Implementing AlignRT and MapRT into a Clinical Workflow

So my talk today is on implementing AlignRT and MapRT into a clinical workflow. So I have no conflicts of interest to disclose. Some of the objectives we’re going to talk about today is to describe a multidisciplinary implementation strategy that we use at CHLA to successfully integrate AlignRT and MapRT into the clinical workflow, evaluate the positive and negative impacts of introducing new SGRT technologies on the radiation therapy workflow, identify common challenges encountered during implementation, which includes staff buy-in, training, time constraints, and describe strategies to overcome these, and also compare different implementation approaches and discuss how strategies may vary across different clinics.

So Children’s Hospital Los Angeles had implemented AlignRT into the clinical practice and achieved a completely markerless workflow within nine months of the initial implementation. This was back in 2012, and then we followed the introduction of MapRT into the clinical workflow in 2015, and we eliminated the need for physical dry runs and expanded the clinical use of non-coplanar fields for complex treatments. The implementation of AlignRT and MapRT at CHLA followed a multidisciplinary team-based approach. Input and feedback from therapists, physicists, and physicians were incorporated throughout the process, and that helped us result in a more effective and sustainable application of surface guided radiation therapy or SGRT.

So these are just a couple pictures of the MapRT cameras that are in our simulation room. And so MapRT is installed in the simulation room in a two-camera configuration, and it uses 3D surface captures of the patient and its immobilization devices, along with the dedicated software to be able to identify and avoid unsafe gantry/couch configurations. So it is a patient-specific surface data which predicts more accurately than a generic database. And again, MapRT was implemented for us in 2025 into clinical use. And we use MapRT on all cases, regardless of the complexity. And so for us, the use of MapRT expanded the clinical use of non-coplanar fields for complex treatments by allowing for more complex beam arrangements. It provides a significantly higher confidence during treatment planning and delivery by accurately identifying potential patient and machine collision risks.

So these are just a couple pictures of the MapRT software. I purposely chose more adult-like patients, so you guys could imagine what it would be like if you guys were using it as well. So I’ll start with this one over here. This is a patient that we had that had to be, you can see, tilted up. They were having some breathing issues, so we tilted them up to be able to breathe better, and then we were able to use MapRT to verify that it would not collide during the treatment. And this is not standard for us. Usually, we try to have our patients lay flat. The middle is just a picture of a collision. I purposely made it collide so you could see how obvious it is that there would be a collision. And then you can see in this picture as well, you can see there’s a couch kick and a gantry kick. And in this, you can see that if we had used a database or didn’t have the patient available to see, as a therapist doing a dry run or even a physicist, you might say it clears. However, let’s say a patient had a larger belly or there was something sticking out, that would not have been identified prior to MapRT. So those are very helpful for us because a dry run cannot capture what if you don’t have a patient in. So sometimes prior to MapRT, we physically would lay get up on the table, see if it would work, just to test it prior to the patient coming.

During simulation, some patients require positioning that falls outside of the standard setup due to anatomy, disease location, immobilization devices, or comfort considerations. So these atypical positions can really increase the risk of gantry, couch, or accessory collisions if they are not evaluated carefully. And so MapRT provides tools to assess clearance and collision risk at multiple stages of the workflow. Clearance can be evaluated by the therapist prior to even scanning the patient, using in-room software to ensure that the proposed setup is safe. Typically, if you’re going to do a setup like these, these are the emergency cases or the cases that are in pain or something that’s a little bit different than your standard, so they’re a little bit more stressful. So we found that having these… Sorry, I’ll go back. Having that in there gives us also that assurance that however we simulate them, we will also be able to treat them. And then MapRT also allows a user to apply a configurable buffer within the software, and that accounts for the expected daily setup shifts and variations in patient positions to ensure collision-free treatment. So you might have a perfect plan that looks great, but then when you bring the patient into the table, let’s say they’ve moved one cm or there’s a couch roll or something, you can actually allow MapRT to put this in, so you can take that into consideration. For us, we have a six-dof couch, which having that extra buffer helps us because then we’re able to see, like if we were to apply a one-degree couch kick, would it still collide, or are we still clear?

