Peter J. Biggs Ph.D. Massachusetts General Hospital Harvard Medical School Boston, MA 02114

Transcription

1 Basic Design Principles and Latest Recommended Data, Neutrons and Structural Concerns Winter on Cape Cod Peter J. Biggs Ph.D. Massachusetts General Hospital Harvard Medical School Boston, MA 02114

2 Outline 1 Basic principles and definitions 2 Calculations 3 Workload, use and occupancy factors 4 Shielding materials 5 Tenth value layers (TVL) 6 Room layout & features, construction details 7 Neutrons & laminated barriers 8 Mazes and doors low energies 9 Ducts 10 Skyshine 11 Examples of calculations

3 Definitions P: Weekly design dose limit (Sv/wk) d: Distance from target to measurement point W: Workload (Gy/wk) U: Use Factor T: Occupancy factor a: Scatter fraction; (θ, E) d sec : Distance from scatterer to measurement point d sca : Distance from target to scatterer d l: Distance from the target to measurement point F: Area of the beam in the plane of the scatterer (cm 2 ) B: Barrier transmission factor

4 Basic Principles The purpose of radiation shielding is to reduce the effective equivalent dose from a linear accelerator to a point outside the room to a level that is determined by individual states. In the US: Public or uncontrolled area msv per week msv in any one hour Controlled area - 1 msv/wk k(in practice 0.1 msv/wk)

5 Basic Shielding Equations: NCRP #151 Methodology Primary: Scatter: Leakage: B p B B s = l = 2 Pd WUT = I in I out P awt d 2 sec 1000* Pd WT d 2 l B x is the barrier transmission factor 2 sca 400 F B x = I I out in The factor 1000 is due to the 0.1% leakage requirement

6 Number of Tenth Value Layers (Primary & Leakage) The number, n, of tenth value layers (TVLs) required to reduce the dose to the value P is given by: n = log10 ( B) and the thickness for primary pimay and leakage radiation is given by: T = TVL + ( n 11 ) TVL p, l 1 p, l e

7 Number of Tenth Value Layers for Scattered dr Radiation Scatter TVLs are also dependent on scattering angle and there is no shoulder to the attenuation curve, so: T = n s s TVL s ( ) ϑ

8 Tenth Value Layers Primary TVLs (cm)* Energy (MV) Concrete 35(30) 37(33) 41(37) 44(41) 45(53) 46(44) Steel Lead Leakage TVLs (90 )* Energy (MV) Concrete 33(28) 34(29) 35(31) 36(33) 36(34) 36(34) Steel** Lead** * First term is 1st TVL and term in brackets is for all other TVLs. Data from NCRP #151 ** Data from McGinley

9 Obliquity Factor -I P

10 Obliquity yf factor - II Rule of thumb: - for θ >45 - and B< Add ~2 HVL for low-energy photons - And ~1 HVL for high-energy photons

11 Obliquity factor - III - Experimental data by Kirn et. al. (1954) - verified by Biggs (1996) using the Monte Carlo for a large range of clinical energies - Monte Carlo results are in good agreement with previous recommendations at low energy, but more detail is given - Caveat: beware of applying obliquity factor corrections for large angles!!

12 Conventional Workload - Number of patients treated per week multiplied by the average dose (MU)/patient at isocenter - Include calibration, service - Recommended dvalues: 500 or 1000 Gywk -1 - Survey shows values from 250 to 450 Gy wk -1 - Balance between highh energy and low energy use - Electron-onlymachines? l

13 Workload TBI - Workload for TBI>>workload for conventional therapy due to the extended distances: W = D TBI 2 TBI - Leakage workload is also higher, but patient- and wall-scattered workload is not. TBI - Radiation is usually directed at one barrier - Treatment time is much longer than for conventional R x d

14 Workload IMRT (1) - Workload for IMRT is complicated by the fact that the #MU is much larger than for conventional therapy. The IMRT factor is defined as: C = I MU MU IMRT conv and varies bt between 5 and 10 or more - Thisonly affects leakage radiation, not primary or scattered radiation.

