ÿÿ 
`ÿ 


`
` 
`ÿÿÿÿ
` 
`!"#ÿ$%&'()*ÿ+,-!%'-#(.ÿ$/$!#+ÿ0ÿ(ÿ-(12#
`.%#-3ÿ$%4#5ÿ"%2"666ÿ7899:;
`ÿ<
`= > ÿ  
`@AÿB CÿDB ÿEDÿFÿGÿ HÿIÿJKÿAHAÿÿD ÿ
`!?


` I I
`(
` @ L CKÿMN
`3O 
`O P QEÿQA
`3 ÿ?RSTTBUTBUV
`(W ÿG
`1X P7; I ÿA EKÿEBDI Iÿ IA EÿYZ=M@[Kÿ ÿ  K
`DB ÿD ÿY=\][KÿD HBHIÿ IA EÿYZ^M@[KÿZ=M@
`_Kÿ ÿIHKÿ D ÿ`IAÿYaTBSTb[Kÿ ÿ`IAKÿ B IBA
`_
`- PO OcHA
`!def ÿ@A
`gh

?
` QAÿQE
`(h Gÿ ÿ  ÿY [ÿÿAÿDÿDDEÿIÿIÿÿ Dÿ
`
` ÿ`Aÿ ÿÿ ÿ Dÿ IA EÿYciM@[Nÿ@Aÿ\ ÿIÿ
`AHABHEÿJHÿ I ÿY Aÿÿ[ÿÿAÿHÿÿA
`
` jÿDNÿ@Aÿ  ÿÿ I ÿÿAÿ Iÿ  ÿI D HÿA
`I IÿNÿ@ÿDDJÿ I ÿÿA AEÿÿ DÿAIÿÿA IN
`]DDEKÿAÿÿ AIÿEÿHÿ ÿDB ÿD ÿY=\][Nÿ@A
`D ÿÿAÿ`Cÿ` ÿHÿEÿAÿ ÿA ÿ ÿ ÿ\ k=\]K
`Eÿ   ÿÿAÿD CKÿ` ÿIHIÿÿD ÿ IA EKÿ
`
`ViewRay Exhibit 1022
`Page 1 of 121
`
`

`

` ÿ  ÿ ÿ
` ÿ

 ÿ  
` ÿ  ÿÿ ÿ
` ÿÿ

ÿ ÿ ÿÿ
`   ÿ ÿÿ
` 

ÿÿÿ

ÿ 
`  ÿ  ÿ ÿÿ  ÿ  ÿÿ  ÿ  ÿ !
` 

ÿÿ

ÿÿ"ÿ 

ÿ    ÿ  ÿÿÿ#  ÿ

ÿÿ   "
` ÿ
`$ ÿ  ÿ  ÿÿÿ 
` ÿ %ÿ& ÿ ÿ
`  "ÿ  
`
`ÿ ÿ #ÿÿ  ÿ ÿ 
`ÿ
` ÿÿ' ÿ ÿ'  
`  ÿ  ÿ   ÿÿ$ÿ ÿ

