Leksell Gamma Knife Perfexion QA Considerations Paula L. Petti, Ph.D. Taylor McAdam Bell Neuroscience Institute Washington Hospital Healthcare System Fremont, CA 1 QA for LGK Perfexion : : Follow NRC licensing guidelines (10( CFR 35.1000) Basic Tests and Measurements: 1) Coincidence of the mechanical isocenter of the patient-positioning system (PPS) with the radiation-focal point (RFP) 2) Agreement of measured beam profiles with Leksell GammaPlan calculations for all collimator sizes in the XY, YZ and XZ planes 3) Measurement of the absolute dose-rate calibration for largest collimator ) Confirmation of the relative output factors (ROFs( ROFs) ) for smaller collimators. 2 Leksell Gamma Knife Perfexion The patient is positioned via precise couch motions A frame adapter attaches the Leksell coordinate frame (affixed to the patient s skull) to the treatment couch Collimation system is built into the unit 3 Collimator sizes: 16-mm, 8-mm and -mm 192 60 Co sources 60 Co sources not fixed in space: they reside on 8 moveable sectors Collimator Patient Frame Adapter Patient Couch: 3-axis positioning system 3 1
8 independent, identical sectors 2 sources per sector Perfexion Sector Design Sectors slide back and forth on outside of collimator Sector Drives Sources are arranged in 5 rings The sources in each sector can be aligned with a different collimator size or blocked completely Coincidence of PPS and RFP for Perfexion No helmets or microswitches to check Collimator settings (, 8 and 16) are independent Therefore, we must check the coincidence of the PPS and RFP for each collimator Elekta uses: Master Diode Tool New Film Holder Tool We use: Diode Test Tool (-mm collimator only) New or old Film Holder 5 Alignment of Patient-Positioning System (PPS) with Radiation Focal Point (RFP) Master Diode Tool: Service instrument used during bi-annual preventative maintenance New Film Holder: Service and Field instrument used for annual QA and acceptance testing Old-Style Film Holder: Field instrument: used for annual QA and acceptance testing Diode Tool: Field instrument used for routine checks (at least monthly, but is usually done 6more often) 2
Master Diode Tool: Used by Elekta during commissioning to align the RFP to the PPS Attaches directly to the PPS Accomodates up to 5 diodes There are programmed scannning sequences for all three collimators Diode Test Tool: Used by GK physicist for routine QA Attaches to patient frame adapter, which then attaches to the PPS One central diode There is one programmed scannning sequence for the -mm collimator No scanning sequences for other collimators. 7 Film Holders New Model Old Model Dosimetry adapter More precise than old-style film holder because: 1) Uses the same frame adapter that is used for patient treatments (not shown above) which is machined to very high tolerance 2) Has larger film compartment easier to interpret 16-mm films 1) Requires special dosimetry adapter, which is not machined to the same high tolerances as the patient frame adapter 2) Provides QA for RFP reproducibility, but does not provide QA for patient setup, since patient adapter is not employed 8 EBT GAFCHROMIC Film Manufactured by International Specialty Products (ISP, Wayne, NJ) Excellent white paper available at the manufacturer s website: http://online1.ispcorp.com/_layouts/gafchromic/index.html Main advantage over earlier GAFCHROMIC films for LGK dosimetry: More sensitive to lower radiation doses: EBT films can measure doses in the range of 0.01Gy to 8 Gy Low energy dependence (<5% between MeV and kev photons) Film density growth complete within < 2 hrs 9 3
