Rounded MLC Offset Table

[Eoin Toomey]

Hi all,

Just as a follow-on from my recent post about having consistently high
planned chamber doses;

I mentioned in my last post that the problem was resolved, with some
of your help, by adjusting the MLC transmission, the additional
interleaf leakage, and the Rounded MLC offset table. I have received a
few emails requesting this "new" offset table. Just so you are aware,
I believe the origins of this table lies in a few centres in the UK,
as users found that the default table was not a particularly good
representation of their machine. Philips then provided a more suitable
theoretical table (see below). I may be completely wrong, but that's
the story I heard! This will probably not suit every machine, but it
certainly suited our Elekta machine better. Enjoy! Eoin


-12.5 -0.15
-12 -0.12
-11 -0.07
-10 -0.02
-9 0.02
-8 0.06
-7 0.09
-6 0.12
-5 0.14
-4 0.16
-3 0.18
-2 0.19
-1 0.20
0 0.20
1 0.20
2 0.19
3 0.18
4 0.16
5 0.14
6 0.12
7 0.09
8 0.06
9 0.02
10 -0.02
11 -0.07
12 -0.12
13 -0.18
14 -0.24
15 -0.30
16 -0.37
17 -0.44
18 -0.52
19 -0.60
20 -0.69

[Randall Miller]

Hi Eoin:

 

Actually anyone can get a copy of the table form Philips tech support. There is some additional info with the table as well

Thanks

Randall

>>> Eoin Toomey <eoint...@gmail.com> 9/9/2011 10:13 AM >>>

--
You received this message because you are subscribed to the Google
Groups "pinnacle3-users" group.
To post to this group, send email to pinnacl...@googlegroups.com
To subscribe to this group, send email to
pinnacle3-us...@googlegroups.com
To unsubscribe from this group, send email to
pinnacle3-use...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/pinnacle3-users?hl=en

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- The HIPAA Privacy Rule requires covered entities to safeguard certain Protected Health Information (PHI) related to a person's healthcare. Information being sent to you may include PHI, after appropriate consent, acknowledgment, or authorization from the patient or under circumstances that do not require patient authorization. You, the recipient, are obligated to maintain PHI in a safe and secure manner. You may not re-disclose this patient information without additional patient consent or as required by law. Unauthorized re-disclosure or failure to safeguard PHI could subject us, or you, to penalties described in federal (HIPAA) and state law. If you, the reader of this message, are not the intended recipient, or the employee or agent responsible to deliver it to the intended recipient, please notify us immediately and destroy the related message. Thanks for your help. ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------   ­­  

[Robert Matthews]

Hello all,

I have a question concerning the employment of the adjusted mlc offset table. Varian provides an offset table with their machines and I have not seen a change in this table since the beginning, although the firmware running the mlcs has changed considerably, e.g.,kissing algorithms that fine tune the table on a daily bases and other adaptations that make leaf collisions and burnt out motors much less of a problem. It's usual practice to take the numbers from this offset table for use in the Philip's Pinnacle RTPS for mlc commissioning, which I've always suspected as leading to wide burn gaps in the picket fence test. There ought not be burn gaps in the picket fence if the mlc offset table places the 50% isodose at the geometric field edge correctly. So, my questions are: is the table only adjusted for the Philips side or is it also adjusted on the linac side? If it is only on the RTPS side, what do you see on the picket fence which may be run independent of the RTPS? Should the tables be the same?

Bob

Robert J. Matthews. Ph.D., DABR
Kootenai Medical Center
CdA 208-666-2529
PF 208-619-4149

[Feygelman, Vladimir]

Pinnacle since v 7.something does not rely on the 50% intensity for
calculations but rather explicitly models the leaf end. Hence it does
not matter what the picket fence looks like, as long as the MLC leaves
follow the table accurately. The picket fence is burned because the
table reflects the optical calibration. Radiation penetrates more than
light.

Vladimir Feygelman, PhD
Physicist
tel: (813) 745 8424 | fax: (813) 745-7231 | email:
vladimir....@moffitt.org

Hello all,

Bob

This transmission may be confidential or protected from disclosure and is only for review and use by the intended recipient. Access by anyone else is unauthorized. Any unauthorized reader is hereby notified that any review, use, dissemination, disclosure or copying of this information, or any act or omission taken in reliance on it, is prohibited and may be unlawful. If you received this transmission in error, please notify the sender immediately. Thank you.

[Randall Miller]

Hi Vladimir:

 

I agree with you.  I would add that sometimes you do have to have the Elekta Service rep confirm the optical alignment of the MLC. I had numerous issues at first with my Synergy MLC.  One issue was one bank was not aligned properly.  I think we were almost 1 mm off. Once, the MLC was realigned, I was able to make corrections to my model.

The table values will work well but the optical setup is critical in the Elekta world.

 

best

 

Randall

 
>>> "Feygelman, Vladimir" <Vladimir....@moffitt.org> 9/9/2011 11:22 AM >>>

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- The HIPAA Privacy Rule requires covered entities to safeguard certain Protected Health Information (PHI) related to a person's healthcare. Information being sent to you may include PHI, after appropriate consent, acknowledgment, or authorization from the patient or under circumstances that do not require patient authorization. You, the recipient, are obligated to maintain PHI in a safe and secure manner. You may not re-disclose this patient information without additional patient consent or as required by law. Unauthorized re-disclosure or failure to safeguard PHI could subject us, or you, to penalties described in federal (HIPAA) and state law. If you, the reader of this message, are not the intended recipient, or the employee or agent responsible to deliver it to the intended recipient, please notify us immediately and destroy the related message. Thanks for your help. ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------   ­­  

[Robert Matthews]

