Transcript
Man 1: Hello, and welcome to our first Novalis Circle technical webinar. Today, we want to show you how to plan an easy case with our brand new Multiple Brain Mets SRS element. We have prepared a video that will guide you through all the necessary steps on how to plan with our new software. Following the video, you'll have the ability to send us questions, and we will gladly respond. As always, our webinars are recorded and should be available for you to watch on the Novalis Circle website at a later date. Also, don't forget to sign up for upcoming webinars on the Novalis Circle website. Thank you very much for logging in, and we hope that you find this tutorial helpful.
Man 2: Hello everyone, and welcome to the first Novalis Circle technical webinar. Today, we're gonna go through step-by-step the planning of a simple multiple brain mets SRS case. As many of you know, the Elements software is our brand new treatment planning platform. Each element has been specifically tailored to treat a specific indication. For multiple mets, this is a novel treatment planning system, that has been tailored to treat multiple targets with a single isocenter.
This is the Multiple Brain Mets SRS treatment planning workflow. As you can see, we have laid out each one of the tools in a way that guides you through the treatment planning process. From top to bottom, left to right, you'll see, we start with the DICOM Viewer tool, where you can review any DICOM images. Then we would go into the Image Fusion element, where we would rigidly register your DICOM datasets. From there, we have the Distortion Correction element, that either uses a CT or an undistorted MR to remove any geometric or patient-specific distortions that you may have in any other MR sequences.
Once that's been completed, we can open the Anatomical Mapping element, where we perform any auto-segmentation of critical structures that you may want. The SmartBrush element then allows you to contour any of your cranial lesions. The Object Manipulation element allows you to perform Boolean operations or add margins to any of your cranial lesions. Next is the actual planning element itself, Multiple Brain Mets SRS. This is where you'll be able to perform the automated treatment planning process, and generate your RT plan. The last element in the workflow is the Patient Specific QA element. This allows you to map your generated treatment plans onto any phantom you have available to you.
We'll start by reviewing the datasets available for this patient in our DICOM Viewer. This is a tool that's been designed to give you full flexibility and power over your imaging data. It's been designed in a way where it lays out different panes, and allows you to customize these views. When we open up this element, you're going to see one pane for every DICOM dataset that you've imported. As you can see here, we have the CT dataset on the left, and the two MRs in the middle and on the right.
As you can see while I'm scrolling through each one of these datasets, the rest of the datasets are not moving. This is because we have not yet fused the datasets. If you'd like to create additional views, you simply click on the head icon at the top of the pane. This will open up the options that you have available for that particular dataset. As you can see here, you can create different reconstructed orientations from axial, sagittal, and coronal, and you can also create image modality-specific 3D representations as well, such as a skin filter or a bone filter with the CT set.
Once that's been created, you can go ahead and zoom or rotate the pane. And once you're finished with that visualization, you can simply click on the x box at the top, and that will remove the pane from your screen. One thing to keep in mind when you're working through the different Elements workflows is all the tools that you would normally use are going to be located on the right-hand column. As you can see here, in the View tab, there's multiple different view options that you can select. You can change the windowing. You can perform image slice thickness averaging. This allows you to see more information on each one of your image slices. You can take a screenshot, and you can also add additional views.
Below the View options are your measurement tools. Here, you can see we can measure distances, circle diameters, angles, and drop points for later viewing. If we wanna measure the distance of this particular metastasis, you can hold down the control key and use your scroll wheel to zoom in. If you hold down control, and click with your left mouse button, you can actually pan the image to center the target. By clicking the Distance button, you activate that tool.
You can start the measurement by clicking on the edge of the particular met you're looking at, and then you drag the second circle to the opposite edge. And then you're gonna see the measurement icon above your tool. To delete that measurement, simply click on the garbage bin on the right column, and then click on that arrow. To measure diameter, click on the Circle tool, then click in the center of the target to create your measurement tool. To move the position, click in the center of that circle, and drag to the middle of your target. To adjust the diameter, click on the periphery, and drag inwards or outwards.
The DICOM Viewer element is a powerful tool that can be very helpful in tumor boards or any presentation when you need to review your DICOM data. Once you've finished reviewing the DICOM data that you have available, we can move on to the next step, which is Image Fusion.
Next, we'll go into our Image Fusion element. When you open the Image Fusion element, your DICOM datasets will automatically be registered together. As you can see here, the two MR datasets have been automatically fused. You can review the fusion with the Spy Glass, moving it around by clicking and dragging. And if you want to change the size of the Spy Glass, you can click and drag from one of the corners. You can change your view orientation by clicking in the upper left-hand corner, on that Axial tab, and now you can change to coronal or sagittal. We also have the ability to review the fusion with the amber blue blending tool that you may be used to from iPlan.
If you're happy with the results, you can go ahead and click on accept the fusion, at the top. Once you accept the fusion, it'll automatically move to the next fusion pair. Here, you can see the CT being fused to the MR. If you'd like to change the windowing, you can click on the windowing icon, and then choose which dataset you would like to change the window level. Here, for the CT, we can select that in the orange, and then you can click and drag your mouse on the screen to change the window level. When you're happy with the result, go ahead and click the windowing button one more time, to deactivate the tool. If you're not happy with the image fusion result, you can click on one of the tools to manually adjust your fusion.
The first one is the Region Of Interest box. Here, you can see a visualization of your region of interest that you're going to define for your fusion. By clicking in the center, you can click and drag the Region Of Interest box to your desired location. You can also adjust the size by clicking on the edge and dragging. For this case, we'd like to remove any moving anatomy, such as the mandible, so we're going to change to the best view for that, which is the sagittal plane. Here, you can click on the upper right-hand edge to rotate your Region Of Interest box, and then resize accordingly. Once you've set the Region Of Interest box that you're happy with, you can click on the yellow fusion button, to automatically fuse based on the anatomy within your region of interest.
After reviewing this result, if you're still not happy, you can click on the Adjust button to make fine manual adjustments to your fusion. As you can see at the very top, the MR is on the left and the CT is on the right. When you click and drag your fusion, you're actually going to be moving your CT in relation to your MR position. If you'd like to restart the image fusion process, you can always click on the Reset button on the right-hand side. This is going to reset your Region Of Interest box, and everything with the fusion. To reengage the automatic fusion process, go ahead and click on Fusion. If you're happy with the fusion result, go ahead and click on the accept button above.
The navigation panel on the right allows you to click through the different fusion pairs, to re-review your image fusion results. The image fusion pairs are automatically selected for you. What we're trying to do is fuse the CT to the largest field-of-view, highest-resolution MR. This is going to be your fusion root, and then we're going to fuse all subsequent MRs to that fusion root. If you're not happy with the way that we're creating this fusion tree, you can click on the Data tab, and then you can click on Edit to actually edit your fusion pairs manually. All you have to do is select dataset from the left and a dataset from the right, and click on Use this Pair. When you're happy with your image fusion results, you can click on Done in the bottom.
After image fusion, we move on to our Cranial Distortion Correction element. Every single MR sequence has some amount of geometric or patient-specific distortions. This is a tool that allows you to minimize these effects. The layout here looks very similar to the Image Fusion element. You can review the rigid fusion with the Spy Glass, or the same amber/blue blending tool, and if you're still happy with the fusion result, we can go ahead and click on the Calculate button to start the correction.
