Transcript
Man 1: Okay. Good mornings. Today, I'm going to talk about the validation and application of the iPlan 4.0 planning systems. First, I will briefly introduce what iPlan 4 is. Then the method of validation such as which one do we use. And the comparison between those calculations and the measurement. And the difference between the Pencil Beam and the Monte Carlo dose calculations. I will also show some of the real patients, Monte Carlo and treatment plans. I'm pretty sure, lots of institutions already have the iPlan 3, also some are still using the BrainSCAN. So, the iPlan 4 is just iPlan 3 which use the Pencil Beam dose calculations. But we use the Monte Carlo...
Man 2: Please speak in the microphone.
Man 1: Oh. But we use with the Monte Carlo dose calculations. iPlan 4, it allows the user to crosscheck the results by the two almost independent dose calculations. It provides the most transition from the clinical experience based on the Pencil Beam dose calculations to a more accurate experience based on the Monte Carlo dose calculations. The faster Pencil Beam dose algorithm can be used for the intermediate planning process. Later the user can switch to Monte Carlo to fine-tune the treatment plan.
This is the menu of the iPlan 4. It is very similar to iPlan 3. There are four treatment types here you can select. After finishing the treatment planning, there is the icon to toggle between the Monte Carlo and the Pencil Beam dose calculations. And this is the phantom we used for our validations. It is a heterogeneous phantom from the CIRS, which lung and tissue densities are 0.21 and 1.06 respectively. We choose the three isocenter in these phantoms to perform different treatment plannings. The isocenter 1 is in the middle of the phantoms, it's relatively bigger and irregular shape. The iso-2 is near the left lung, and iso-3 is in the center of the tissue-equivalent rod inserted in the left lung. I should emphasize though, here for this phantom, we can, for any of these point in the lung, I can measure the in-lung dose. But I can't replace some of the tissue equivalent material. So for this iso-3, this is the tumor in lung.
So for each isocenter, we perform the 3D conformal plan, dynamic arc plan, and IMRT plans. This is the measurement that was compared to the Monte Carlo and the Pencil Beam. I already mentioned for the iPlan 4, we have Monte Carlo and the Pencil Beam. So for this, this is for the isocenter 1. It's in the middle of the tissues. For Monte Carlo, we can find for the Monte Carlo and even Pencil Beam, and those calculations compared to the measurements are very good. So we don't care, this is the 1% off or 2% off, 1%, 2% off is the same. So that's about 1% off maybe looks bad, but it's the same as from the physics' point of views. But if you notice, for this plan, I measured the point dose in lung. For the Monte Carlo, still is achievable. There's about a 3% off for the 3D conformal and IMRTs, but for the dynamic arc it's with 1%. But for the Pencil Beam dose calculations, for the 3D conformal plan, more than 30% off. For the dynamic arc, 20% off. IMRT plans, more than 10% off. But those calculations remain for the Pencil Beam, remember. And this is for the iso-2. And the tumor is near the lung. So both the Monte Carlo calculations and Pencil Beam calculations are quite good compared to the measurement. And for the tumor in length, the point dose, this dose is still in the tissue. So, the point dose, the calculation from the Monte Carlo and Pencil Beam also seems achievable. But for the IMRT, it's a little off, but we still think it's still okay, the dose calculation in tissue.
And this is the film measurement, compare the MC Monte Carlo isodose distributions. The colored isodose lines are for the film measurement and the black lines are for the Monte Carlo dose calculations. And the left side is the 3D conformal plan for the iso-1, and the right side is dynamic arc plans. And their Ting factor is quite good. The Ting factor is we use... The definition of the Ting factor is the cumulative dose of the film divided by the cumulative dose of the plan. So usually for the IMRT 2A, you compare that isodose, and I used the same software. So the Ting factor, one is the Profact [SP]. But we think 5% is acceptable. So this is IMRT film measurement compared to Monte Carlo dose calculations. And this is for the iso-1, this is for the iso-2. So, that is quite good.
So, we're interested for those calculation in lung. So this is the tumor in lung. And this show the Pencil Beam isodose lines. The yellow line is the 90% isodose lines. Now, this is the Pencil Beam isodose lines. I use the same setup and the same MU, the Monte Carlo dose isolines. For the 90% isodose line, the shift between the Monte Carlo and Pencil Beam dose, those calculations is more than 3 millimeters. For this case, for the 80% isodose line, the shift is about 2 millimeter. These graphs show the Monte Carlo and the Pencil Beam dose calculations, eight different distance from the peripheral of the PTVs. And this is... Yeah, for each direction, I choose the anterior, posterior, and the lateral. They're similar magnitude of the difference between the Monte Carlo and the Pencil Beam calculations of mean about 13% overestimation of the dose at the margin of the PTV. So this is the distance from peripheral of the PTV.
