cooperation between the different disciplines between the radiocoologists, new research and medical physicists and technicians because you only can do if you really work together and therefore it is a good thing to have a good organization of the team of the radiotherapy team. Therefore I made sometimes a suggestion that the introduction of radio surgery is something which may help to improve the relation between the medical physicists and the doctor. I don't, I do not know your experience but I have my own experience and I know from many, many hospitals that medical physicists was at the beginning something very low level and the meaning and importance of doing good medical physics was, has not this reputation which it should have. In my eyes it should be on the same level of eyes between the medical physicists and the radiocoologists. This is not always the case and it takes time to develop it and just to say that it has nothing to do with imaging that's never. I found it, especially in our institute it was extremely good occasion to grow together in such a way that we are really crewlings. So to immunize the patient during treatment to provide reliable, has two fundamental roles to immobilize the patient to provide reliable means of reproducing the patient position from treatment planning and simulation to treatment and from other treatment to another. So these are some systems here which are from our institute has been developed in our institute and it's also from our institute we have such mask system which can be applied to the head but they are also applied to the total body. On the other hand, this is very time consuming and I think it cannot be recommended to do it everywhere but we had some very precise irrigation say in panel cord or even for prostate which was quite useful to introduce such a system. This is quite normal system which can be, this is the simplest way that you can use some head rest which can be used as another mask system which is commercially available. And this shows again this, yeah, let me just say that okay, this shows this radio surgery immobilization device where we had a ring which is really fixed on the table and it has sharp needles which goes into the brain and people normally say, oh God, there's something. But it's only a single treatment irradiation and it can be well tolerated with local anesthesia and it's extremely accurate. Here, this one, this is something it's commercially available. It may be used for if you broke your arm and then you can use, instead of, in Germany we say Gibbs so the old technique is Gibbs, yeah. And then you can buy that thing which is expensive but they are flexible and they will buy warming and that, I think, yeah. Expensive but well doable and we had implemented such a system with everything and it works very well, it was extremely accurate. But again, it needs personal for that that it can be recommended in general. It was a special technique which we have implemented in Heidelberg in our institute when we started with the radio surgery and then we found that fractionated radiotherapy, that not only in one fraction but in several fractions is a very good way also for tumors. The radio surgery is not so well suited for tumors. It's good for anterior venous malformation or for brain metastasis, it's wonderful. If you treat a metastasis, a liver metastasis, it will disappear in two weeks, it's strange. If it's not coming again, you can really cure a patient at least for several years. So now I come to the Convention Treatment Simulation process which has been introduced to have a better planning tool for the subsequent irradiation. This is an example for a very simple technique. It's called the double exposure technique. A film is irradiated with a treatment field first that is here and then the collimator are open with a wider setting and a second exposure is given to the film. So you can see some structures in the film and you can see how well the irradiated film matches with the structures. It needs again experience to use with that but it's a very simple technique. But I've seen that and I think it's a very valuable technique to use to convince yourself where the position is correct. Very simple. But of course it requires film and that is maybe another problem. We cannot get film anymore. So who are using film and development machines? Yeah, of course. I like this technique. I still like it and I was very disappointed that we cannot buy anymore such films like the therapy verification film which was an excellent film. It was produced by Kodak but you cannot get it anymore. Presently treatment simulation has a more expanded role in the treatment patients consistent of the donation of patient treatment position, target volumes on organs at risk and the donation verification of treatment field geometry. Generation of simulation is for each human being with treatment to produce such graphs. And this is again our check film and this is now a more modern radiograph with a conventional simulation machine, a simulator using KV radiation. Of course we have a much better contrast and what's much better imaging in KV imaging compared with a mega voltage imaging. Modern simulators provide the ability to mimic them many treatment geometries and this is one example. So we can hear, we have adjustable bars which can mimic the field sizes and they are made of tungsteners I think so you can really see them in the film and that was a technique which has been used in our institute still 20 years ago. Now again this is old fashioned but I still think it's a very good technique and such conventional KV simulators are still in use in many hospitals. Is that true? Yes. Yeah, yeah. So in the vast majority of sites the disease is not visible on similar radiographs therefore we need things like the bony structures or lead wire clinically placed on the surface of the patient or even surgically implanted fiducial markers. There is one system it's called the Calypso system is someone using it? It's a very nice thing, it's a small probe which can be virtually put into the tumor and it is sending a signal which can be detected and the position can be detected and can be used for, it's a very perfect fiducial marker but again it's very expensive and I know only a few centers like in America the big centers were using such a system and still it's not clear whether it can overcome the problem of what I taught at the beginning. We don't see the tumor, we don't see the beam and we have to measure. This is again an old technique which is still in use and quite nice so this is a radiograph and we can easily draw the blocking in its use of course as a matter of record but we can also use to determine the shielding directly here and they are drawn by the doctor and then they can be made of this information. Now we are using the modern CT simulators this is a picture, it's a dedicated radiotherapy CT simulator which are now generally available and these are pictures made of that and it's well known that the position of each slice and therefore the tag can be related to bone in the 20 months through the use of scout or pilot images. What is interesting is that using such CT simulators a virtual simulation can be done solely