Basic Requirements for EBT Film Analysis Peak absorbance of EBT film is 636 nm, response enhanced by measurement with red light (match absorbance of film with spectral response of scanner) Flatbed scanners: Epson and Microtek scanners are designed to scan color film According to the ISP white paper, Epson scanners produce the optimum results with EBT films 10 Basic Requirements for EBT Film Analysis Software ImageJ: Available for free download at the website: http://rsb.info.nih.gov/ij/ Microsoft Excel More sophisticated software can be employed 11 EBT Film Handling Must keep track of film orientation: Before cutting film into appropriate size pieces, number the same corner on each sub-section of film: When analyzing films with flatbed scanner, the amount of scattered light depends on the orientation of the active particles in the film, which depends on the direction in which the film was coated. 1 2 3 5 6 7 8 9 10 11 12 Film coating for EBT film is parallel to the short edge of film For RFP PPS coincidence measurements, cut film to fit holder For Profile and ROF measurements cut film into pieces between about 6.0 cm and 6.5 cm 12
EBT Film Handling (continued) Place film in holder such that the marked corner is in the same place for all films When scanning, place films in the same area of the flatbed scanner with the marked corner of the film in the same place Avoid fingerprints and mechanical stress Avoid long exposures to daylight 13 Recommended Procedures for GK EBT Film Measurements To obtain background subtraction for each piece of cut film, pre-scan each film on flatbed scanner prior to irradiation. Scanning parameters: Scan as positive film Use either 200 or 00 dpi to obtain resolution of the order of 0.1 mm (0.127 mm for 200 dpi and 0.06 mm for 00 dpi) 8-bit color Save as PNG files (or equivalent lossless format) 1 RFP - PPS Coincidence Measurements Not necessary to convert readings to absolute dose (i.e., calibration curve not required), since we are only looking at symmetry about the pin-point. If using ImageJ software to analyze films Extract red channel (Image color RGB split) Invert image (Edit Invert) (so that profile will be positive - personal preference) Using line drawing tool, draw line through center of scan Extract Profile (Analyze Plot Profile) List Profile Data, select it and copy to Excel (or other) worksheet Subtract background 15 5
Coincidence of RFP and PPS: Film Analysis with ImageJ Software Raw Data Red Channel Extracted Red Channel Inverted -mm irradiation Old-style film holder XZ Plane Background Reading for Film # 16 Coincidence of RFP and PPS: Film Analysis with ImageJ Software Reading for Background Subtraction Using Ellipse Tool, outline central portion un-irradiated film Under Analyze options, select Histogram Background reading for this film = mean value in central region of unirradiated film 17 Coincidence of RFP and PPS: Film Analysis with ImageJ Software Sup Use Line tool to indicate axis for profile Under Analyze select Plot Profile L R Inf Note: Draw line in the direction of increasing LGK coordinate, in this case from superior to inferior List, select and copy the data to Excel (or other) worksheet 18 6
Coincidence of RFP and PPS: Film Analysis with ImageJ Software Identify channel containing pin point By definition, this corresponds to X, Y or Z = 100 in LGK coordinate system Assign LGK coordinates to other channels by the formula: LGK Coord = 100 " (C 100 " C) # P Where, C 100 = the channel that contains the pin point C = the channel in question P = conversion factor, e.g., 0.0635 mm/pixel for films scanned at 00 dpi 19 -mm Z Axis: Coincidence of PPS and RFP Counts (Background Subtracted) 200 180 160 10 120 100 80 60 0 20 Central Channel (pin point) defines center of LGK coordinate system (magenta line) Compute the channel value for the center of the FWHM from the endpoints of the horizontal line defining the FWHM (green line) mm Z axis from XZ film PPS Center Center of FWHM FWHM!Z = -0.03 mm = 100 - (center of FWHM) 0 85 90 95 100 105 110 115 Z Coordinate of PPS (mm) 20 Example of Pin-Point Films to Check RFP - PPS Coincidence for 16-mm Collimator New-Style Film Holder Old-Style Film Holder Y Z X Y 21 7
