Monaco

Monaco Static MLC Sequencer Technical Reference

Document ID: LRMMON0004

Language: English
 Monaco Static MLC Sequencer Technical Reference

Monaco Static MLC Sequencer

Technical Reference
Version 2.0

Copyright 2014
All Rights Reserved
IMPAC Medical Systems, Inc.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 2 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Copyright statement
© 2014 IMPAC Medical Systems, Inc. All rights reserved. Do not make printed or electronic copies of this
document, or parts of it, without written authority from IMPAC Medical Systems, Inc.
The information contained in this document is for the sole use of IMPAC Medical Systems, Inc. personnel,
authorized users of the Equipment, and Licensees of IMPAC Medical Systems, Inc. and for no other purpose.

Use of trademarks and trade names statement

The Elekta® trademarks, service marks, logos and trade names that we use in this document are the registered and
unregistered trademarks and trade names of Elekta AB (publ.), its affiliates or a third party that has licensed its
trademarks and trade names to Elekta AB (publ.) or its affiliates. Do not make copies, show, or use trademarks or
trade names without written authority from IMPAC Medical Systems, Inc. an affiliate of Elekta AB (publ.).

Acknowledgement of other trademarks

Elekta acknowledges the registered trademarks and trade names of other manufacturer that we use in this document.

Referenced documents
Elekta does not supply all documents that we refer to in this document with the equipment. Elekta reserves the right
to make the decision on which of the documents it supplies with the equipment.

Contact information
GLOBAL ELEKTA SOFTWARE SUPPORT

Online Support: www.elekta.com/supportplus

Contact information (phone, email, WebEx): www.elekta.com/software-support
Training Calendar: www.elekta.com/training

ELEKTA REGIONAL OFFICES

13723 Riverport Drive, Suite 100 Sunnyvale, California Europe, Latin America, Africa and Middle East
Maryland Heights, MO 63043 100 Mathilda Place, Fifth Floor Contact Information
Sales Sunnyvale, CA 94086 Elekta, Inc.
Toll Free: +1 800 878 4276 Phone: +1 408 830 8000 Linac House
Fax: +1 314 993 0075 Fax: +1 408 830 8003 Fleming Way, Crawley, West Sussex RH10 9RR
Email: [email protected] Email: [email protected] Phone: +44 1293 544 422
Customer Support Fax: +44 1293 654 321
Phone: +1 800 878 4CMS (4276)
Email: [email protected]

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 3 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

Table of Contents
Introduction .............................................................................................................................................................. 7
Overview of the Algorithm ......................................................................................................................................... 8
1. Profile Filtering .................................................................................................................................................. 8
2. Definition of Jaw Frames ................................................................................................................................... 9
3. Profile Decomposition ..................................................................................................................................... 10
4. Profile Clustering ............................................................................................................................................. 10
5. Shape Extraction .............................................................................................................................................. 11
6. Segmentation .................................................................................................................................................. 13
7. Recursion of Segment Extraction ..................................................................................................................... 13
8. Recursion of Jaw Frame Processing ................................................................................................................. 13
9. Segment Refinement ....................................................................................................................................... 13
10. Segment Sorting ............................................................................................................................................. 16
11. Post-Processing of Open Leaf Gaps ................................................................................................................ 16
Handling of MLC types requiring special treatments, Input-Output filters ......................................................... 20