on CT information and one of these is that the CT data can be manipulated to render synthetic radiographs of the patient for arbitrary geometries and such radiographs are called digitally reconstructed radiographs it's well known and in many cases now they are used. They are produced mathematically by tracing rail lines from virtual source positions with the CT data of the patient to a virtual film plane and simulating saturation of X-rays and the advantages and anatomics which can be used directly for the donation of triple field parameters as a transfer error, patient position a second time, it can be awarded. This is an example of an idea and what is interesting that the gray levels, the brightness and contrast all the things now can be adjusted in such a way that it fits the needs to make good decisions on that. That's quite different from a KV simulator. Another thing which has been also quite often used in our insert was the beam eye views which are projected through the patient onto a virtual film plane which is perpendicular to the beam direction and I will show here some of the beam eye views here which are shown here. It's again the prostate irradiation with a regular field but from different directions which are really applied so we have not planar irradiation but from several sites in order to overcome the problem of a very irregular shape of the beam of the tumor. From each side we can easily see, construct such beam eye views and also then compare with associated port films. So this is a comparison of conventional simulator. This is an advantage useful to perform a fluoroscopic simulation in order to verify center position in feed limits as well as to mark the patient for treatment. I have seen such techniques of course is a little bit quite a dose to the patient. I think it's now the old fashioned and not done anymore. Disadvantages, limited soft tissue contrast with conventional simulator. The tumor is not visible requires knowledge of tumor position with respect to visible landmarks so it requires knowledge of a well educated radio oncologist restricted to setting film limits with respect to both landmarks or with structures with a little bit of a contrast. So a CT simulator, it has an increased soft tissue contrast. Anatomical, axi-anthomical information is available. Delineation of target and organs of this can be done directly on the CT slices with those typical reconstructed radiographs or beam eye views. There's also some disadvantages, limitations used for some treatment setups where patient motion effects are involved and requires additional training and qualification in 3D planning. So if I'm staying here and telling you things on this and this, it's fine. But the most important or much more important is the training which you should get somewhere. And I made one example with an exercise at Monday evening. I would like to do a much more exercise with that but which cannot be organized. But maybe in the future we can simulate on a computer such equipment, much easier. I know companies are now offering such software which can simulate. So my dream is to do this really in a computer lab. Would be good. So these are goals and tools in quenching and CT simulation, so treatment position. In conventional CT simulation, many things can only rely to bony landmarks. In CT simulation we have a better way to get out the information which is really needed. Now last few words on magnetic imaging, are imaging plays an increasingly more important role because of course MR is a soft tissue contrast which is much more better. And even such small leasons can be seen with greater ease. The MR-E cannot direct lose the serrated P simulation because the physical dimension of the MR-E device, equipment and accessories may limit the use of a modulation device in compromised treatment position. So it's a rather narrow hole that is not so easy to use. On the other hand we have also other devices which are open devices. Bone signal is absent and therefore digitally reconstructed radiographs cannot be generated. And there is no electron density information available for heterogeneity correction of those calculations. We know that for high energy photons, if you approximate the model of the patient just with water, in many, many cases it works quite well. If the lung is not involved, in many cases it does also work. So what you can do, you can use the picture of the image from MR and you can exchange the value of the pixels with that of water and then do treatment planning. And in many cases this would really also work. What is still maybe a problem that MR is prone to geomotorical artifacts and distortion. These, especially if you do it for very precise radiotherapy, it's a problem. There are methods to correct for distortions but it's very time consuming. So I think you can use this really for treatment planning for the data surgery. To overcome these problems, many modern visual simulation treatment plans that have the ability to combine the information from different imaging studies using image using or core registration. So this is one example. Here we have an MR image where the clinical target volume may be well seen by this wide area and then we segment this area, we can overtake this in the CT slice where we do see nothing at all. And by doing that we can identify the target volume and can use now this for treatment planning but those calculations and having the information of the CT. On the other hand I know from studies where you can see tumor volumes or clinical target volume in CT on a pet and then MR and they make quite different. Again, what is it to us? We don't know exactly. Oh, so it seems to be a very simple thing, very plausible to do that like that but again it has some problems which we as a medical physicist we cannot solve this problem. That we need again a good education from the radio oncologist. So this is my summary. Patient dimension are always required for treatment time or monitor calculations whether obtained with a caliper that's very old fashioned or from the T slices. Those calculations again is a part only with a treatment planning system. Almost any image modernity can be used and is used for data acquisition for patient undergoing therapy even ultrasound can be used in some cases. The process of distinguishing relevant structures or volumes from background is called segmentation. Different methods are developed for that. The XYZ coordinate system of images of different images must be correlated to each other. This is called registration, matching or image correlation. The display of different registered images data sets simultaneous called image fusion. Immobilization have two fundamental roles. I have already told that to immobilize the patient during treatment to provide reliable means for reproducing the patient position. There are many methods enabled. Treatment simulation is a major component in patient data acquisition. It started from with portal imaging developed to dedicated X-ray simulators, CT simulators and recently combi imaging or MR imaging. It may be summarized by image-guided radiotherapy and this is now the topic of the next talk by my colleague and friend, Pavel Kukulovich. So thank you for information.