15 Workload IMRT (2) - Serial tomotherapy has the highest C values, >10. Helical tomotherapy is much less. - Linac based IMRT at MGH gives an average C value of about 6 - Anecdote: For a 6/18 MV machine the energy use prior to IMRT was 20%/80% (MU). With 28% IMRT patient load, the use was 70%/30%

16 Copnventional Use Factors 90 interval: - 0 (31%); 90 and 270 (21.3 % each); 180 (26.3 %) - previously, all barriers were assigned interval: - 0 (25.6 %); 45 and 315 (5.8 % each); 90 and 270 (15.9 % each); 135 and 225 (4 % each); 180 (23 %)

17 Use Factor for Conventional Therapy

18 Use Factors Special Cases - A significant TBI load will require one wall to have an increased use factor - IMRT may also require a change in the values assigned to the use factor for the primary barrier

19 Use Factor Dedicated Machines - One should pay py attention to the types of procedures to be carried out in new room - A room dedicated to SRS and SRT requires different considerations from a general therapy room - A room dedicated to breast treatments will use specific c ranges of primary ybeam angles ges

20 Example: Use Factor for Stereotactic Radiosurgery

21 Typical Angles Used in Breast Therapy

22 Workload - Summation (1) Primary: WU p = ( W U W U + W U + W U +...) conv = WU wall scat conv Patient scatter: W + TBI TBI IMRT IMRT QA QA scatter, iso = conv IMRT QA + = ( W + W + W +...) (note that the TBI contribution is calculated separately source not at isocenter)

23 Workload - Summation (2) Leakage: W L = conv TBI I IMRT QA QA ( W + W + C W + C W +...)

24 Occupancy Factors - Occupancy factors down to 1/40 are now allowed (previously 1/16) - Lowest occupancy allows 0.02 msv in any one hour and 0,02 msv/wk - Of note, the occupancy outside a treatment room door is 1/8, instead of the conventional unity for a controlled area

25 If the Dose Rate is High, Should I Freak ko Out? Note that at the weekly limit of msv/wk, the dose rate outside a barrier could be quite high. e.g. for a primary barrier, an occupancy of 1/40 would imply a weekly dose of 3.2 msv and for a beam on time of 2.8 hr, h,the instantaneous dose rate is 1.1 msv/hr!! Moral: Choose factors wisely!!

26 TADR - TADR is the dose rate averaged over a week, a day or an hour: Daily beam on time TADR = IDR * Length of working week - The TADR (1 hour) or dose limit it in any one hour is: WU R * h h = IDR DR 0

27 TADR - NCRP recommends use of 1 hr as the minimum period for measuring IDR - Not an issue for occupancy of f1 - Only an issue when treatments t t are not given uniformly throughout the 40 hr week

28 Structural Details

29 Shielding Materials Materials Density Comments (g/cm 2 ) Concrete 235; ; 3.85 High density concrete is very expensive Concrete blocks 2.35; 3.85; 4.62 Lack structural integrity of concrete Lead Great for photons; bad for neutrons. Needs structural support Steel 7.87 Not as efficient as lead for photons Earth 1.5 Cheap! Therefore, build underground Brick Polyethylene; l ~ Used to shield against neutrons in doors, ducts etc. borated polyethylene ** Note: For h.d. concrete blocks, mortar of equiv. density should be used.