So this is a patient picture of what we were setting up. You can see right here their arm is up, and then their other arm was down. We were trying to treat in the lung. And so you can actually see you could get the gantry closer, and if we had just simulated them, we may not have known that ahead of time. You can actually, as a therapist, go into the MapRT setup and if you know, okay, I want to treat at a gantry angle of, let’s say, 20 and the couch at 10, for whatever reason. Let’s say that’s your standard setup for something, a breast patient with their arm up. You can actually go into the software and find it and see if it’ll collide or not, and find if your arm needs to go further up or further down. We typically don’t do that, but it is something that’s available. We don’t have a lot of cases where the arms are up like this, but it is available.

And so some of the benefits of MapRT for us is it’s eliminated the need for physical dry runs that are required by a therapist prior to treatment. For those that run a very tight schedule, you don’t have a lot of time for extra dry runs, even if it’s only five, 10 minutes. Or physics says, “Hey, can you check one gantry angle for me?” You still have to go in the room, takes at least five to 10 minutes. With MapRT, you can take that out, and then you can just use it from the software. And so this workflow modification allows therapists to complete new start chart QA independent of machine availability. So again, at busier centers, this is a huge time saving. Implementation of MapRT had a negligible impact to both therapists as well as physics and dosimetry workflow.

So for a therapist during the simulation process, we create our immobilization device. We position the patient. We turn on MapRT. It adds maybe two to three minutes tops. It’s just a couple quick clicks, and I have a couple captures to show you what that’s like. The simulation therapist captures an image, multiple images if necessary, if it’s a little more complex. You merge them into a final surface, select Finish, and you’re done. Two to three minutes is being very generous. I would say adds maybe a minute tops, if that. The planning process, it just requires an export to a MapRT software to check for those collisions. It’s a little more than that, but I’ll let our physicists talk about them later. By shifting collision evaluation to a software-based work, MapRT supports safe planning practices while preserving valuable machine availability for patient care. Again, we’re not having to take these extra slots now to do dry runs. Therapists aren’t having to stay late to check them. Physics isn’t needing the machine at any point. It’s been huge for us.

And so these are just pictures, if you were a therapist, what you might see. And so when you take your first capture—oh, sorry, it’s on this side. You’ll take a capture of the room. I just put the plate on so you guys could see it, and you’ll just select two little dots, and you hit Generate Setup Surface. Simple as that. And then if you want to take a second one, you can move the couch along, do a second one, hit Capture Setup. And then you’ll end up with—this is again, just the phantom plate—but you’ll end up with completed images, and then you’ll select Finish. If there are holes in it, you can see there’s a little bit right here, but let’s say if it was bigger, you can move the table in and out to fully capture. So if you’re doing a breast setup and the arm was up, but you were closer to the CT, you can bring the table out, capture, and then you’ll get a full capture of the arm and everything else that’s included.

And so this is just the MapRT software real quick for you guys to see. You just select the patient and how you want to treat them. There’s a whole lot of different things that can be customized to your clinic, and then it comes up with some software. And I just selected this one. So you can move the gantry however you want by dragging this point. So you could see where it would collide and where it might not. And then you can see the gantry angle. It’ll actually give you the gantry and the couch angle. So if you were doing planning, you could decide what those might be.

So some quick tips for MapRT. As a therapist, when they wanted to introduce new technology, once again, we were like: We don’t have time. We’re pretty busy. I don’t know how this is going to work. Turns out it was pretty simple for us. Warm-up is one to two minutes. Actual treatment is one to two minutes, and then we don’t have to do dry runs, which I keep repeating myself, but it is huge as a therapist. So some of the things that we found that can be a little bit tricky is when you’re merging the data sets. There’s an outermost CT travel table limit, where if you go too far out, those white dots are no longer visible, and you won’t be able to take a capture. So what we did, we added a reference line to the end of the table that indicates the maximum allowable position where the camera can still visualize the white dots for registration, so you don’t even have to think about it. And I have a picture in one second.