In this screen, you're asked a question, do you want to transfer the correction to all MR datasets, or do you want to transfer the correction to one MR dataset and break the fusions? The first option should be used if you have a 3D MR acquisition and you have different reconstructions. The second option should be used if you have different MR acquisitions, then you should go ahead and correct each scan individually. When clicking OK, the calculation will start. During the calculation, the entire cranium is broken up into hundreds of different volumes of interest. All of these different volumes are individually fused together. And then at the very end, all of these separate fusions are brought into a global solution. This gives you the elastic deformation and the corrected MR dataset.
When the correction completes, it's actually going to generate a brand new DICOM dataset that you can use for the rest of the planning process, and you could also export to any other treatment planning system or application that you'd like. In this case, we're using the CT to undistort a MP-RAGE MR sequence. The second T1 sequence would then have to go through the same process, and be corrected individually. Through our own validation processes, we've noticed that in T1 MR sequences, there can be up to around a millimeter of distortion detected.
This effect can be amplified with some of the other MR sequences like DTI, or diffusion tensor imaging, if you're trying to look at fiber tracks. We found that you can actually decrease the amount of distortion in these types of images from maybe around three millimeters to one. This difference might be clinically relevant if you're trying to identify certain structures like fiber tracks. This process can also be helpful for the gamma knife users who are only contouring their targets based on MR.
On a clinical system, this correction takes about two minutes time. Once the calculation finishes, you're going to be able to visualize the deformation matrix. As you can see here, the MR has been segmented into a grid, and in the upper right-hand corner, you see red pixels. This is actually coinciding to the voxels that were elastically deformed. If we zoom in, this grid actually adapts and becomes finer and finer as you look at anatomy of interest.
In the bottom right-hand corner of the screen, you'll see the Original button. This actually allows you to toggle between the original scan, that has no corrections done, and the corrected scan. If you click on and off this button, you'll be able to see the differences in the anatomy. If you deem the results of the correction to be acceptable, you can click the accept button above. After you've accepted the correction, you can click Done in the bottom right.
The next step in our treatment planning process is the auto-segmentation tool. Here, we have the Anatomical Mapping element. Brainlab introduced Atlas-based segmentation more than a decade ago. Atlas segmentation does a pretty good job at contouring critical structures. However, it can't account for all the variability in patient anatomy and it also can't account for all the variability in imaging modalities. We have fat suppression MRs, we have T1, T2 sequences, and the tissues are constantly changing what their values are.
When you're using golden data from an Atlas, this makes it very difficult to arrive at a accurate contour. Moving forward into anatomical mapping, we've moved away from Atlas segmentation and into a tissue classification-based model. What this means is we've gone through every single tissue type, every single structure, in CT and MR imaging modalities, and we've mapped out what are those gray values and what are those Hounsfield units associated with each type of tissue.
In doing this, we can create a full 3D model of the human body in the background of the software. Once you've corrected your MR datasets, you bring them into the Anatomical Mapping element. What you see here is the actual tissue classification that happens on your particular dataset. Here, this is a T1 dataset. So the next step, we're going to use the 3D model that we have, create a T1 dataset with all the tissues classified, and elastically morph that model onto your patient dataset.
When we click on Next, you can see the default structures are automatically created. These structures are specified based on a template that you can modify. If you look at the name of each one of your structures, you see a little empty circle beside it. This is actually the review state. The software forces you to review each one of the structures, to make sure they're accurate, before you can move on to the next step.
Here, we click on Brainstem for example, and the empty circle turns into an orange with a checkbox. This means that the structure has been reviewed, and when we click on Done, it'll actually bring that into the next step of the treatment planning process. Each structure must be reviewed before you move on to the next step. If you'd like to manually add an additional object that wasn't already automatically created by the software, click on the Data tab in the upper right-hand corner. Next you'd click on the Additional Objects drop-down, and now we can specify the image basis.
You can select the MR or the CT scan where the contour will reside. Below, you see a list of commonly-used critical structures. If you can't find an object in this default list, you can click on the search box, and manually type in a structure that you're looking for. To add this structure, you click on the plus sign beside it. Once you go back to the main screen, you'll see that that structure has been added to the list, and needs to be reviewed. As you can see in the list, Fornix Left has not been reviewed. If you don't want to use an object moving forward in the planning process, when we click Next, it'll ask you, do you want to review that object or do you want to delete it? If we move forward, it'll delete it.
The next step in the treatment planning process is target delineation. Here, we're gonna use our SmartBrush element to accomplish this. The SmartBrush technology in the Elements has been much evolved from the iPlan days. This new SmartBrush adds a lot more anatomical intelligence to the algorithm. To start with, you see our side-by-side view with the CT and the MR. After utilizing this software thoroughly, I like to start in the ACS view that you can find by clicking on the Data tab, and then the first visualization, ACS.
Here, I'm going to scroll through the dataset to find our first target. Again, if we hold down the control key and click, you're going to pan in all views to that location. And if you hold the control key and scroll your mouse wheel, you'll be able to zoom in. Once we've zoomed into a met that we're going to contour, we can select the Brush tool from the right-hand column.
Above the Brush button, you're going to see the SmartBrush toggle. This is the off position, and this is the on position. When the SmartBrush toggle is on, we're actually going to be able to do 3D interpolation and edge detection while we're contouring. To contour your 3D volume, start with a contour on one orthogonal plane, start contouring on one axial slice, and then contour on an another slice of an orthogonal plane. Once you do this, the third plane and the full 3D structure will be automatically created. After creating the 3D volume, I like to make fine adjustments on the side-by-side view. In this view, you can control which datasets you'd like to visualize. In this case, I want to visualize both MR datasets simultaneously.
Here, if I zoom in to these mets, and turn off the SmartBrush toggle, I can utilize the simple brush to go through slice by slice, and make any kind of fine adjustments that the physician sees fit. These fine adjustments can also be made in any plane. If you'd like to blow up a different reconstructed view, you'd click on the upper arrow icon, and select whatever plan you'd like to visualize. Keep in mind if you're contouring on reconstructed views, you may see a little bit of jumpiness in your adjustments, because you're limited by your slice thicknesses.
Our recommendation is to pick a plane and stick with it, go through every single slice in that plane for your final result. If you're happy with your final contour, you can go ahead and click on the Data tab, click on the drop-down toggle for that met, and you can rename it based on standard nomenclature in your clinic. We recommend that you use anatomical naming conventions. You may notice that the tumor Type is automatically set to "Tumor." This is on purpose, and this setting should not be changed, or else the planning element will not be able to read your contour as a target.
The contour Role in this particular element does not really matter. You can select PTV if you choose to, but this is mostly used for our cranial and spine SRS elements. You can also change the color of your contour, and see a volumetric report, as shown here. In this report, you can review unidimensional, bidimensional, and volumetric size specifications of your tumor. When you're happy with your tumor definition here, you can close down that tumor, and click on New Object to repeat the process for your next lesion.
