运动管理

主 讲 人:

Ed Taylor PhD

以下为课程中英字幕：

Hello! My name is Ed Taylor. I'm a medical physicist at Princess Margaret Cancer Center, and today I will be speaking about motion management - primarily in liver and lung SBRT and our experience with it.
你好！我叫艾德·泰勒，我是玛格丽特公主癌症中心的医学物理师，今天我要讲的题目是——肝脏和肺部SBRT治疗下的运动管理以及分享我们在这方面的经验。
So, I'm going to start by reviewing some of the application sites in the thorax and abdomen. I'll move on to give you some characteristic examples of respiratory motion artifacts. I will also introduce you to 4D CT, one of the primary ways of dealing with and quantifying motion. I will discuss some motion management guidelines, in particular, in the context of clinical studies, and then go on to discuss some motion management strategies for lung SBRT and liver SBRT, 2 sites that we are commonly treating here.
首先，我将回顾一些关于胸部和腹部的临床案例，然后讲一些关于呼吸运动伪影的典型例子。再然后我将介绍4D CT，考虑到4D CT是处理和量化运动的主要方法之一，讨论一些运动管理方面的指南，特别是在临床研究背景下的指南，最后探讨肺部SBRT和肝脏SBRT的一些运动管理策略，这是在我们这里通常治疗的两个部位。
So on the left, this is a classic example of a rapid acquisition helical CT scan of the upper abdomen and what you can see is as a result of the fast scan, the liver dome or diaphragm has been clipped because it's being captured in different phases of breathing. And on the right, you see a cone-beam CT, which is typically over-acquired over a much longer period of time, with the result that rather than being clipped, the liver dome is blurred.
在左边你可以看到，这是上腹部快速螺旋CT扫描的一个经典案例，作为快速扫描的结果是肝膈顶或横膈膜被切除了，因为它是在不同的呼吸时相被捕捉到的。而在右边看到的图像是锥形束CT（CBCT），它通常是在很长一段时间内获取的，其结果不是肝膈顶被剪掉，而是变模糊了。
And so this presents a problem when one is planning on CT and treating on cone-beam CT for image guidance because there's a disconnect between the 2 sets of images, and how we deal with this sort of disconnect is a central theme of today's talk.
所以这就出现了一个问题，当我们在利用CT做计划而利用CBCT进行影像引导治疗时，由于这两组图像之间存在有不连贯的地方，我们如何处理这种不连贯情况是我今天演讲的主题。
So, some of the impact of respiratory artifacts, such as this one, include incorrect target definition, in particular the size of the target. If you acquire a long duration helical scan, the targets size will be enlarged because you are capturing the target through different phases of motion. And also, if you're capturing a very rapid CT scan, you can miss the location of the target. This can lead to systematic delivery errors.
呼吸伪影带来的一些影响，如：不正确的靶区定义，尤其是靶区大小的定义。如果你获得的是一幅长时间的螺旋扫描重建图像，捕捉到了不同时相下靶区的图像，这样重建的靶区的大小也会相应的增加。如果你获得的是一幅非常快速的CT扫描重建图像，可能会忽略靶区的部分位置信息，进而导致系统性的治疗错误。
So an example is shown here on these pages. You can see on the left a large ITV margin. Normally this tumor would be moving, but you have a large amplitude of motion or large ITV margin to accommodate tumor motion, and on the right, you have other examples of tumor motion.
这里有个例子，你可以看到左边有很大的ITV外放。正常情况下，这个肿瘤会移动，需要一个很大的运动幅度或者很大的ITV外放边界来适应肿瘤的运动，在右边可以看到其他的肿瘤运动的例子。
So the central question of how to freeze motion, how to deal with motion is really one of the time duration of the motion.
因此如何控制运动以及处理运动的核心问题是运动的持续时间
So at the far left here we have the most rapid motion on the anatomical scales, which are cardiac breathing, which is 1 second. We have respiratory motion, which tends to occur over a period of 5 seconds. On the far right, we have intrafraction motion here. I'm specifically referring to the types of anatomical motion, one sees during a free breathing treatment fraction in radiotherapy. So, for instance, prostate being moved by bladder filling or rectum filling. And then finally inter-fraction motion, the results from similar effects of bowel, different bowel filling at different days, different stomach filling. All these can impact the position of the target from day to day.
在最左边，我们有解剖学尺度上最快速的运动是心脏搏动，为1秒。呼吸运动通常在5秒内发生。在最右边，首先是分次内运动，指的是在放疗中自由呼吸治疗时看到的解剖运动如前列腺随着膀胱或直肠的充盈而移动。最后是分次间的运动，不同的时间肠道和胃的充盈度不同，所有这些都会对靶区的位置产生影响。
And if you think about our 2 primary modalities for image guided radiotherapy, in the middle, we have cone-beam which requires images on the duration of 0.5 to 2 min, and on the far left we have more proactive breathing management strategies, such as infrared tracking, and those involve time scales on the order of milliseconds. And so what that means is cone beam is very good at capturing day to day anatomic differences between day to day anatomical motion, arising from intra and interfraction motion. We're essentially taking a single time snap of the patient's anatomy at that moment, but is not very good for capturing respiratory or cardiac motion. It tends to get a time averaged scan of these motions. In contrast, infrared tracking, for instance, can give us time resolved images of cardiac and respiratory motion on those time scales. And so we can begin to develop strategies to deal with those types of motion.
考虑一下我们的影像引导放疗的2种主要模式，CBCT获取图像持续时间0.5到2分钟，在最左边我们有更主动的呼吸运动管理策略如红外追踪，这些涉及到毫秒量级的时间，这意味着CBCT非常善于捕获每天解剖运动之间的差异，每天解剖运动来自于分次内运动和分次间运动，我们基本上是对病人当时的解剖结构进行单个的时间捕捉，但对于捕捉患者呼吸或心脏运动不是很好。这倾向于采用获得这些运动的平均时间扫描，相比之下，红外跟踪可以在这些时间范围上为我们提供心脏和呼吸运动的时间分辨图像，我们可以开始制定策略来处理这些类型的运动。
An advantage of dealing with the simpler modality, such as cone-beam CT, and dealing with long fractional motion times is the increased stability, whereas with shorter tracking times, stability is decreased. But a key advantage is we can exploit the “near reproducibility”. Well, I emphasize “near” reproducible, because not all motion - particularly breathing and cardiac motion - we know that there are variations from complete periodicity, but for the most part we can exploit the reproducible nature of the motion to sort of increase that stability.