  "ÿ ÿÿ ÿÿ$ ÿ
`  ÿÿ
` ÿ  ÿ  ÿ

 ÿ %ÿÿÿ  
` ÿ

 ÿ  "
`ÿ   ÿ ÿ   ÿ  ÿ

ÿÿ
`  

ÿ  "ÿ ÿÿ ($ #
`ÿÿÿ  ÿ  ÿ

ÿ ÿ  ÿ

ÿÿ ÿ 

ÿ$ # ÿ
`  ÿ ÿ ÿ)*ÿ  ÿ   ÿÿÿ ÿÿ   ÿ ÿÿ$ #ÿ 
`ÿ   ÿÿ

ÿ+ÿÿ ,ÿÿÿ   ÿ ÿ ÿÿ   ÿ ÿ
`ÿ(  ÿ   ÿÿ ÿ ÿ ÿ  ÿ

 ÿ %ÿÿ ÿ  
`
`

 ÿ  "ÿ ÿ  ÿ

 ÿ %ÿÿÿ  
` ÿ

 ÿ  ÿ& ÿÿ (
`$ #ÿÿÿ
` ÿ

`  ÿ ÿÿ
` 
`  ÿ  "ÿ
`$ ÿ

 ÿ
` ÿ

ÿ   "ÿÿ ÿ

 ÿ  ÿÿ ÿ  ÿ ÿÿ 
`  
` ÿÿ   ÿ   ÿ ÿ$ÿ  ÿ ÿÿÿ    ÿ
`$ #ÿÿÿÿ 

ÿÿ

ÿ+ÿÿÿ

  ÿ ÿ  
` ÿÿ
` $ ÿÿÿ

ÿÿ "ÿ$
`ÿÿ 

ÿÿÿ$ÿÿ$
`
` ÿ ÿ
`$ ÿ

ÿ   ÿ&  "ÿ ÿ   ÿ ÿ

ÿÿ
`ÿ   ÿ  ÿ ÿ"ÿ ÿ ÿ
` ÿÿ

ÿÿ ÿ 

ÿ 
`  ÿ

 ÿÿ ($ #ÿÿ ÿ 
`  "ÿ ÿ  ÿ ÿ

ÿ 
`  ÿ

 ÿÿ  ÿ

ÿ  ÿ-ÿ

`  ÿ   ÿ

ÿ
`
`ÿ  ÿ.ÿ ÿ  ÿ

ÿ  ÿ  ÿ 

 ÿ ÿ
` ÿ$ "ÿ  ÿ 
` ÿ

ÿ  ÿ ÿ$ÿ $ ÿ ÿ
` ÿ ÿÿ
` ÿ$ 

ÿ   ÿ

ÿÿ ($ #ÿÿ  ÿÿ    ÿ 
`/0123456378!  $  
`
`   +9:;+<<)*
`A*<A<;<)<B
`=>?@
`/CDE3E6C86
`=FC863G3CDHI!I-I!I!H!+*+9:;+<<)*
`JC65F565K
`>70D4CH ÿH  
`ÿ
`
`ViewRay Exhibit 1022
`Page 2 of 121
`
`

`

`The six-bank multi-leaf system
`
`A large fleld size, high resolution collimator for advanced
`radiotherapy
`
`Rajko Topolnjak
`
`ViewRay Exhibit 1022
`Page 3 of 121
`
`

`

`ii
`
`Colofon:
`This text was set using the freely available LATEX 2" typesetting and text for-
`matting system. The line drawings were made using the freely available Xflg
`program.
`
`ISBN-10: 90-393-4363-2
`ISBN-13: 978-90-393-4363-0
`Druk: PrintPartners Ipskamp, Enschede
`
`Copyright:
`Chapter 3 copyright 2004 IOP Publishing Ltd.
`Chapter 4 copyright 2005 IOP Publishing Ltd.
`
`ViewRay Exhibit 1022
`Page 4 of 121
`
`

`

`The six-bank multi-leaf system
`
`A large fleld size, high resolution collimator for advanced
`radiotherapy
`
`Het zes-bank multi-leaf systeem
`Een collimator voor grote velden, en een hoge resolutie ten behoeve van
`geavanceerde radiotherapie
`(met een samenvatting in het Nederlands)
`
`Proefschrift ter verkrijging van de graad van doctor aan de Universiteit
`Utrecht
`op gezag van de Rector Magniflcus, prof. dr. W.H. Gispen
`ingevolge het besluit van het college voor promoties
`in het openbaar te verdedigen
`op dinsdag 17 oktober 2006 des middags te 2:30 uur
`
`door
`Rajko Topolnjak
`geboren op 25 januari 1974 te Zagreb, Kroati˜e
`
`ViewRay Exhibit 1022
`Page 5 of 121
`
`

`

`iv
`
`promotor:
`
`co-promotor:
`
`Prof. dr. ir. J.J.W. Lagendijk
`Faculteit der Geneeskunde, Universiteit Utrecht
`Dr. U.A. van der Heide
`Faculteit der Geneeskunde, Universiteit Utrecht
`
`Het beschreven werk werd verricht op de afdeling Radiotherapie van het Uni-
`versitair Medisch Centrum Utrecht, participerend in het Image Sciences Insti-
`tute en de onderzoekschool voor biomedische beeldwetenschappen ImagO en
`werd mogelijk gemaakt met flnanci˜ele steun van Elekta Ltd. Deze uitgave is
`tot stand gekomen met flnanci˜ele steun van ImagO en Elekta Ltd, Crawley,
`UK.
`
`ViewRay Exhibit 1022
`Page 6 of 121
`
`