30 Possible Beam Orientations for Room with Maze

31 Width of Primary Barrier -I Allow 0.3 m (1 foot) either side of the primary beam w = d allow for fixed primary collimator diameter (0.5 m) w = 0.5 d Allow for greater width at the top of the wall w = 0.5d cos( ϑ) where d is the distance between the source and a point in the barrier and θ is the angle subtended by the central axis with the horizontal

32 B A C D

33 Contouring the Primary Shielding in the Ceiling

34 Physics Conduit

35 Lintel over Maze Wall

36 Lead-Only Room: Groundshine

37 Construction Techniques for Upgrading Linear Accelerator Rooms

38 A B C

39

40

41 Neutrons

42 Neutron Photoproduction Cross-Sections on Pb Cross: (γ,xn) Plus: (γ,1n)

43 Neutron Yields vs Energy and Z kw -1 ) ns sec -1 k Y (neutron Electron energy (MeV)

44 Photoneutron Energy Spectrum Re elative nu umber of neutron ns per ½ MeV Neutron energy, E n (MeV)

45 Shielding Thickness Required to Halve the Average Neutron Energy Half Ene ergy Lay yer (cm) most probable energy E(MeV)

46 Neutron Spectra from Medical Linacs - Neutrons produced d in the head of the linac are first moderated by the x-ray shielding - Neutrons are further moderated by scattering off the concrete walls of the therapy room

47 Neutron Spectra from Medical Linacs - The total neutron fluence therefore consists of direct ( fast ) fast) neutrons, scattered neutrons and thermal neutrons Φtotal = Φdir + Φsca + Φth - Fast neutrons obey the inverse law, but scattered and thermal neutrons are isotropically distributed; hence the neutron fluence drops off less fast than inverse square

48 Facure et. al.: 10 MV

49 Production of Neutrons by Primary Beam in a Laminated ated Barrier Incoming photon t t m t2 1 neutron Capture γ ray For high energy x-ray beams, the x-ray dose is enhanced by a factor of 2.7 to be conservative

50 Production of Neutrons by Primary Beam in a Laminated Barrier H DRF 0 max 1 x 2 = *10 *10 tm + t t / TVL t / TVL n where e H is the neutron dose equiv.(µsv wk -1 ) D 0 is the x-ray dose at isocenter (cgy wk -1 ) R is the neutron prod n. rate (µsv cgy -1 m -2 ) and F max is the max. beam area at isocenter (m 2 )

51 Total Dose Behind a Laminated Barrier (1) H tot = H n + H ph = H n1 +H n2 +2.7*H tr The 2.7 factor accounts for the production of capture gamma rays by the photo-produced neutrons. This is an conservative, empirical figure givenbymcginley.h n1 and H n2 represent direct neutrons and photo-produced neutrons respectively..

52 Total Dose Behind a Laminated Barrier (2) - Photoneutron production data is given for 18 MV for lead and steel and 15 MV for lead - The capture gamma ray enhancement factor was derived empirically for 18 MV on lead. No data for lower energies. However, stated to be good for all energies.

53 Mazes & Doors Low Energies

54 Contributions to the Dose at the End of a Maze 1. Primary scatter 2. Patient scatter 3. Leakage along maze 4. Leakage through maze wall

55 Secondary Radiation at Door due to Wall Scatter of Primary Beam S s = W U B α Aα A ( d d d ) 2 i r1 r 2 2

56 Secondary Radiation at Door due to Patient Scatter S p = aw U B ( F / 400) ( d d d ) 2 sca sec s α A 1 1

57 Secondary Radiation at Door due to Leakage L = L 0 W U B α1 ( d d ) 2 d s d 1 A 1

58 Secondary Radiation at Door due to Leakage Through Maze Wall L d = L0 W U B10 ( d ) 2 dl d ( t / TVL) Note that this radiation is more energetic than radiation along the maze

59 Mazes The total dose when the beam is directed towards the wall is the sum of the four components: D t = f*s s + S p + L + L d Note that the S s component is reduced by a factor f to account for patient attenuation. Assuming use factors of ¼ for the 4 gantry angles, McGinley gives an average total dose of: D T = 264D 2.64 t

60 Comments on NCRP Formalism Several authors have noted that the foregoing formalism underestimates the true dose at the door. This problem has been partly rectified by redefining the geometry of the scatter and assuming lower energies for the scattered radiation which, in turn, increase the scatter coefficients Numark &Kase;Al-Affan; McGinley & James