处理更简单的模式，如：CBCT和处理较长的分次运动时间的一个优点是增加稳定性，然而跟踪时间较短，稳定性下降。但一个关键的优势是我们可以利用“近似可重复性”，我强调的"近似“可重复性，是因为不是所有的运动特别是呼吸和心脏运动，有完整的周期性的变化，在大多数情况下，我们可以利用运动的可重复性来增加稳定性。
So some examples of intrafraction motion are listed here. This is a well-known paper by Langen & Jones in the Red Journal, in 2001. Lung tends to move 12 plus or minus 5 mm, that's 1.2cm. Pancreas moves up to 2 cm. Liver is probably the largest motion, without any type of motion management, or compression, it can move up to 2.5 cm and that's mostly due to respiratory motion, same with kidney. It also moves, but not quite as much. It tends to be substantially less than 2 cm. And finally, organs such as prostate or cervix, also deformed by as much as 3mm over a longer time period, that's not so much due to respiratory motion, but that can be due to differences in the filling of organs surrounding the prostate and cervix, so bladder and rectum being the principle examples.
这里列出了一些分次内运动的例子。这是Langen & Jones 2001年在红皮书上发表的一篇著名论文，肺趋向于移动了12±5毫米，也就是1.2厘米。胰腺移动到2厘米。肝脏可能是运动最大的，如果没有任何运动管理类型，或腹压，它可以移动到2.5厘米，这主要是由于呼吸运动，肾脏也是如此，它也会移动，但没有那么多，它通常小于2厘米。最后像前列腺和子宫颈这样的器官，也会在较长时间内变形达3毫米，这不是由于呼吸运动造成的，但可能是由于前列腺和子宫颈周围器官充盈的差异造成的，如膀胱和直肠。
So the primary means we have of analyzing 4 dimensional motion and quantifying 4 dimensional motion is the 4D CT. So in the next few slides I'll introduce you to a cartoon of how 4D CT works. This was created by my colleague Doug Moseley, and he had this little clock which called the Crocodile Clock. And this is basically a clock affixed to a little swinging crocodile. And what that represents from a physics standpoint is 2 weakly coupled harmonic oscillators, and the thing with 2 weakly coupled harmonic oscillators is that they will exhibit periodic behavior until they don't. They will exhibit signs of chaos, and occasionally deviate from periodicity. And so in this sense, the crocodile clock makes a good surrogate, a good example of the type of experience we see with respiratory breathing. Patients will normally breathe fairly periodically, but then occasionally they'll have a hiccup. They will suddenly breathe irregularly for a moment before returning to regular periodic motion.
分析四维运动和量化四维运动的主要方法是4D CT。在接下来的几张幻灯片中，我会向你们介绍4D CT是如何工作的漫画。这是我的同事Doug Moseley发明的，他有一个小钟叫做鳄鱼钟，这基本上是一个固定在摇摆的小鳄鱼上的钟，从物理学的角度来看这代表了两个弱耦合谐振子，两个弱耦合谐振子的问题是它们会表现出周期性的行为直到它们没有，它们会表现出混乱的迹象，偶尔会偏离周期性，所以在这个意义上，鳄鱼钟是一个很好的替代品，是那种一个很好的呼吸体验的例子，病人通常会有规律地呼吸，但偶尔会打嗝，他们会突然不规律地呼吸一会儿，然后恢复正常的周期性运动。
And so what Doug did is he took this crocodile clock, which you can imagine, you have the clock moving back and forth. In addition, there's this periodicity of the crocodile itself. It goes back in about 10 s period, and Doug took a video footage of the crocodile clock in motion at 30 frames per second, and then created a phase plot, a phase plot which you see on the right side here, so you can imagine the angle corresponds to the degree of deviation from the bottom. So if the bottom represents 0, the crocodile oscillates from 90 to 270 degrees. That's why this sort of phase plot is asymmetric. It's primarily to the left side - it's not a complete circle. If it were a complete circle you'd see a uniform ring in the phase plot. But instead, here you see this sort of squished kidney bean shaped phase plot because it's primarily in the 90 and 270 degrees and it's not completely sinusoidal. That's why it's not a complete ellipsoid. The other interesting thing about this clock, as I mentioned, is there is this aperiodic motion at 5 s, which is due again to this chaotic interaction between 2 weakly coupled harmonic oscillators. And you can see that as this sort of filling in of the trajectories inside the otherwise kidney bean shape that represents the aperiodicity. So they're fundamentally 2. If we have a quasi harmonic motion, such as our crocodile clock or the breathing of a lung cancer patient, we have 2 options if we're requiring images in terms of reconstructing the image retrospectively, we can bin the images, so a 4D CT, or 4D MR works by acquiring lots of images throughout breathing cycle. And that's able to retrospectively reconstruct the image for different breathing cycles. And there's fundamentally 2 ways it can do this, it can do this either by what's called amplitude binning, or phase binning.
Doug所做的是他拿这个鳄鱼钟，你可以想象可以让钟来回移动，另外鳄鱼钟本身也有周期性。它追溯到大约10秒的时间段，Doug拍摄了鳄鱼钟以每秒30帧的速度运动的视频片段，然后创建了一个相位图，你在右边看到的相位图，你可以想象角度对应于从底部偏移的程度，如果底部代表0，鳄鱼的摆动角度从90度到270度，这就是为什么这种相位图是不对称的，它主要在左边，不是一个完整的圆，如果它是一个完整的圆，你会在相位图中看到一个均匀的环，但相反，这里你看到的是这种压扁的芸豆形状的相位图，因为它主要是在90度和270度之间，它不是完全正弦的，这就是为什么它不是一个完整的椭球体。关于这个时钟的另一个有趣的事情是，就像我提到的，在5秒的时候有这个非周期运动，是由于两个弱耦合谐振子之间的混乱的相互作用，你可以看到在代表非周期性的芸豆形状中，有这样一种轨迹的填充。如果我们有一个准谐波运动，比如鳄鱼钟或者肺癌患者的呼吸，如果我们需要图像进行回顾性重建，我们有两种选择，我们可以将图像分类，4D CT或4D MR通过在呼吸周期中获取大量图像，这样就可以对不同的呼吸周期进行回顾性重建。基本上有两种方法可以做到这一点，它可以通过所谓的振幅归统，或相位归统来实现。
So amplitude binning, as the name implies, works by triggering a set of images, or rather collecting a set of images based on a measured deviation from some point, so an amplitude deviation. So if you imagine you're taking a dynamic CT of a lung tumor you could say “Okay, Well, every time the tumor moves more than 1 cm from its median point that I'm going to start acquiring images. When it comes back I’m going to stop acquiring images and i'll call that the exhale phase or the inhale phase of breathing.