`

`to Ivana and Ema
`
`ViewRay Exhibit 1022
`Page 7 of 121
`
`

`

`ViewRay Exhibit 1022
`Page 8 of 121
`
`ViewRay Exhibit 1022
`Page 8 of 121
`
`

`

`Contents
`
`1 General introduction
`1.1 New developments in radiotherapy . . . . . . . . . . . . . . . .
`1.2 Design equipment . . . . . . . . . . . . . . . . . . . . . . . . . .
`1.2.1 Linear accelerator and multi-leaf collimator (MLC) . . .
`1.2.2 MRI-linear accelerator . . . . . . . . . . . . . . . . . . .
`1.2.3 A six-bank multi-leaf system . . . . . . . . . . . . . . .
`1.3 This thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`1.3.1 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . .
`1.3.2 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`2 In(cid:176)uence of the Linac design on intensity-modulated radio-
`therapy of head-and-neck plans
`2.1
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`2.2 Methods and Materials . . . . . . . . . . . . . . . . . . . . . . .
`2.2.1 Patients . . . . . . . . . . . . . . . . . . . . . . . . . . .
`2.2.2 Plan criteria
`. . . . . . . . . . . . . . . . . . . . . . . .
`2.2.3
`Inverse treatment planning . . . . . . . . . . . . . . . .
`2.2.4 Linac/MLC design parameters
`. . . . . . . . . . . . . .
`2.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`2.3.1 Parameters check . . . . . . . . . . . . . . . . . . . . . .
`2.3.2
`Ideal plans
`. . . . . . . . . . . . . . . . . . . . . . . . .
`2.3.3 Leaf width . . . . . . . . . . . . . . . . . . . . . . . . .
`2.3.4 Transmission . . . . . . . . . . . . . . . . . . . . . . . .
`2.3.5
`Source size
`. . . . . . . . . . . . . . . . . . . . . . . . .
`2.3.6 Middle linac/MLC . . . . . . . . . . . . . . . . . . . . .
`2.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`1
`1
`3
`3
`4
`5
`6
`6
`7
`
`9
`10
`11
`11
`11
`12
`15
`17
`17
`18
`20
`20
`22
`22
`23
`26
`
`vii
`
`ViewRay Exhibit 1022
`Page 9 of 121
`
`

`

`viii Contents
`
`3 A six-bank multi-leaf system for high precision shaping of
`large flelds
`3.1
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`3.2 Ray tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`3.2.1 Computer simulation . . . . . . . . . . . . . . . . . . . .
`3.2.2 Validation of the computer simulations . . . . . . . . . .
`3.3 The six-bank multi-leaf system . . . . . . . . . . . . . . . . . .
`3.3.1 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`3.3.2 Characteristics . . . . . . . . . . . . . . . . . . . . . . .
`3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`4 IMRT sequencing for a six-bank multi-leaf system
`4.1
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`4.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . .
`4.2.1 Conventional sequencing . . . . . . . . . . . . . . . . . .
`4.2.2 High-resolution sequencing . . . . . . . . . . . . . . . .
`4.2.3 Parameters for evaluating IMRT sequences
`. . . . . . .
`4.2.4 Fluence matrices . . . . . . . . . . . . . . . . . . . . . .
`4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`4.3.1 Conventional sequencing . . . . . . . . . . . . . . . . . .
`4.3.2 High-resolution sequencing . . . . . . . . . . . . . . . .
`4.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`
`5 MLC Leaf Design
`5.1
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`5.2 Materials and methods
`. . . . . . . . . . . . . . . . . . . . . .
`5.2.1 Radius of the leaf tip.
`. . . . . . . . . . . . . . . . . . .
`5.2.2 Geometric penumbra.
`. . . . . . . . . . . . . . . . . . .
`5.2.3 Transmission penumbra. . . . . . . . . . . . . . . . . . .
`5.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`5.3.1 Radius of the leaf tip . . . . . . . . . . . . . . . . . . . .
`5.3.2 Penumbra width (80-20%) for Elekta, Varian and Siemens
`MLCs . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`5.3.3 Penumbra width (80-20%) for a general MLC . . . . . .
`5.3.4 A six-bank MLC . . . . . . . . . . . . . . . . . . . . . .
`
`27
`28
`30
`30
`32
`34
`34
`36
`42
`
`45
`46
`47
`47
`49
`54
`55
`56
`56
`57
`63
`67
`
`69
`70
`71
`72
`74
`75
`75
`75
`
`76
`76
`79
`
`ViewRay Exhibit 1022
`Page 10 of 121
`
`

`

`Contents
`
`ix
`
`5.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`5.A Radius of the leaf tip . . . . . . . . . . . . . . . . . . . . . . . .
`5.B Geometric penumbra . . . . . . . . . . . . . . . . . . . . . . . .
`5.C Transmission penumbra . . . . . . . . . . . . . . . . . . . . . .
`
`6 Summary and general discussion
`
`7 Samenvatting
`
`References
`
`Publications
`
`Acknowledgments
`
`Curriculum vitae
`
`82
`83
`84
`85
`87
`
`89
`
`93
`
`97
`
`103
`
`105
`
`109
`
`ViewRay Exhibit 1022
`Page 11 of 121
`
`