Dear All,

I'm providing a synopsis of the responses I've received concerning the rounded mlc leaf end problem and corrections that may be applied, at least for the Elekta/Pinnacle system. Thanks to Dr Feygelman for taking the time to respond, although I tend to disagree with his assessment of the picket fence. My understanding is more in line with what Dr Mohr posted in Feb, 2011, which is attached below. I believe Eoin Toomey has a grad student looking at this issue, although I'm not sure which system; I am concerned with the Varian/Pinnacle combination. It is a big problem that leads to large errors percentage wise for small mlc fields and is directly related to the picket fence results. I think that most Pinnacle/Varian users are restricting the size of mlc fields to not less than 2 cm^2 to avoid the size of the error. Randall Miller provided data similar to what Dr Mohr suggested. Miller's correction data is a rather flat parabolic that is positive at isocenter and goes negative at large distances from isocenter. Mohr's data, as I recall, was positive at isocenter and negative at large distances but recovered to resemble something like a handle-bar mustache. The Varian mlc offset table is zero at isocenter and is sharply parabolic to the negative by comparison. I think it should look more like what Mohr/Miller provided. What I don't understand is the interplay between the offset table in Pinnacle vs the offset table on the Varian linac. If someone has looked at this, I would appreciate them sharing their findings. TIA

Bob

Robert J. Matthews, Ph.D., DABR

Kootenai Medical Center
CdA 208-666-2529
PF 208-619-4149

I posted the questions:

So, my questions are: is the table only adjusted for the Philips side or is
it also adjusted on the linac side? If it is only on the RTPS side, what
do you see on the picket fence which may be run independent of the RTPS?
Should the tables be the same?

Vladimir Feygelman <Vladimir....@moffitt.org> response

Pinnacle since v 7.something does not rely on the 50% intensity for
calculations but rather explicitly models the leaf end. Hence it does
not matter what the picket fence looks like, as long as the MLC leaves
follow the table accurately. The picket fence is burned because the
table reflects the optical calibration. Radiation penetrates more than
light.

Dr. Peter Mohr wrote Feb 2011:

Dear Pinnacle/Elekta users,

Let me comment on the picket fence / rounded leaf problem:
1) The usual AutoCal MLC calibration is a DOSIMETRIC calibration, i.e. the picket fence test should not
show 1mm stripes (neither overdose nor underdose).
2) The pinnacle leaf position is defined as GEOMETRIC position.
Obviously, these definitions are different! The rounded leaf offsets in pinnacle are used to adjust the
geometric definition to the dosimetric calibration. I have plotted (approximately to scale!) the geometry
at different leaf positions: 200 mm (l200.pdf), 125 mm (l125.pdf), ... -125 mm (overtravel, lm125.pdf).
There you can easily see what the offset in pinnacle has to do:
A) At positions close to 0: the leaf tip is too thin to provide a 50% isodose. You have to move the leaf
in (0.5 - 1mm) to find the calculated 50% isodose at position 0.
B) At increasing leaf positions from a few mm up to about 125 mm: the leaf cuts more and more from
the edge of the beam: you have to move the leaf out (approximately increasing with the square of the
position)
C) At further increasing leaf positions beyond 125 mm: the rounded leaf offsets further increase, but
because of the finite leaf height the slope of the rounded leaf does not further increase quadratically.
This leads to my suggestions for the "Leaf Offset Calibration" in Pinnacle (see attached "rlo_pinn.pdf",
red line). The blue line was recommended by our Philips consultant in 2005; in 2007 this suggestion
was slightly modified (green line). The new curve (red line) is a fit to the red data points; these points
have been calculated from the shift of the 50% isodose for an empty leaf offset table (offsets = 0 at all
positions). The results have been presented at the German Pinnacle Users Meeting in 2009.
Please contact me if you are interested in the numerical values of the red curve. From my point of
view, this curve should work for all Elekta LINACs with the Elekta standard 80-leaf MLC (as used in our
LINAC which was manufactured in 2005).
Please note (i) that the offsets are large (several mm) for large leaf positions, and (ii) the significant
change between the different curves at large positions.

Peter

Dr. Peter Mohr
Strahlentherapie
Diakonie-Klinikum Schwäbisch Hall gGmbH
Tel. 0791 753 4845
Fax 0791 753 4911
Peter...@dasdiak.de
www.dasdiak.de

[Robert Matthews]

Correction: Eoin Toomey provided the correctio table data. Bob

Dear All,

Bob

I posted the questions:

Vladimir Feygelman <Vladimir....@moffitt.org> response

Peter

[Anton Eagle]

Hi all,
I just had a follow-up question regarding this.

I admit that I am not sure about this... but I was under the
impression that there is no "offset table" on the linac or in the
actual MLC control system. I thought the offset table was only in the
treatment planning system and was an attempt to account for changes in
leaf-end thickness on the rounded leaves. But, I didn't think the MLC
system tried to adjust for this... I thought it just "is what it is"
so to speak. Am I wrong on this? Does the Varian MLC system tweak
leaf positions to account for the dosimetric leaf gap? If so, how
does it pull this off when opposing leaf ends are in contact?... since
the leaves would have to move closer together but would be unable to?

I just want to make sure that I'm not missing something or have a
misconception about how the MLC operates. Thanks.

-Anton Eagle

[Feygelman, Vladimir]

Varian MLC is physically offset form its calibration (by light) position to produce approximately equal penumbra throughout the range of MLC positions.