振幅归统，顾名思义，是通过触发一组图像来工作的，或者更确切地说，是根据某个点的测量偏差来收集一组图像，也就是振幅偏差。所以想象你正在做一个肺部肿瘤的动态CT，每次肿瘤从它的中点移动超过1厘米时我就开始获取图像，当它回来时，我将停止获取图像，我将称之为呼吸运动的呼气时相或吸气时相。
The other option is through a phase binning. So here, you're just simply taking a constant time from a given point in space. So you take the full periodicity of the breathing cycle. You break it up into 2 pi radians, and so you say “Maybe I'll bin it from, in this case, about 45 degrees to 315 degrees, and I'll call that a phase, and every time the motion enters that particular phase, I will collect the images and I will bin those. I'll call that my inhale or exhale phase, and reconstruct the image that way.”
另一种选择是通过相位归统，从空间中的一个给定点取恒定的时间，你可以取呼吸循环的完整周期，把它分成2弧度大约是 45度到315度，我把它叫做一个相位，每次运动进入到那个特定的相位，就会收集图像，然后把它们存档，我把这称为吸气或呼气，并以这种方式重建图像。”
So here's an example of amplitude binning at work, so you can imagine the crocodile is moving back and forth now, and once a crocodile exceeds a certain amplitude which I'm going to define, then I'm going to acquire images. So here I can define a very small amplitude deviation. So this indicates a crocodile is just moved past the bottom point, and during that phase while it's say, between 1 and maybe 2 cm past that bottom point, I will acquire images, and I will call that my sort of let's say inhale image. And they're not going to acquire images also for, say, 2 to 3 cm away. That's sort of a midpoint breathing cycle. And I can acquire images, maybe 3 to 4 cm away from the bottom of the crocodile displacement, and then, finally, I come back to the outermost point of displacement. This would correspond, say to an exhale point of breathing, and I can acquire the images during that phase. And what you can notice from the projections is that there's more images acquired during the extremes of motion - at the bottom, at the inhale phase and also at the exhale phase - that gives me better “signal to noise” in those phases at least.
这是一个振幅归统的例子，你可以想象鳄鱼在来回移动，一旦鳄鱼超过了我将要定义的某个振幅，我就会获取图像，在这里我可以定义一个非常小的振幅偏差，这表示鳄鱼刚刚被移过了最低点，在这一阶段，在距离最低点1到2厘米之间，我将获取图像，称为吸气图像。他们也不会获取2到3厘米外的图像，这是一种中位点呼吸循环。我可以在距离鳄鱼底部大概3到4厘米的地方获取图像，最后回到最外面的位移点，这与呼气时相对应，我就能获得那个阶段的图像。你可以从投影中注意到，在运动的极端阶段——在底部，在吸气阶段，在呼气阶段——获得了更多的图像，这至少在这些阶段给了更好的“信噪比”。
So that's an advantage of amplitude binning. Conversely, if I do phase binning, then I'm acquiring images for this phase, while, you know, so I trace out a fixed amount of time and from the point of, say, minimum displacement, outwards and I measure the phase of motion during that phase. And for the next phase I can similarly acquire motion for the next phase of motion, and then so on, and so forth to the end. One of the characteristics of phase binning is it gives you almost uniform numbers of images per phase of binning, so that's different from the amplitude binning where it was very strongly weighted to both the inhale and exhale phase, where the patient tends to spend the most time in breathing.
这就是振幅归统的一个优点。相反地，如果做相位归统，那么获取的是这个相位的图像，追踪一个固定的时间点，从最小位移点向外，测量在这个相位阶段的运动。对于下一阶段，我同样可以获得下一阶段的运动，以此类推，一直到最后。相位归统的一个特点是它在每个归统阶段给你几乎相同数量的图像，这和振幅归统是不同的，振幅归统对吸气和呼气时相都有很强的权重，病人往往花最多的时间在呼吸上。
The advantage of phase binning, a key advantage, is that it's less susceptible to kind of artifacts that arise from irregular breathing motion. So it's less sensitive, whereas for amplitude binning, if a patient skips a breath, or it's out of phase, that can really damage the reconstruction. So phase binning is much less sensitive to those sorts of errors.
相位归统的优点，一个关键的优点是它不太容易受到不规则呼吸运动产生的伪影的影响，所以它不那么敏感。而对于振幅归统，如果病人漏了一次呼吸，或者是不同步，就会对重建造成很大的影响，因此相位归类对这类错误不太敏感。
But amplitude binning has the key advantages that it more accurately represents actual anatomic motion. So because i'm acquiring more images during the inhale and exhale phase, and those tend to be the phases of motion that are of greater interest, so typically we tend to plan for lung SBRT on the exhale phase images. In actual clinical practice, I don't think anyone has ever rigorously established that either phase or amplitude binning are better than each other. It's just important to be aware of the differences between the 2 means of binning.
但振幅归统的关键优势在于它更准确地反映了实际的解剖运动，因为在吸气末和呼气末时相获得了更多的图像，而那些往往是更感兴趣的运动时相，所以通常我们会倾向于在呼气末时相的图像中设计SBRT计划。在实际的临床实践中，我认为没有人曾经严格地确定相位或振幅归统孰好孰坏，重要的是要知道这两种方法的区别。
So there's a number of ways that we can assess breathing motion. This is a varian RPM system and this uses an external surrogate which is placed on the patient's chest, and there's an infrared camera which detects the position of the infrared surrogate and correlates that to motion of the target.
我们有很多方法来评估呼吸运动，这是一个瓦里安RPM系统，它使用一个外部替代物放置在病人的胸部，有一个红外摄像机可以探测到红外替代物的位置并将其与靶区的运动关联起来。
So that's an external target - the advantage of this approach is that there's no radiation and it's not invasive doesn't have to go into the patient, the patient doesn't have to breathe anything, have to insert anything that their mouth. So this is advantageous in particular for lung cancer patients for whom other means of motion management may not be ideal. At the same time, it's important to emphasize that this is only an indirect surrogate of actual tumor motion. So this relies on the position of the surrogate being well correlated with the position of the tumor. As I will show you shortly, that is not always the case.