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 4 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Table of Figures
Figure 1: Three fluence weight profiles. Left: original; Center: segmentated; Right: difference between
original and segmentated ................................................................................................................................. 7
Figure 2: Profile filtering operations performed as a pre-processing step......................................................... 8
Figure 3: Additional profile filtering operations performed as a pre-processing step. .................................... 9
Figure 4: Profile decomposition into peaks that require delivery in separate segments. ............................. 10
Figure 5: Clustering translates the previously smooth fluence profile into a piecewise constant one. ....... 11
Figure 6: Shape Extraction....................................................................................................................................... 12
Figure 7: Shape Extraction with various minimum MU per Segment ............................................................... 12
Figure 8: Translation of a field shape into deliverable segments. .................................................................... 13
Figure 9: Mechanisms of segment shape optimization – Minimum area difference – MU merge............... 14
Figure 10: Mechanisms of segment shape refinement – Shape merge. ........................................................... 14
Figure 11: Mechanisms of segment shape refinement – Edge swap................................................................. 15
Figure 12: Mechanisms of segment shape refinement – Edge fuse&split. ...................................................... 15
Figure 13: Mechanisms of segment shape refinement – Incremental Leaf Position Changes. ..................... 16
Figure 14: Option ParkLeafGaps for the Elekta BeamModulator MLC. Center: full fluence profile. Left: one
segment with the leaf gaps parked under the jaws (shaded areas). Right: same segment with gaps
parked along the crest of the fluence profile. .............................................................................................. 17
Figure 15: Schematic correspondence between the number of segments and plan quality for a variation
of the min MU per segment parameter. Typical features are: 1. The number of segments needs to
exceed a certain patient-dependent threshold to achieve a sufficient plan quality; 2. The number of
segments depends sensitively on the min MU per segment parameter. Plots of this type are known as
“Pareto fronts”: there is no treatment plan with an equivalent number of segments that has a higher
plan quality than the curve............................................................................................................................. 18
Figure 16: Schematic correspondence between the number of segments and plan quality for a variation
of the min area per segment parameter. The dependency is weaker as for minimum MU per
segment, but the behavior is similar. Notice that the critical value for the minimum area is highly
patient specific, especially when tight dose gradients need to be drawn precisely around high risk
organs. ................................................................................................................................................................ 19

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 5 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 6 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Introduction
The Monaco Static MLC Sequencer (StatSeq) translates fluence weight profiles into deliverable, static
MLC field shapes. The translation aims to minimize the difference between the original fluence weight
profile and the resulting profile. It also maintains a number of constraints on the field shapes and all MLC
design constraints. Further, the process makes any effort to produce the minimum number of apertures
with shapes as regular as possible. This facilitates accurate dose computation and rapid delivery due to
minimized leaf travel.
You must notice the difference between a fluence distribution and a fluence weight profile, which is the
input to the sequencer. The former describes the number of photons passing through any given point of
the radiation field. The latter describes the aperture time, commonly expressed in terms of Monitor Units
(MU). The dose computation algorithms perform the translation from aperture time to fluence.
Therefore, it is intuitive to think of a fluence weight profile as a distribution of exposure times.
The StatSeq does not explicitly recognize the actual MLC type; all relevant constraints and parameters for
the sequencing action are provided implicitly in a MlcGeometry data container.

Figure 1: Three fluence weight profiles. Left: original; Center: segmentated; Right: difference between
original and segmentated

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 7 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

Overview of the Algorithm

The sequencing algorithm employs a multi-stage process. Here, the sequence of this process is described.
Each stage is described in detail below.

1. Profile Filtering
The fluence weight profile as obtained from the dose optimizer passes through a filter that removes
features of the profile that cannot be delivered with the given MLC. These features include:
• Fluence elements of very low weight.
• Fluence peaks that are narrower than the minimum leaf gap or segment width.
• Fluence depressions that are narrower than the minimum segment width.
• Fluence islands that are smaller than the minimum segment area.

Fill&Skim

Leaf fill Leaf

fill
direction direction
skim

Fills fluence dips narrower than a specified width (e.g. 5 mm)

Removes fluence peaks narrower than a specified width (e.g. 5 mm)

Figure 2: Profile filtering operations performed as a pre-processing step

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 8 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Truncate

truncate

Truncate removes fluence weights that cannot be delivered, because

they are below min MU per segment (dotted line).

Figure 3: Additional profile filtering operations performed as a pre-processing step.