Okay, we go to some of the real case. This is a patient that we treat last month in our institution. We still use Brain... Oh, I have the... Sorry. So I use the BrainSCAN. So this study, for this lung patient that we used SBRT 45 gray with 5 fractions. This is the PTV for the single fractions we already treated. So, I use the iPlan 4 to be planning. This is the DVHs of the Pencil Beam calculations. I used the 9 conformal beams for this patient. When we treated it, we use the 10 beam per rays [SP]. I got the same DVHs. I used the same setup and the same MU, Monte Carlo. There's the difference of the mean dose between the Monte Carlo, Pencil Beam, more than 10%. The PTV coverage, I think about a 20% difference. I use the different... I also practiced some dynamic arc plan. This is the DVHs of calculated Pencil Beam DVHs. Same setup, same MU, Monte Carlo. The difference of the mean dose is more than 10%. And the PTV coverage is about 12%, 13%, is not big a difference. As a physicist I can plan, so, I want to give you very good DVHs. Use the 3D IMRT. This is the Pencil Beam, Pencil Beam calculations. The DVHs are very good. The DVHs have the same MU, same setup. So the difference of the mean dose between the Monte Carlo and Pencil Beam calculation is more than 10%. And the PTV coverage I think are more than 20%. Yeah, also I mentioned the DVHs are very good. But if accurate that dose calculations Monte Carlo show is not good.
So this is the isodose for the IMRT plan. And this is, of course, to show you the Pencil Beam. And we can find that the 90% isodose line is totally covered PTVs. Same MU, same setup, the Monte Carlo. So even the 80% isodose line not totally covered the PTV, almost covered the PTVs. So, for the lung calculations, so far, every way for the BrainLab customers, either way, use the Pencil Beam dose calculations for the following calculations. For the lung patient, this calculation. We'll think about it because even RTOG, they do not recommend some kind of heterogeneity correction for the lung. Yeah, so far there's...only the Monte Carlo can give the reasonable dose calculations for lung. So yeah, I think for the other side of the tumor, they have the other case. So this seems is okay from our point of views. So this is the DVHs for the brain tumor. So use the Pencil Beam. Use the Pencil Beam. This is DVHs of Pencil Beam, same MU, same setup, the Monte Carlo is really close. The Pencil Beam calculations is acceptable for this case, I said for this case. And for the liver cancer patient. So this is the DVHs particularly for the Pencil Beams, same MU, same setup. So, and the difference of the mean dose is about 2%, 2% or 3%. And this coverage, I think about 2% or 1%. So for the liver it's still acceptable.
So our conclusions, so for both Monte Carlo and the Pencil Beam dose algorithms in the iPlan 4 planning systems are accurate for the dose calculation for tumor in tissues. For those calculations in lung, only the Monte Carlo dose algorithms achieved reasonable accuracy. For lung tumor patients, the Monte Carlo planning method can avoid underdose of the margins of the planning tumor volume that occur with the Pencil Beam dose algorithm. Thank you.
Man 2: Please speak in the microphone.
Man 1: Oh. But we use with the Monte Carlo dose calculations. iPlan 4, it allows the user to crosscheck the results by the two almost independent dose calculations. It provides the most transition from the clinical experience based on the Pencil Beam dose calculations to a more accurate experience based on the Monte Carlo dose calculations. The faster Pencil Beam dose algorithm can be used for the intermediate planning process. Later the user can switch to Monte Carlo to fine-tune the treatment plan.
This is the menu of the iPlan 4. It is very similar to iPlan 3. There are four treatment types here you can select. After finishing the treatment planning, there is the icon to toggle between the Monte Carlo and the Pencil Beam dose calculations. And this is the phantom we used for our validations. It is a heterogeneous phantom from the CIRS, which lung and tissue densities are 0.21 and 1.06 respectively. We choose the three isocenter in these phantoms to perform different treatment plannings. The isocenter 1 is in the middle of the phantoms, it's relatively bigger and irregular shape. The iso-2 is near the left lung, and iso-3 is in the center of the tissue-equivalent rod inserted in the left lung. I should emphasize though, here for this phantom, we can, for any of these point in the lung, I can measure the in-lung dose. But I can't replace some of the tissue equivalent material. So for this iso-3, this is the tumor in lung.
So for each isocenter, we perform the 3D conformal plan, dynamic arc plan, and IMRT plans. This is the measurement that was compared to the Monte Carlo and the Pencil Beam. I already mentioned for the iPlan 4, we have Monte Carlo and the Pencil Beam. So for this, this is for the isocenter 1. It's in the middle of the tissues. For Monte Carlo, we can find for the Monte Carlo and even Pencil Beam, and those calculations compared to the measurements are very good. So we don't care, this is the 1% off or 2% off, 1%, 2% off is the same. So that's about 1% off maybe looks bad, but it's the same as from the physics' point of views. But if you notice, for this plan, I measured the point dose in lung. For the Monte Carlo, still is achievable. There's about a 3% off for the 3D conformal and IMRTs, but for the dynamic arc it's with 1%. But for the Pencil Beam dose calculations, for the 3D conformal plan, more than 30% off. For the dynamic arc, 20% off. IMRT plans, more than 10% off. But those calculations remain for the Pencil Beam, remember. And this is for the iso-2. And the tumor is near the lung. So both the Monte Carlo calculations and Pencil Beam calculations are quite good compared to the measurement. And for the tumor in length, the point dose, this dose is still in the tissue. So, the point dose, the calculation from the Monte Carlo and Pencil Beam also seems achievable. But for the IMRT, it's a little off, but we still think it's still okay, the dose calculation in tissue.