所以这是一个外部目标，这种方法的优点是没有辐射，没有侵入性不需要进入病人体内，病人不需要限制呼吸，不需要把任何东西插入他们的口腔，这对肺癌患者尤其有利，因为其他运动管理方法可能并不理想。需要强调的是这只是肿瘤实际运动的间接替代方法，所以这依赖于替代物的位置与肿瘤的位置很好地相关性。我很快就会告诉你，情况并非总是如此。
Another system is this: the Philips Pulmo scan, and this constitutes a fine pitch, helical scan with perspective signogram reconstruction, and it uses phase sorting using external bellows. Advantage of this over the varian RPM system is that it's not an optical measure so there are no “line of sight” issues, and again it's noninvasive, but, like the RPM system, it relies on external surrogate.
另一个系统是:飞利浦Pulmo扫描，它包含了一个精细的螺距螺旋扫描仪和透视信号重建装置，使用外部波纹管进行相位分类，与瓦里安RPM系统相比，该系统的优势在于它不是光学测量，不存在“视线”问题，并且它是无创的，但与RPM系统一样的地方是它也依赖于外部替代物。
And as an example of another means of motion management, this is real time MR guns on our Elekta Unity MR Linac system for liver cancer. So when we started this, we used to do acquire 3D navigator-triggered MR Scan on liver patients. So this means using an internal surrogate, such as the liver dome or diaphragm itself. Images would be binned based on exhale and inhale, and we would plan on an exhale image. Our full workflow actually required us to take both an exhale as well as a free breathing scan.
作为另一种运动管理方法的例子，这是在我们的Elekta Unity 磁共振加速器上用于治疗肝癌的实时MR检测。当我们开始这项研究时，曾经对肝脏病人进行3D导航触发MR扫描，这意味着要使用一个内部替代品，比如肝脏膈顶或横膈膜本身。图像将根据呼气和吸气进行归档，然后将计划设计在呼气末图像上。我们的整个工作流程实际上需要我们同时进行呼气末和自由呼吸扫描。
The free breathing scan would be a high resolution, but blurred image. The exhale would be a sharper crisper image, showing the actual anatomical position of the liver during the exhale breathing phase. And we would normally plan on whichever imaging data set gave us the better image. So the better quality image. A concern with liver tumours, in particular hepatocellular carcinomas, is their visibility on MR. They don't always have great contrast, and so our workflow dictated that we would choose whichever image set gave us the best image, and if we were using the triggered image because that represents a single snapshot of the breathing cycle, we would have an ITV margins, we would add a special margin to the PTV to encompass breathing, so we'd also be able to monitor breathing actively using a cine-MR, which would allow us to directly assess the motion of the target. And with this cine-MR, as long as the lesion is within the PTV, then we were okay to treat. More recently we've been using just the high resolution free breathing MR acquisition, we found it give slightly better contrast. And so in that case we don't use an ITV margin. We still have a PTV margin of between 2 and 4mm but it doesn't include an explicit margin for motion because the MR is acquired over a long period of time, up to 2-3 minutes, and so it includes a breathing motion we have a blurred lesion. But either way the structure, whether we have an ITV margin or not, the general workflow is to plan on one of these image sets, and then verify the position before delivery on the cine-MR to make sure we're covering the target.
自由呼吸扫描将得到一个高分辨率，但模糊的图像，呼气末会是一个更清晰的图像，在呼气末时相下显示肝脏的实际解剖位置，通常会选择能给我们提供清晰图像的成像数据的时相进行计划设计。关于肝肿瘤，特别是肝癌的一个问题是，它们在MR上的可见度并不总是有很大的对比度，所以我们的工作流程决定了我们需要选择最清晰的图像集，如果我们使用触发图像，因为它代表呼吸周期的单个快照，我们会勾画一个ITV边界，在PTV中添加一个特殊的边界来兼顾呼吸运动，所以我们也可以用cine-MR来监测呼吸，这可以让我们直接评估靶区的运动，有了磁共振成像，只要病灶在PTV以内，我们就可以放心治疗。最近我们一直在采集自由呼吸下的高分辨率的MR影像，发现这些图像能提供更好的对比度，在这种情况下，我们不用ITV外放，但仍然有2到4mm的PTV外放边界，它不包括明确的运动边界，因为MR是在很长一段时间内获得的，长达2-3分钟，因为它包括呼吸运动，所以涉及到一个模糊的病变边界。无论哪种结构，不管我们是否有ITV外放边界，一般的工作流程是在这些图像集中的一组图像进行计划设计，然后在治疗之前使用cine-MR进行位置验证以确保我们靶区得到了充分覆盖。
In the future, and I think this is coming imminently now, Elekta is developing software solutions to actually be able to gate and automatically turn off the radiation or signal to the therapist at the unit, whether or not the target has left some predefined volume, so the MR linac will be able to actively track motion in real time based on respiratory motion.
在未来，我认为这一点能实现了，因为Elekta正在开发整套的软件解决方案，来控制机器实际通过门控技术自动开关束流并反馈实时信号给到治疗师，同时预估靶区是否偏移定义的位置， 磁共振加速器都能根据呼吸运动情况进行主动实时跟踪治疗。
So in general, there are a number of options when you have moving tumors to treat them. Of course, if you have a static tumor, you can just beam on, and with appropriate image guidance, and not have to be worried about the tumor falling in and out of your treatment fields. If you have a moving target, you can add a margin that encompasses the motion. So this is the ITV Margin, internal target volume.
所以一般来说，针对运动的肿瘤我们有很多治疗方案选择。当然，如果你有一个静态的肿瘤，你可以直接照射，在合适的影像引导下，不必担心肿瘤是否在治疗野里面。如果你有一个移动的靶区，你可以添加一个外放边界来兼顾这个运动范围，这就是我们说的ITV外放，也叫内靶区。
This is typically assessed on fluoroscopy, or in the case of the MR linac that I showed you on the previous slide, the real-time cine-MR image. So in this case we have a larger treatment field that encompasses the motion of the target, and that has the advantage that it ensures you irradiate the target as long as there's no irregular breathing. But you're going to irradiate a larger fraction of normal tissue.
这是一个通过影像透视来评估的案例，我在上一张幻灯片上有给大家展示了磁共振直线加速器的实时cine-MR图像检测。在这种情况下，会有一个更大的治疗范围包括靶区运动，它的好处是，能确保只要没有不规则的呼吸运动你就能照射到靶区，但是正常组织接受的照射范围也会增大。
You could also, conversely, not have a margin, but the clear disadvantage of that is that at certain points in the breathing cycle the tumor will be outside the field of radiation and hence the external aspects of the tumor will not be fully irradiated.
