 Before I start with my lecture, well, again, a few words of introduction. Well, all my life I was medical physicists. I started in 1980 when I finished my master of science, physics. And, well, what was important for me and I think it would be important also for you that to be a medical physicist is not only to be a physicist, but also to be a human who serves for other people. And that was really important for me because all my life I was human, not mathematician, not physicist. But I was enough good in mathematics and enough good in physics. So I told myself I can be, in the same time, I can use my skills, the mathematics and the physics and also to serve to people. So please remember that your work is, well, focused on people. Well, what was my major interest? Major interest was treatment, planning and dosimetry. However, I'd like to emphasize strongly that dosimetry is, I think, the most important task for physicists. Because your task is to measure. First of all, it's to measure. The second to us is to model. So please remember that if you think about, because people are really very crazy about treatment planning because they think it's very nice to play with computer to prepare plans and so on. But the most important task is dosimetry. So dosimetry should be in the center of your heart. So starting from dosimetry, well, her name was the lecturer in Virginia. I told you about the spoiler. I just, well, took from my computer the example of spoiler, how spoiler works. Because this is, these are measurements made by one of my master of science students many years ago. Where you have, it's much easier just to show to you. You have this set up of measurements, this water phantom and the spoiler above of the phantom. And you have several curves. Here is the first lesson you polish. Spuito means just with spoiler and this is open field, okay? And you may see how the depth dose curve, it's here, okay, thank you. How depth dose curve are changed by moving the spoilers to the surface of the, of the phantom. Why? Because you create electrons. Create, well, just interaction, compound interaction makes that you create a fluence of electrons. They reach the surface of the phantom and that will change the dose distribution. So that's the way you can change the dose distribution of your, of your bills. So that's just for good start. And now, well, my presentation is devoted to IGRT. Well, but because I see that not everybody is very well waken up, so let's make such an exercise. What is prerequisite for good, efficient, not radiotherapy, but for skiing? It's clear. We should have skis. We have, we should have snow. Skill. We should have mountains. Okay, but what is needed for radiotherapy? Okay, let's start. I will draw right down on the table. What is prerequisite for good radiotherapy? What do you think? Good machine. Okay. Okay. What more? Stuff. Well trained stuff. Okay. What more? Louder. Dosimetry. Good dosimetry devices. Okay. Okay. And for quality assurance. Okay, let's, let's leave it with dosimetry device. It's part of, let's call it that it's part of dosimetry devices, but I agree with you. Yeah. Communication. Well defined protocols. Patient. Of course, without the patient, there is no need for radiotherapy. Then, then we will be really lucky. I have a friend of mine, he's a doctor and he says that hospital is a perfect place for work, but the problem are patients, so. Training. Well, let's just leave with well trained stuff. Okay. Protocols. Diagnosis. Oh, that's, that's important, I think. But okay. Well, communication. Yes. You said? Diagnostic. Diagnostic. Sorry. What is, what is treatment planning system? In general, I would call it ability to calculate those distributions, okay? But because it doesn't matter, it's treatment planning system. Now we all, we think about computers, but when I start, everything was done manually. So that was also good for radiotherapy. So ability, okay, let's say, let's say treatment planning system, call it, okay. Now try to arrange it. What would be the first? Diagnostics. That's my choice. That stuff is at the first place. Because without stuff, we can do anything. We can have machines, we can have dosimetry, we can have treatment planning system, but without stuff, we cannot do anything. What stuff? Where is stuff? Well, I think about, at the first place, physicists. Without good physicists, there is no good radiotherapy. You may have ingenious doctors, ingenious radiation technologists, but if you don't measure, if you don't prepare tools without tools, there is no radiotherapy. It cannot be good radiotherapy. So remember, we are the most important chain in radiotherapy. Of course, it might be, it's not a good way to tell to doctors then, but anyway, remember about that. So we have medical physicists, we have doctors, and we have radiation technologies. All of you, all groups of people are very important in radiotherapy, okay? Please remember about that. So don't, for example, don't forget about radiation technologies. They also have to be very well trained. And it's also your responsibility to train these people. The second one, well, it might be diagnostic or it might be good machine. Yeah? But it's difficult to say, be, let's say, be free, okay? But let's discuss a little bit what doesn't mean good machine. What for you is good machine, calibrated machine? Well, modulated, and measured modulated and calibrated. Okay? Well, I tell you my opinion about that. I will call that point as a radiation. Modulated machine, but first of all, radiation. And to have good radiotherapy, we should have photos of enough high energy. If you have such a source