New Film Holder: Example Pin-Point Test of Coincidence of PPS and RFP, New Film Holder 10 Reading (background subtracted) 120 100 80 60 0 20 0!X = 0.02 mm 16mm X Axis FWHM PPS Center RFP Center 80 85 90 95 100 105 110 115 120-20 X Coordinate of PPS (mm) 22 Old Film Holder: Example 200 16-mm X Axis (XY Plane): Coincidence of PPS and RFP Counts (Background subtracted) 180 160 10 120 100 80 60 0!x = 0.13 mm 16-mm X Axis XY Film PPS Center RFP Center 80% Width 20 0 80 85 90 95 100 105 110 115 120 X Coordinate of PPS (mm) 23 Test Specifications and Frequency Elekta s specification for the -mm collimator is that Δx, Δy and Δz are all 0.3mm and that Frequency of tests: " r = 2 2 " x + " y + " z! 0. mm 2 1) Master Diode test is done bi-annually by Elekta service engineers as part preventative maintenance 2) Films are usually irradiated annually by the on-site GK physicist 2 8
Example of Results using Different Tools to Check PPS and RFP Coincidence for WHHS Perfexion GK Collimator = Master - Diode Tool! 2 2 2 r =! x +! y +! z (mm) -mm 8-mm 16-mm 0.098 0.16 0.16 New-Style Film Holder 0.09 0.15 0.30 Old-Style Film Holder 0.11 0.20 0.38 25 Profile Measurements using Same film handling requirements as for pin-point films Measurements performed in spherical phantom (80 mm radius) Must measure dose versus film response calibration curve 26 Beam Profiles (film): Measurement Tools New-style Spherical Phantom Old-style Spherical Phantom 1) Attaches to patient frame adapter 2) Film positioned between 2 rods 3) Can irradiate a 3D stack of films ) Composed of certified Therapy Grade Solid Water 5) 3 adapters provided for ion chambers or other detectors 1) Attaches to dosimetry adapter 2) Film positioned in central insert 3) Can irradiate only one film at a time ) Presumably composed of polystyrene 5) Additional inserts are supplied for ion chambers and other detectors 27 9
EBT Film Calibration Calibration curve obtained by irradiating films at several energy levels between 0.5 Gy and 8 Gy with the 16-mm collimator Pre-scan all films to obtain background subtraction Dose on the horizontal axis equals the dose at (100, 100, 100) plus the transit dose for the 16-mm collimator. Film Reading Calibration Curve Reading minus Background (inverted) 160 10 120 100 80 60 0 20 0 Calibration Curve from 16mm Data 0 1 2 3 5 6 7 8 9 Dose (Gy) at (100, 100, 100) Film reading in red channel 28 Transit Dose for 16-mm Collimator Sources move to the blocked position at the initiation of a treatment and between shots. Therefore, to achieve a 16-mm shot, the sources must pass over the -mm collimator, and thereby deliver a so-called transit dose. 16 B 8 Blocked source position between - and 8-mm collimators Sources Sector Collimator 29 Transit Dose for 16-mm Collimator One way to measure 16-mm transit dose is to deliver the same prescribed dose via two treatment plans to isocenter in the spherical phantom, one plan consisting of a single 16-mm shot, and the second plan consisting of four 16-mm shots. The transit dose is then: Transit Dose = difference in the number of times the source travels back and forth over the -mm collimator D! D 1 3 difference in measured dose between the two runs Gy/(16-mm shot) 30 10