Vladimir Feygelman, PhD
Physicist
tel: (813) 745 8424 | fax: (813) 745-7231 | email: vladimir....@moffitt.org

-----Original Message-----
From: pinnacl...@googlegroups.com [mailto:pinnacl...@googlegroups.com] On Behalf Of Anton Eagle
Sent: Monday, September 12, 2011 1:22 PM
To: pinnacl...@googlegroups.com
Subject: Re: [p3rtp] Rounded MLC Offset Table

-Anton Eagle

[hugo.t...@ssss.gouv.qc.ca]

@Anton. The MLC offset table is located on the 4D workstation.

Hugo


"Feygelman, Vladimir" <Vladimir....@moffitt.org>

"Feygelman, Vladimir" <Vladimir....@moffitt.org> Envoyé par : pinnacl...@googlegroups.com 2011-09-12 13:25 Veuillez répondre à pinnacl...@googlegroups.com

A <pinnacl...@googlegroups.com> cc Objet RE: [p3rtp] Rounded MLC Offset Table

[Greg Gibbs]

This was touched on March 10, on this list by Vadim Kuperman.

Basically there is a file on the mlc workstation for older machines and on
the 4D console for newer machines named mlctable.txt. (HDTable.txt for high
def versions). They are in the directory c:\program
files\varian\oncology\mlc\controller\exec

On Truebeam the file is integrated into the software, but to my knowledge it
is either not accessible or imbedded in the code.

Greg Gibbs

Greg Gibbs MS
Colorado Associates in Medical Physics

[Curran, Bruce]

Although physically on the 4D workstation for Varian, it is actually a networked disk volume used by the MLC Controller.

 

Bruce

 

Bruce Curran

Associate Chief Medical Physicist

Radiation Oncology

Rhode Island Hospital        (401)444-6249

593 Eddy Street               (401)444-5335 (F)

Providence, RI  02903       (401) 265-1303 (C)

                  bcur...@lifespan.org

[Phil Vial]

Hi all,
the Varian/Pinnacle combination seems reasonably straight foward:
calibrate to leaf tip at or near central axis, TPS models radiation
penumbra at all positions, geometric off-axis correction
(mlctable.txt) applied at linac and therefore required at TPS also
(geometric off-axis offsets approximately equal to radiation off-axis
offsets). Plenty of literature supporting this. Some simple
restrictions to avoid negative leaf gaps/collisions etc. Not sure if
the Varian MLC leaf end model in Pinnacle is geometrically accurate
enough to simply apply Varian's offset table or is it an approximation
that might allow for some small tweaking for optimisation of the
offset table(?).

Its the Elekta/Pinnacle combination I find hard to conceptualise or
get a straight answer on. A simple question:
- Does Elekta implement an off-axis correction at the linac (analogous
to Varian's mlctable.txt)?

Can anyone put me straight on this seemingly simple question
(preferably with references)? I'm talking about MLCi if that makes any
difference.

I understand the calibration procedure is usually based on radiation
field edge defined at a couple of points (different to Varian) but
unless it is based on many points across the entire MLC range of
motion I do not understand how the MLC controller corrects for the
continuously varying (non-linear) geometric offset between mechanical
leaf (or reflector) position and projected radiation field edge as a
function of off-axis position. Interpolating between two or even four
points would not suffice if the offsets are similar to the varian ones
(and the Philips/Elekta tables provided below would indicate they at
least follow a similar trend). Perhaps the leaf end shape or optical
system somehow avoids or manages this problem??? I'm obviously missing
some fundamental piece of info. Worrying thing is that I get different
answers about this from everyone I've asked (most common answer is a
blank stare), including within Elekta.

Regards
Phil Vial
medical physicist
liverpool and macarthur cancer therapy centres
NSW, Australia
email: phili...@sswahs.nsw.gov.au

[Mohr, Dr. Peter]

Dear colleagues:

Let me add some further comments on the Elekta/Pinnacle systems (unfortunately without references):

First of all, I have plotted the new Philips recommendation (see attached). It is still the old parabola from 2005, but now shifted by 2 mm instead of 1 mm (as recommended in 2007)- this is not really convincing to me. So I continue to "advertise" my offset table (as posted in February 2011; further explanations see there).

I fully agree with Vladimir Feygelman: "radiation penetrates more than light" - but I disagree with his conclusions. Picket fence should look fine (and does look fine for our LINAC!) if you calibrate the MLC using a dosimetric procedure (like the widely used AutoCal). The offset table in Pinnacle is then used to match the geometrical leaf position in Pinnacle to the dosimetric leaf position in the real MLC. My recommended offset table was determined in this way; i.e. it is valid only for a dosimetrically calibrated Elekta MLC.

The situation becomes different if you calibrate your MLC by light (Vladimir, this is how I understand your sentence "The picket fence is burned because the table reflects the optical calibration"). Also in this case the offset table in Pinnacle must be used to match the real MLC; however, because the optically calibrated MLC is different from the dosimetrically calibrated MLC, you will need a different offset table for an optically calibrated MLC.

I think that there is no right or wrong (or better or worse) calibration of the MLC - but in any case the output of Pinnacle should match your real LINAC!

By the way, such an optical calibration of the MLC (using the light field instead of the radiation field) should not be mixed up with the optical readout and control of the Elekta MLC. The optical control reads the positions of the reflectors on each leaf (i.e. pixel number on a camera view), and the usual Elekta gain/offset calibration translates the pixel position of the reflector to the leaf position in mm. This should be a (more or less) linear procedure, and as far as I understand, there is no mlctable hidden in the Elekta system. In reply to Phil Vial, the proper dosimetric properties of the MLC result mainly from the leaf shape.