2. Definition of Jaw Frames

In case the collimator jaws cannot move during the delivery of the MLC segments in one sequence, the
fluence weight profile may have to be split into sub-profiles. These sub-profiles can be delivered in one
sequence of segments with constant jaw positions (carriage group). Whether a fluence weight profile
needs to be split depends on both:
• A number of MLC constraints and the shape of the maximum aperture of the fluence weight
profiles
• The particulars of its modulation
These constraints include the maximum projection of a leaf relative to its neighbors and relative to the
Parallel Jaw (the jaws that travel parallel to the leaf direction) of the same side and the minimum leaf
separation. The actual need to extend one leaf far from its bank arises when the leaf has to perform
shielding towards the opposing field edge because of a narrow fluence weight peak close to the field edge.
If splitting is necessary, initially two jaw frames (a set of all 4 jaw positions) with maximum separation of
the Parallel Jaws are generated. If their overlap is smaller than a specifiable parameter (carriageOverlap,
currently defaulted to 15% of the maximum width of a jaw frame), a third frame centered on the midline
of the profile is created. Thereby, a maximum of three jaw frames is created, at the expense of not being
able to cover the maximum field width if the jaw frame width is not large enough (Notice, that the
maximum covered field width is 2.7 x maximum jaw frame width with the above default for
carriageOverlap).
Notice, that during a later post-processing the final jaw positions of a group of segments can be altered if
this improves shielding. Notice further, that automatic profile splitting can be deactivated by another

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 9 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

sequencer flag. singleFrameWidth, which overwrites the MLC constraint-derived jaw frame width, i.e., the
sequencer acts as if the jaw frame width can be wider than normally possible. This can reduce the
capability to perform fluence modulation at the field edges.

3. Profile Decomposition
Monaco then decomposes, or splits, the profile into a series of individual peaks. The split lines are put
along the fluence valleys, where leaves have to be placed to produce the gradients. This ensures that
ultimately segments are created in groups that have similar shapes. Since it is more difficult to post-
process the segment shapes to ensure “organic” shapes, the profile is segmented piece-by–piece. These
pieces are generated such as to have a regular shape. This process will also result in minimized leaf travel.
The decomposition uses a modified, anisotropic watershed algorithm that distinguishes between fluence
valleys in leaf travel direction (that require leaves being placed at the slopes on either side of the valley to
produce the gradient) and valleys perpendicular to the leaf travel direction (which require jaws be placed
or leaf pairs be closed).

Decomposition

The decomposition aims to identify fluence peaks that need to be

delivered in separate shots. If the peaks are segmented separately,
the segment shapes are more regular and intuitive, and gratuitous
leaf travel is avoided.

Figure 4: Profile decomposition into peaks that require delivery in separate segments.

4. Profile Clustering
Each sub-profile initially has continuous fluence weight values, which are clustered into a small number of
discrete values in this step. The clustering operation transforms the continuous values of the smooth
profile into a small set of discrete values. Only the superposition of a few static MLC segments can deliver
these discrete values. This results in a stair-stepped fluence weight profile (see Figure 1 center). The
clustering algorithm establishes these discrete fluence steps to:
• Fulfill the minimum MU per segment parameter
• Minimize the deviation between the initial profile and the clustered profile.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 10 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Clustering

Min MU per Segment

The continuous fluence values are translated into discrete steps.

This operation considers the minimum MU per segment, which equals
the minimum discrete increment between two fluence levels. Further,
the neighbouring leaf strips also need to be clustered to the same
fluence levels!
MLC constraints are not considered at this stage, other than min MU
per segment, and the direction of leaf movement.

Figure 5: Clustering translates the previously smooth fluence profile into a piecewise constant one.

5. Shape Extraction
The clustered sub-profiles are processed by a shape extraction search. This algorithm latches onto the
discrete fluence values. It then decomposes the clustered profile into a set of field shapes, which are
basically constant profiles. The superposition of these field shapes with their appropriate MU weights
equals the stair-stepped input profile. At the end of this stage, the initial profile has been decomposed into
a set of stacked field shapes, which do not yet obey MLC delivery constraints. The difference between the
initial, filtered profile and the stacked shapes is a difference profile, which can have positive and negative
entries (see Figure 1 right). The modulus of each of these is normally smaller than 0.5 * minimum MU per
segment, i.e. the maximum translation error is half the MU per segment. Larger translation errors can
occur due to the initial filtering of the input fluence profile.
Up to this stage, the MLC design constraints have not been taken into account, i.e. there is no guarantee
that the field shapes are deliverable at all, or that a field shape can be delivered with a single MLC
segment.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 11 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

Shape Extraction

The shape extraction algorithm tries to associate upward steps and

downward steps of the same height. The movement of left and
right leaves need not be uniformly in one direction. In the above example,
the topmost shape on the right produces a right gradient which originates
at a lower fluence weight level.
The three peaks would be delivered in 11 segment shapes.
Some of these shapes are very similar. This is not considered here, but
at the level of MLC shapes, where the minimum area difference constraint
is enforced.