Also, objects and people can block the camera system, so always double-check before you take. I work at a children’s hospital. Sometimes we have anesthesia, so we’ll just tell everyone, “Hey, go to the back corner.” They’ll go stand there for 30 seconds. We take a couple captures. We’ve taken them with people in the way. It really hasn’t changed that much. As long as you look at that final merge surface and it doesn’t have any holes, you’re good to go. Also, a big thing that we found is make sure that there are no blankets or clothing or anything that is blocking it because it could be interpreted by the system as a collision, even though it’s just a blanket. So if you’re covering the patient and it kind of bulges up, it’s actually going to think that’s a patient. So what we do is routinely before we take a capture, move everything out of the way as if we were going to do a treatment. Anything bulky, we just take out of the way so it doesn’t later say you’re having a collision even though it technically was not there.

So this is our handy little mark that we made. Here it is, sorry. So I didn’t want to get in trouble, so I used a piece of paper tape first, and then literally just brought the CT out and drew a line. And you can see it really merges in, and the paper tape doesn’t get up or down. But now you know when I’m just releasing the table, I go exactly to this point, and I know exactly how far to go. Because one of the frustrating things in the beginning is you bring the table all the way out, you couldn’t find it. And so as a therapist, we said, “How can we make this faster?” We just made a little mark. You could do it however you wanted, but tape, nobody got mad at, so we did that.

Okay. So the warm-up, this kind of includes both of it, but for AlignRT and MapRT, it adds three to five minutes daily for QA. And again, this is tops. If you’re still learning, it doesn’t really take that long. So AlignRT requires a daily setup with the plate. We also added an extra daily warm-up set, which I’ll show in a second, that we wanted to check that when the tolerance is out, so it’s red, that the beam actually shuts off, and we want to practice that on a patient every day. Also, a quick tip on here, you can see on the picture for MapRT, for those that have it, you do a quick capture, you move the table in 250, and then you recapture. So what I did was—it’s kind of hard to… oh, sorry, good, go back. There’s a little dot, it’s probably hard to see, right here, and that indicates right around that 250 value. So I literally, I’ll set up to the zero. I zero the couch, go to the dot, and I should be pretty close to 250 and fine tune it, just to make it faster for us. Again, even if you didn’t do that, it still does not take very long.

So the AlignRT warm-up consists of the daily plate, and we added one additional test. Part of this was due to APEX accreditation. They were asking about it, so we just came up with something on our own. So we added an additional test that verifies the functional integration between AlignRT with the linear accelerator, ensuring that the motion outside of tolerance correctly triggers a beam hold interlock, providing an added layer of safety. It’s not super high tech. So during warm-up, the therapist positions a motion device, tracking, which is our old Varian RPM that we don’t use anymore, and then we just found—this is actually from Astro—it’s just like a stress ball that was a CT. It was still not picking up everything great, so working at Children’s Hospital, I found a little rubber duck that we added to it, and we just taped them all together. So now when the motion goes, it’s very easy for AlignRT to set up.

And you can see on here, this is like… sorry. Before we capture every morning, you set it up wherever, take a quick capture, then you turn on the device. It measures, you go to around 100 MU, and when it comes out of tolerance, should shut the beam off, and then we move forward. So it’s 100 MU, so it’s maybe a two-minute extra set, and then it verifies that every morning. We were trying other things first. So we did find that if you utilize high-contrast white objects, it works much better. So we didn’t purchase anything to do this test. So you can look around your clinic, see what you want, but it was pretty straightforward.

So then for AlignRT implementation, that one was a little bit longer because we were going from marked setups. We didn’t use tattoo, but we did use Sharpies to mark less. During the implementation of AlignRT, we actually collected clinical data before, during, and after integration of SGRT. Metrics evaluated included our patient shift magnitudes, the frequency of repeat imaging, and overall patient setup time. All of these were important working in pediatrics, especially when they’re under anesthesia. We wanted it to be faster, but also still safe. So the data was used to assess the workflow efficiency as well as the clinical impact.