Best regards,
Peter

Dr. Peter Mohr
Strahlentherapie
Diakonie-Klinikum Schwäbisch Hall gGmbH
Tel. 0791 753 4845
Fax 0791 753 4911
Peter...@dasdiak.de
www.dasdiak.de

Diakonie-Klinikum Schwäbisch Hall gGmbH
Diakoniestraße 10, D-74523 Schwäbisch Hall
Geschäftsführer: Dr. Peter Haun (Sprecher), Rainer Münch
Registergericht: Amtsgericht Stuttgart, HRB 571831

Die Diakonie-Klinikum Schwäbisch Hall gGmbH mit den Klinikstandorten Schwäbisch Hall und Gaildorf ist eine hundertprozentige Tochter der Gesundheitsholding Schwäbisch Hall gGmbH.
Zur Gesundheitsholding Schwäbisch Hall gGmbH gehört neben der Diakonie-Klinikum Schwäbisch Hall gGmbH mit den Klinikstandorten Schwäbisch Hall und Gaildorf auch die Landkreis Schwäbisch Hall Klinikum gGmbH mit dem Klinikstandort Crailsheim.

Diese Information ist ausschließlich für den Adressaten bestimmt und kann vertraulich oder gesetzlich geschützte Informationen enthalten. Wenn Sie nicht der bestimmungsgemäße Adressat sind, unterrichten Sie bitte den Absender und vernichten Sie diese Mail. Anderen als dem bestimmungsgemäßen Adressaten ist es untersagt, diese E-Mail zu lesen, zu speichern, weiterzuleiten oder ihren Inhalt auf welche Weise auch immer zu verwenden. Wir verwenden aktuelle Virenschutzprogramme. Für Schäden, die dem Empfänger gleichwohl durch von uns zugesandte mit Viren befallene E-Mails entstehen, schließen wir jede Haftung aus.

[Eoin Toomey]

Hi Bob,

Just to let you know, I did indeed have a grad student working on the
rounded leaf end table for an Elekta Precise linac. Due to unforeseen
circumstances, the student had to leave quite early on in the project.
Ultimately, I decided to jump right in and use the alternative table
that I posted earlier, as opposed to coming up with a table via
measurements. As mentioned, it has yielded much better results than
the previous model. It does mean, though, that the project will be put
on the back-burner for a while.

Cheers,

Eoin



On Sep 12, 5:54 pm, Robert Matthews <lightningri...@imaxmail.net>
wrote:

> Dear All,
>
> I'm providing a synopsis of the responses I've received concerning the rounded mlc leaf end problem and corrections that may be applied, at least for the Elekta/Pinnacle system. Thanks to Dr Feygelman for taking the time to respond, although I tend to disagree with his assessment of the picket fence. My understanding is more in line with what Dr Mohr posted in Feb, 2011, which is attached below. I believe Eoin Toomey has a grad student looking at this issue, although I'm not sure which system; I am concerned with the Varian/Pinnacle combination. It is a big problem that leads to large errors percentage wise for small mlc fields and is directly related to the picket fence results. I think that most Pinnacle/Varian users are restricting the size of mlc fields to not less than 2 cm^2 to avoid the size of the error. Randall Miller provided data similar to what Dr Mohr suggested. Miller's correction data is a rather flat parabolic that is positive at isocenter and goes negative at large distances from isocenter. Mohr's data, as I recall, was positive at isocenter and negative at large distances but recovered to resemble something like a handle-bar mustache. The Varian mlc offset table is zero at isocenter and is sharply parabolic to the negative by comparison. I think it should look more like what Mohr/Miller provided. What I don't understand is the interplay between the offset table in Pinnacle vs the offset table on the Varian linac. If someone has looked at this, I would appreciate them sharing their findings. TIA
>
> Bob
>
> Robert J. Matthews, Ph.D., DABR
> Kootenai Medical Center
> CdA 208-666-2529
> PF  208-619-4149
>
> I posted the questions:
>
> So, my questions are: is the table only adjusted for the Philips side or is
> it also adjusted on the linac side? If it is only on the RTPS side, what
> do you see on the picket fence which may be run independent of the RTPS?
> Should the tables be the same?
>

> Vladimir Feygelman <Vladimir.Feygel...@moffitt.org> response

> Peter.M...@dasdiak.dewww.dasdiak.de

[Anton Eagle]

Hi Listers,

It sounds like much of the questions regarding Elekta MLCs have been
resolved in this thread.

I would like to ask you all to indulge me for a bit regarding how the
Varian MLC operates, as I am still a bit confused about the interplay
between the various offsets.

I have looked at the MLCTABLE.TXT file, and do indeed see that there
are what appear to be position offsets that vary according to absolute
leaf position. All the offsets seem to be in the more-extended
direction... that is the offset seem to be intended to push the leaf
out into the field a little bit, which is precisely what we would
expect if this were compensating for the dosimetric edge of the leaf
vs. the physical edge of the leaf.

However, I am puzzled by a couple things, and am appealing to the
collective wisdom of this email list to set me straight.

First, when I look at the actual leaf position that actually occurs
when I tell the leaf to go to a certain position... I don't see any
evidence that an offset is applied. For example, in my MLCTABLE.TXT
file, the entry for 15 cm is... 15.0000 : 14.8285. This would seem
to indicate that when I request a position of 15 cm, it is actually
going to 14.8285 cm, which would compensate somewhat for the rounded
leaf end. However, when I actually measure where the leaf ends up...
it ends up at exactly 15 cm. An offset of 1.7 mm should be readily
apparent, but simply isn't there. Is there a simple explanation for
this?

Second, I am puzzled by the interplay of linac applied offsets and TPS
offsets. If the linac were moving the leaves slightly into the field
to compensate for rounded leaf tips, and the treatment planning system
where adjusting the dosimetric field edge for the same reason, then it
seems to me that we would be double-correcting.