相反地，你也可以不用外放边界，这样会导致肿瘤在呼吸周期的某个位置时脱靶，部分肿瘤将得不到足量照射。
Another more advanced option is to gate, based on some either external surrogate or internal motion-derived surrogate motion. The beam only turns on when the tumors is in the appropriate position. The disadvantage of this is it lengthens the treatment time, because you're only treating during a particular portion of the breathing cycle. So now treatment can take 2 to 3 times longer.
另一种更高级的选择是门控技术，它是基于一些外部替代物或内部运动派生替代物。只有当肿瘤处于合适的位置时，射束才会打开，这样做的缺点是它延长了治疗时间，因为你只在呼吸周期的特定时间进行治疗，所以现在治疗时间是原来的2到3倍。
Another option is to actually have the couch move in time with the target motion to ensure that the target is always being covered by the radiation field. Equivalently you can get the MLC leaves to move in real time to gain cover the target.
另一种选择是让治疗床与靶区保持同步运动，以确保靶区总是被照射野覆盖。同样地，你可以让MLC叶片实时移动以获得对靶区的覆盖。
Finally, you have modality such as a cyber knife, in which the whole radiation delivery device is moving in real time to ensure that the target is being covered. And you can also have deformable MLC if your target is actually not just being translated in real time. But of course they never are. These are elastic media, and so the target will deform, and you can incorporate that deformation of motion with real-time tracking and closing of MLC leaves. And again, that has the advantage over just moving rigidly back and forth that you can potentially spare greater normal tissue
最后，有一种治疗技术如赛博刀，整个辐射传输装置实时移动来确保靶区被覆盖，如果靶区不是实时变化的可以使用可形变的MLC。当然，他们从来都不是。这些是弹性介质，所以靶区会变形，你可以将运动形变与实时跟踪MLC叶片结合起来。再次强调，这比僵硬地来回移动的优势在于你可以避让更多的正常组织。
So this is a well-known slide from Paul Keall. The gold standard for 4D radio therapy is to explicitly include all the temporal changes in anatomy during the imaging planning and delivery radiotherapy. So all 3 components that starts with 4D CT or 4D MR imaging, which means we acquire CT over different image sets acquired during the breathing cycle. so we can both quantify the motion to the case, and we can, in principle, then go and actually develop treatment plans for the different phases of breathing. So that we can have one treatment plan, say, for the inhale phase, one treatment plan for something intermediate between those phases, and then another treatment plan for the exhale phase and then 4D treatment delivery using some of the techniques I showed you on the other slide, including motion of MLC leaves that we can continuously deliver treatment throughout the entire breathing cycle.
这是来自Paul Keall的一张著名的幻灯片。4D放射治疗的金标准明确地包括在所有呼吸周期的解剖变化情况下进行4DCT定位影像，4D计划设计和执行4D放射治疗过程。这3个组成部分都是从4D CT或4D MR成像开始的，这意味着我们在呼吸周期中通过不同的图像集获取CT/MR来量化病人的运动，原则上，我们可以针对呼吸的不同时相制定治疗方案，我们可以有一个治疗计划如吸气阶段，一个治疗计划在中间时相上，还有一个治疗计划在呼气阶段，最后使用4D放疗执行，在另一张幻灯片上展示的一些技术包括MLC叶片的运动，我们可以在整个呼吸周期中持续执行治疗。
So this is very much the gold standard of 4D radiotherapy. It requires an awful lot of physics support, it's very difficult to do, and challenging to do. So gating, or other forms of respiratory management, are by far the most common means of dealing with motion. But this absolutely represents the most efficient sort of ideal 4D radiotherapy option.
4D放射治疗的金标准需要大量的物理支持，这是非常困难和具有挑战性的。所以门控技术，或其它形式的呼吸管理，是目前为止最常见的运动管理的方法，这绝对代表了最有效的理想4D放射治疗选择。
So the AAPM Task Group Report-76 gives a nice decision tree to determine what is the appropriate or when we should use motion management. So you can start by asking, do we really need, or is there a method of measuring respiratory motion? Is it available? So it may be that in your clinic you do not have a 4D CT, or you do not have a fluoroscopy device, in which case you can't even assess motion, in which case you can't do respiratory motion management.
因此，AAPM TG-76提供了一个很好的决策来决定什么是合适的，或者什么时候我们应该使用运动管理。所以你可以先问，我们真的需要或者有测量呼吸运动的方法吗？它可用吗？可能在你的诊所里没有4D CT，或者没有影像透视设备，在这种情况下你不能评估运动，甚至不能做呼吸运动管理。
But if you do have a 4D CT, and if the tumor motion is significant - defined here as being greater than 5mm - then you can do motion management. At the same time we have to ask, is this really worthwhile? There's a cost associated with doing motion management. It takes time, it takes resources, it takes both in terms of personnel and time, and at the end of the day, if there's no significant clinical benefit, then why bother? Often in the case of lung and liver, which are largely parallel organs, there is a benefit to reducing the amount of dose received by radiotherapy to the normal tissues, and so we opt for motion management strategy. But of course at the same time, if we can achieve our full clinical goals without any kind of motion management, then we won't do it. But assuming we can do better dosimetrically, with motion management, and we have the capacity to do it, then we can do it. And of course, a critical component as well, is the patient able to comply with the respiratory management procedure? So here are those same critical questions just repeated.
但如果你有4D CT，而且肿瘤移动很明显——这里定义为大于5mm——那么你就可以做运动管理。与此同时，我们不得不问，这真的值得吗？做运动管理是有成本的，是需要资源，人力和时间的，到最后如果没有显著的临床效益，那为什么还要麻烦呢，通常在肺癌和肝癌治疗的情况下，它们基本上是并行的器官，减少放射治疗对正常组织的剂量是有好处的，所以我们选择运动管理策略。但与此同时，如果我们能在没有任何运动管理的情况下实现我们的临床目标，我们就不会这么做了。假设我们可以通过运动管理，我们有能力在剂量计算上做得更好，那么我们就可以做到。当然还有一个关键因素就是病人是否能够遵守呼吸管理流程，这是非常关键的问题
Do we have the capacity to measure respiratory motion? Is the tumor motion greater than 5mm or is there a significant normal tissue sparing? What type of respiratory management is available at your clinic? And then finally, do the clinical goals really require respiratory management, and what is the patient compliance?