`

`ViewRay Exhibit 1022
`Page 12 of 121
`
`ViewRay Exhibit 1022
`Page 12 of 121
`
`

`

`Chapter 1
`
`General introduction
`
`1.1 New developments in radiotherapy
`
`The quality of radiotherapy for treating patients with cancer has improved
`signiflcantly in the past couple of years due to technological developments in
`three areas. The use of multiple imaging modalities make it possible to visual-
`ize the tumor and deflne the target for irradiation. With intensity-modulated
`radiotherapy (IMRT) a dose distribution can be tailored to this target volume
`and with improved position veriflcation technology it is guaranteed that the
`dose is indeed deposited in the right place.
`
`Before the radiotherapy treatment, it is important to know the exact bound-
`aries of volume to be irradiated. Therefore, the regions of interest (ROIs),
`such as target volumes and organs at risk (OAR) are delineated on a plan-
`ning computer tomography (CT). This is particularly important for IMRT,
`because with this technique the dose distribution is shaped accurately around
`the target volumes. Consequently, if the targets and organs at risk are not
`delineated correctly, (Austin-Seymour et al., 1995), the dose distribution will
`be inappropriate for the patient.
`
`CT is most commonly used for radiotherapy treatment planning, but magnetic
`resonance imaging (MRI) is of increasing importance (Huch B˜oni et al., 1996;
`Cruz et al., 2002). The advantage of MRI is its good soft tissue contrast. CT
`provides information on the electron density and is therefore necessary for dose
`calculation.
`
`Magnetic resonance spectroscopy (MRS) and positron emission tomography
`(PET) have the potential to characterize the tumor and provide information
`
`1
`
`ViewRay Exhibit 1022
`Page 13 of 121
`
`

`

`2 Chapter 1. Introduction
`
`about the tumor metabolism (Ling et al., 2000). Application of multiple imag-
`ing modalities and subsequent registration of the obtained images, enables
`clinicians to deflne targets with higher reliability. This may lead to a reduc-
`tion of margins around the target volumes and prescription of higher doses
`(Mohan et al., 2000; Zelefsky et al., 2000; Nederveen et al., 2001; Vicini et al.,
`2001) to certain parts of the target. At the same time, the organs at risk can
`be spared better.
`
`With IMRT it is possible to shape the dose distribution to the anatomical
`and biological characteristics of the tumor, a concept called dose painting
`(Ling et al., 2000; Webb, 2001b). To this end a non-uniform beam intensity is
`used. This variation in intensity can be achieved by using compensators that
`attenuate the beam at speciflc positions, or by using a multi-leaf collimator
`(MLC) with which a sequence of irregular fleld shapes can be delivered (Jordan
`and Williams, 1994; Galvin et al., 1993b; Das et al., 1998).
`
`Determining the optimal way of modulating the radiation beams for IMRT
`is generally done by inverse treatment planning (ITP). ITP works backward
`through the complex range of radiation delivery options. The most important
`inputs for ITP are number of beams, their orientations, ROIs and prescriptions
`to the ROIs. The output is the calculated dose to the ROIs and a description
`of the treatment flelds (beam segments) which have to be delivered.
`
`The flnal important step in radiotherapy treatment is position veriflcation
`during the therapy. To avoid underdosing the target volume due to geometrical
`uncertainties (Jafiray et al., 1999; Langen and Jones, 2001) it is common
`practice to apply a margin around the target volume. A drawback of this
`approach is that healthy tissue is irradiated as well. To deliver the dose to
`the correct location it is essential that positions of the target(s) and OAR(s)
`during the treatment are the same as on the IMRT plan.
`
`Stroom et al. (1999) and Van Herk et al. (2000) presented a ‘margin recipe‘
`in which the relation between random and systematic positioning errors and
`the required margin was established. In order to minimize the size of these
`margins in particular the systematic positioning uncertainty must be reduced
`as much as possible.
`
`In adaptive radiotherapy (ART)(Martinez et al. (2001)) the target motion and
`systematic variation in patient setup is estimated by daily imaging during the
`
`ViewRay Exhibit 1022
`Page 14 of 121
`
`