of radiation, of course, not in very poor condition. But then, do you hear me? Yes? Or there is, yeah, then you are able to prepare a good plan. When I start my career, we have only cobalt units. So if you come from the country where there is no very modern accelerators, but there are cobalt units, you may prepare a very good treatment plan. You may treat properly your patients. Because what, let's say, treats your patients, it's not the machine itself, but the radiation. And radiation from cobalt units is exactly the same as from very modern electa-variance Siemens machine, okay? Next I think that the very important is dosimetry and dosimetry devices, okay? Yeah, and well, communication and treatment planning is also important. Yeah, okay, yeah, I agree with you. I fully agree with you. It's quality control, so I just place it in my presentation all this information. But certainly modern radiotherapy and good radiotherapy is based on image information. Why so important? If you can't see, you don't know where ionizing radiation should be delivered, and to deliver precisely the ionizing radiation, you must have a dosimetric description of the absorbent. From physicist's point of view, these two things are very important. Because if you have images from city, you have density, and you have the position of different anatomical structures. So then you are able to prepare good, good plan. But remember that not only topography is important, but also dosimetric information about your patient. So you should calibrate your city. You should have the transfer from Hansfield units to electron density. That's also one of the tasks of medical physicists. Ah, I can change from here. So we should know where ionizing radiation should be delivered, and we must be able to check if what we do is what we had planned to do. And this is the narrow meaning of IGRT, image-guided radiotherapy. So according to Wikipedia, IGRT is the process of frequent two- and three-dimensional imaging during a course of radiation treatment used to direct radiation therapy utilizing the imaging coordinates of the actual radiation treatment plan, just simply the utilizing the images to make the actual plan as much as possible identical with what had been planned. This is IGRT. However, in a broad sense, modern, oh sorry, it's a mistake, modern radiotherapy, the entire modern radiotherapy is driven by images. So images are very important. When I started there was no city, we had not set data, we just used the anatomical atlas of the patients, and doctors draw the contours based on that atlas, but the precise information was not accessible for us. That was really dangerous for our patients. Well, so the aim of IGRT. When we plan, and usually the plan is beautiful, those distribution is very conformal. Well everything is nice, but when we come to realization, it's not as beautiful as before. Because everything can be destroyed. Everything can be destroyed, yeah? Well optimist, pessimist says it cannot be worse, but optimist says it can be so. Even plan can be destroyed by our activities. So plan with IGRT is much better. It's not exactly as we had planned, because always we make some mistakes, some at least very small errors in our work. So we have a cycle. We have plan, beautiful plan. We have realization without IGRT, and we have plan which is corrected by IGRT. What images we use for IGRT? We use 3D images, which is computerized tomography. What in general images we use? From magnetic resonance, now it comes area of magnetic resonance, I think also to radiotherapy. Positron emission tomography. In our country very often we use positron emission tomography images just for drawing contours. Ultra sound, and we use 2D images, which mostly are electronic portal imaging images. So what does it mean to make the actual plan as much as possible identical with what had been planned? We have reference object, which we have at this stage of planning, and we have actual object. We got during the treatment, please remember that well most of protocols are based on conventional fractionation when we deliver 20, 25, even 30 times the same plan. So every day we have to very precisely follow our plan. Of course this reference object and actual object are related, are placed with respect to coordinate system, so the coordinate system plays a very important role in your daily routine work and the coordinate system is defined by laser system. So laser systems should be checked regularly. We in our hospital every day our radiation technologist, I obliged to check whether the laser coordinate system in some extent is set up properly. Every week we do it ourselves, medical physicists. So you can imagine that we have planned image and we have actual image and immediately you realize that there is a difference because the bones are placed in different place with respect to beam edges. So that's the way we compare. So if we overlay these two images using bones we can measure the distance along two perpendicular axis and we can measure the angle of rotation. And that's the result of quality control which says us how well the plan is being performed. And in my opinion this is one of the most important quality control during the irradiation. I will come back to that problem in my second lecture, okay? Okay, but how objects are recognized? What doesn't mean we recognize objects? And now the question to you. Do you know what's, what is in that image? Tree. Tree? Yes, there is a tree. That's small animal, what? Well, that's also always shocked me because in my country everybody recognizes it immediately because this is taken from very famous book by Milner, Winnie