Consider the following example. Using a position of 15 cm again, lets
follow through the corrections. At the linac, an offset would push
the leaf into the field... say the above mentioned amount of 1.7 mm.
So, the leaf ends up at 14.83 cm which puts the dosimetric field edge
at 15 cm. The TPS however, takes the nominal position of 15 cm, and
chops it bit wider because it knows that the dosimetric edge is not at
the physical tip (dosimetric leaf gap), but is slightly into the leaf.
So, using that same 1.7 mm, it would put the dosimetric field edge at
15.17 cm.

So, if these offsets are being applied at both the linac and the TPS,
it looks to me like there would be a mismatch. It seems like what we
would want is for the linac to just put the leaf at its correct
physical position (15 cm in this case) and allow the TPS to take care
of the effects of the rounded leaf ends. And, interestingly, this
does seem to be what is happening in reality. As I stated above, I
could find no actual evidence that an offset was being applied at the
linac... which begs the question... what is MLCTABLE.TXT for?

As I said... I am confused. Can anyone help me out?

-Anton Eagle.

[Jeffrey V. Siebers]

Anton and others interested in how the Varian mlctable.txt file works --

There are several excellent articles/ book chapters by Art Boyer which
describe the difference between the light field and the radiation field
-- and the relationship between the two for rounded leaf ends. As I
recall, the book chapters were in a Summer School proceedings.
Sorry I am not giving specific citations, but I am working out of the
office today.
However, in order to get our Monte Carlo dose calculation to work with
moving MLCs, we had to understand precisely how to translate the
positions specified in the treatment planning system (which are then
transferred as .dml or .dma files to the MLC controller) into physical
leaf positions at the plane of the MLC.
The appendix in the paper http://iopscience.iop.org/0031-9155/47/17/312
gives you the translation --

Hope this is helpful

Jeff

[Feygelman, Vladimir]

As somebody who has encountered unexplainably changed mlctable.txt file
not once but twice, I can assure you that it does affect the leaf
position. If you want to convince yourself, shoot a picket fence, backup
the table, change the offsets by whatever amount you want and shoot the
picket fence again.

Now to the TPS. Pinnacle does not have an offset as of version 7.? It
takes the leaf positioned at 14.8285 cm from isocenter and calculates
the dose at the end. This is why you need to feed it a table in the
first place.

If you look at the typical picket fence, the junctions are hot on
average. So if you want the junctions to be OD neutral, with less
sophisticated penumbra modeling relying on just an offset, you introduce
a small offset in a system like XiO (0.3 mm would do).

Regards

VF

Vladimir Feygelman, PhD
Physicist
tel: (813) 745 8424 | fax: (813) 745-7231 | email:
vladimir....@moffitt.org

-----Original Message-----
From: pinnacl...@googlegroups.com
[mailto:pinnacl...@googlegroups.com] On Behalf Of Anton Eagle
Sent: Tuesday, September 13, 2011 1:21 PM
To: pinnacl...@googlegroups.com

Hi Listers,

-Anton Eagle.

--

[Elizabeth Shiner]

Ummmm, my Pinn v8.0m does have a leaf offset calibration table - are you saying that Pinn does not use this table??

Elizabeth A. Shiner, M.Sc., DABR
Medical Physics and Radiation Safety
Samaritan Regional Cancer Center and
Good Samaritan Regional Medical Center, SHS
Corvallis, Oregon
(541) 768-5839
fax (541) 768-5303

Regards

VF

Hi Listers,

-Anton Eagle.

Confidentiality Notice: This e-mail message, including any attachments, is
for the sole use of the intended recipient(s) and may contain confidential
and privileged information. Any unauthorized review, use, disclosure or
distribution is prohibited. If you are not the intended recipient, please
contact the sender by reply e-mail and destroy all copies of the original
message.

[Feygelman, Vladimir]

No, I am saying that Pinnacle uses the exact leaf position based on the
nominal + offset from The Table and then explicitly calculates the dose
through the rounded edge based on the MLC transmission and the leaf end
radius that you also have to specify. This is different from say XiO
that simply says that in order to get the 50% intensity line at X I need
to open the leaf by X - const.

[Robert Matthews]

Anton,

I think you can be assured that the mlctables are implemented on the linac end as well as the Pinnacle end. Question is: how sancrosanct are they? And if you think they may not be doing the job as well as you would prefer, what do you do?

Many thanks to Phil Vial for the following references that promise to be most helpful:
Vial et al PMB 51(2006), 5517-5538, and Graves et al Med Phys 28, 2227-2233. And to Jeff Siebers for the link
http://iopscience.iop.org/0031-9155/47/17/312. The PMB articles are on order through my medical library. The Med Phys article is very practical and I anticipate the PMB articles will complete the detail fill in. My plan is to build a virtual machine in Pinnacle and experiment with the mlc commission in order to more accurately model the mlc fields for my machine. Hopefully, I can have more confidence in Pinnacle planned SRS type treatments. The Med Phys article does point out that a second table may be involved, MLCXCAL.TXT, so far as the discussion has gone, MLCTABLE.TXT has been the focus of interest. MLCXCAL.TXT table contains a leaf gap value that was designed to prevent leaf collisions and apparently there are limitations to making changes to it. My service engineer has informed me that my Trilogy has circuit mods that are designed to prevent motor burn out should the leaves collide, this may not be the case on my 2100.

Bon venue.
Bob

Robert J. Matthews, Ph.D, DABR
Kootenai Medical Center
PF 208-619-4149
CdA 208-666-2529

----- Original Message -----
From: Anton Eagle <anton...@gmail.com>
To: pinnacl...@googlegroups.com
Sent: Tue, 13 Sep 2011 10:20:34 -0700 (PDT)
Subject: Re: [p3rtp] Re: Rounded MLC Offset Table

Hi Listers,

-Anton Eagle.