我们有能力测量呼吸运动吗？肿瘤移动是否大于5mm或是否有明显要保护的正常组织？你的诊所有什么呼吸管理方法？最后临床目标真的需要呼吸管理吗？病人依从性如何？
So here's some examples of some respiratory motion management devices. Probably the most simple is this wooden frame plastic compression device. The nice things about this it's very easy with indexing. The downside of this in any sort of compression device is that you see a lot of contour deformation. So I'm very sensitive to this because I have a research study in which I take patients on our MR linac, and they're routinely treated with a compression belt, such as the MR Compression belt you see in the middle panel, and that for pancreatic and liver cancers, and the tumors have a certain shape, and then we take them up to our diagnostic MR scanner to have repeat MR Images. And what I found is that there's a lot of deformation that target, in particular pancreatic cancer targets, there's a lot of deformation just from the compression belt, and it's difficult to map the contours over from one device to the other.
这是一些呼吸运动管理设备的例子，可能最简单的就是这个木框塑料腹压装置，这样做的好处是控制肿瘤位置非常简单。这种腹压装置的缺点是你能看到很多轮廓形变。我对此非常敏感，因为我有一个研究，把病人放在磁共振线加速器上使用腹压带治疗，就像你们在中间看到的MR腹压带，对于胰腺癌和肝癌，肿瘤有一定的形状，然后我们把他们带到诊断磁共振扫描仪进行重复磁共振成像。我发现靶区有很多变形，特别是胰腺癌靶区，由于腹压带导致有很多变形，很难把一个设备的轮廓线映射到另一个设备上。
A third option is an active breathing device which offers to improve the therapeutic ratio, because you're not dosing normal tissue. However, this typically requires extensive education and monitoring, and also patient compliance. And this goes back to the practicality issue. Do we have the resources to do this, and is a patient able to do it?
第三种选择是主动呼吸装置，它可以提高治疗增益比，因为你不需要做额外的。然而这通常需要广泛的教育和监控，以及病人的依从性，这又回到了实际的使用问题上，我们有人力资源来做这个吗？病人有能力做吗？
So for lung SBRT, at our institution we typically acquire 4D CTs. There are a number of options in terms of the data sets which you plan on. So again, the Paul Keall ideal of having a plan for each phase of breathing is typically not done. It's certainly not done at our institution but you have the option normally, of either planning on the end exhalation, the end inhalation, or maximum intensity projection, or an average phase, or, alternatively, you could just have a long duration free breathing helical CT with a 4D CT. Of course, the first 4 options are only available if you have 4D CT. At our institution we use the exhalation image for a couple of reasons. One is that that represents the most conservative from a lung dosimetric point of view. That's the most conservative phase of breathing because that's when your lung is the smallest. And so any volumetric lung clinical goal will be sort of maximally bad during that cycle.
对于肺SBRT，在我们的机构我们通常获取4D CT，就会有许多图像集需要进行计划设计，所以保罗·基尔的理想是为每个呼吸阶段制定计划通常是不能完成的。我们的机构当然不会这样做，但通常可以选择在呼气末，吸气末，最大强度投影，平均时相，或者可以只做带有4D CT螺旋CT一个长时间的自由呼吸，前4个选项只有当你有4D CT时才可用。在我们的机构，我们使用呼气末图像有几个原因，一是从肺剂量学的角度来看这是最保守的，因为这时肺最小，所以任何肺体积的临床目标在这个周期中都是最不好的。
The other thing is because patients spend the most amount of time at the exhale phase breathing, the images, in particular the normal tissue images tend to be a little crisper.
另一件事是，因为病人在呼气时相花的时间最多，图像采集时特别是正常组织会更清晰
So, as I mentioned before, most of our image guidance is done using external surrogates. The problem is, these sometimes fail. So this is a lovely paper by Jeremy Hoisak and others in 2004 from Princess Margaret, and what you're seeing here is for a single patient, the respiratory volume on the left, so that's the kind of metric you get with an active breathing control device. You see that in the lower left hand panel the patient with a breathing apparatus, or hose, in their mouth, and so that represents a surrogate for the respiratory volume. At the same time this patient had external surrogate monitoring system, and that's shown on the right, which is measuring the abdominal displacement. And then the middle finally with fluoroscopy, this study was able to assess the actual tumor displacement, and so an ideal surrogate would be well correlated with the actual tumor displacement. And you can see for this patient that both the respiratory volume as well as the abdominal displacement, so that's the external surrogate, and on the left the active breathing control are very well correlated with the actual displacement of the tumor. There is a phase line which needs to be accounted for. But if you're gating, but otherwise this constitutes, a robust way of assessing tumor motion.
所以，正如我之前提到的，我们大部分的影像引导都是使用外部替代完成的。问题是这些方法有时会失败。这是来自玛格丽特公主医院的Jeremy Hoisak和其他人在2004年写的一篇论文，你在这里看到的是单个病人的呼吸量，左边的是呼吸量是用主动呼吸控制装置得到的度量，你可以看到在左下方的图示中病人的嘴里含着呼吸设备，或者说是软管，这代表了呼吸量的替代品，与此同时，这个病人有外部替代监测系统，如右边所示，它在测量腹部的位移，中间用影像透视，这项研究能够评估实际的肿瘤移位，所以一个理想的替代物将与实际的肿瘤移位有很好的相关性，你可以看到这个病人的呼吸量和腹部位移，这是外部替代物，左边的主动呼吸控制与肿瘤的实际位移相关，有一条相位线需要考虑。除非你使用门控技术，否则这就构成了一种评估肿瘤运动的鲁棒方式。
However, this is the same plot shown for different patients with more irregular breathing, and what you see is that the abdominal displacement, the external surrogate, is a very poor surrogate for actual tumor displacement. The respiratory volume continues to be a good surrogate. And that makes sense, because respiratory volume should be more closely aligned with what this, of course, is a lung tumor so respiratory volume should be more in line with what the actual tumor, where the tumor position is, because it's a direct measure of where you are in the breathing cycle, whereas abdominal displacement is really an indirect surrogate of breathing cycle measures.
然而，这是相同的图显示的是不同的病人不规则呼吸，你看到的是腹部移位，外部替代是一个非常糟糕的替代实际的肿瘤移位。呼吸量仍然是一个很好的替代品，这是有意义的，因为呼吸量应该与这个更接近。一个肺肿瘤呼吸量应该更接近实际肿瘤的位置，因为它是一个直接测量你在呼吸周期的位置，而腹部移位实际上是一个间接的替代呼吸周期测量。
Alternatively, you can use an internal surrogate. So this is what we tend to do on the MR linac as well. You can actually assess motion of the tumor itself, if it's clearly visible on 4D CT or a 4D cone beam CT. Or on, say in this case a surrogate which is the liver dome, and you can measure the motion, using that liver dome, and that can be used as a surrogate for lung tumors or liver tumors.