the Pooh. This is the Winnie the Pooh and how it's strongly related to the culture. In our culture it's clear that it's Winnie the Pooh, everybody recognizes it. But why I show that image? Because if you look at that animal you may immediately well notice there is only one line and there is no doubt that is Winnie the Pooh because how we recognize objects? We recognize objects by edges. If you look at people, if you look at different objects, we don't look at colors. We don't look at, I don't know what structure, we look at edges. And we have to be familiar with edges to recognize an object. For example, that's why for white people all black people are the same because we have no experience looking at these faces and vice versa. For black people it's difficult, they say, well everybody is the same. Have exactly the same face because it's the way we learn. And please remember that exactly is the same if you use medical images. You have to spend some time to be acquainted with images to recognize them. Sometimes physicists also involve using images, medical images. What is edge? Edge is a second derivative of intensity. So mathematically it's calculated that way. And the problem is that it's always very noisy. So whenever we overlay two images, it's some kind of mathematically based, driven minimization what is at one object and what is on the other object. So we always make very small, let's say, errors. We don't know how to overlay them very precisely. So verification of treatment plan involves comparison, mostly, comparison of portal images acquired during a treatment fraction with a reference image. But tell me what should be a reference image when you prepare your plan? What's your opinion? DRR, are there any other ideas? Digitally reconstructed radiograph. You know what's that. It's image which is reconstructed from city data. Well, I ask you that question because due to history, sometimes we use as a reference object simulator films. So the sequence is patient, city data, treatment plan, simulator, from simulator to accelerator. And then we treat simulator film, simulator images as a reference. Why that solution is not the best one? Because I agree with you that to take digitally reconstructed radiographs is much better than to take simulator images. Why? On DRRs, you don't see some tissue. No, on 2D images, you cannot see. You can have a DCR computerized radiograph on it. You cannot see the tissue on the DRR because the data is not so good for it. OK. Why? I don't criticize. I just talk to you. Because what is simulator? Simulator is another type of accelerator. Why are we going to treat simulator differently from the accelerator? We just transfer the object from planning to another machine. And in that sense, the another machine is just simulator. So we can make the so-called transfer error. So if we get reference images from the unit where we made transfer error, we start from the wrong images. So it's much better to start from sitting directly to accelerator, from digitally reconstructed radiographs to accelerator itself. OK. I see that you don't understand me fully, but I cannot explain that better. So might be later on if you have questions, come to me and we can discuss. So, but remember that you should always limit a number of transfer points. Yeah. So directly from planning to treatment, then your references is right one. OK. Of course, you mentioned about the quality of images, soft tissue, bones and so on. We can improve the quality of our images with different ways of mathematical methods with different filters. And remember that we have different filters even well delivered to you as a user in your treatment planning system, in your verify and record system. And please be acquainted with that. Just learn how to correct your images to make easier the work which is performed by radiation technologists because mostly they overlay, they match images. OK. But remembering about Ipits, about electronic portal images, which you use to get images during irradiation, please remember that it's also the part of your quality control. These images should be also controlled. I mean electronic portal imaging devices itself. So we have following tests, mechanical electrical safety, safety of mounting of the Ipits, operation of collision system. Well, you should check whether that anti-collision system works properly because really you can damage your patient with that tool. The very important is geometrical reproducibility. If the center of, if it should conform to the central axis and that should be checked with very simple tests. Image quality, spatial and contrast resolution and software performance. And please really treat that seriously. In my experience there was a mistake and error in software which was delivered by the company and we found it making such a test. So we see the differences. Well, there was no error but the software informed us that there is error of about 4 to 5 millimeters. We found that error and we corrected, send a letter to the company and they corrected. Also vendors usually recommend some tests and calibration should be made regularly. For example, dark current or noise, uniformity of image should be checked but they have their own test that should be followed regularly. Once a week, once a month, once a year. Okay? Also linearity distortion of images should be eliminated. Sympathons can be used for that. You have very good document on that published in Journal of Platinum Medical Physics. How to use these phantoms? It's free. Access to that journal is free so we can read that paper. There are spatial phantoms that can be used for that task. Of course they