--

[Anton Eagle]

Hi all,

I have spent quite a bit of time on the phone with Varian regarding
thesse issues, and I think I have a better understanding of what is
going on. I think what we are running into in this discussion is
maybe a little bit of a problem with semantics.

Here's the official description of MLCTABLE.TXT and its use. This
table is NOT intended to correct for the difference in the physical
position of the leaf versus the dosimetric position of the leaf. In
other words, this table has nothing to do with dosimetric leaf gap,
rounded leaf ends, or dosimetric offsets.

What this table is used for is simply to correct for errors in the
drive system so that when you request a particular leaf position...
say 5.0 cm... the leaf ends up at exactly 5.0 cm. Without
MLCTABLE.TXT it won't end up there because there are errors in the
motor system, etc.

So, I think everyone is right that has said that this table IS
implemented. However, I think its somewhat misleading to call these
corrections "offsets"... because they are not intended to "offset" the
leaf, but rather to... well... make sure that there aren't any offsets
(errors).

So, back to the original discussion... corrections for the effects of
rounded leaf ends. Corrections for these effects are only being
applied at the treatment planning system, and are not being applied at
the linac. This makes sense... as you would end up double correcting
if offsets were being applied both at the linac and the TPS.

I hope this helps, and I hope this clarifies rather than adds to the confusion.

-Anton Eagle

[Feygelman, Vladimir]

So you are saying there is a progressive "error" in the MLC positioning,
continuously increasing from 0 in the center to some 3 mm at the edge,
the same in pretty much all Varian Linacs?

Then it follows that we are all wrong by using the offset table in
Pinnacle since v 7.4. Should we abandon using the offset table? Pretty
good results are being reported with the table, starting with Cadman et
al JACMP 2009.

As was pointed out before, how it all works is very well explained by
Vial et al PMB 2006.

I think you are mixing up the two tables: mlctable which is based on
dosimetric offsets and can be very well used for pretty much any Clinac
and mlccalib (or something to that effect) which is indeed used to
account for errors in individual leaf calibrations and general skewness
of the bank.

VF

Vladimir Feygelman, PhD
Physicist
tel: (813) 745 8424 | fax: (813) 745-7231 | email:
vladimir....@moffitt.org

-----Original Message-----
From: pinnacl...@googlegroups.com
[mailto:pinnacl...@googlegroups.com] On Behalf Of Anton Eagle

Sent: Wednesday, September 14, 2011 1:41 PM
To: pinnacl...@googlegroups.com
Subject: Re: [p3rtp] Re: Rounded MLC Offset Table

Hi all,

-Anton Eagle

--

You received this message because you are subscribed to the Google
Groups "pinnacle3-users" group.
To post to this group, send email to pinnacl...@googlegroups.com
To subscribe to this group, send email to
pinnacle3-us...@googlegroups.com
To unsubscribe from this group, send email to
pinnacle3-use...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/pinnacle3-users?hl=en

[Feygelman, Vladimir]

To further clarify, quoting from Vial et al:

"1.4. MLCTABLE or leaf position offset file (LPOF)
The nonlinearity of the geometric leaf position is corrected by the MLC
control system using a table of leaf position offsets (LPO) that covers
the range of leaf positions (LoSasso et al 1998, Graves et al 2001). The
offsets are read from a text file (MLCTABLE.TXT) which we shall refer to
as the leaf position offset file (LPOF). The vendor supplies the LPOF
with a common
set of offset values for all MLCs of a given model (LPOFvendor). This is
purely a geometric correction."

Ad you can see, it has nothing to do with the "drive errors" but is
rather a result of using the light shadow to calibrate a rounded leaf
away from the center.

Then, again from Vial et al

"The leaf position is calibrated so that the geometric leaf position
corresponds to the
nominal leaf position setting. The LPO adjusts the leaf position so that
the geometric leaf
positionmatches the nominal leaf position at all positions (see figure
2, LPOFvendor curve)."

Finally, for Eclipse (and XiO - VF):

"The RFO is corrected for by incorporating the increased width of the
radiation field in
the RTPS dose calculation. This is the method endorsed by the vendor and
implemented
in this work. The Eclipse RTPS (Varian Medical Systems, Palo Alto, CA,
USA) models
the radiation field edge at the nominal leaf position, and then
decreases the actual leaf
separation by a constant value. This method incorporates the RFO into
the leaf position
specified in the MLC file."

Pinnacle is different (and most would say better). Instead of
incorporating a blanket Radiation Field Offset (RFO), it actually
calculates the physical leaf position (nominal + MLCTABLE) and then
propagates the ray through the rounded end.

From Cadman et al 2005:

....

"MLC parameters:
* MLC transmission
* rounded leaf tip radius
* MLC leaf position offset as a function of the nominal MLC leaf
position
* tongue-and-groove width
* additional interleaf leakage transmission...

...The MLCs are modeled by creating an effective transmission array.
This array is created to
model the presence of the MLC leaf, the rounded leaf ends, and the
tongue-and-groove effects.
In regions below the full thickness of a leaf, the MLC transmission
parameter is used. Below
the leaf tip, the rounded leaf tip radius and the leaf position offset
are used to generate the
increase in transmission in the transition from the full thickness MLC
leaf to the leaf tip."

There is no need to reinvent the wheel on the subject that has been
investigated in detail since 1998...