或者，您可以使用内部替代，这也是我们在磁共振直线加速器上做的。你可以评估肿瘤本身的运动，如果它在4D CT或4D CBCT上清晰可见，或者说在这个例子中是肝膈顶，你可以用肝膈顶来测量运动，它可以被用作肺肿瘤或肝肿瘤的替代物。
So in general, 4D CT and 4D cone-beam CT, we tend to use, the typical amplitudes of motion for free breathing. We can use some kind of respiratory monitor system using internal surrogate if the tumor motion is less than a centimeter for free breathing. If we have tumor motion greater than 1 cm, then we can use abdominal compression. Finally there's active breathing control. The problem with active breathing control, even though it offers better robustness as a surrogate of lesion position, it's not routinely used in lung patients for the simple reason that lung patients tend to have impaired lung activity.
一般来说，我们倾向于使用4DCT和4DCBCT，用于自由呼吸的典型运动振幅。如果自由呼吸状态下肿瘤移动小于1厘米，我们可以使用内部替代物作为呼吸监测系统。如果肿瘤移动大于1厘米，我们可以使用腹部压迫技术。最后是ABC主动呼吸控制系统，主动呼吸控制的问题是尽管它作为病变位置的替代物提供了更好的稳健性，但它并不常用于肺部患者，原因很简单，肺部患者往往有肺功能受损。
Conversely for liver patients, we routinely use ABC in our clinical practice. So for liver SBRT, motion is typically assessed using kV fluoroscopy, and at that point a motion management decision will be made. We try for every patient that comes through, we put them on the active breathing control system. This is the preferred modality for reasons that i've already described, and up to 60% of our patients will get ABC.
相反，对于肝脏患者，我们在临床实践中习惯性的使用ABC。对于肝脏SBRT，通常使用kV透视来评估运动，然后做出运动管理方面决策。我们试着给每一个来的病人用上主动呼吸控制系统。ABC是我们首选的治疗方式，原因我已经描述过了，我们这高达60%的患者会用到ABC。
Alternatively, they can get abdominal compression if they're not suitable - often they can't understand the instructions for ABC, or they're not compliant, or they breathe regularly then we can submit them to abdominal compression. And finally in a worst case scenario we'll use free breathing. If we have free breathing, then we have to make a decision on margin - ITV margins if we're not using 4D CT. If we have a 4D CT but we're not capable of using ABC, we acquire the 4D CT, and we measure the maximum displacement between phases of the tumor for liver and that allows us to define an ITV. Alternatively, we can use cine-MR for image guidance, and we'll use 4D cone beam CT. Or if we're not using ABC, our patients will often go on the MR linac. A key component here is that there be consistent management of motion from simulation to the plan to treatment. So whatever motion management option we use for treatment, we have to have the imaging capacity to do that.
另外，对于ABC不适合的患者如不理解ABC的使用流程的患者，或者不配合的患者，如果他们呼吸有规律，然后我们会考虑给他们用到腹部压迫。最后，在最坏的情况下（以上两方面都不满足）我们才会考虑给患者使用自由呼吸，如果是在自由呼吸的且没有4D CT的情况，我们必须决定ITV外放边界。在有4D CT但不能使用ABC的情况下，我们先扫描4D CT得到图像，然后通过测量肝脏肿瘤在不同相位图像之间的最大位移，指导我们定义ITV，或者我们可以使用cine-MR或者使用4D CBCT来进行图像引导，或者如果我们不使用ABC，我们的病人通常会使用医科达的磁共振加速器。这里的一个关键点就是运动管理从模拟到计划再到治疗要做到一致，无论我们在治疗中使用什么运动管理方式，都必须有相应的影像引导方案。
So a major phase 3 study of liver hepatocellular carcinoma, led by Laura Dawson at Princess Margaret Cancer Center is the RTOG1112 study, stipulated that a major factor of great importance for randomized studies such as this is the quality of radiotherapy. And in particular for liver radiotherapy we know that motion management has a major impact. Breath hold is the preferred modality so that patients are assessed based on their ability to undergo the active breath hold management system. So patients are identified for how reproducible the position is, and also the stability of the position. So patients try and undergo ABC at their sim sessions and their ability to comply, and the reproducibility of the motion is assessed to that point.
这里列举一项针对肝癌III期研究，是由玛格丽特公主癌症中心的劳拉·道森主导的RTOG-1112号研究，列举了随机研究中非常重要的一个主要因素就是放疗的质量，特别是在肝脏放射治疗过程中，我们都知道运动管理会带来很大的影响，屏气呼吸是首选的方式，以便根据患者接受自主屏住呼吸的能力进行评估，我们会重点评估患者体位的重复性以及稳定性，因此我们会让病人在模拟训练过程中尽量适应ABC流程和呼吸掌控能力，达到重复性运动稳定的评估点。
Alternatively free breathing is the other option, in which case at the imaging session, the magnitude of target, or liver motion is assessed. So, of course, target is always ideal to assess the motion of, but sometimes the target is not visible on 4D CT or even MR, in which case we use the liver itself as a surrogate for motion. RTOG11123 strongly recommends some form of motion management if the motion is greater than 5mm, which is typically the case.
另一种选择是自由呼吸，在成像过程中我们需要评估靶区或肝脏运动的范围大小。靶区的运动评估总是最优先考虑的，可有时靶区在4D CT甚至MR上是看不到的，在这种情况下我们使用肝脏本身作为运动评估对象。RTOG1112强烈建议，通常情况下如果运动大于5mm，必须采用某种形式的运动管理。
So, of course, the goal is to minimize liver motion in this case, because you want to irradiate as little of the liver as possible. But there are a number of considerations that lead to that decision.
在这种情况下，我们的目标是尽量减少肝脏的运动，如果想要照射尽可能少的肝脏就必须在做决策时需要权衡并考虑很多因素。
First and foremost, we have to be reasonable. Do we have the resources to do this? And of course, patient compliance and patient desires are a major part of this decision making process, critically for ABC, is the fact that the patient has to hold their breath for up to 15 seconds, more than 15seconds. If they can't do that then ABC is strongly discouraged. The number of lesions is also an important factor as well as imaging for setup and verification.
首先，我们必须认识到是否有足够的资源来做这件事， 病人的配合度和治疗意愿也是决策过程中重要的考虑因子，对ABC至关重要的是病人必须屏住呼吸长达15秒或者15秒以上，如果达不到15秒，ABC的使用就会受到限制，病灶的数目同摆位成像验证一样也是一个重要的因素。
So the recommended margins for RTOG1112 and again, this is a very important fact when assessing randomized trials in radiotherapy, is the quality of radiotherapy. So the recommended motion management depend on patient motion, and also the reproducibility.