Transit Dose for 16-mm Collimator Using ion-chamber measurements as described on previous slide, transit dose for 16-mm collimator = 0.03 Gy/shot for WHHS Perfexion in July 2008 Note that the transit dose changes as a function of time as the Co-60 sources decay. 31 Determination of Transit Dose by Film Measurements Irradiate 1 st film with 1 shot, dose = 5 Gy Irradiate 2 nd film with 18 shots, dose = 5 Gy Number of shots: large enough s.t. dose difference in measurable, small enough s.t. irradiation times are valid for LGK PFX Subtract background reading determined from unirradiated film Use Dose-to-reading calibration curve to convert readings to dose Transit dose = (D 18 - D 1 )/17 Results (July 2008): 16-mm Collimator: 0.03 Gy/shot (same as result obtained with ion chamber!) 8-mm and -mm Collimators: 0.01 Gy/shot 32 Determining Beam Profiles using ImageJ Using Line Tool, draw line in direction of increasing coordinate value Under Analyze options, select Profile List profile data, Select it, and Copy to Excel (or similar) program for analysis 33 11
Beam Profile Analysis Raw Data (from ImageJ,, red channel, inverted) Subtract background Apply dose calibration to data points Identify FWHM Define X (or Y or Z) = 100 at center of FWHM (note that we have already confirmed that the center of the RFP is at (100,100,100) with the pinpoint measurement.) Define X (or Y or Z) coordinate value on horizontal axis (00 dpi) Calculate FWHM in mm 3 7 6 16-mm Y Profile from YZ Film 16 mm Y Profile from YZ film Center FWHM LGP Data 5 Dose (Gy) 3 2 1 0 60 70 80 90 100 110 120 130 10 LGP Y Coordinate (mm) Raw Data Converted to Plot of Dose vs LGP Coordinate (example): Measured FWHM = 21.7 mm, LGP FWHM = 21.3 mm, Diff = 0. mm35 (where LGP stands for Leksell GammaPlan) Simplified Procedure to Check only the FWHM Use Calibration curve to calculate R 1/2, the film reading (with background subtracted) that corresponds to 50% of maximum profile dose Locate R 1/2 on the raw profile curve (background subtracted), and compute: FWHM = (C 2 C 1 ) P Reading minus Background (inverted) 160 10 120 100 80 60 0 20 0 R 1/2 Film Reading Calibration Curve ½ Max Dose 0 1 2 3 5 6 7 8 9 Dose (Gy) at (100, 100, 100) Calibration Curve from 16mm Data Raw Data from ImageJ R 1/2 Where, P = conversion factor, e.g., 0.0635 mm/pixel for films scanned at 00 dpi C 1 C 2 36 12
Specifications and Frequency for Beam Profile measurements Specification: According to Elekta: Measured and LGP values for FWHM should be within ± 1 mm of each other Frequency: Beam profiles should be measured upon acceptance of the GK unit and annually thereafter 37 Results from last Annual QA: All FWHM were between ± 0.1 mm and ± 0. mm of Leksell GammaPlan 38 16-mm Dose Rate Measurement Currently no official calibration protocol specific to the LGK Charges of AAPM Task Group 178: Suggest a protocol for calibration with ionization chambers calibrated at an ADCL that can be used with all Gamma Stereotactic Radiosurgery (GSR) devices Work with the working group on dosimetry calibration protocol for beams that are not compliant with TG-51 39 13
16-mm Dose Rate Measurement Calibration Protocols: Performed in spherical phantom TG-21: Ion chamber calibrated 60 Co in-air Can be used for various phantom materials Can be used for various geometrical setups TG-51: Ion chamber calibrated for 60 Co in water. Designed to facilitate linear accelerator calibration and QA phantom must be water, 10 cm 10 cm field size 0 LGK Dose Rate Measurement: Some Recent Publications R Drzymala R Wood, J Levy: Calibration of the Gamma Knife using a new phantom following AAPM TG51 and TG21 protocols, Med. Phys. 35:51-521;2008 Compared the 2 protocols in the Elekta old-style spherical phantom and in a newly designed water phantom TG-51 in water phantom results were 1.