You can quickly go back to the ACS view, find your target, use the hot keys to pan and zoom in to that target, and then create your 3D volume using the SmartBrush tool. Again, you only have to contour one slice in the axial plane and another slice in another reconstructed view. From there, you can go back to the side-by-side view and make any fine adjustments you may need. Once you're happy with the contour, by going back into the Data tab and selecting the drop-down for that particular met, you can change the name again, the color, or review the volumetric data report. Perform this for all the lesions in the brain, and then we can continue to the next step, which is Object Manipulation.
The next step is Object Manipulation. In this element, we can prep all of our RT structures for treatment. We can create margins for our targets, or we can perform Boolean operations with any of the other structures. For example, if we want to create a margin for this particular met, select the met from the list, and then the Operations drop-down on the right. From here, we're going to select the Margin tool, and it's automatically going to give you a pop-up where you have a slide bar that you can create positive or negative margins.
If you wanna create an asymmetric margin, you can click on the Advanced button, and you can make adjustments, head, foot, left, right, or AP. In this case, we're going to use a one-millimeter symmetric margin. At the top of the right column, you see "Left Parietal (Copy 01)." This is the new object with your margin. When you wanna save your new structure, click on the Store button. This will open up a pop-up window where you can rename the structure. If you'll notice, the Type is set to Undefined, and the Role is set to PTV. This is on purpose and should not be modified. The software will know that this is the PTV, and the original met, that has no margin, will not be considered in the planning algorithm.
You will also notice that in the notes section, it says object "+ 1mm." This is a nice way to keep track of what you actually did to that particular object. When you're ready, click on OK, and the object will save. Make sure you go through and repeat this for all targets that you plan on treating with a margin.
Now we've defined PTV margins on each one of your targets, to account for setup uncertainties. In the Multiple Brain Mets SRS planning element, all targets are treated with a single virtual isocenter. The relative distance between that virtualized isocenter and your targets may force you to reevaluate what planning margins you'd like to use. If you need to change your planning margin, you can come back to the Object Manipulation page, and create a new object with a new margin definition. However, you need to keep in mind if you had already defined a margin, you need to select that PTV, go to the Data tab, and select the dropdown for that particular previous PTV. And we need to change the Role from PTV to Undefined. This is very important so that we don't have multiple concentric targets trying to be treated by the planning element. For now, we've decided on a one-millimeter margin, so when you're finished, click Done and proceed to the planning element.
Now we can move into the actual planning element, Multiple Brain Mets SRS. When opening up this element, you'll be brought to a preparation page. On this page, you can select the machine profile that you'd like to use. Multiple Brain Mets SRS supports both flattened and unflattened beams. After that, you select the table model that you use clinically, and adjust the position of that table based on the patient's scan. Once that's been done, you can select your Hounsfield to ED unit table, and verify the treatment orientation is head to the gantry.
From here, if you're happy with the placement of that table, we can click Next and see the next two views. Again, check the alignment and continue. On the next page, we see the automatically-generated tissue model for the plan. This version of Multiple Brain Mets SRS does not allow you to modify the outer contour, but we do have heterogeneity correction enabled within the element.
Once you've reviewed the outer contour, you can click Next to continue. We've specifically designed this view to allow you to visualize the PTVs and the OARs in relation to one another. As you click through on the 3D objects, it'll open up the planes associated within your image sets. This planning system has been designed to be extremely automated. The only manual things you have to accomplish on this page are to select your clinical templates.
As you see, when we click here on the template selector, there is a Protocol and a Setup column. The Protocol defines your prescription settings, such as your prescription dose value, your coverage, or your fractionation. The Setup column allows you to select the templates that define your arc geometries. For this case, the protocol that we're gonna select is a standard RTOG protocol. This is going to automatically define prescription doses based on volumetric size of each target.
Under the Setup column, we're going to select the five arc template. This is a default template that ships with the system that actually defines a symmetric arc distribution around the skull. Please note that these default templates can be modified by the user, based on your clinical needs. We'll cover template editing guidelines in a more advanced course. Once you've selected a protocol and a setup protocol, you can go ahead and click Calculate to start the optimization. As a delivery modality here, we chose to use what we're calling an enhanced dynamic conformal arc. What this means is we're calculating every five degrees, however, we're limiting the modulation that the MLCs can perform for every 10 degrees. In this way, the treatment console will read this as a standard dynamic conformal arc, but as you'll soon see, the apertures are not very standard.
If you look at what treatment modalities have been used for multiple brain mets in the past, a lot of it is conformal type treatments. And if we're using a margin, it's usually a one-millimeter margin. We chose specifically to use dynamic conformal arcs rather than maybe volumetric modulated arc therapy, because we wanted to give you something that's very MU-efficient, and something that can QA at the highest degree possible. So on purpose, we chose not to go with a VMAT-type delivery.
During the optimization, two main steps really happen. First, the system is going to look at the delivery freedom that you gave it, based on the setup protocol. In this case, we have a five arcs. At each table angle, you can have a forward and a reverse pass, which totals a potential plan that ends up with 10 arcs. The system will determine first, if it needs to use all 10 arcs, and out of the remaining arcs, it'll try and determine what tumors should be treated with each arc.
Once it defines this, then we'd go into figuring out what the arc weighting is going to be to most efficiently deliver your prescription doses, and result in the best conformity indices possible. Novel to this software, we use the conformity index as an active penalizer to the objective function. As a result, you can expect to see the dose heterogeneity within the tumors to be slightly higher. As a result, you should expect to see a little more dose heterogeneity within each one of your targets, at least when compared to iPlan's dynamic conformal arc deliveries. This has been done intentionally, because the treatment of metastatic disease requires a higher dose heterogeneity within the targets.
Now that the plan has finished calculating, we can evaluate each target in the plan. If you click on the 3D object, it will automatically zoom in so we can review the isodose distribution in all planes. You can scroll through here and see if there's any dose bleeding into any critical organs or any adjacent targets. As you click through each one of your targets, you can notice in the upper left-hand corner the conformity index. Here, we achieved a 1.24 conformity index, next one, we have a 1.17 conformity, and the final tumor, we have a 1.18. Keep in mind, this is an inverse Paddick Conformity Index.
To analyze how these dosimetric results are possible, let's review the arc geometries generated for this plan. Click on the Data tab in the upper right-hand corner, and then click on the second view, Beam's Eye View. Here, you'll see a visualization of the MLC apertures in the main window. And then, on the left-hand corner, you'll see a 3D representation, as well as the arc geometries that were actually used. These arrows indicate arcs going in the forward or reverse directions.
As you can see here on the lower left, we used a total of five table angles, and at only two of those table angles, we used a forward and a reverse arc. When clicking through each one of these arcs, you can scroll through the entire MLC apertures. Special to this Brainlab optimization, we specify which targets should be treated with which arcs. And one technique when doing this is we never allow apertures to open up to cover two tumors that may line up in the MLC leaf direction. This can be visualized here with the green and the yellow met.
Traditional VMAT techniques will treat all targets with all arcs. And in doing this, you're going to get a lot of this bridging dose in between these two tumors, irradiating all that normal tissue in between. Here, as you can see, we focused the MLC apertures around that green met only, and then we're going to treat the yellow met with some of the other arcs in the plan. If you have access to an HDMLC, obviously the conformity numbers will become better if you use the micro MLC leaves.