Vladimir Feygelman, PhD
Physicist
tel: (813) 745 8424 | fax: (813) 745-7231 | email:
vladimir....@moffitt.org

-----Original Message-----
From: pinnacl...@googlegroups.com
[mailto:pinnacl...@googlegroups.com] On Behalf Of Anton Eagle
Sent: Wednesday, September 14, 2011 1:41 PM
To: pinnacl...@googlegroups.com

Subject: Re: [p3rtp] Re: Rounded MLC Offset Table

Hi all,

-Anton Eagle

--

You received this message because you are subscribed to the Google
Groups "pinnacle3-users" group.
To post to this group, send email to pinnacl...@googlegroups.com
To subscribe to this group, send email to
pinnacle3-us...@googlegroups.com
To unsubscribe from this group, send email to
pinnacle3-use...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/pinnacle3-users?hl=en

[Anton Eagle]

Hi Vladimir,
I think we are actually both saying the same thing in different ways,
and are essentially in agreement about most of this... and are maybe
just getting bogged down in the wording.

I wasn't trying to be specific about the sources of error the results
in needing MCLTABLE.TXT to get to the right position. And, regardless
of the sources of error, the ultimate purpose is to ensure that the
MLC leaf ends up with the physical tip at the required position.
Also, we seem to be in agreement that this correction has nothing to
do with the difference in the dosimetric edge of a MLC field vs. the
physical edge.

To me, it doesn't really matter how Varian gets the MLC leaves to the
correct spot... so long as we know that the ARE at the correct
(nominal) spot. Once that's taken care of, we can then focus in on
correcting the dosimetric effects in our treatment planning system.

I think the only place we disagree is with the statement "Pinnacle ...
calculates the physical leaf position (nominal + MLCTABLE)". From my
point of view, I would say that Pinnalce uses ONLY the nominal
position, and MLCTABLE just ensures that the actual position equals
the nominal position. In fact, one of the quotes you supplied even
backs this up...

"The leaf position is calibrated so that the geometric leaf
position corresponds to the nominal leaf position setting."

The only reason this is an issue, is that by stating "(nominal +
MLCTABLE)" one might be led to believe that the actual position is
somehow different than the nominal. This is what caused me some
confusion.

Instead, the way I would put it is this... Pinnalce takes the nominal
position and corrects it to essentially come up with a "dosimetric"
position. Again, maybe its all just a matter of semantics.

I do appreciate you being willing to work through this. Thanks.

Regards,

-Anton

[Feygelman, Vladimir]

It would be the same thing as long as two points are taken into account:

1. Varian calibrates the MLC by light
2. "Drive errors" have nothing to do with the MLCTABLE.TXT

Then it becomes rather simple. Because of the rounded leaf, the physical leaf extension changes non-linearly across the field compared to the light field which is used for calibration and is reported on the linac digital readout.

This reported position is modified by the offset to get to the correct physical position. Depending on whether you specify the MLC position (conventional) or dosimetric margin (IMRT), Pinnacle dose the same or the reverse, but in every case it knows the relationship between the linac (light field) readout and the physical position of the leaf provided by the mlctable file. It calculates the dose based on the physical position of the leaf, not the digital readout that is based on the shadow in the light field

Regards

VF

[Hobie Shackford]

Could part of the correction be due to the fact that the MLC's for Elekta and Varian do not travel in an arc about the source?

Hobie Shackford

Providence, RI

---

From: "Feygelman, Vladimir" <Vladimir....@moffitt.org>
To: pinnacl...@googlegroups.com
Sent: Wednesday, September 14, 2011 2:33 PM
Subject: RE: [p3rtp] Re: Rounded MLC Offset Table

pinnacle3-users+subs...@googlegroups.com

To unsubscribe from this group, send email to

pinnacle3-users+unsub...@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/pinnacle3-users?hl=en


This transmission may be confidential or protected from disclosure and is only for review and use by the intended recipient. Access by anyone else is unauthorized. Any unauthorized reader is hereby notified that any review, use, dissemination, disclosure or copying of this information, or any act or omission taken in reliance on it, is prohibited and may be unlawful. If you received this transmission in error, please notify the sender immediately. Thank you.

--
You received this message because you are subscribed to the Google
Groups "pinnacle3-users" group.
To post to this group, send email to pinnacl...@googlegroups.com
To subscribe to this group, send email to

pinnacle3-users+subs...@googlegroups.com

To unsubscribe from this group, send email to

pinnacle3-users+unsub...@googlegroups.com

[Phil Vial]

Sorry to keep this tired topic going but I am still not clear on the
Pinnacle/Elekta configuration.

My original question was

"Does Elekta implement an off-axis correction at the linac (analogous
to Varian's mlctable.txt)?"

Thanks to Dr Peter Mohr who responded to my question with:

"In reply to Phil Vial, the proper dosimetric properties of the MLC
result mainly from the leaf shape."

I think that means no.

So if the Elekta MLC controller does not apply any off-axis offsets, I
am now wondering why Pinnacle does?

Unlike the well documented Varian case, I can't find any literature
that addresses this issue in depth for the Elekta/Pinnacle case.

Would greatly appreciate a definitive explanation to this.

Regards,
Phil

[Douglas Drake]

I have been following this topic with great interest, but have hesitated to reply publically due to my own continued lack of detailed understanding. One high level person within Elekta has been kind enough to clarify that their MLC control software is intended to display the nominal radiological position of the leaf tip, not the mechanical position. Another person who was involved in designing their MLC system many years ago implied that the correction for mechanical vs. radiological position (somewhat equivalent to Varian mlctable.txt) is essentially coded into their software. Perhaps this is the source of the curve in the 2004 sample MLCi machine in Pinnacle. I have made them aware of this discussion and will certainly share anything new I learn. Perhaps someone from Elekta will join this group and reply directly.