61 Reflection Coefficients for Monoenergetic X- Rays on Concrete NCRP #51

62 Maze Special Case

63 Ducts

64 Principle of Minimizing Loss of Shielding due to Presence of Duct Outside room Shielding wall Inside room Duct Leakage or scatter radiation Plane of gantry rotation Isocenter

65 A Ideal arrangement Additional shielding Ceiling Duct B C Ceiling duct McGinley s recommendation Isocenter

66 Skyshine Photons Neutrons

67 Skyshine: Measurements etc. - McGinley has shown that, for x-rays, the calculations are an underestimate at short distances (factor of 4) and an overestimate at long distances (factor of 5). - McGinley has shown that, for neutrons, the calculations are a gross underestimate at all distances. - Zavgorodni has presented a formalism for side scattered radiation for radiation penetrating thin barriers.

68 Practical Problems with NCRP #151 - Use of dual energy workloads, dual TVLs, laminated barriers creates confusion in the calculations - Use of the obliquity factor for scattered radiation (not leakage) is unclear two citations in the text give opposing views. Experts are divided on this issue, but it can be used, but with caution!

69 Handling Dual Energies in NCRP #151 - A dual energy machine having say 6 and 18 MV will have different workloads for (P,S) and L radiation for 3D CRT and IMRT (6 MV assumed). In general, a conservative approach assumes: - use of the maximum energy in a full 3D CRT environment for the primary and - the IMRT workload for the secondary barrier orthogonal to the plane of gantry rotation

70 Handling Dual Energies in NCRP #151 - in the secondary barrier region adjacent to the primary, the dominance of 3D CRT vs 3D CRT+IMRT depends on the explicit workloads - one should be careful to consider 20 scatter radiation for laminated barriers since the primary beam angle is only 14 (exclusive of extra 2 barrier width recommended by NCRP)

71 Handling Dual Energies in NCRP #151 - In that same angular zone in the secondary barrier, there will be be TVLs for two energies for scattered and leakage radiation - Rule is that: - scatter should be added to scatter to determine if an extra HVL is required - leakage should be added to leakage to determine if an extra HVL is required - Scatter is then compared to leakage to determine if an additional HVL is required

72 Handling Dual Energies in NCRP #151 - To recap: (i) S1 + S2 S (ii) L1 + L2 L (iii) S +L Total secondary thickness - Question remains as to which of the TVLs to use; higher or lower for L (ii); S or L for (iii) - However, because of these multiple additions, this may overestimate t the required shielding

73 Example: Barrier Adjacent to Primary 3D CRT + IMRT Energy Radiation Workload (Gy) Thickness Comment 15 MV Leakage MV Leakage Combined MV Leakage Use TVL 1 for 15 MV 15 MV Scatter MV Scatter Combined MV Scatter Use TVL sca for 15 MV Combined MV Leakage + scatter 53.3 Use TVL 1 for 15 MV N o t e Note: and only one shielding material is used!

74 Example: Barrier Orthogonal to Gantry Rotation Plane 3D CRT + IMRT Energy Radiation Workload (Gy) Thickness Comment 15 MV Leakage MV Leakage Combined MV Leakage Use TVL 1 for 15 MV 15 MV Scatter MV Scatter Combined MV Scatter Use TVL sca for 15 MV Combined MV Leakage + scatter 47.8 Use TVL 1 for 15 MV

75 TVLs for Laminated Secondary Barriers

76 Summary - The basic formalism of the calculations is well explained - NCRP #151 addresses the shielding issues raised by modern radiotherapy accelerators and techniques hi and allows the designer to perform the necessary calculations (but see second talk) - Data is up-to date - A complete set of examples are given for every type of barrier plus robotic arm (~30% of book

77 Thank Thank you foryou for your attention! your attention! Muito obrigado pela sua atenção