In this case, you can see the yellow met being treated by both some of the fine leaves, as well as the coarse five-millimeter leaves on the outside. During the optimization, we try and minimize the amount of arcs that are treating targets with these outer leaves. However, sometimes this is unavoidable. As you can see by the conformity, we're still doing very well. However, if you want to completely avoid the use of any outside leaves, you may want to consider clustering your mets so that the field can capture all mets within the finer leaves. Keep in mind the 2.5-millimeter leaves in the HD120 MLC only span eight centimeters.
So if you'd like to only use these finer leaves, you may need to cluster your mets in a way that only allows these leaves to be used. We've found that for spherical targets, the five-millimeter leaves do a very adequate job at creating nice dosimetry. And the 2.5-millimeter leaves really come into play if you have a highly irregularly shaped object. Now that you have a better understanding of how the arc geometry has resulted in this dosimetry, let's go ahead and review the DVH curves.
First, you'd click on the Data tab, and then click on the next view, 3D view. On this page, you can visualize the DVH curves for each target individually, by selecting them in the upper left-hand corner. You can also visualize all DVH curves for all targets and all OARs, by clicking on that structure twice. If you scroll your mouse over the DVH curve, you can visualize the numeric values.
Here, we have 24 gray being delivered to 99.5% of the volume. In the upper right-hand corner, you can also see the mean, min, and max doses applied to this target, as well as a conformity and gradient index. If you'll notice in the upper right-hand box, you see another entry that's called "NT (15.0 mm)." This is the blue line that you see visualized here in the DVH curve. This represents a ring object that's been created around your target, which is used to calculate the gradient and conformity indices.
In the bottom row, we've provided additional visualizations to help you evaluate the plan. On the left, we have the maximum intensity projection, that allows you to review your isodose distribution in any plane that you choose. You can click and drag that view around in any direction. The second visualization is dose to surface. If you zoom in here, you can see where the prescription isodose level has been delivered to the surface, and where it has not.
And the final view is the dose cloud view, which actually shows you a transparent cloud, that represents the 50% isodose line around each target. If we click on the Data tab, we have one final visualization, called Fusion Review. This tab allows you to analyze your tumor edges on multiple datasets in relation to your isodose distributions. Because we're treating a distribution of mets in the brain, you may see slight changes in the positioning of your met versus a dataset. Hopefully we've corrected any distortions that come within the cranial distortion correction element. But if not, this provides a final reviewing step for your physicians.
If you'd like to make a prescription change to any one of the tumors, you can select it in the 3D view, and then click in the box where the prescription is defined. Here, you can define on the fly any prescription level that you choose for every single met. If you don't wanna treat this met with this particular isocenter, you can de-select the Treat Metastasis box, and then create a completely separate plan for this target. If you make any changes to the prescription dose level, or selection of mets that are going to be treated, you'll have to recalculate the plan by selecting the Calculate button. This will re-optimize the plan based on the changes that you made, resulting in a brand new plan.
If you're not happy with the dosimetric results of the resultant plan, the easiest thing you can do is try and change the volumetric coverage that is specified in the clinical protocol. In each protocol, you can specify one coverage for all your mets. The default protocol has coverage set at 99.5%. To make a quick improvement, you can try reducing that coverage from 99.5% to maybe 99% or 98%, and then recalculate. More advanced tips on refining your dosimetric results will be covered in future webinars.
For now let's go ahead and save this plan by clicking the Save button in the bottom right-hand corner. Here, it's going to ask you to name your plan and specify a Plan Intent. For you, it'll be Clinical, but for this system, it will be set at "MACHINE_QA." You can also specify a customized plan description. On the left-hand side, you can see a preview of the PDF that will be generated. To continue, click on Save Plan, and then you need to either save or print the PDF. Next, you can click on Export to export the plan. In the center, you'll see the Export Target options. These will be set up for you in the beginning in the installation, and will specify the different locations that you'd like to export your DICOM RT plans too.
You must also specify a course ID, and then check the box for your final review. Once you've done that, you can click on the yellow box for export. If the export is successful, you'll get a message that says the export has been completed successfully. Once you've completed the plan, you can go ahead and click on Exit in the bottom right. This will bring you back out to the workflow.
Now we can move on to patient-specific QA. Utilizing the phantom definition element, you can specify every single phantom that you have available in your clinic. These are the options that you'll see here. For this example, we're going to select the Quick Phantom that has already been defined, and click on Next. Once you've selected the phantom you're going to use, your patient plan will be transferred to it. In the left column, you can see every treatment field that was created for your patient plan. You can specify a new table angle for each one of these treatment fields. Some clinics like to specify a zero table angle for all fields, and deliver the plan coplanar.
You also have the option to zero out the monitor units for each one of the treatment arcs. By default, the isocenter position is mapped to the landmark position that you defined in phantom definition. You can also change the isocenter position by clicking on Set Isocenter, and then moving the isocenter within your different planes. You can also type in isocenter coordinates manually below. Once the isocenter has been defined, you can go ahead and calculate the plan. Once the plan has been calculated, you can review the isodose distribution to the phantom by scrolling through the slices.
If you'd like to review a different reconstruction, you can click on the arrow tabs above, and change your orientation. You can also review DVHs for any objects that you created within each phantom. For this phantom, the only object we created was the measurement volume of the ion chamber in the center. Remember, we're using a single virtual isocenter here, so you shouldn't expect to receive much dose to the isocenter position. If you're making absolute dose measurements with an ion chamber, you really want to line up the hotspots of the plan with the ion chamber measurement volume.
If you're performing absolute dose measurements with an ion chamber, we're gonna want to line up the hotspots of your target to the ion chamber measurement volume. In order to do that, first we must note down what the isocenter coordinates are. You can click on the Isocenter tab here on the right column, and read out what the isocenter coordinates are. Next, we need to find the coordinates of a treatment hotspot. Once you know the isocenter and the hotspot coordinates, you need to figure out what the directional vector is for each axis, X, Y, and Z. Once you know this vector, you need to move the isocenter in reverse of this vector to the new coordinates, to line up the hotspot to the ion chamber.
I realize this isn't the easiest thing to do, and we are addressing this in future versions of the software. However, we can also provide you an Excel spreadsheet that easily calculates these numbers. Once you have the new isocenter coordinates, go ahead and type them in in the boxes below. Once you've done that, you'll need to click Recalculate. You should see the hotspot line up directly on top of your ion chamber measurement volume, as shown here.
This is Ashland's QuiCk Phantom. And as you can see, there is no distinguishable anatomy that can be picked up with kV imaging. So this makes this phantom very difficult to position on the table. One technique we used to make the positioning very easy is we scan this phantom with ExacTrac infrared markers. We can use ExacTrac's pre-positioning feature to automatically drive the table to your measurement isocenter position.
Once you've moved the isocenter, you can save this plan and send it out to ExacTrac in the treatment console. Keep in mind, you're gonna have to do this for each target that you want to measure. Once you click the Save button, you'll specify the Plan Name, any Plan Description that you'd like, save the plan, save the PDF, and then export. And that concludes the treatment planning process. As you can see, the Elements platform offers many tools that helps automate and streamline your planning process. We hope that this video will help you plan the majority of your simple multiple mets cases. We appreciate your time, and we thank you for your business. Good luck with your planning efforts.