are very costly so not everybody can have them but anyway you can prepare your own phantoms if you try to do that. I'm surprised because there is, I don't see it here but there is also very good document, recommendation written by AAPM and you can find it in the internet. There is free access to that. Quality controls recommendations. So mostly we, to control the position of the patient, we get to orthogonal images. It can be taken with megavoltage radiation and kilovoltage radiation. Of course kilovoltage radiation gives much better quality of images in comparison to megavoltage radiation but why? My dear physicists. Okay, I see that you know that in case of kilovoltage radiation we have more photoelectric effect which depends on atomic number but Compton effect doesn't depend on atomic number so we can't see very well bones. With kilovoltage radiation we can see bones. Okay, so of course it's much better to use kilovoltage images but not always it's, we have kilovoltage machines well at the place so we have to use megavoltage images and that's also very good quality control. But what is, what the problem we have with cobalt unit if we want to make portal images? And that's the real difference between cobalt units and accelerators. All our differences are not so important. That really makes the difference between cobalt units and accelerator. My time passed okay but I have four minutes more because it was okay. Yes, that's right but that's not something which makes difference. Yeah, that's right because at cobalt units it's also, in general it's impossible to say where is the edge of the beam but you measure the position of your bones of your internal structures with respect to edges of the beam. So if you don't see you can't make a decision. If you don't know where is your edges, where are your edges, sorry, where are your edges? You cannot place your object because well I think for you it's clear that for accelerator penumbra is about four to five millimeters but for cobalt units it's 12 to 16 millimeters. So here you may make a mistake of let's say one, two millimeter but we don't make it but anyway. But here the room for making a mistake is really very, very wide. So that was, well, that is really, well, disadvantage of cobalt, cobalt unit. Not energy, not other well features of cobalt unit but wide penumbra. Not because the dose distribution is worse. It's worse a little bit but not so much but because we cannot make portal control of our patients with cobalt units. Okay, that was, here are two slides telling what is contrast, definition of contrast. You should know that it's difficult to talk about that. You should just read, I leave my presentation so if somebody would like to have it you can copy it and you can use it, of course. Oh, here is the pointer to that report. Is that report? You can find it in the internet at that website, okay. Here is the definition of signal to noise ratio and talking about kilovoltage and megavoltage images I should say that contrast is better for kilovoltage radiation and signal to noise ratio is much better. If the noise is very high then contrast is much poorer. So that's interrelated values. Okay, and that's also another problem. Quantum efficiency, how well radiation is to say that forget one English word, is detected, oh, how well radiation is detected. To get the enough good quality of image you should deliver a given dose. So the smaller dose, the better. So the quantum efficiency described that process. Yes? The mdf of the ET devices in later. Modulation transfer function, okay. To be honest, I'm not ready to answer such a complicated question but telling in a simple way. Modulation transfer function is a fantastic function which describes the contrast and all these features of images. It tells us how different frequency of images because image can be with Fourier transform transferred to the space of frequency are transferred from the object to image itself. We want to have all frequencies would be transferred in a perfect way but it's not that way. High frequencies are usually transferred in a very poor way. So modulation transfer function describes how well we transfer different frequencies from image object to image. If you have an access to the book written by Jones and Cunningham, Physics for Radiology. I, that's my personal opinion, that's one of the best book about physics. It's very old but anyway still is, it's really very good from physics point of view. There is a very nice chapter in that book while describing modulation transfer function. It's not easy to understand but if you read that chapter very carefully I think you will be able to understand better that. Okay, that's, I can tell right now about it. Okay, so how we improve quality of our images? We just add some source of kilovolted radiation to our accelerators. This is, here is the oldest solution proposed by Haines and Radiation. Just two sources of kilovolted radiation and two detectors and at once we take two images, perpendicular images and we can compare these images with DRR, with simulator images. That time mostly we use simulator images. The same idea is used at CyberKnife. We have two source of kilovolted images. For kilovolted images we have here in the floor we have two detectors. The same is proposed by BrainLab. The same is, okay that's I have already talking about that. So 3D technology, 3D technology, the idea is exactly the same. From two perpendicular images you have three coordinates. You have X, Y and Z. But from 3D images, immediately you have 3D images, however you have more problems. Because then you immediately will see that one object, if that