Figure 6: Shape Extraction

Shape Extraction with various minimum MU per

Segment
11 Shapes

Min MU per Segment

The number of fluence levels and shapes is efficiently controlled

with the minimum MU per segment parameter.

5 Shapes

Min MU per Segment

Figure 7: Shape Extraction with various minimum MU per Segment

Figures 6 and 7 show the translation of clustered profiles into individual field shapes whose superposition
produces a modulated beam. The difference between the original profile (blue line) and the clustered
profile is the difference weight grid, which is subsequently used for segment shape refinement.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 12 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

6. Segmentation
Each field shape is translated into a static MLC segment. If necessary, it will be translated into more than
one. The segments are subjected to the minimum area constraint and put into a repository. At this stage,
MLC constraints are finally considered.

Shape translation into MLC segments

The green segment

The field shape is smaller than the
as extracted from minimum field area
the profile and is deleted
(thick black line)
is decomposed
into three
MLC segments
red, green, blue

leaf

Figure 8: Translation of a field shape into deliverable segments.

7. Recursion of Segment Extraction

After all field shapes have been segmented, the total fluence profile resulting from all the segments in the
repository is computed by superposition. The total fluence is subtracted from the input profile to yield a
new difference profile. The latter is filtered (step 1) and fed into step 4 for clustering. Some more segments
may be created in step 5 and 6. This process repeats until no more segments can be extracted.

8. Recursion of Jaw Frame Processing

If multiple jaw frames need to be processed, extract all segments in one frame and advance to the next
frame. The system filters and feeds the residual profile into stage 3.

9. Segment Refinement
At this stage, there is a repository of valid MLC segments and a final difference profile. Segment
refinement aims to alter the segment shapes and combine segments in order to reduce:
• The irregularity of their shapes
• Their number, or

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 13 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

• The maximum error of the translation difference.

This is where the system applies the minimum area difference constraint (this leads to MU merging, i.e.
segment superposition). Non-overlapping segments can also be merged into one segment with combined
shape or multiple apertures per segment (shape merging). Leaf positions can be swapped between
segments if this reduces the irregularity of the segment outline. The difference grid is used to initiate
incremental changes of the leaf positions.

Figure 9: Mechanisms of segment shape optimization – Minimum area difference – MU merge.

Figure 10: Mechanisms of segment shape refinement – Shape merge.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 14 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Figure 11: Mechanisms of segment shape refinement – Edge swap.

Figure 12: Mechanisms of segment shape refinement – Edge fuse&split.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 15 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

Figure 13: Mechanisms of segment shape refinement – Incremental Leaf Position Changes.

10. Segment Sorting

The system sorts segments into an order that reduces the maximum leaf travel between them. Sometimes,
a leaf that does not actively participate in field shaping can cause the maximum leaf travel between two
segments. This leaf may be closed or hidden by a collimator jaw in one of the segments. When the system
sorts segments, it ignores closed leaf strips if you do not enable the Park Leaf Gaps field.

11. Post-Processing of Open Leaf Gaps

Normally, the sequencer parks leaves in order to hide their leaf gap under a Parallel Jaw or a Transverse
Jaw. The only exception to this rule is when you set the Minimum Leaf Separation parameter to 0. This
signifies minimum transmission through the perfectly-closed leaves of some MLC types. However, parked
leaves may incur large amounts of leaf travel, especially when the distance to the opposing Parallel Jaw is
large.
If you do not enable the Park Leaf Gaps field, the leaves are not parked so that a Parallel Jaw shields the
opening between two leaves. This sets the leaf gap to a position where it ideally sits in a region of high
fluence. The leaf gaps of multiple segments are not placed over the same position so as to avoid hot
streaks. For this reason, the gaps can only be placed after the entire segment sequence has been created.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 16 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Figure 14: Option ParkLeafGaps for the Elekta BeamModulator MLC. Center: full fluence profile. Left:
one segment with the leaf gaps parked under the jaws (shaded areas). Right: same segment with gaps
parked along the crest of the fluence profile.