We started small. Everything that we did were in small chunks, because we felt like if we took too big of a bite, we were not going to be as ready to implement, so we started small. So what we did, we just had a printed spreadsheet we kept next to the machine. We actually set up first to the patient marks, then we imaged, and then after we applied shifts, what we would then do is turn on AlignRT and record the data of the shifts. So that gave us the magnitude of shifts from AlignRT into imaging. That way, in the beginning, we were just going with our normal workflow, we weren’t adding a lot of extra steps. This is just an example of data—it’s not real data, sorry, guys. But you can see we found that when we were doing it, the data did show that it was comparable or less than; sorry. So we then moved forward.

As a therapist, this kind of talked toward me, was initial reluctance. At the time, I did not want to. So there could be initial reluctance to transition to AlignRT-only setup, leading to continued reliance on standard skin marks. So we took our time, and this approach was maintained until it became evident that the shifts between the CBCTs and AlignRT were consistently smaller magnitude, which gave us the increased confidence to use AlignRT only. So after that, that was maybe about two to three weeks, and then we moved over. We transitioned to AlignRT-only setup. We did, for a short time, retain the skin marks as backup until we regained confidence in the system. Shift data continued to be recorded, and then once as a multidisciplinary team, we reached consensus, skin markings at simulation were discontinued. So all of this, I would say, was less than six months. It took us nine months total because we do CSIs frequently, and we held onto those marks for a little bit longer. But after nine months, we no longer mark any of our patients, including electron or keloids as well.

Like I said before, therapists were the most hesitant to remove any marking on patients. This was our comfort item. Sometimes people say, “Just try it. Just try it.” And as like Mama Bear, we’re like, “No, no, no, we need to do that.” But once we saw the data, we go, “Okay, this is pretty good.” And after positive feedback from patients and not having to keep their marks, seeing the minimal shifts and less imaging that we had to do, we felt confident in removing all the marks.

Again, so it took us approximately nine months; that’s including CSI. We implemented it back in 2012. At that time, there were no other pediatric centers or people using it in pediatrics, so we had to develop our own workflow and tailor it to the custom needs of CHLA. So that did add a little bit of time and complexity to the implementation strategy or process. So although the underlying technology is the same for both adult and pediatric patients, the application in pediatrics presents unique considerations. However, once the initial implementation changes were overcome, it became clear that SGRT provided a greater range of benefits than we as therapists and the whole team originally expected.

So some of the examples of differences or changes that we had to encounter were, we had to change some of our ROIs. For anesthesia patients, we had to move their monitors, uncovering them, or undoing the blankets or just moving things. All of these, again, after a couple of weeks, we figured out how to do it, and now smooth sailing. But that initial bump was a little bit harder. Again, most of ours were six months. CSI added a little bit longer because they were a multi-isocenter. However, CSI patients, once we actually implemented them, we’re getting them down to 30 to 40 minutes for CSIs, even the adult-size CSIs.

Concurrent installation and implementation of TrueBeam and AlignRT required staff to undergo training on two new technologies and proved to be a bit overwhelming in the beginning. And so one of the things that we learned from that was to maybe just try one at a time or to focus on one first. After we gained confidence in the TrueBeam, then we added AlignRT and began collecting data, which included the shift magnitudes, frequency of repeat imaging, and patient setup time. Again, we did this in a phased approach because, at the time, we did not want to add too much extra workload to the therapist, because we didn’t have a lot of extra time. We did it phased and that seemed to be really helpful. After we analyzed the shift magnitudes, we expanded the data to include repeat imaging frequency, overall patient setup times, and the comparison between all those. As therapists, as we became more familiar with it, we saw a decrease in the metrics as well. So over time, they kept getting smaller and smaller.