`

`1.2. Design equipment 3
`
`flrst few treatments fractions. Subsequently, a new treatment plan is made in
`which an individual margin is applied that encompasses the positions found
`in these treatment fractions.
`
`Imaging during the treatment fraction has the possibility to recognize changes
`in patient anatomy due to internal organ motion. If the imaging technique
`provides su–cient soft-tissue contrast also the response of the tumor to the
`treatment can be observed. This approach is called image-guided radiotherapy
`(IGRT) and its potential is being explored clinically to date using a cone-beam
`CT mounted on an accelerator (Jafiray et al., 2002) and with the tomotherapy
`system (Mackie et al., 1993). Raaymakers et al. (2004) are working to integrate
`an MRI system with a linear accelerator so use the superior soft-tissue contrast
`of an MRI for treatment guidance in radiotherapy.
`
`1.2 Design equipment
`
`Technological improvements in design of linear accelerators have made IMRT
`and IGRT possible. Nevertheless, many linacs that are currently used have not
`been designed speciflcally for this purpose. In this paragraph we describe the
`available equipment for IMRT and IGRT and discuss possible improvements
`that would beneflt the quality of the treatment.
`
`1.2.1 Linear accelerator and multi-leaf collimator (MLC)
`
`A linear accelerator (linac)(Khan, 1994) is the most commonly used device
`for treatment of patients with cancer in external beam radiotherapy (EBRT).
`The Linac delivers a high-energy ionization radiation (photons or electrons) to
`the region of the patient’s tumor. The absorption of radiation in the treated
`area damages the diseased cells.
`
`To minimize irradiation of healthy tissue beams should be shaped. Commonly,
`this is achieved by using an MLC. The MLC (Jordan and Williams, 1994;
`Galvin et al., 1993b; Das et al., 1998) is located in Linac’s treatment head
`and is composed of computer controlled tungsten leaves. The various types of
`MLCs that are currently available commercially have difierent leaf widths and
`number of leaves. Originally, they have been introduced as a substitute for alloy
`block fleld shaping. Now, they are used for intensity modulated radiotherapy
`
`ViewRay Exhibit 1022
`Page 15 of 121
`
`

`

`4 Chapter 1. Introduction
`
`as well. With an MLC IMRT can be performed in two ways (Chui et al.,
`2001). A step-and-shoot approach of delivering IMRT can be applied (Bortfeld
`et al., 1994; Webb, 2001b). This usually produces a big number of segments.
`Therefore, MLCs should be capable of delivering many segments quickly. A
`second approach is dynamic IMRT. Here, the MLC leaves need to allow fast
`movement and accurate positioning.
`
`MLCs typically have leaves of 1 cm width projected at the isocenter and can
`produce a maximum fleld size of 40 x 40 cm2 (Jordan and Williams, 1994;
`Galvin et al., 1993b; Das et al., 1998). While the fleld size is su–cient for most
`treatments, the leaf width limits the precision at which flelds can be shaped.
`By introducing add-on mini-MLC (Meeks et al., 1999; Xia et al., 1999; Cos-
`grove et al., 1999; Hartmann and F˜ohlisch, 2002) with leaf widths ranging from
`1.6 to 4.5 mm at isocenter, the problem of undulations is decreased. However,
`a drawback is that none of them can achieve fleld sizes larger than 10 x 12 cm2.
`Varian produces an MLC with 60 leaf pairs and a fleld size of 40 x 40 cm2. In
`the central 20 cm of the fleld the leaf width is 0.5 mm while in the outer 10 cm
`on both sides of the fleld the leaf width is 1 cm. With such a design a high
`resolution is achieved in the central part of the fleld, but unfortunately not
`in the outer part. Elekta produces a mini-MLC integrated in the accelerator
`head with 40 leaf pairs and a fleld size of 16 x 21 cm2. The maximum fleld
`size of the mini-MLCs is reduced because, at maximum overtravel and a ’hor-
`izontal’ position the leaves may bend under their weight. The displacement
`is proportional to the fourth power of overtravel and inversely proportional
`to the square of the leaf width (Shigley and Mischke, 1989). Thus, by using
`0.4 cm leaves rather than 1 cm the maximum overtravel for the leaf would
`be decreased by a factor 0.63 without increasing the inter-leaf distance. It is
`important to flnd the optimal balance between fleld size and leaf width.
`
`Figure 1.1 shows an MLC manufactured by Elekta Ltd, Crawley, UK.
`
`1.2.2 MRI-linear accelerator
`
`From the currently available imaging modalities the MRI scanner has the
`best performance in soft tissue contrast. This is important in deflning tumor
`boundaries relative to surrounding healthy tissue. Moreover, MRI and MRS
`can characterize the tumor itself. Hypoxia, vascularity or blood (cid:176)ow inside the
`tumor can be investigated and characterized.
`
`ViewRay Exhibit 1022
`Page 16 of 121
`
`