% lower than TG-21 in polystyrene phantom 1 LGK Dose Rate Measurement: Some Recent Publications S Griffin Meltsner and LA DeWerd: Air Kerma based dosimetry calibration for the Leksell Gamma Knife, Med Phys 36:339-350;2009 Proposes an air-kerma-based dosimetry protocol using either an in-air or in-acrylic phantom measurement Modified version of TG-21 specific to LGK calibration geometry With new protocol, measured dose rates were between 1.5% and 2.9% higher than those used clinically by at 7 LGK sites (Models B and C) 2 1
16-mm Dose Rate Measurement: Practical Issues: Choice of new or old-style phantom Old-style phantom requires dosimetry adapter Bhatnagar, et al. (Med Phys 36:1208-1211;2009): The dosimetry adapter attenuates some beams in the lateral (3 and 7) sectors of the Perfexion unit, causing the overall 16-mm dose rate to be underestimated by approximately 1%. 3 New spherical phantom Patient Frame Adapter Choice of phantom Does not require dosimetry adapter Attaches to patient adapter More precise Provides better check of entire system 16-mm Dose Rate Measurement: Practical Issues Checking measured dose rate: compare results to TLD: either in-house or outside service (e.g., RPC SRS phantom) EBT GafChromic 5 15
Practical Method for Checking GK Dose Rate Calibration using EBT Film Irradiate films in 6-MV linear accelerator beam at 3 dose levels, e.g., 5 and 6 Gy to obtain a mini calibration curve Note that EBT film exhibits very little energy dependence Net Counts in Red Channel (Background subtracted) 6-MV Mini-Calibration 10 130 120 110 100 6-MV Mini-Calibration 90 80 3 5 6 7 Dose (Gy).0 Gy 5.0 Gy 6.0 Gy Red channel extracted, Background subtracted, film intensity inverted 6 Practical Method for Checking Dose Rate Calibration Irradiate film with 16-mm collimator, 5.0 Gy @ maximum Net Counts in Red Channel (Background subtracted) 6-MV Mini-Calibration 10 130 120 110 100 6-MV Mini-Calibration 90 80 3 5 6 7 Dose (Gy) Determine dose at center of peak region from minicalibration curve Compare to expected value (e.g..96 Gy) 7 Expect ± 2% to 3% agreement Relative Output Factors (ROF) for the 8- and -mm Collimators Relative output factor for Gamma Knife is defined as: dd C / dt dd dt 16 / (100,100,100) i.e., ROF = dose rate of collimator C relative to 16- mm collimator, where both are measured at isocenter = (100,100,100) in spherical, 80-mm radius phantom 8 16
Unique Features of Perfexion Geometry Co-60 sources distributed in 5 rings for each identical sector 1 2 3 5 Ring Number Number of Sources 1 8 2 32 3 0 32 5 0 9 ROF: Unique Feature of Perfexion Geometry Each collimator within each ring has a different beam geometry 16 8 50 Beam Geometry: There are, therefore, 15 distinct beam geometries: 5 rings multiplied by 3 beam-on positions per ring. This is in contrast to previous LGK designs for which all of the beam channels were identical, and there was only one type of beam. Each of the 15 beam types has a different: 1) Virtual source-to-isocenter distance (SAD) 2) Output factor Elekta determined these values by fitting a beam model to Monte Carlo generated data. 51 17