Throughout the learning and implementation process, several key factors were identified as critical to the success of adoption of new technology, which includes AlignRT or SGRT. These included early education prior to product installation, strong staff buy-in, identification of clinical champions, a small rollout rather than a broad implementation, and direct observation of positive patient impacts. Those all really helped us. If staff members have a basic knowledge of SGRT prior to installation through online learning modules or in-person training, it’ll greatly also help your implementation. Staff buy-in is critical for all team members to have buy-in. I hear from different sites that sometimes a physician says, “Just do it.” That doesn’t always work for therapists when they have a busy workflow. So sometimes having extra champions like we talked about earlier, or having their buy-in or having them see what the differences could be. So therapists may be resistant at first to new technology because it’s going to impact their workflow, and many centers don’t have a lot of flexibility in their time. Twelve minutes in, out, in, out, and I know some places if you get behind, you’re never going to catch up. So saying, “Just add this in,” while seemingly small, can have a huge impact.

Champion or super users: if you can identify staff that are willing to serve as super users to support adoption training and troubleshooting, this will greatly help the therapist as well. It can identify possible issues that maybe other people weren’t thinking of, and you can help find solutions when necessary. Identifying knowledgeable, well-respected therapists within the group can also help increase your buy-in.

Small incremental rollouts: this was huge for us. It gave us time as staff to learn the new technology before implementing it to all patients. So due to short staffing, strict schedules, implementing a new technology such as SGRT for all patients at once can be challenging. So starting with a small number and expanding as staff become more confident is effective. We began with just chest and abdomen, and then we went to the next extremity, to CSI and pelvis, and now we do it on every patient from brain all the way to CSIs, TBIs—everything gets it now. A stepwise approach was used, adding new body sites once staff demonstrated confidence with the previous one.

Direct observation of positive patient outcomes encouraged therapists to continue implementation, which played a key role in our early success and improving the overall buy-in of therapists. We found patients did not like to have visible marks on their skin, and they did not like to keep them in place during treatment. I think that’s pretty standard across the board for every patient. Also, direct monitoring during treatment made it possible for us to safely lower the age of anesthesia or the age a patient required anesthesia because now we could actually quantify how much they were moving.

Challenges associated with AlignRT and MapRT implementation include the staff hesitancy, a steep learning curve, mostly for AlignRT, and increased setup times during the early adoptions. Resistance to change and unfamiliarity with surface-guided radiation therapy workflow contributed to the slower uptake, while additional training requirements and workflow adjustments temporarily affected our efficiency. However, these challenges diminished as staff gained experience and confidence in the system. And with all that being said, we work much faster now than we did prior. But there is that when you first start it, you might see a little bit of slowdown, but it gets better.

So given an already demanding schedule for treatments, therapists found it difficult to incorporate the additional time required per patient during the early stages. So we would select a small number of patients and add five minutes to their treatment schedule. I know that’s not always possible, but if you can pick one or two, expand their treatment, that gives you a little time, and it kind of avoids the stress. I know, not always possible. We also had some hesitancy because we were one of the early adopters, so it took us a little longer, as well as we already kind of talked about the additional time, which I know I’m repeating myself, but it does have an impact.

Just some quick tips while we’re finishing up. So I don’t know how many of you guys have the remote system that came with a lanyard. We love the remote, didn’t love the lanyard, and our computer for it is on the other side of the room. So what we did is I took a little Velcro, and I put some Velcro on the back of it, and it’s just on the side of our TrueBeam. So now I can just reach over, hit play, take a capture, put it back, and we’re good to go. So again, just little minor adjustments, depending on what your room looks like, can have a huge impact. So one therapist can easily reach a computer, and one therapist can reach the remote. Have regular trainings, check your Vision RT contract because I don’t know how that all works, but I know we have yearly training that we can use. Also, any new staff or even if we have students come through, we just had a practicum one, we just throw on the phantom and say, “Have fun. Practice because practice makes perfect.” So we’ll just let them do whatever they want on Emily or whatever phantom you have.

Conclusion is just we’ve successfully implemented AlignRT in the clinical workflow through a collaborative team-based approach. While challenges were encountered, the multidisciplinary team ultimately developed a workflow that best met the needs of our institution.

*This transcript has been AI-generated. Contact us at secretary@sgrt.org if there are any issues.