Transcript
The basic... How many of you in the audience have what we call Novalis Classic Brainlab Mask? Most of you. Okay, good, good number of you. So I think it'll make sense what I'm trying to say here. It's that when you're trying to do the mask-based stereotactic radiotherapy, multi-fraction, conventional fractionation, what really happens is this is the basic steps we follow. We do a CT scan with localizer box. I think most of you know what I'm talking about. A coordinate system is defined by the six traditional rods embedded in the box.
You do the scan, you have a coordinate system of what we call Brainlab CT System, right? Then you have...a computer planning system does the planning and then speaks out printed templates. You mount the templates on the box, and I think I might have the box. By the way, this is the mask I'm talking about, not the new mask, but what we call classic. I was gonna say old one, but someone told me, "Use the word classic." What we are talking about here is that this is a typical mask that most of the users so far in the world have been using.
And this mask is like a head ring in the lower and the upper parts. Here is the box. You mount the box on the arm to the head ring and then line up the lasers with the marks on the template. There is a slight problem with it. And there are two problems. First problem is that there is no anatomical verification. And I'm going to show you in just a minute what really happened when we started our program about three and a half years ago. The setup box really is totally independent of the mask.
And what I'm trying to say here is that people who really use this know what I'm talking about. Here, if there are five slots on both sides of the ring... And for some patients, the mask can be here or the supports can be in this slot, for some patients, support could be underneath, one below.
And so you could have a patient on the treatment table. You could have one wonderful, fantastic planning and the physicist is the most brilliant guy in the world. However, you go ahead and set up with the box and you got the smartest therapy in the world. But if they forgot the fact that here, this patient was scanned with just one slot and the treatment is done with another slot, the box will not share that knowledge with you.
And so you have treated the patient 2 1/2 centimeters, roughly about 2 centimeters away from the target. Here goes your plus-minus, what we call 1-millimeter uncertainty we talk about. These are the errors that we should be very careful about. So the study objective really started way back for us, way back about...during the first month when we were starting to treat the patient and the third patient that we are treating.
And I'll maybe share something with you at the risk of maybe going back to 14 minutes. However, what happened was, we had a patient in those days, for the first six months, we had the physicist go at the treatment machine, Jack [SP], the first days set up. I went to the treatment machine and said, "Something doesn't make sense." I said, "Don't treat the patient." I went back to the brain scan planning computer, looked at the isocenter in relation to the eyeballs, said, "Something doesn't look right."
I didn't know what was going on. It turned out that the therapist had really adjusted that between the previous patient and this patient. So what...you know, that really prompted me to think that maybe I want to really do the ExacTrac X-rays on the patients that have the box as the treatment setup, you know, system.
So I decided, actually, brain scan computer does not allow you to do this. So what I really say that, you know, I'm going to really do something that... You know, you don't have to really follow everything that company tells you. You can really kind of try to, you know, do something a little more with the tools they give you. So what we do is we create. We start with a fresh CT data set and we recreate the isocenter to within a tenth of a millimeter.
And I'm going to show you that you...in X and Y coordinates, you can do that. But even in the...here, we started with a brand-new CT data set. And then we put a fake isocenter. And this is the CT data set I'm going to export to the ExacTrac X-ray. Because I have a reference point geometry here. And let me show you even in the... We do use at sharp. We have 2-millimeter thick slices and they are plenty. I had some studies done on it. If slice 55 and 56, you can tell that isocenter may be in the middle of this...in between two slices, and you can still reproduce it to within usually a 10th of a millimeter.
So what we do is here. We go ahead and take for Brainlab Mask patient fractionated radiotherapy. We go ahead and do the conventional setup with the box. And then we go ahead and take an ExacTrac X-ray. And I don't need to go into detail about how good or how great are the ExacTrac X-rays. However, I want to show you just one set of results. These are the numbers that we have taken for the frame-based patients.
I wanted to see how good is my ExacTrac X-ray system. All right, so these are not the numbers that some physics geeks produced in a phantom. These are from the patients measured by the therapist at the machine, real life. Turns out that in ExacTrac X-ray, it's within 1 millimeter. Forty-seven data points in the last three and a half years.
So really, what is a good mask, in my opinion? A good mask is where day after day after day, the ExacTrac X-rays show me that I'm within 2 millimeters. And back then, I accepted 2 millimeters. Now, we accept 1 millimeter. Otherwise, we make an adjustment. Acceptable mask is where on the first day, patient is nervous, mask doesn't fit, second, third day, it's fine.
The sub-optimal mask is where the shapes are all over. And most of the people in the audience will agree with me that most of the time the shifts here, the positioning errors are in the longitudinal direction. Really, patients don't have much of... Brainlab masks are good even though someone... Javid [SP] told me that it's $200 a piece. They are very good masks as compared to the X-ray [SP] masks.
My 13 minutes aren't over yet, right? Thank you. Okay. Another thing I can guarantee, I have only one conclusion per talk, and that is the ExacTrac X-rays are, for us, have become quite indispensable to us. We do not ever at Sharp Hospital, we do not do a single fraction treatment, whether it's a frame-based or frameless without X-rays.
And part of the reason is that even frame-based, if I have a... You know, in my previous job at the university, we did about 500 frame-based stereotactic cases over a 10-year time period back in '90s. It turned out that I had 2 out of those 500 patients where the patient really moved between the time the neurosurgeon placed the frame and by the time we...