[00:38:35]
[music]
[00:39:01]
Man 2: Hello everyone, and welcome to the first Novalis Circle technical webinar. Today, we're gonna go through step-by-step the planning of a simple multiple brain mets SRS case. As many of you know, the Elements software is our brand new treatment planning platform. Each element has been specifically tailored to treat a specific indication. For multiple mets, this is a novel treatment planning system, that has been tailored to treat multiple targets with a single isocenter.
This is the Multiple Brain Mets SRS treatment planning workflow. As you can see, we have laid out each one of the tools in a way that guides you through the treatment planning process. From top to bottom, left to right, you'll see, we start with the DICOM Viewer tool, where you can review any DICOM images. Then we would go into the Image Fusion element, where we would rigidly register your DICOM datasets. From there, we have the Distortion Correction element, that either uses a CT or an undistorted MR to remove any geometric or patient-specific distortions that you may have in any other MR sequences.
Once that's been completed, we can open the Anatomical Mapping element, where we perform any auto-segmentation of critical structures that you may want. The SmartBrush element then allows you to contour any of your cranial lesions. The Object Manipulation element allows you to perform Boolean operations or add margins to any of your cranial lesions. Next is the actual planning element itself, Multiple Brain Mets SRS. This is where you'll be able to perform the automated treatment planning process, and generate your RT plan. The last element in the workflow is the Patient Specific QA element. This allows you to map your generated treatment plans onto any phantom you have available to you.
We'll start by reviewing the datasets available for this patient in our DICOM Viewer. This is a tool that's been designed to give you full flexibility and power over your imaging data. It's been designed in a way where it lays out different panes, and allows you to customize these views. When we open up this element, you're going to see one pane for every DICOM dataset that you've imported. As you can see here, we have the CT dataset on the left, and the two MRs in the middle and on the right.
As you can see while I'm scrolling through each one of these datasets, the rest of the datasets are not moving. This is because we have not yet fused the datasets. If you'd like to create additional views, you simply click on the head icon at the top of the pane. This will open up the options that you have available for that particular dataset. As you can see here, you can create different reconstructed orientations from axial, sagittal, and coronal, and you can also create image modality-specific 3D representations as well, such as a skin filter or a bone filter with the CT set.
Once that's been created, you can go ahead and zoom or rotate the pane. And once you're finished with that visualization, you can simply click on the x box at the top, and that will remove the pane from your screen. One thing to keep in mind when you're working through the different Elements workflows is all the tools that you would normally use are going to be located on the right-hand column. As you can see here, in the View tab, there's multiple different view options that you can select. You can change the windowing. You can perform image slice thickness averaging. This allows you to see more information on each one of your image slices. You can take a screenshot, and you can also add additional views.
Below the View options are your measurement tools. Here, you can see we can measure distances, circle diameters, angles, and drop points for later viewing. If we wanna measure the distance of this particular metastasis, you can hold down the control key and use your scroll wheel to zoom in. If you hold down control, and click with your left mouse button, you can actually pan the image to center the target. By clicking the Distance button, you activate that tool.
You can start the measurement by clicking on the edge of the particular met you're looking at, and then you drag the second circle to the opposite edge. And then you're gonna see the measurement icon above your tool. To delete that measurement, simply click on the garbage bin on the right column, and then click on that arrow. To measure diameter, click on the Circle tool, then click in the center of the target to create your measurement tool. To move the position, click in the center of that circle, and drag to the middle of your target. To adjust the diameter, click on the periphery, and drag inwards or outwards.
The DICOM Viewer element is a powerful tool that can be very helpful in tumor boards or any presentation when you need to review your DICOM data. Once you've finished reviewing the DICOM data that you have available, we can move on to the next step, which is Image Fusion.
Next, we'll go into our Image Fusion element. When you open the Image Fusion element, your DICOM datasets will automatically be registered together. As you can see here, the two MR datasets have been automatically fused. You can review the fusion with the Spy Glass, moving it around by clicking and dragging. And if you want to change the size of the Spy Glass, you can click and drag from one of the corners. You can change your view orientation by clicking in the upper left-hand corner, on that Axial tab, and now you can change to coronal or sagittal. We also have the ability to review the fusion with the amber blue blending tool that you may be used to from iPlan.
If you're happy with the results, you can go ahead and click on accept the fusion, at the top. Once you accept the fusion, it'll automatically move to the next fusion pair. Here, you can see the CT being fused to the MR. If you'd like to change the windowing, you can click on the windowing icon, and then choose which dataset you would like to change the window level. Here, for the CT, we can select that in the orange, and then you can click and drag your mouse on the screen to change the window level. When you're happy with the result, go ahead and click the windowing button one more time, to deactivate the tool. If you're not happy with the image fusion result, you can click on one of the tools to manually adjust your fusion.
The first one is the Region Of Interest box. Here, you can see a visualization of your region of interest that you're going to define for your fusion. By clicking in the center, you can click and drag the Region Of Interest box to your desired location. You can also adjust the size by clicking on the edge and dragging. For this case, we'd like to remove any moving anatomy, such as the mandible, so we're going to change to the best view for that, which is the sagittal plane. Here, you can click on the upper right-hand edge to rotate your Region Of Interest box, and then resize accordingly. Once you've set the Region Of Interest box that you're happy with, you can click on the yellow fusion button, to automatically fuse based on the anatomy within your region of interest.
After reviewing this result, if you're still not happy, you can click on the Adjust button to make fine manual adjustments to your fusion. As you can see at the very top, the MR is on the left and the CT is on the right. When you click and drag your fusion, you're actually going to be moving your CT in relation to your MR position. If you'd like to restart the image fusion process, you can always click on the Reset button on the right-hand side. This is going to reset your Region Of Interest box, and everything with the fusion. To reengage the automatic fusion process, go ahead and click on Fusion. If you're happy with the fusion result, go ahead and click on the accept button above.
The navigation panel on the right allows you to click through the different fusion pairs, to re-review your image fusion results. The image fusion pairs are automatically selected for you. What we're trying to do is fuse the CT to the largest field-of-view, highest-resolution MR. This is going to be your fusion root, and then we're going to fuse all subsequent MRs to that fusion root. If you're not happy with the way that we're creating this fusion tree, you can click on the Data tab, and then you can click on Edit to actually edit your fusion pairs manually. All you have to do is select dataset from the left and a dataset from the right, and click on Use this Pair. When you're happy with your image fusion results, you can click on Done in the bottom.
After image fusion, we move on to our Cranial Distortion Correction element. Every single MR sequence has some amount of geometric or patient-specific distortions. This is a tool that allows you to minimize these effects. The layout here looks very similar to the Image Fusion element. You can review the rigid fusion with the Spy Glass, or the same amber/blue blending tool, and if you're still happy with the fusion result, we can go ahead and click on the Calculate button to start the correction.