One thing I've noticed from involvement in multiple new machine projects and sites is lack of consistency between how various personnel calibrate major and minor gains and offsets (leaves as well as jaws)- various installation engineers and field service engineers (not to mention physicists) use various methods. Light field/graph paper, film, AutoCAL, picket fences, etc. I wonder of this could be the source of the variation in "recommended curves" from Pinnacle support that I've seen posted here. I have tried to standardize our machines via AutoCAL, and found that the "2007 recommended curve" (1mm vertical offset from the sample in Pinnacle) I've seen here matches quite well (Mapcheck IMRT QA, and penumbra coincidence in Pinnacle physics modeling with crossplane "MLC only" diode scans offset inplane by half the leaf width). However, as long as variation exists between users in setting up MLC major gain and offsets, a universally recommended curve for Pinnacle seems of limited use.

I would also like to understand Elekta's recommended AutoCAL leaf tip threshold setting of 44% (rather than 50%). Anyone have any insight? I know of one large institution that uses 50%, so my MLC major offsets are set up systematically differently than theirs, therefore we should not be using the same Pinnacle MLC offset curve.

Regards,
Doug

=====
Douglas G. Drake, M.S., DABR
Medical Physicist
Dept. of Radiation Oncology
William Beaumont Hospital
3601 West 13 Mile Rd.
Royal Oak, MI 48073
Phone: 248-551-7035 Fax: 248-551-3784

>>> Phil Vial <phili...@sswahs.nsw.gov.au> 9/19/2011 4:24 AM >>>

Regards,
Phil

--

You received this message because you are subscribed to the Google
Groups "pinnacle3-users" group.
To post to this group, send email to pinnacl...@googlegroups.com
To subscribe to this group, send email to
pinnacle3-us...@googlegroups.com
To unsubscribe from this group, send email to
pinnacle3-use...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/pinnacle3-users?hl=en

As part of Beaumont's enhanced security and to reduce the amount of SPAM coming into the organization, please click https://www.mailcontrol.com/sr/XmdkjFHhjbDTndxI!oX7UlFqrXJaieUTb0XomCq!kbR9AGV!3AibVAnaUQTEPPneXStBmSRlfiqz5oFEicUPXg== to report this email as spam.

[Anton Eagle]

Hi Phil,
I think there is still a misconception regarding the off-axis
correction applied at the linac, and the rounded-leaf end correction
applied in the treatment planning system.

The correction applied at the linac simply gets the leaf at the
nominal position... so that when you tell the MLC system to put the
leaf at... say... 10.00 cm, it truly ends up at 10.00 cm. Without the
MLCTABLE.TXT... it will not. This is not a correction for the
dosimetric difference between the physical end of the leaf, and the
dosimetric end of the leaf (which are different).

The correction applied at the treatment planning system, however, does
attempt to correct for the difference between the physical end of the
leaf and the dosimetric end of the leaf.

The corrections are entirely unrelated, other than the fact that the
rounded end of the leaf plays a role in both cases. The reasons for
the corrections are different, and the results of the corrections are
different. In the first case, the result is to get the leaf to the
correct nominal position. In the second case, you are calculating the
difference between the physical field edge and the dosimetric field
edge, because the dosimetric end of the leaf is different than the
nominal position.

So, in general, you would not expect there to be a link between the
two. Theoretically, if you had some kind of MLC system that didn't
have any discrepancies between where it wanted the leaf to be and
where it actually ended up... then you wouldn't need a table like
MLCTABLE.TXT. But even in that hypothetical case, you would still
need a rounded leaf-end correction in the TPS.

So, if Elekta leaves have rounded ends... you need to correct for that
in the TPS regardless of what is happening at the linac.

I hope this helps.

-Anton Eagle

[Feygelman, Vladimir]

I think Phil has a very keen understanding of the Varian MLCTABLE as evidenced by his definitive paper on the subject of RFO...

----- Original Message -----
From: Anton Eagle [mailto:anton...@gmail.com]
Sent: Monday, September 19, 2011 10:27 AM
To: pinnacl...@googlegroups.com <pinnacl...@googlegroups.com>
Subject: Re: [p3rtp] Re: Rounded MLC Offset Table

I hope this helps.

-Anton Eagle

[Anton Eagle]

I apologize if my reply was offensive... that certainly wasn't my
intent. I was just trying to be thorough with my reply.

-Anton

[Sau...@klinik.uni-wuerzburg.de]

why 44%?
Doug and *, field sizes measured with film or in a water tank were compared with those measured with Elekta's portal imager. The best match was found using 44% in ACAL. This value certainly may vary with detector type etc.
Regards Otto Sauer

[Douglas Drake]

Thanks! I figured as much; didn't get around to trying it myself. Found nothing in Elekta or AutoCAL documentation. Any paper, etc. on this?
Best regards,
Doug

>>> <Sau...@klinik.uni-wuerzburg.de> 9/20/2011 3:22 PM >>>

Regards,
Doug

--

You received this message because you are subscribed to the Google
Groups "pinnacle3-users" group.
To post to this group, send email to pinnacl...@googlegroups.com
To subscribe to this group, send email to
pinnacle3-us...@googlegroups.com
To unsubscribe from this group, send email to
pinnacle3-use...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/pinnacle3-users?hl=en

As part of Beaumont's enhanced security and to reduce the amount of SPAM coming into the organization, please click https://www.mailcontrol.com/sr/o02tVZKtchvTndxI!oX7Ulvwl1GqFAPrtn6NXxGgxatWzCnBk2YVWQGZMwDw3dulAoCJ4GgJwFxca1YcpG8Zdg== to report this email as spam.