And this is the film measurement, compare the MC Monte Carlo isodose distributions. The colored isodose lines are for the film measurement and the black lines are for the Monte Carlo dose calculations. And the left side is the 3D conformal plan for the iso-1, and the right side is dynamic arc plans. And their Ting factor is quite good. The Ting factor is we use... The definition of the Ting factor is the cumulative dose of the film divided by the cumulative dose of the plan. So usually for the IMRT 2A, you compare that isodose, and I used the same software. So the Ting factor, one is the Profact [SP]. But we think 5% is acceptable. So this is IMRT film measurement compared to Monte Carlo dose calculations. And this is for the iso-1, this is for the iso-2. So, that is quite good.
So, we're interested for those calculation in lung. So this is the tumor in lung. And this show the Pencil Beam isodose lines. The yellow line is the 90% isodose lines. Now, this is the Pencil Beam isodose lines. I use the same setup and the same MU, the Monte Carlo dose isolines. For the 90% isodose line, the shift between the Monte Carlo and Pencil Beam dose, those calculations is more than 3 millimeters. For this case, for the 80% isodose line, the shift is about 2 millimeter. These graphs show the Monte Carlo and the Pencil Beam dose calculations, eight different distance from the peripheral of the PTVs. And this is... Yeah, for each direction, I choose the anterior, posterior, and the lateral. They're similar magnitude of the difference between the Monte Carlo and the Pencil Beam calculations of mean about 13% overestimation of the dose at the margin of the PTV. So this is the distance from peripheral of the PTV.
Okay, we go to some of the real case. This is a patient that we treat last month in our institution. We still use Brain... Oh, I have the... Sorry. So I use the BrainSCAN. So this study, for this lung patient that we used SBRT 45 gray with 5 fractions. This is the PTV for the single fractions we already treated. So, I use the iPlan 4 to be planning. This is the DVHs of the Pencil Beam calculations. I used the 9 conformal beams for this patient. When we treated it, we use the 10 beam per rays [SP]. I got the same DVHs. I used the same setup and the same MU, Monte Carlo. There's the difference of the mean dose between the Monte Carlo, Pencil Beam, more than 10%. The PTV coverage, I think about a 20% difference. I use the different... I also practiced some dynamic arc plan. This is the DVHs of calculated Pencil Beam DVHs. Same setup, same MU, Monte Carlo. The difference of the mean dose is more than 10%. And the PTV coverage is about 12%, 13%, is not big a difference. As a physicist I can plan, so, I want to give you very good DVHs. Use the 3D IMRT. This is the Pencil Beam, Pencil Beam calculations. The DVHs are very good. The DVHs have the same MU, same setup. So the difference of the mean dose between the Monte Carlo and Pencil Beam calculation is more than 10%. And the PTV coverage I think are more than 20%. Yeah, also I mentioned the DVHs are very good. But if accurate that dose calculations Monte Carlo show is not good.
So this is the isodose for the IMRT plan. And this is, of course, to show you the Pencil Beam. And we can find that the 90% isodose line is totally covered PTVs. Same MU, same setup, the Monte Carlo. So even the 80% isodose line not totally covered the PTV, almost covered the PTVs. So, for the lung calculations, so far, every way for the BrainLab customers, either way, use the Pencil Beam dose calculations for the following calculations. For the lung patient, this calculation. We'll think about it because even RTOG, they do not recommend some kind of heterogeneity correction for the lung. Yeah, so far there's...only the Monte Carlo can give the reasonable dose calculations for lung. So yeah, I think for the other side of the tumor, they have the other case. So this seems is okay from our point of views. So this is the DVHs for the brain tumor. So use the Pencil Beam. Use the Pencil Beam. This is DVHs of Pencil Beam, same MU, same setup, the Monte Carlo is really close. The Pencil Beam calculations is acceptable for this case, I said for this case. And for the liver cancer patient. So this is the DVHs particularly for the Pencil Beams, same MU, same setup. So, and the difference of the mean dose is about 2%, 2% or 3%. And this coverage, I think about 2% or 1%. So for the liver it's still acceptable.
So our conclusions, so for both Monte Carlo and the Pencil Beam dose algorithms in the iPlan 4 planning systems are accurate for the dose calculation for tumor in tissues. For those calculations in lung, only the Monte Carlo dose algorithms achieved reasonable accuracy. For lung tumor patients, the Monte Carlo planning method can avoid underdose of the margins of the planning tumor volume that occur with the Pencil Beam dose algorithm. Thank you.