IMRT Delivery System QA. IMRT Dose Delivery. Acceptance testing. Why: specific tests for IMRT? Accuracy of leaf positioning (gaps) MLC Alignment

Transcription

1 1 IMRT Delivery System Q Thomas LoSasso, PhD Memorial Sloan Kettering Cancer Center IMRT Dose Delivery cceptance testing Commissioning Quality assurance Verification Q Why: specific tests for IMRT? 2. Gap error Dose error ccuracy of leaf positioning (gaps) Static IMRT Leaf affects borders only affects dose within target position 1-2 mm OK.2 mm OK % Dose error Range of gap width.5 Gap error (mm) Nominal gap (cm) MLC lignment cceptance testing MLC alignment Film Field Field MLC leaf / gap calibration Timing errors 1 1 eam characteristics at small MU Film 2 6 Field 9 1 Field

2 2 Doubly - exposed films MLCtable.txt corrections Film 1 Film 2 perfect alignment Correction (mm) MLCtable measurement "precise" Varian.5 mm misalignments Distance from central axis (cm) Gap measurement - mechanical DMLC - Symmetry and Flatness Gap at isocenter upper lower.95.9 open 5mm 5mm/open cm 12 cm DMLC - Symmetry and Flatness Gap correction - measurement open 5mm 5mm/open transmission gap only MV.2 y =.9945x R 2 = MV.2 y =.9954x R 2 = Nominal gap (mm)

3 3 Effective gap offset variation Mark 1 Mark 2 Millenium Phelps X 15X 6X 15X 6X 15X 6X 6X 15X Off-axis variation - "cupping" Millennium 12 Effective gap offset (mm) CX 5cm off 1cm off Series4 Series5 Series6 +/- 1% 15 MV 6 MV Eff gap offset iso) 2 4% Dose Off-axis variation - corrected Eff gap offset 1.2 S =.47 mm with cupping correction Eff gap offset 1.2 S =.27 mm ~1% dose with individual leaf corrections Frequency of deviations (%) # Global logged - prescribed leaf iso (mm) checks RMS Field Room ngle < LO Log file Leaf sequence file dose distribution LPO Sup P Post nt RPO RO cm Inf Overlay jaws Dose differences (log-dva) # checks = # moving leaves * OT / 55 msec

4 4 Difference (%) Timing error SMLC (Varian) MU/seg 16 MU/seg 25 MU/seg 4 MU/seg -5 Delivered with dose mode at 4MU/min 1 MU/seg (CL23, 6MV, 1 SSD, cm depth) MU/seg # Segment 16 MU/seg 25 MU/seg 4 mu/min Courtesy of Ping Xia, UCSF Timing error SMLC (Varian) Courtesy of Gary Ezzell and the Journal of pplied Clinical Medical Physics. Overlay 11 MU Dose differences (log-dva) 11 MU 55 MU 2X Timing / communication errors Distance from axis (mm) 28 MU 4X 14 MU 8X MU (%) 15 MU Timing / communication errors Distance from axis (mm) Distance from axis (mm) Comparison of prescribed and delivered doses Log file Leaf sequence file dose distribution 15 MU prescribed / 6 MU delivered F G F G MU (%) 15 MU MU (%) 6 MU D Overlay plan log C E D C Difference (log - plan) E

5 5 Relative dose (corrected to 15 MU) MU decrease dose errors increase Timing error Is there an analogous problem for DMLC? C D E F G Prescribed MU = MU delivered Trailing leaves (-side) compensate for the leading leaves (-side) Dose (MU) error = position error () position error leaf speed () leaf speed () However, some leaf sequencers actively utilize the beam hold-off to control leaf speed eam Stability: Flatness, Symmetry Stability of flatness and symmetry affects dose rate for small fields directed off the central axis. For an open 2x2cm 2 field, measure profiles for irradiations ranging from 1 to 1 MU. Sun Nuclear Profiler (46 diodes, 1 profiles/sec). Flatness is +/-3% if more than 5MU delivered. Symmetry +/-3% if more than 4MU delivered. Courtesy of M. Sharpe, William eaumont Hospital Transmission Transmission profiles across leaves Mark 1 (445) Millenium (444) average average Mark 2 (245) Mark 2 (442) average average Respiratory gating Parallel Splitting large fields Y Perpendicular Leaf motion X