This concludes the initial sequencing algorithm.

Segment shape optimization follows this sequence:
1. Application of segment constraints: if some segments have dropped below the minimum MU per
Segment threshold in the dose optimization, they are deleted. Similarly for the minimum segment
area, should the segments have been reduced in size by segment shape optimization.
2. Segment shape optimization as described in the separate SSO document.
3. Segment sorting.

Sequencing Parameters
Apart from MLC design constraints, the Static Sequencer accepts the following parameters:
• Minimum Number of MU per Segment: This parameter controls the coarseness of the
translation of the initial fluence profile into segments. The sequencer will generate a new field
shape only if the depth of the fluence weight modulation exceeds this parameter. So, if this
parameter is large, fewer segments with greater differences in shape will be produced. If it is small,
more segments with potentially smaller differences in shape (between consecutive segments) will
be produced (see Figures 6 and 7). This is the most potent parameter to control the number of
segments obtained from a given fluence weight profile.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 17 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

Figure 15: Schematic correspondence between the number of segments and plan quality for a variation
of the min MU per segment parameter. Typical features are: 1. The number of segments needs to exceed
a certain patient-dependent threshold to achieve a sufficient plan quality; 2. The number of segments
depends sensitively on the min MU per segment parameter. Plots of this type are known as “Pareto
fronts”: there is no treatment plan with an equivalent number of segments that has a higher plan quality
than the curve.

• Minimum Area per Segment: This parameter guarantees that the area of each segment is greater
than this value in cm2. Apart from having an impact on dose computation, this also reduces the
number of segments by removing the smallest ones. It can be detrimental to the overall quality of
the dose distribution to set this parameter too large. The dependency between segment number
and this parameter is not as strong as with minimum MU per segment.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 18 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

Figure 16: Schematic correspondence between the number of segments and plan quality for a
variation of the min area per segment parameter. The dependency is weaker as for minimum
MU per segment, but the behavior is similar. Notice that the critical value for the minimum area
is highly patient specific, especially when tight dose gradients need to be drawn precisely around
high risk organs.

• Maximum Number of Segments Per Plan: This parameter controls the total number of
segments, summed up over all beams. It therefore needs to be applied in an external loop over all
beams, which is located in the treatment plan (TP) object. This loop calls the sequencing multiple
times with increasing min MU per Segment parameters until the desired number of segments is
reached. The loop starts with the used provided parameter of min MU per Segment, and increases
it by 0.5 MU at every iteration.
• Minimum Segment Width: This parameter controls the minimum width of the segments, such
as the minimum separation of a leaf pair. You can lessen the effect of bad MLC calibration, leaf
positioning accuracy or dose computation model with this parameter. It does usually not
influence the number of segments. However, it may have a detrimental effect on overall plan
quality if you set it too large. Notice that the minimum leaf separation is also controlled by a MLC
design constraint.
• Minimum Overlap between Carriage Groups: If profile splitting becomes necessary, the carriage
groups need to have a minimum overlap of this parameter times the maximum carriage group
width (see below for more details). Defaults to 15%.
• Single Jaw Frame Width for Sequencing: Turns off the profile splitting for field widths up to this
parameter. Defaults to 0.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 19 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
 Monaco Static MLC Sequencer Technical Reference

Handling of MLC types requiring special treatments, Input-Output filters

The static sequencer is by design universal and governed solely by the MLC constraint set defined in
MlcGeometry. However, some MLC types require special handling of leaves not participating in field
shaping. Also, the DICOM interface specification may allow leaf position prescriptions that violate some
MLC constraints. In this case, segments need to be passed through input filters on plan loading before
they comply with all MLC constraints. The same applies when segments are exported.
The special requirements of some MLC types are dealt with in a post-processing step, when MlcNotePads
produced by the sequencer are converted into segments. The function doSpecials() is sensitive to the MLC
model by means of a tag, which is provided with the MlcGeometry. MLC types requiring special
treatment are currently:

MLC Types Description

Elekta MLCi/MLCi2 Parking of the guard leaves. The first set of leaves shielded by a Transverse Jaw
is set to the same position as the last exposed pair.
Elekta APEX and Application of an interlacing leaf parking position for parked leaves (leaf gaps
miniMLCs cannot be hidden and have some residual leakage when closed). Leaves are set to
a left-right-left parking position if interdigitation and overtravel constraints
allow.
Elekta Agility 1. Leaves parked under the Transverse Jaw need to be set to special parking
positions forced by the T-shaped jaw design. For details, refer to Elekta
documentation.
2. Exposed leaf gaps need to be set to an off-axis dependent minimum width.
The details are explained below. Leaves are moved to off-axis parking
positions if the functionality is enabled.
3. Parking of guard leaves. The first TWO sets of leaves shielded by a
Transversal Jaw are set to the same position as the last exposed pair.