所以再次回到RTOG1112的推荐的外放边界情况，一个很重要的事实是在随机放疗中的重要评估点还是要落到放疗的质量或者疗效上。因此，运动管理方式的推荐往往要考虑到患者的运动情况和重复性。
So the minimum margin, which is just the PTV margin and includes the ITV, is 4mm. There can be up to a maximum margin of 2 cm, which would be in the case of a free breathing case. So if the motion amplitude was, say, greater than 1.9mm, because you also have to include it in your PTV margin, of course, room for systematic errors and random errors, not just, not just breathing artifacts, the ITV. If the breathing motion is too large, then we will not be able to give liver SBRT.
所以最小的外放边界，即PTV外放边界包括ITV部分是4mm。在自由呼吸的情况下，最大的外放边界可达2厘米，如果运动振幅大于1.9毫米，必须还把它包括在PTV外放边界中，当然还要结合系统误差和随机误差带来的影响，呼吸运动伪影，ITV范围等等因素，如果呼吸运动实在过大，我们则不能够给肝脏进行SBRT照射。
Ideally, we want to keep the margins as small as possible, again to reduce dose in the normal tissue, and they can be asymmetric depending on the pattern of breathing and the motion observed during the fluoroscopy session for these patients. Another critical point is that the margins cannot be edited to proximal normal tissues which, of course, is the ICRU specification. A PTV margin is just that, it is a margin that delineates our potential error in the location the tissue. It should never be edited to reflect the proximity of normal tissues, and yet that's often done.
理想情况下，我们希望保持外放边界尽可能小来进一步减少正常组织的剂量，外放边界可以是不对称的，这取决于患者的呼吸模式和在透视期间观察到的患者运动情况。另一个关键点也是ICRU的规范提到的，当外放边界扩张到就近的正常组织里时不允许再编辑，PTV外放边界就是这样，它描述了我们在组织定位上的潜在误差，它不应该被编辑用来反映与正常组织的接近范围，但实际中我们却常常修改。
So here's, 2 examples of liver HCC patients with different motion assessments. On the left the motion was assessed to be 15mm sup/inf during free breathing with ABC. There's still a residual drift during the ABC. So even though the patient is not actively breathing, their little microbreaths prompt the liver to move slightly up and down. And so that was estimated to be 2 to 3mm during the 20 second breath hold duration. Based on that 2 to 3mm, in this case, a 5mm PTV margin, which is quite small, was felt to be appropriate. On the right, you have a case which was not eligible for ABC. This was free breathing situation, under free breathing, without any compression. The motion amplitude was 9mm sup/inf. With compression, it was reduced to 6 mm, but based on the pattern of motion, an asymmetric ITV margin, and hence an asymmetric PTV margin, based on the varying position of tumor during the 4D CT, was created. So in this case, 5 or 6mm superior, 7mm anterior, and 10mm interior margins were defined.
这是两个采用不同的运动评估情况的肝癌患者的案例。在左侧是患者使用的ABC自由呼吸控制技术，运动幅度评估为头脚方向15mm，考虑到ABC过程中还是有些许偏移，即使病人没有主动呼吸，他们的微小呼吸也会促使肝脏轻微上下移动，在屏气20秒的时间里估计有2-3毫米，2 - 3mm运动幅度应该算是非常小的，此时PTV外放5mm应该是合理的。右边是一个不符合ABC情况的案例，患者在自由呼吸的情况下，没有任何腹部压迫，运动振幅在头脚方向为9mm，增加腹压装置后，运动幅度减少到6mm，基于运动管理模式并结合4D CT中肿瘤的不同位置信息，可以创建一个非对称的ITV边界，进而得到了一个非对称的PTV外放。此患者PTV外放情况为头方向外放5-6mm，患者前（面部）方向外放7mm, 而脚方向则外放10mm.
Or course, the advantage of the ABC plan which you see here, if you can make out the PTV, which is that yellow structure surrounded by a set of isodose lines, is the entirety of the PTV can be covered by the prescription dose, while sparing the bowel which sits just below this particular tumor.
当然，ABC方案的优势就在这里，如果你能分辨出PTV，也就是图中被一组等剂量线包围的黄色结构，整个PTV都可以被处方剂量覆盖，同时保留了位于肿瘤下方的肠道。
Conversely for the free breathing plan, the PTV is larger, and not only are you irradiating a greater portion of the normal liver, but you end up having to undercover the tumor to spare the bowel.
相反，对于自由呼吸方案，PTV更大，你不仅要照射更多的正常肝脏而且最终必须放弃部分肿瘤来保护肠道
So to summarize, if i've convinced you of anything today, I hope that it's motion management needs will vary for different clinical sites. Lung is very different from liver because most lung cancer patients have impaired breathing. We cannot use our preferred motion management method, which is active breathing control - instead, we have to use 4D CT and ITV margins. Whereas in liver, most of our patients are able to undergo automatic breathing control.
总结一下，如果我今天说服了你们什么，我希望运动管理的需求会因不同的临床部位而不同。肺和肝有很大的不同，因为大多数肺癌患者都有呼吸障碍，我们不能使用我们首选的运动管理方法即主动呼吸控制，相应的，我们必须使用4D CT和ITV外放边界。然而在肝脏，我们的大多数病人能够进行自动呼吸控制。
Motion management decisions are patient-specific. Every patient is different and we have to take into account their compliance, and the nature of their breathing.
运动管理的决定是因人而异的，每个病人都是不同的，我们必须考虑到他们的配合度，以及他们呼吸的规则度。
Also, as we've seen in particular on our MR linac, a lot of patients have high levels of anxiety, and that needs to be managed as well. That can lead to irregular breathing pattern so everything is really very patient centered.
此外，正如我们在磁共振直线加速器上看到的，很多患者都会感觉到高度焦虑，这个也需要处理，焦虑会导致呼吸不规律，所以一切都是以患者为中心。
Finally, it’s very important to have a consistent motion management throughout the entire SBRT process. So whatever your initial reference imaging, has to be appropriate for your actual treatment modality. So if you have, you know gating, for instance, then you absolutely need 4D CT imaging.
最后，在整个SBRT过程中保持一致的运动管理是非常重要的，无论你最初的参考影像是什么，都必须适合你实际的治疗方式，如果你有门控装置且知道门控技术的使用方法，那么你绝对会需要4D CT成像。
Finally, decisions impact planning and treatment approaches. Thank you very much.
最后，想说的是，运动管理决策将影响着计划设计和治疗方法。非常感谢。