`

`1.2. Design equipment 5
`
`Figure 1.1: The multi-leaf collimator manufactured by Elekta Ltd. (From
`Elekta website.)
`
`From a clinical point of view having a precise on-line, soft-tissue based posi-
`tion veriflcation system for IGRT would be greatly beneflcial. The feasibility
`of integrating MRI system with a linear accelerator has been investigated by
`Lagendijk and Bakker (2000); Lagendijk et al. (2002); Raaymakers et al. (2002,
`2003, 2004). Figure 1.2(a) shows an illustration of an integrated MRI accel-
`erator. The system combines a 1.5 T MRI scanner and single energy (6 MV)
`radiotherapy accelerator rotating around it. Figure 1.2(b) shows a schematic
`view of such a system.
`
`1.2.3 A six-bank multi-leaf system
`
`We propose an alternative design for an MLC. The idea comes from the project
`at our department of integrating an MRI system with a linear accelerator.
`Figure 1.2(b) shows a schematic view of such a system. The space available
`for an MLC is small and the currently available MLCs are not compact enough
`or they do not have the necessary performance speciflcations.
`
`The six-bank MLC is an alternative design of a multi-leaf collimator that com-
`bines high-resolution fleld shaping with a large fleld size. The system consists
`of three layers of bank pairs, positioned at 60o relative to each other. The
`
`ViewRay Exhibit 1022
`Page 17 of 121
`
`

`

`6 Chapter 1. Introduction
`
`(a)
`
`(b)
`
`Figure 1.2: (a) An illustration and (b) a schematic view of a MRI-linear
`accelerator.
`
`bank pairs each have 40 single-focussed leaves with a standard 1 cm width
`at isocenter. With the three layers of leaf banks a fleld-shaping precision can
`be achieved comparable to that of mini-MLCs with a leaf width of 4 mm at
`isocenter. The distance between the leaves and the source vary for each layer,
`which will as a consequence produce a difierent geometric penumbra. Further-
`more, by varying the leaf thickness the in(cid:176)uence on transmission penumbra
`can be made in order to have the same total penumbra for each layer.
`
`Although, the initial idea was to make MLC suitable for MRI-accelerator only,
`the MLC was developed further as a multipurpose MLC available for any
`Linac. It could be used as a conventional 1 cm leaf width MLC or as a mini-
`MLC. Furthermore, a six-bank MLC will have small transmission, and its
`compact size will allow more clearance for the patient. Collimator rotation
`would not be required and it will not sufier from the small maximal fleld sizes
`of conventional mini-MLCs. Figure 1.3 shows a six-bank MLC.
`
`1.3 This thesis
`
`1.3.1 Purpose
`
`The purpose of this thesis is to investigate if linac/MLC design can be im-
`proved to achieve a better radiotherapy treatment of cancer patients. We will
`
`ViewRay Exhibit 1022
`Page 18 of 121
`
`

`

`1.3. This thesis 7
`
`S1a (1.5mm FWHM)
`S3
`
`S4
`
` 12 cm
`
` 100 cm
`
`B3
`
`S2a (40mm FWHM)
`
` 5 cm
`
` 4 cm
`
` 3.5 cm
`
` 26 cm
`
` 33.5 cm
`
`B2
`
` 40 cm
`
`B5
`
` 8 cm
`
`B4
`(77.2 1/m)
`
` 6cm
`
` 4cm
`
`Figure 1.3: Schematic view of the six-bank MLC system and the computer
`simulation parameters from the lateral view (all parameters are listed in
`table 3.1).
`
`examine which design parameters are the most important. Finally, we will
`propose a new design of MLC and investigate its feasibility.
`
`1.3.2 Outline
`
`In chapter 2 we characterize and quantify the impact of linac/MLC design
`parameters on IMRT treatment plans. The investigated parameters were: leaf
`width in the MLC, leaf transmission related to the thickness of the leaves, and
`penumbra related primarily to the source size. We used CT scans from head-
`and-neck cancer patients. For each patient treatment plans were made with a
`difierent set of linac/MLC parameters. A hypothetical ideal linac/MLC was
`
`ViewRay Exhibit 1022
`Page 19 of 121
`
`