In this screen, you're asked a question, do you want to transfer the correction to all MR datasets, or do you want to transfer the correction to one MR dataset and break the fusions? The first option should be used if you have a 3D MR acquisition and you have different reconstructions. The second option should be used if you have different MR acquisitions, then you should go ahead and correct each scan individually. When clicking OK, the calculation will start. During the calculation, the entire cranium is broken up into hundreds of different volumes of interest. All of these different volumes are individually fused together. And then at the very end, all of these separate fusions are brought into a global solution. This gives you the elastic deformation and the corrected MR dataset.
When the correction completes, it's actually going to generate a brand new DICOM dataset that you can use for the rest of the planning process, and you could also export to any other treatment planning system or application that you'd like. In this case, we're using the CT to undistort a MP-RAGE MR sequence. The second T1 sequence would then have to go through the same process, and be corrected individually. Through our own validation processes, we've noticed that in T1 MR sequences, there can be up to around a millimeter of distortion detected.
This effect can be amplified with some of the other MR sequences like DTI, or diffusion tensor imaging, if you're trying to look at fiber tracks. We found that you can actually decrease the amount of distortion in these types of images from maybe around three millimeters to one. This difference might be clinically relevant if you're trying to identify certain structures like fiber tracks. This process can also be helpful for the gamma knife users who are only contouring their targets based on MR.
On a clinical system, this correction takes about two minutes time. Once the calculation finishes, you're going to be able to visualize the deformation matrix. As you can see here, the MR has been segmented into a grid, and in the upper right-hand corner, you see red pixels. This is actually coinciding to the voxels that were elastically deformed. If we zoom in, this grid actually adapts and becomes finer and finer as you look at anatomy of interest.
In the bottom right-hand corner of the screen, you'll see the Original button. This actually allows you to toggle between the original scan, that has no corrections done, and the corrected scan. If you click on and off this button, you'll be able to see the differences in the anatomy. If you deem the results of the correction to be acceptable, you can click the accept button above. After you've accepted the correction, you can click Done in the bottom right.
The next step in our treatment planning process is the auto-segmentation tool. Here, we have the Anatomical Mapping element. Brainlab introduced Atlas-based segmentation more than a decade ago. Atlas segmentation does a pretty good job at contouring critical structures. However, it can't account for all the variability in patient anatomy and it also can't account for all the variability in imaging modalities. We have fat suppression MRs, we have T1, T2 sequences, and the tissues are constantly changing what their values are.
When you're using golden data from an Atlas, this makes it very difficult to arrive at a accurate contour. Moving forward into anatomical mapping, we've moved away from Atlas segmentation and into a tissue classification-based model. What this means is we've gone through every single tissue type, every single structure, in CT and MR imaging modalities, and we've mapped out what are those gray values and what are those Hounsfield units associated with each type of tissue.
In doing this, we can create a full 3D model of the human body in the background of the software. Once you've corrected your MR datasets, you bring them into the Anatomical Mapping element. What you see here is the actual tissue classification that happens on your particular dataset. Here, this is a T1 dataset. So the next step, we're going to use the 3D model that we have, create a T1 dataset with all the tissues classified, and elastically morph that model onto your patient dataset.
When we click on Next, you can see the default structures are automatically created. These structures are specified based on a template that you can modify. If you look at the name of each one of your structures, you see a little empty circle beside it. This is actually the review state. The software forces you to review each one of the structures, to make sure they're accurate, before you can move on to the next step.
Here, we click on Brainstem for example, and the empty circle turns into an orange with a checkbox. This means that the structure has been reviewed, and when we click on Done, it'll actually bring that into the next step of the treatment planning process. Each structure must be reviewed before you move on to the next step. If you'd like to manually add an additional object that wasn't already automatically created by the software, click on the Data tab in the upper right-hand corner. Next you'd click on the Additional Objects drop-down, and now we can specify the image basis.
You can select the MR or the CT scan where the contour will reside. Below, you see a list of commonly-used critical structures. If you can't find an object in this default list, you can click on the search box, and manually type in a structure that you're looking for. To add this structure, you click on the plus sign beside it. Once you go back to the main screen, you'll see that that structure has been added to the list, and needs to be reviewed. As you can see in the list, Fornix Left has not been reviewed. If you don't want to use an object moving forward in the planning process, when we click Next, it'll ask you, do you want to review that object or do you want to delete it? If we move forward, it'll delete it.
The next step in the treatment planning process is target delineation. Here, we're gonna use our SmartBrush element to accomplish this. The SmartBrush technology in the Elements has been much evolved from the iPlan days. This new SmartBrush adds a lot more anatomical intelligence to the algorithm. To start with, you see our side-by-side view with the CT and the MR. After utilizing this software thoroughly, I like to start in the ACS view that you can find by clicking on the Data tab, and then the first visualization, ACS.
Here, I'm going to scroll through the dataset to find our first target. Again, if we hold down the control key and click, you're going to pan in all views to that location. And if you hold the control key and scroll your mouse wheel, you'll be able to zoom in. Once we've zoomed into a met that we're going to contour, we can select the Brush tool from the right-hand column.
Above the Brush button, you're going to see the SmartBrush toggle. This is the off position, and this is the on position. When the SmartBrush toggle is on, we're actually going to be able to do 3D interpolation and edge detection while we're contouring. To contour your 3D volume, start with a contour on one orthogonal plane, start contouring on one axial slice, and then contour on an another slice of an orthogonal plane. Once you do this, the third plane and the full 3D structure will be automatically created. After creating the 3D volume, I like to make fine adjustments on the side-by-side view. In this view, you can control which datasets you'd like to visualize. In this case, I want to visualize both MR datasets simultaneously.
Here, if I zoom in to these mets, and turn off the SmartBrush toggle, I can utilize the simple brush to go through slice by slice, and make any kind of fine adjustments that the physician sees fit. These fine adjustments can also be made in any plane. If you'd like to blow up a different reconstructed view, you'd click on the upper arrow icon, and select whatever plan you'd like to visualize. Keep in mind if you're contouring on reconstructed views, you may see a little bit of jumpiness in your adjustments, because you're limited by your slice thicknesses.
Our recommendation is to pick a plane and stick with it, go through every single slice in that plane for your final result. If you're happy with your final contour, you can go ahead and click on the Data tab, click on the drop-down toggle for that met, and you can rename it based on standard nomenclature in your clinic. We recommend that you use anatomical naming conventions. You may notice that the tumor Type is automatically set to "Tumor." This is on purpose, and this setting should not be changed, or else the planning element will not be able to read your contour as a target.
The contour Role in this particular element does not really matter. You can select PTV if you choose to, but this is mostly used for our cranial and spine SRS elements. You can also change the color of your contour, and see a volumetric report, as shown here. In this report, you can review unidimensional, bidimensional, and volumetric size specifications of your tumor. When you're happy with your tumor definition here, you can close down that tumor, and click on New Object to repeat the process for your next lesion.
You can quickly go back to the ACS view, find your target, use the hot keys to pan and zoom in to that target, and then create your 3D volume using the SmartBrush tool. Again, you only have to contour one slice in the axial plane and another slice in another reconstructed view. From there, you can go back to the side-by-side view and make any fine adjustments you may need. Once you're happy with the contour, by going back into the Data tab and selecting the drop-down for that particular met, you can change the name again, the color, or review the volumetric data report. Perform this for all the lesions in the brain, and then we can continue to the next step, which is Object Manipulation.