The only special requirement affecting exposed leaves is point 2 of the Elekta Agility MLC. The
requirement is that the effective minimum leaf gap prescription (i.e. leaf position projected to isocentre
plane) varies with off-axis position. The algorithm employed here is as follows:
1. The sequencer uses the overall minimum leaf gap as constraint for all leaves at all positions. This
parameter must be greater than 0, e.g. 0.1 mm. Therefore, leaf positions are created whose gap is
too narrow if it is in a central location of the field.
2. The required minimum leaf gap is computed by applying a parabolic formula.
3. The leaf positions are then opened up symmetrically around the centre of the leaf gap.
Special input/output filters are also provided. Internal segments always comply with all constraints. These
functions are currently provided for the following MLC types:

MLC Types Description

Elekta BeamModulator This MLC allows maximum leaf positions to be +-105 mm, but if a leaf pair is
prescribed as “closed”, i.e. both leaves at the same position, the real parking

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 20 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
Monaco Static MLC Sequencer Technical Reference

position is retracted further, behind a fixed backup jaw.

Monaco detects “closed” leaf prescriptions and moves the leaves to the actual
parking position, currently at +- 113 mm.
Monaco converts the positions of “closed“ leaves to +- 105 mm.
Elekta APEX This MLC allows the prescription of fully closed leaf gaps while the actual
minimum separation is greater than zero, currently 0.1 mm.
Monaco detects “closed” leaves and opens their gaps to the minimum
Monaco detects leaves set to minimum gap position and converts them to
“closed”
Elekta Agility This MLC has a mobile leaf guide whose position is not prescribed, but that
influences attainable leaf positions. In fact, it can be thought of a very high
transmission Parallel Jaw. For example, the leaves must not project further
than 20 cm from this “jaw”.
Monaco computes the position of the leaf guide and sets it as Parallel Jaw
position.
Siemens 27, 41 These MLCs disallow leaves that are at their minimum gap position next to a
Transversal Jaw. Further, leaves may be prescribed as “closed”, i.e. with a zero
leaf gap.
Monaco opens “closed” leaves to their minimum gap position.
Monaco pulls the Transversal Jaws in if they need to shield a set of leaves with
minimum gap opening.
Varian, all types These MLCs allow the prescription of zero leaf gap on “closed” leaves.
Monaco creates these zero leaf gaps for closed leaf pairs.

LRMMON0004 / 2.0 Monaco

2014-06-06 Monaco Static MLC Sequencer Technical Reference
Page 21 ©2014 IMPAC Medical Systems, Inc. All rights reserved.
13723 Riverport Drive, Suite 100, Maryland Heights, MO 63043
 European Union Authorized Representative
Manufacturer
Elekta Business Area Software Systems
Elekta Limited
IMPAC Medical Systems, Inc.
Linac House, Fleming Way
13723 Riverport Drive
Crawley, West Sussex
Suite 100
RH10 9RR,
Maryland Heights, MO 63043
United Kingdom
USA
www.elekta.com Human Care Makes the Future Possible
Corporate Head Office: Regional Sales, Marketing and
Service:

Elekta AB (publ) North America Europe, Latin America, Asia Pacific

Box 7593, SE-103 93 Stockholm, Atlanta, USA Africa, Middle East & India Hong Kong, China
Sweden Tel +1 770 300 9725 Tel +44 1293 544 422 Tel: +852 2891 2208
Tel +46 8 587 254 00 Fax +1 770 448 6338 Fax +44 1293 654 321 Fax: +852 2575 7133
Fax +46 8 587 255 00 [email protected] [email protected] [email protected]
[email protected]