`

`8 Chapter 1. Introduction
`
`introduced to investigate the in(cid:176)uence of one parameter at the time without
`the interaction with other parameters. As a reference to clinical practice, we
`also optimized the plans with the clinically used Elekta SLi15.
`
`In chapter 3 we present a design of an alternative multi-leaf collimator, called
`a six-bank MLC, that combines high-resolution fleld shaping with a large fleld
`size. The six-bank collimator will enable shaping flelds of about 40 cm diam-
`eter, with a precision comparable to that of existing mini-MLCs with a leaf
`width of 4 mm.
`
`For the six-bank MLC which would function as a multi-purpose collimator,
`suitable for all types of treatments, it is important that IMRT can be delivered
`as well. Chapter 4 presents a sequencer for delivering step-and-shoot IMRT
`using a six-bank MLC. Two methods for delivering IMRT with a six-bank MLC
`were developed. In a low-resolution mode similar segments can be delivered as
`with a conventional two-bank MLC with a leaf width of 1 cm. The performance
`in high-resolution mode is comparable to that of a mini-MLC with a leaf width
`of 4 mm, but a trade-ofi had to be made between accuracy and number of
`segments.
`
`Chapter 5 presents an analytical model of an optimal MLC leaf design for a
`given setup of parameters. This was of importance because the six-bank MLC
`consists of three layers of two opposing leaf banks. The leaves in the banks
`that are closest to the source produce the largest geometric penumbra. This
`efiect was compensated by reducing the transmission penumbra of the higher
`banks.
`
`Summary and general discussion are presented in chapter 6. Here, we discuss
`beneflt of using difierent Linac design on IMRT. We summarize the perfor-
`mance and characteristics of a six-bank MLC and IMRT sequencer developed
`for it. Finally, the analytical model for leaf design is discussed.
`
`ViewRay Exhibit 1022
`Page 20 of 121
`
`

`

`Chapter 2
`
`In(cid:176)uence of the Linac design on intensity-modulated
`
`radiotherapy of head-and-neck plans
`
`This chapter has been submitted as:
`R. Topolnjak, U. A. van der Heide, G.J. Meijer, B. van Asselen, C. P. J. Raai-
`jmakers and J. J. W. Lagendijk. In(cid:176)uence of the Linac design on intensity-
`modulated radiotherapy of head-and-neck plans.
`
`Abstract
`In this study, we quantify the impact of linac/MLC design parameters on IMRT treat-
`ment plans. The investigated parameters were: leaf width in the MLC, leaf transmis-
`sion, related to the thickness of the leaves, and penumbra related primarily to the
`source size. Seven head-and-neck patients with stage T1-T3N0-N2cM0 oropharyn-
`geal cancer were studied. For each patient nine plans were made with a difierent set
`of linac/MLC parameters. The plans were optimized in Pinnacle3 v7.6c and PLATO
`RTS v2.6.4, ITP v1.1.8. A hypothetical ideal linac/MLC was introduced to investigate
`the in(cid:176)uence of one parameter at the time without interaction of other parameters.
`When any of the three parameters was increased from the ideal setup values (leaf
`width 2.5 mm, transmission 0%, penumbra 3 mm), the mean dose to the parotid
`glands increased, given the same tumor coverage. The largest increase was found for
`increasing leaf transmission. The investigation showed that by changing more than one
`parameter of the ideal linac/MLC setup, the increase in the mean dose was smaller
`than the sum of dose increments for each parameter separately. As a reference to
`clinical practice, we also optimized the plans of the seven patients with the clinically
`used Elekta SLi 15, equipped with a standard MLC with a leaf width of 10 mm. As
`compared to the ideal linac this resulted in an increase of the average dose to the
`parotid glands of 5.8 Gy.
`
`9
`
`ViewRay Exhibit 1022
`Page 21 of 121
`
`

`

`10 Chapter 2. In(cid:176)uence of the Linac design on IMRT of head-and-neck plans
`
`2.1
`
`Introduction
`
`A novel approach of delivering intensity-modulated radiotherapy (IMRT) is
`known as rotation IMRT or dynamic helical tomotherapy as presented by
`Mackie et al. (1993). The system is equipped with a multi-leaf collimator with
`leaves of 6.25 mm width and 10 cm thickness. Van Vulpen et al. (2005) com-
`pared step-and-shoot IMRT delivered by a conventional MLC with dynamic
`helical tomotherapy IMRT plans for patients with head-and-neck cancer. The
`results showed that the mean dose to the parotid glands can be reduced with
`about 6.5 Gy by using the tomotherapy system. This raises the question as to
`what is the in(cid:176)uence of the linac/MLC design parameters on IMRT.
`
`Several authors investigated the efiect of the leaf width on IMRT plans (Fiveash
`et al., 2002; Burmeister et al., 2004; Leal et al., 2004; Wang et al., 2004, 2005;
`Nill et al., 2005; Zhu et al., 2005). An overall conclusion is that when the size
`of the target is small, it is beneflcial to use MLCs with smaller leaf width. By
`using sampling theory Bortfeld et al. (2000) showed that optimum leaf width
`is 1.5-1.8 mm for the leaves without transmission.
`
`The main purpose of this study is to extend the investigation of linac/MLC
`design to the impact of MLC leaf thickness and source size. As the starting
`point we chose a hypothetical ideal linac. By increasing the MLC leaf width
`and the source size, and by decreasing the MLC leaf thickness the impact of
`these parameters on the IMRT plans was characterized.
`
`In this planing study CT scans of seven head-and-neck patients with stage
`T1-T3N0-N2cM0 oropharyngeal cancer were used. The flrst reason for this
`choice is that head-and-neck patients are complex and the dose requirements
`are challenging because of the variation in target dose and the overlap of
`targets with organs at risk (OAR). Secondly, flve patients from our group of
`seven were already investigated (Van Vulpen et al., 2005), which gave us the
`opportunity to compare our results with that study.
`
`Two treatment planning systems (TPSs) have been used so that the bias of
`speciflc planning system on the results could be minimized. Computations were
`made on both Pinnacle3 v7.6c (Philips Medical Systems, Best, The Nether-
`lands) and on PLATO RTS v2.6.4 (Nucletron BV, Veenendaal, The Nether-
`lands).
`
`ViewRay Exhibit 1022
`Page 22 of 121
`
`