The next step is Object Manipulation. In this element, we can prep all of our RT structures for treatment. We can create margins for our targets, or we can perform Boolean operations with any of the other structures. For example, if we want to create a margin for this particular met, select the met from the list, and then the Operations drop-down on the right. From here, we're going to select the Margin tool, and it's automatically going to give you a pop-up where you have a slide bar that you can create positive or negative margins.
If you wanna create an asymmetric margin, you can click on the Advanced button, and you can make adjustments, head, foot, left, right, or AP. In this case, we're going to use a one-millimeter symmetric margin. At the top of the right column, you see "Left Parietal (Copy 01)." This is the new object with your margin. When you wanna save your new structure, click on the Store button. This will open up a pop-up window where you can rename the structure. If you'll notice, the Type is set to Undefined, and the Role is set to PTV. This is on purpose and should not be modified. The software will know that this is the PTV, and the original met, that has no margin, will not be considered in the planning algorithm.
You will also notice that in the notes section, it says object "+ 1mm." This is a nice way to keep track of what you actually did to that particular object. When you're ready, click on OK, and the object will save. Make sure you go through and repeat this for all targets that you plan on treating with a margin.
Now we've defined PTV margins on each one of your targets, to account for setup uncertainties. In the Multiple Brain Mets SRS planning element, all targets are treated with a single virtual isocenter. The relative distance between that virtualized isocenter and your targets may force you to reevaluate what planning margins you'd like to use. If you need to change your planning margin, you can come back to the Object Manipulation page, and create a new object with a new margin definition. However, you need to keep in mind if you had already defined a margin, you need to select that PTV, go to the Data tab, and select the dropdown for that particular previous PTV. And we need to change the Role from PTV to Undefined. This is very important so that we don't have multiple concentric targets trying to be treated by the planning element. For now, we've decided on a one-millimeter margin, so when you're finished, click Done and proceed to the planning element.
Now we can move into the actual planning element, Multiple Brain Mets SRS. When opening up this element, you'll be brought to a preparation page. On this page, you can select the machine profile that you'd like to use. Multiple Brain Mets SRS supports both flattened and unflattened beams. After that, you select the table model that you use clinically, and adjust the position of that table based on the patient's scan. Once that's been done, you can select your Hounsfield to ED unit table, and verify the treatment orientation is head to the gantry.
From here, if you're happy with the placement of that table, we can click Next and see the next two views. Again, check the alignment and continue. On the next page, we see the automatically-generated tissue model for the plan. This version of Multiple Brain Mets SRS does not allow you to modify the outer contour, but we do have heterogeneity correction enabled within the element.
Once you've reviewed the outer contour, you can click Next to continue. We've specifically designed this view to allow you to visualize the PTVs and the OARs in relation to one another. As you click through on the 3D objects, it'll open up the planes associated within your image sets. This planning system has been designed to be extremely automated. The only manual things you have to accomplish on this page are to select your clinical templates.
As you see, when we click here on the template selector, there is a Protocol and a Setup column. The Protocol defines your prescription settings, such as your prescription dose value, your coverage, or your fractionation. The Setup column allows you to select the templates that define your arc geometries. For this case, the protocol that we're gonna select is a standard RTOG protocol. This is going to automatically define prescription doses based on volumetric size of each target.
Under the Setup column, we're going to select the five arc template. This is a default template that ships with the system that actually defines a symmetric arc distribution around the skull. Please note that these default templates can be modified by the user, based on your clinical needs. We'll cover template editing guidelines in a more advanced course. Once you've selected a protocol and a setup protocol, you can go ahead and click Calculate to start the optimization. As a delivery modality here, we chose to use what we're calling an enhanced dynamic conformal arc. What this means is we're calculating every five degrees, however, we're limiting the modulation that the MLCs can perform for every 10 degrees. In this way, the treatment console will read this as a standard dynamic conformal arc, but as you'll soon see, the apertures are not very standard.
If you look at what treatment modalities have been used for multiple brain mets in the past, a lot of it is conformal type treatments. And if we're using a margin, it's usually a one-millimeter margin. We chose specifically to use dynamic conformal arcs rather than maybe volumetric modulated arc therapy, because we wanted to give you something that's very MU-efficient, and something that can QA at the highest degree possible. So on purpose, we chose not to go with a VMAT-type delivery.
During the optimization, two main steps really happen. First, the system is going to look at the delivery freedom that you gave it, based on the setup protocol. In this case, we have a five arcs. At each table angle, you can have a forward and a reverse pass, which totals a potential plan that ends up with 10 arcs. The system will determine first, if it needs to use all 10 arcs, and out of the remaining arcs, it'll try and determine what tumors should be treated with each arc.
Once it defines this, then we'd go into figuring out what the arc weighting is going to be to most efficiently deliver your prescription doses, and result in the best conformity indices possible. Novel to this software, we use the conformity index as an active penalizer to the objective function. As a result, you can expect to see the dose heterogeneity within the tumors to be slightly higher. As a result, you should expect to see a little more dose heterogeneity within each one of your targets, at least when compared to iPlan's dynamic conformal arc deliveries. This has been done intentionally, because the treatment of metastatic disease requires a higher dose heterogeneity within the targets.
Now that the plan has finished calculating, we can evaluate each target in the plan. If you click on the 3D object, it will automatically zoom in so we can review the isodose distribution in all planes. You can scroll through here and see if there's any dose bleeding into any critical organs or any adjacent targets. As you click through each one of your targets, you can notice in the upper left-hand corner the conformity index. Here, we achieved a 1.24 conformity index, next one, we have a 1.17 conformity, and the final tumor, we have a 1.18. Keep in mind, this is an inverse Paddick Conformity Index.
To analyze how these dosimetric results are possible, let's review the arc geometries generated for this plan. Click on the Data tab in the upper right-hand corner, and then click on the second view, Beam's Eye View. Here, you'll see a visualization of the MLC apertures in the main window. And then, on the left-hand corner, you'll see a 3D representation, as well as the arc geometries that were actually used. These arrows indicate arcs going in the forward or reverse directions.
As you can see here on the lower left, we used a total of five table angles, and at only two of those table angles, we used a forward and a reverse arc. When clicking through each one of these arcs, you can scroll through the entire MLC apertures. Special to this Brainlab optimization, we specify which targets should be treated with which arcs. And one technique when doing this is we never allow apertures to open up to cover two tumors that may line up in the MLC leaf direction. This can be visualized here with the green and the yellow met.
Traditional VMAT techniques will treat all targets with all arcs. And in doing this, you're going to get a lot of this bridging dose in between these two tumors, irradiating all that normal tissue in between. Here, as you can see, we focused the MLC apertures around that green met only, and then we're going to treat the yellow met with some of the other arcs in the plan. If you have access to an HDMLC, obviously the conformity numbers will become better if you use the micro MLC leaves.