6 6 Routine Q Quality assurance MLC Q MLC specific Mechanical stability ccuracy of leaf positioning (gaps) Verification Patient / field specific Tx plan is transferred and delivered correctly Motor fatigue MLC calibration acklash Grease Film pattern (semi-weekly) Dosimetry (monthly) Dosimetry (monthly) lcohol (semi-annually) Motors Motor replaced replacement history MLC motors replaced # Jan- 92 Jan- 93 MLC installed 1st MLC Tx Room Jan- Jan- Jan- Jan- Jan Date 1st dmlc Tx Jan- Jan- Jan Jan- Jan a Field Size Defined by MLC at 1 cm SD by Light Field: Leaf / carriage positioning vs off-axis distance Date: Tolerance:.5 mm Initials: Digital setting (cm) carriage carriage leaves leaves Note: If leaf positions exceed tolerance, reinitialize and recheck. If this does not correct the problem, verify that the correction parameters in files MLCTLE.TXT and MLCXCL.TXT have not changed. 4 cm

7 7 1 mm bands Leaf positioning errors introduced 17b DMLC Output (Gap width) Check 15 MV x-rays, File: DMLC_Q2.dva, Dose = 1 MU, Dose rate = 24 MU/min SD: 1cm, Depth = 3.1cm, ackup = 1cm, C.S. = 1x1cm2 Tolerance: (The ratios should be within 2% (<.2 mm gap) of the annual calibration.) -.5 mm -.2 mm Measure the output for the scanned field before and after reinitializing the MLC. Measure the output for a 1x1cm2 reference field size defined by the jaws (MLC retracted). Calculate the ratio of the scanned field to the reference field outputs. Reference 1x1 DMLC - before reinitialization DMLC - after reinitialization Reference 1x1 +.2 mm +.5 mm verage Ratio 17c eam Holdoff Functionality Date: Initials Same setup as 17b, Dose = 5 MU Record the output and the Tx time. Check if the holdoff is being applied. Reading Functioning nnual DMLC / 1x DMLC Output vs Time Jan-98 Jan-99 Jan- Jan-1 Jan-2 Jan-3 Jan-4 Date.2 mm ~1% 17d Output vs Gantry and Collimator ngles Date: Initials 15 MV x-rays, File: DMLC_Q2.dva, Dose = 1 MU, Dose rate = 24 MU/min SD: 1cm, Spokas in air w/buildup, C.S. = 1x1cm2 Tolerance: (The ratios should be within 3% (<.2 mm gap) of each other.) Position a cylindrical chamber in air at isocenter. For each gantry angle, measure the open field and the scanned field outputs. t gantry angles of 9 and 27, also measure output for collimator angles of 9. Calculate the ratio of the scanned field to the reference field outputs. Collimator angle Gantry Reference DMLC DMLC / Reference angle DMLC output vs gantry and collimator angle Mark 1 (445) Millennium (444) DMLC output vs gantry angle (diode array) 3 DMLC / 1x Mark 2 (245) Millennium (241) ±.2 mm (± 1%) mm Year

8 8 Summary 1. cceptance testing MLC capable of IMRT delivery Improved calculations needed for IMRT fields 2. Commissioning MLC modeled accurately in TPS 3. Routine MLC Q leaves and carriages - mechanics 4. Patient specific Q independent verification IMRT Dose ccuracy 2 cm Plan Film Overlay 2 cm Film - Plan Film - Plan Difference ~25% Dosimetry (film plan) T&G Source MLC scatter Lung P field Corrections (plan plan) x1 Tongue and Groove

9 9 Relative dose Tongue and groove Distance from CX (cm) -1 cm cm 1 cm 2 cm DMLC / 1x1 (norm to -) DMLC output vs gantry and collimator angle adjusted bearings replaced bearings adjusted bearings adjusted bearings Mark 1 (445) adjusted drive belts adjusted bearings