`

`2.2. Methods and Materials 11
`
`2.2 Methods and Materials
`
`2.2.1 Patients
`
`Seven head-and-neck patients with stage T1-T3N0-N2cM0 oropharyngeal can-
`cer were studied. A planning CT scan (Philips Aura) with 3 mm slice thickness
`was made and CT-images were transformed to the treatment planing system
`(TPS).
`
`The regions of interest (ROI) were split in two groups: the targets and the
`organs at risk (OAR). The primary tumor GT Vprim, the positive lymph nodes
`GT Vnodes and a volume for elective irradiation of lymph nodes were contoured
`by a physician as targets. CT Vprim=nodes were deflned at 10 mm around the
`GT Vprim=nodes. Organs at risks were the parotid glands and the spinal cord.
`
`Five patients from our group of seven were already investigated in another
`study (Van Vulpen et al., 2005). Two of them had positive lymph nodes on
`both sides and three on only one side. Two patients without positive lymph
`nodes were added to the investigation to obtain a complete range of indica-
`tions.
`
`2.2.2 Plan criteria
`
`According to the protocol used clinically at our department, a dose of 66 Gy
`extending 5 mm around the
`was prescribed to the P T V66(prim=nodes),
`CT Vprim=nodes. The dose was boosted further to 69 Gy in the P T V69(prim=nodes),
`extending 5 mm around the GT Vprim=nodes. For the elective lymph nodes a dose
`of 54 Gy was prescribed to the P T V54(elec: nodes), extending 5 mm around the
`CT Velec: nodes. Figure 2.1 shows the OARs, targets and the prescription to the
`targets for one patient.
`
`In our investigation the mean dose to the parotid glands (Eisbruch et al., 1999;
`Roesink et al., 2001) was used as the quantitative measure for sparing of OARs
`and minimized as much as possible. In addition, the maximum volume of the
`spinal cord which received more than 46 Gy had to be smaller than 2 cc and
`care was taken to avoid ’hot spots’ in the body. Because the P T V54(elec: nodes)
`and the parotid glands were overlapping, a trade-ofi between target coverage
`and parotid gland sparing had to be made. In order to quantitatively compare
`plans, hard constraints were imposed on the target coverage: 99% of the target
`
`ViewRay Exhibit 1022
`Page 23 of 121
`
`

`

`12 Chapter 2. In(cid:176)uence of the Linac design on IMRT of head-and-neck plans
`
`Figure 2.1: The ROIs for one patient. The prescribed dose to the primary
`tumor and positive lymph nodes in the boost region was 69 Gy, and 66 Gy
`elsewhere. For the elective lymph nodes the prescribed dose was 54 Gy. Note
`the overlap of the targets with parotid glands.
`
`volume had to receive at least 95% of the prescribed dose. The volume that
`received more than 107% of the prescribed dose should be smaller than 2 cc.
`
`With this procedure we had a well deflned target coverage for all plans, so that
`the mean dose to the parotid glands could be used as a quantitative measure
`for comparing the plans.
`
`2.2.3 Inverse treatment planning
`
`The planning was performed on two difierent TPSs, Pinnacle3 v7.6c (Philips
`Medical Systems, Best, The Netherlands) and PLATO RTS v2.6.4 with ITP
`v1.1.8 (Nucletron BV, Veenendaal, The Netherlands). The purpose of using
`two TPSs was to reduce the dependency of this planning study on one partic-
`ular TPS. The modelling of the beam and the multi-leaf collimator is imple-
`mented difierently in Pinnacle3 and in PLATO. The optimization engine in
`the IMRT modules is difierent as well.
`
`For both TPSs, all calculations were performed with an identical beam set-up.
`
`ViewRay Exhibit 1022
`Pa