fits another one at one part of the body, it doesn't fit to another part of the body. Why? I see that you become tired and that's also, yeah, I understand that. When you listen very carefully then, you know. Distortions, because patient body is not rigid object. So we want to fit just telling, in general, apples to plums, yeah, because we have the object and we have the body placed on the table and patient was not only transferred, was only moved from the right position but also was distorted. If it's distorted, if you match images at one part of the body, you cannot match them at another one. You have to decide how to match these two objects. But coming back to the, well, to real work, then you should place your patient again on the table. If you see the big difference between object and what is actually obtained, you should tell your technologist, you should place your patient again because you cannot correct with moving the table. Because you just want to do something which is impossible to do. So we have 3D technology which is combing CT. All modern accelerators today are usually equipped with kilovoltage combing CT, which is just CT but performed made with combing, not with narrow beam. What makes that signal to nice ratio is higher because there is more scatter radiation from such a beam. Why I answered that question, okay, so you have variant and you have electa machines. And what is interesting, it's possible also to make combing CT with mega-voltage radiation. And for several years I worked with such a solution in Kielce because in program is that I'm the worker of Kielce Center, but currently I'm the worker, I'm the physicist working Warsaw Cancer Center, so sorry, that's okay. And but then we use mega-voltage combing CT, mega-voltage combing CT, the image quality was worse in comparison to kilovoltage combing CT, but anyway we could see very well bones. So bones registration could be made very well with that solution which is much cheaper than with kilovoltage combing CT. Kilovoltage needs some extra equipment. For mega-voltage you have radiation, you have detector, but all accelerators are equipped with Ipits, so you can measure your images with that solution. Here you have image quality from that accelerator. I think that you see very well bones, of course you can't see soft tissue. So the advantage of kilovoltage combing CT is that you can also overlay your images using the soft tissue, which is not simple. And decisions are not very simple if you try to match these images with soft tissue. And what is interesting, we delivered only three monitor units for getting in that good image, so those was not very, very, very big. And also another very interesting solution is CT on rays. CT on rays is CT. When you get images from CT, you move the patient on the table. Table moves, okay, and you scan your patients. But with CT on rays, patient is at one place, but we move the gun. And that solution was installed in Kielce. You have perfect images. The only problem is that you have to rotate the table. So there is some doubts about the positioning of a patient. But anyway, the quality of images is the best one you can have, okay? Okay, just leave that. You can read it. And very modern solutions for positioning of the patient are markers, markers placed inside of any organ. You can make two perpendicular images. You can overlay these images, and you can really say where is your object. Because please remember that we are not interested in bones, whether bones are well related. In the right place with respect to beam edges. We are interested about the target itself, okay? So that's the way how we can trace the target. And we have also better markers, which are called transponders, which are let's say passive active markers. There is electromagnetic wave. It activates these transponders. They respond and it's working like GPS. So we know in time with frequency of 10 times per second the position of these transponders. So we can instantly trace position of our object. The problem is that that solution is very, very expensive. So only in some centers that solution is used. There is another one which is Sentinel. Sentinel is laser system which trace the external surface of the body. In some sense, external surface of the body is, tells us how well patient was set up. So in that way you can say whether your patient was set up properly or not properly. How good was your set up? Coming to the summary, the modern radiotherapy is certainly image based. And city information is the most important for radiotherapy. So city information should be, well cities you use should be checked, should be well not only left to diagnostic department, it's also your task to be careful about these tools. We have several solutions. We can visualize high contrast objects, I mean bones, gold markers, and we have with modern solution also visualize also soft tissue with kilo voltage or city on rates. We can pre-irradiation get information. We don't know what's going on later on, but right now we can also just look at the patient all the time during irradiation, making images several seconds or with the GPS with transponders with markers and with the skin as a surrogate of position of a patient. But good news is that in more than 80% of cases it's my estimation, conventional portal control with if it is enough. The most important are good protocols for that. About protocols I will talk in the next part of that presentations. So please remember that again, not equipment is the most important. You are the most important in applying that solutions into radiotherapy. Well, thank you very much for your attention. Thank you.