In this case, you can see the yellow met being treated by both some of the fine leaves, as well as the coarse five-millimeter leaves on the outside. During the optimization, we try and minimize the amount of arcs that are treating targets with these outer leaves. However, sometimes this is unavoidable. As you can see by the conformity, we're still doing very well. However, if you want to completely avoid the use of any outside leaves, you may want to consider clustering your mets so that the field can capture all mets within the finer leaves. Keep in mind the 2.5-millimeter leaves in the HD120 MLC only span eight centimeters.
So if you'd like to only use these finer leaves, you may need to cluster your mets in a way that only allows these leaves to be used. We've found that for spherical targets, the five-millimeter leaves do a very adequate job at creating nice dosimetry. And the 2.5-millimeter leaves really come into play if you have a highly irregularly shaped object. Now that you have a better understanding of how the arc geometry has resulted in this dosimetry, let's go ahead and review the DVH curves.
First, you'd click on the Data tab, and then click on the next view, 3D view. On this page, you can visualize the DVH curves for each target individually, by selecting them in the upper left-hand corner. You can also visualize all DVH curves for all targets and all OARs, by clicking on that structure twice. If you scroll your mouse over the DVH curve, you can visualize the numeric values.
Here, we have 24 gray being delivered to 99.5% of the volume. In the upper right-hand corner, you can also see the mean, min, and max doses applied to this target, as well as a conformity and gradient index. If you'll notice in the upper right-hand box, you see another entry that's called "NT (15.0 mm)." This is the blue line that you see visualized here in the DVH curve. This represents a ring object that's been created around your target, which is used to calculate the gradient and conformity indices.
In the bottom row, we've provided additional visualizations to help you evaluate the plan. On the left, we have the maximum intensity projection, that allows you to review your isodose distribution in any plane that you choose. You can click and drag that view around in any direction. The second visualization is dose to surface. If you zoom in here, you can see where the prescription isodose level has been delivered to the surface, and where it has not.
And the final view is the dose cloud view, which actually shows you a transparent cloud, that represents the 50% isodose line around each target. If we click on the Data tab, we have one final visualization, called Fusion Review. This tab allows you to analyze your tumor edges on multiple datasets in relation to your isodose distributions. Because we're treating a distribution of mets in the brain, you may see slight changes in the positioning of your met versus a dataset. Hopefully we've corrected any distortions that come within the cranial distortion correction element. But if not, this provides a final reviewing step for your physicians.
If you'd like to make a prescription change to any one of the tumors, you can select it in the 3D view, and then click in the box where the prescription is defined. Here, you can define on the fly any prescription level that you choose for every single met. If you don't wanna treat this met with this particular isocenter, you can de-select the Treat Metastasis box, and then create a completely separate plan for this target. If you make any changes to the prescription dose level, or selection of mets that are going to be treated, you'll have to recalculate the plan by selecting the Calculate button. This will re-optimize the plan based on the changes that you made, resulting in a brand new plan.
If you're not happy with the dosimetric results of the resultant plan, the easiest thing you can do is try and change the volumetric coverage that is specified in the clinical protocol. In each protocol, you can specify one coverage for all your mets. The default protocol has coverage set at 99.5%. To make a quick improvement, you can try reducing that coverage from 99.5% to maybe 99% or 98%, and then recalculate. More advanced tips on refining your dosimetric results will be covered in future webinars.
For now let's go ahead and save this plan by clicking the Save button in the bottom right-hand corner. Here, it's going to ask you to name your plan and specify a Plan Intent. For you, it'll be Clinical, but for this system, it will be set at "MACHINE_QA." You can also specify a customized plan description. On the left-hand side, you can see a preview of the PDF that will be generated. To continue, click on Save Plan, and then you need to either save or print the PDF. Next, you can click on Export to export the plan. In the center, you'll see the Export Target options. These will be set up for you in the beginning in the installation, and will specify the different locations that you'd like to export your DICOM RT plans too.
You must also specify a course ID, and then check the box for your final review. Once you've done that, you can click on the yellow box for export. If the export is successful, you'll get a message that says the export has been completed successfully. Once you've completed the plan, you can go ahead and click on Exit in the bottom right. This will bring you back out to the workflow.
Now we can move on to patient-specific QA. Utilizing the phantom definition element, you can specify every single phantom that you have available in your clinic. These are the options that you'll see here. For this example, we're going to select the Quick Phantom that has already been defined, and click on Next. Once you've selected the phantom you're going to use, your patient plan will be transferred to it. In the left column, you can see every treatment field that was created for your patient plan. You can specify a new table angle for each one of these treatment fields. Some clinics like to specify a zero table angle for all fields, and deliver the plan coplanar.
You also have the option to zero out the monitor units for each one of the treatment arcs. By default, the isocenter position is mapped to the landmark position that you defined in phantom definition. You can also change the isocenter position by clicking on Set Isocenter, and then moving the isocenter within your different planes. You can also type in isocenter coordinates manually below. Once the isocenter has been defined, you can go ahead and calculate the plan. Once the plan has been calculated, you can review the isodose distribution to the phantom by scrolling through the slices.
If you'd like to review a different reconstruction, you can click on the arrow tabs above, and change your orientation. You can also review DVHs for any objects that you created within each phantom. For this phantom, the only object we created was the measurement volume of the ion chamber in the center. Remember, we're using a single virtual isocenter here, so you shouldn't expect to receive much dose to the isocenter position. If you're making absolute dose measurements with an ion chamber, you really want to line up the hotspots of the plan with the ion chamber measurement volume.
If you're performing absolute dose measurements with an ion chamber, we're gonna want to line up the hotspots of your target to the ion chamber measurement volume. In order to do that, first we must note down what the isocenter coordinates are. You can click on the Isocenter tab here on the right column, and read out what the isocenter coordinates are. Next, we need to find the coordinates of a treatment hotspot. Once you know the isocenter and the hotspot coordinates, you need to figure out what the directional vector is for each axis, X, Y, and Z. Once you know this vector, you need to move the isocenter in reverse of this vector to the new coordinates, to line up the hotspot to the ion chamber.
I realize this isn't the easiest thing to do, and we are addressing this in future versions of the software. However, we can also provide you an Excel spreadsheet that easily calculates these numbers. Once you have the new isocenter coordinates, go ahead and type them in in the boxes below. Once you've done that, you'll need to click Recalculate. You should see the hotspot line up directly on top of your ion chamber measurement volume, as shown here.
This is Ashland's QuiCk Phantom. And as you can see, there is no distinguishable anatomy that can be picked up with kV imaging. So this makes this phantom very difficult to position on the table. One technique we used to make the positioning very easy is we scan this phantom with ExacTrac infrared markers. We can use ExacTrac's pre-positioning feature to automatically drive the table to your measurement isocenter position.
Once you've moved the isocenter, you can save this plan and send it out to ExacTrac in the treatment console. Keep in mind, you're gonna have to do this for each target that you want to measure. Once you click the Save button, you'll specify the Plan Name, any Plan Description that you'd like, save the plan, save the PDF, and then export. And that concludes the treatment planning process. As you can see, the Elements platform offers many tools that helps automate and streamline your planning process. We hope that this video will help you plan the majority of your simple multiple mets cases. We appreciate your time, and we thank you for your business. Good luck with your planning efforts.
[00:38:35]
[music]
[00:39:01]