You will normally monkey around all the planning business and monkey around all the, you know, 1/2 a millimeter plus 1 millimeter. Then the patient moves, then how do you know the patient has moved? ExacTrac X-ays would tell us. If I have got a shift on a frame-based patient more than 3 millimeters, I do not recommend the oncologist to proceed with the treatment because that's how much I trust the ExacTrac X-rays. Thank you very much. Thank you.
You do the scan, you have a coordinate system of what we call Brainlab CT System, right? Then you have...a computer planning system does the planning and then speaks out printed templates. You mount the templates on the box, and I think I might have the box. By the way, this is the mask I'm talking about, not the new mask, but what we call classic. I was gonna say old one, but someone told me, "Use the word classic." What we are talking about here is that this is a typical mask that most of the users so far in the world have been using.
And this mask is like a head ring in the lower and the upper parts. Here is the box. You mount the box on the arm to the head ring and then line up the lasers with the marks on the template. There is a slight problem with it. And there are two problems. First problem is that there is no anatomical verification. And I'm going to show you in just a minute what really happened when we started our program about three and a half years ago. The setup box really is totally independent of the mask.
And what I'm trying to say here is that people who really use this know what I'm talking about. Here, if there are five slots on both sides of the ring... And for some patients, the mask can be here or the supports can be in this slot, for some patients, support could be underneath, one below.
And so you could have a patient on the treatment table. You could have one wonderful, fantastic planning and the physicist is the most brilliant guy in the world. However, you go ahead and set up with the box and you got the smartest therapy in the world. But if they forgot the fact that here, this patient was scanned with just one slot and the treatment is done with another slot, the box will not share that knowledge with you.
And so you have treated the patient 2 1/2 centimeters, roughly about 2 centimeters away from the target. Here goes your plus-minus, what we call 1-millimeter uncertainty we talk about. These are the errors that we should be very careful about. So the study objective really started way back for us, way back about...during the first month when we were starting to treat the patient and the third patient that we are treating.
And I'll maybe share something with you at the risk of maybe going back to 14 minutes. However, what happened was, we had a patient in those days, for the first six months, we had the physicist go at the treatment machine, Jack [SP], the first days set up. I went to the treatment machine and said, "Something doesn't make sense." I said, "Don't treat the patient." I went back to the brain scan planning computer, looked at the isocenter in relation to the eyeballs, said, "Something doesn't look right."
I didn't know what was going on. It turned out that the therapist had really adjusted that between the previous patient and this patient. So what...you know, that really prompted me to think that maybe I want to really do the ExacTrac X-rays on the patients that have the box as the treatment setup, you know, system.
So I decided, actually, brain scan computer does not allow you to do this. So what I really say that, you know, I'm going to really do something that... You know, you don't have to really follow everything that company tells you. You can really kind of try to, you know, do something a little more with the tools they give you. So what we do is we create. We start with a fresh CT data set and we recreate the isocenter to within a tenth of a millimeter.
And I'm going to show you that you...in X and Y coordinates, you can do that. But even in the...here, we started with a brand-new CT data set. And then we put a fake isocenter. And this is the CT data set I'm going to export to the ExacTrac X-ray. Because I have a reference point geometry here. And let me show you even in the... We do use at sharp. We have 2-millimeter thick slices and they are plenty. I had some studies done on it. If slice 55 and 56, you can tell that isocenter may be in the middle of this...in between two slices, and you can still reproduce it to within usually a 10th of a millimeter.
So what we do is here. We go ahead and take for Brainlab Mask patient fractionated radiotherapy. We go ahead and do the conventional setup with the box. And then we go ahead and take an ExacTrac X-ray. And I don't need to go into detail about how good or how great are the ExacTrac X-rays. However, I want to show you just one set of results. These are the numbers that we have taken for the frame-based patients.
I wanted to see how good is my ExacTrac X-ray system. All right, so these are not the numbers that some physics geeks produced in a phantom. These are from the patients measured by the therapist at the machine, real life. Turns out that in ExacTrac X-ray, it's within 1 millimeter. Forty-seven data points in the last three and a half years.
So really, what is a good mask, in my opinion? A good mask is where day after day after day, the ExacTrac X-rays show me that I'm within 2 millimeters. And back then, I accepted 2 millimeters. Now, we accept 1 millimeter. Otherwise, we make an adjustment. Acceptable mask is where on the first day, patient is nervous, mask doesn't fit, second, third day, it's fine.
The sub-optimal mask is where the shapes are all over. And most of the people in the audience will agree with me that most of the time the shifts here, the positioning errors are in the longitudinal direction. Really, patients don't have much of... Brainlab masks are good even though someone... Javid [SP] told me that it's $200 a piece. They are very good masks as compared to the X-ray [SP] masks.
My 13 minutes aren't over yet, right? Thank you. Okay. Another thing I can guarantee, I have only one conclusion per talk, and that is the ExacTrac X-rays are, for us, have become quite indispensable to us. We do not ever at Sharp Hospital, we do not do a single fraction treatment, whether it's a frame-based or frameless without X-rays.
And part of the reason is that even frame-based, if I have a... You know, in my previous job at the university, we did about 500 frame-based stereotactic cases over a 10-year time period back in '90s. It turned out that I had 2 out of those 500 patients where the patient really moved between the time the neurosurgeon placed the frame and by the time we...
You will normally monkey around all the planning business and monkey around all the, you know, 1/2 a millimeter plus 1 millimeter. Then the patient moves, then how do you know the patient has moved? ExacTrac X-ays would tell us. If I have got a shift on a frame-based patient more than 3 millimeters, I do not recommend the oncologist to proceed with the treatment because that's how much I trust the ExacTrac X-rays. Thank you very much. Thank you.