Fitted Values for Relative Output Factor (ROF) and Virtual Source-to-focus distance (R( vsf ) Collimator Size Ring ROF R vsf (mm) Collimator Size Ring ROF R vsf (mm) 1 0.799 521 8 0.808 80 2 0.815 56 8 5 0.730 522 3 0.792 533 16 1 0.961 81 0.725 595 16 2 1.000 59 5 0.663 607 16 3 0.986 55 8 1 0.957 31 16 0.920 88 8 2 0.96 37 16 5 0.851 519 8 3 0.901 68 52 Relative Output Factor for Each Collimator Size We cannot measure the 15 ROFs individually The ROF for the 8- and -mm collimators relative to the 16-mm collimator is determined from the equation: ROF( c) = 5 " i = 1 5 " ni! OFi ( c) i =, ni n! OF ( c = 16mm) i 1 = i Where the sum is taken over all 5 rings (8,32,0,32,0) n i represents the number of sources in each ring ROF(8mm) = 0.92 ROF(mm) = 0.805 53 Two Ways to Determine ROF Measure dose, D c, delivered by each collimator at (100, 100, 100) for: A given treatment time: ROF c = D c /D 16 The same prescription dose: ROF c = D c / T c D 16 / T 16 ROF c " D c no min al # ROF c D 16 T c is the irradiation time for each collimator The approximately equal sign is replaced by an equality if the dose is prescribed to (100,100,100) instead of the point of 5 maximum dose 18
Relative Output Factors: Measurement Techniques Pin-point ion chamber GafChromic Film TLDs (rods and LiF microcubes) Glass Rods (Perks, et al.) Some References: Mack et al., Precision dosimetry for narrow photon beams used in radiosurgery - determination of Gamma Knife output factors, Med. Phys. 29: 2080-9; 2002 Perks et al., Glass rod detectors for small field, stereotactic radiosurgery dosimetric audit, Med. Phys. 32:726-32; 2005 Novotny et al. Measurement of relative output factors for the 8 and mm collimators of the Leksell Gamma Knife Perfexion by film dosimetry, Med Phys. 36:1768-177;2009 55 ROF Meas: : EBT Film and Fixed Dose Cut and mark Films Scan un-irradiated films to obtain background correction Irradiate 2 films at the 3 dose levels, e.g.,.5, 5.0 and 5.5 Gy for 16 collimator to obtain mini-calibration curve (choose either axial, coronal or sagittal plane). It is reasonable to assume that the calibration curve is piece-wise linear between measured points (6 films) Irradiate 2 films the - and 8-mm collimators to a dose in the middle of the minicalibration range in the same plane as calibration films ( films) Process films with ImageJ: extract red channel, invert intensity values, subtract background 56 Sample intensity values in center of films, convert to dose, calculate ROF My Results: Average of 7 sets of ROF Measurements 8-mm Collimator: 0.888 ± 0.012 ( 3.9% lower than Elekta) -mm Collimator: 0.792 ± 0.007 ( 2.5% lower than Elekta) Used constant time for 3 sets of measurements, constant dose for sets Used different measurement planes (axial, coronal or sagittal) Did not change values in LGP 57 19
Results Reported in Literature: Average of 5 sets of ROF Measurements Collimator 8 mm EDR 2 Film 0.90 ± 0.012 (-2.1%) EBT Film 0.917 ± 0.01 (-0.8%) MD-V2-55 Film 0.906 ± 0.018 (-2.0%) mm 0.769 ± 0.010 (-.5%) 0.810 ± 0.007 (+0.6%) 0.819 ± 0.009 (+1.7%) Novotny et al., Med. Phys.2009 58 Error levels in ROF Measurement Neglecting transit dose: Neglecting the difference between maximum dose and dose at (100,100,100) 0.03Gy = 0.6% 5Gy < 1% Standard deviation in film pixel values around point of measurement Between 0.5 and 1 count 59 What order of magnitude error in film reading causes a % error in dose? 160 Film Reading Calibration Curve Reading minus Background (inverted) 10 120 100 Slope in vicinity of 5 Gy ~ 7 counts/gy 80 Y-intercept ~ 85 counts 60 0 Calibration Curve from 16mm Data 20 0 0 1 2 3 5 6 7 8 9 Dose (Gy) at (100, 100, 100) 60 20
What order of magnitude error in film reading causes a % error in dose? From calibration curve on previous slide: 5 Gy a reading of 120 counts.8 Gy a reading of 118.6 counts A difference of only 1. counts results in a % dose discrepancy 61 Summary The AAPM is working towards a calibration protocol for GSR units based on ADCL calibrated ionization chambers PPS/RFP coincidence and beam profile measurements are facilitated by using EBT GAFCHROMIC film ROF measurements remain the most challenging aspect of QA for GSR devices 62 Thank you for your attention! 63 21