Can you be more specific on this "gamma pass range"? If it is related to the accuracy with which you compute the predicted dose, then I could provide you some insight. I believe you want you delivered dose to be as close as possible to the predicted value, assuming that your model is describing your situation accurately.

The generally accepted gamma pass rate for IMRT dose verification is 3% - 3mm. The criteria may be different depending what type of treatment delivery you are attempting to verify.

In the relative 2D dose distribution comparision,accepted gamma pass rate for IMRT dose verification is >95% points within 3% dose difference or 3mm DTA.It is also depends on tretment modality.

Indeed, accepted gamma pass rate, either for 2D or 3D gamma index is 3% of the dose and +/-3mm. But you have to be aware that this must be for the local gamma index, if you use normalised dose or maximum dose as a reference, than you should be more strict because local errors could be more than 3% even if you set up the threshold at 3% for the gamma index.

I guess you mean the dose range that is used to compute the gamma value and gamma passing rates. Usually we ignore points with doses less than 10% of the reference dose, i.e. we don't include them in the gamma analysis, but other people/centres may work differently.

Juan is correct. Doses below 10% of the reference dose are excluded. Low doses delivered to uninvolved areas are also important in evaluating the distribution of dose delivered during the treatment. Doses greater than 10% of the reference up to the maximum dose delivered would answer your original question. The criteria used is 3%-3mm with the expectation that >90% of the points tested passing the 3%-3mm criterion would be a minimum acceptable threshold. Although clinically my facility uses a 95% minimum acceptable pass rate.

The question is very specify for gamma pass, not the dose range. Yuan is asking about gamma pass rate. So, imaging that you have a plan and you want to verify that plan by measuring the dose distribution in phantom in order to see how many points in the plan meet the gamma index. For instance, you check 100 points in the plan with 100 appropriate points in the measured dose distribution. If 95 pair of points have the gamma index less than 1, we say that plan has 95% gamma pass (or gamma pass rate is 95%). Some body said that there is no exact answer because it is depend on you decision for gamma index, the gamma index is defined by % dose difference and the distance between two points that you want to compare. For example, you have 2 points, point A in treatment plan, point B in measured dose distribution map, now you want to compare point A to point B. Actually poin A and point B is not the same coordinate (one from TPS, one from measurement) so that they have some distance between 2 points, that is distance-to-agreement (DTA) value, let say this value is 3mm. Additionally, point A and point B is different in dose, let say the dose difference is 3%. So, if you agree distance-to-agreement is 3mm and dose difference is 3%, the gamma index of this pair of point (A and B) is equal to 1. If you agree with DTA=2mm and dose diff = 2%, then the gamma index of A & B is greater than 1, if you agree with DTA=5mm and dose diff=5%, then the gamma index of A&B less than 1.

Go back the 100 pairs of A and B, if you agree with DTA=3mm and dose diff=3%, maybe 95 pairs have gamma index less than or equal to 1 (gamma pass rate =95%). If you agree with DTA = 5mm and dose diff =3%, maybe 98 pairs have gamma index less than or equal to 1 (gamma pass rate = 98%).

You can search the formula of gamma index on internet.

I believe the question is about dose range.

10% low dose threshold is suggested in that nice speech:

http://vimeo.com/75797576

It is mentioned at 10:52, but the whole presentation is interesting. Maybe the 10% recommendation will be put "officially" in the APPM TG218 Report, but it is not ready yet:

http://www.aapm.org/org/structure/?committee_code=TG218

You can also have a look at paper by Palta et al:

http://www.medikalfizik.org/images/MedicalFizik/SANAL%20KÜTÜPHANE/AAPM%202003/AAPM2003/PaltaAAPM2003.pdf

They propose different criteria for high-dose and low-dose regions, which may make sense (Table 2, page 602).

I'm not much into radiotherapy, but feel that not-veryfying points with dose below 60% would be completely wrong. Yes, we want to have precise information on "high" doses, cause they are needed in the target to cure patient. But at the same time we want to know everything about the "low" doses, especially in organs at risk, so we know that we will not kill the patient...

If you read the following paper, uncertainty becomes worse (>3%) below 0.2 optical density. That should give you a lower bound for the dose range.

Accurate dosimetry with GafChromic™ EBT film of a 6MV photon beam in water: What level is achievable. http://scitation.aip.org/content/aapm/journal/medphys/35/2/10.1118/1.2828196

I suggest you consult the recent review (Low, D. A.,et al. (2013) in Medical Physics Vol40 Iss 7 Article Number: 071722 ) in which the creators of the concept review all published values used for the dose and distance scaling factors.

I recommend to read the earlier contribution of the same author:

Low D. A., et al. Med. Phys. 25, 656 (1998)

Thank you very much. All these answers are very helpful.

A 60% low dose threshold for the gamma pass value is sometimes too high. You have to validate both, the dose in the PTV and the dose in the organ of risk. To avoid an overdosage in the organ of risk lower dose values have also to be validated.

This is a pure clinical question. Here are some suggestions:

(1): Look at your patient plans. For example, for SRS cases, the surface dose might be around 30%. For body cases, the surface dose might be 10%. This is the minium "cut" value that you are going to use.

(2): The dose range depends on your own interests. Say an organ at risk is around 50%. Then you of course should at least consider the dose range from 50% to 100%.

Therefore, there is no definite answer to this question.

I agree that it depends somewhat on the clinical application - but note it also depends on the measurement technique. As Moises pointed out, for film measurements you are limited by the film shade, so the dose range depends on how far above the shade your dose goes. We often use the PTW Octavius with 2D-Array and rotation unit, where you don't run into this problem - but here, you need to look out for sampling effects due to the space between the detectors, and need to adjust your dose grid for the TPS plan calculation accordingly. Depending on the case, we consider between 10 % and 30 % of the maximum dose as the cut-off value. You can also allow relaxed pass criteria for doses below a certain threshold, but we normally don't do this.

Dear Change

Dose rate it's not from 60% to 100% , it's depend clinically . As well as , as a physitist I think the dose rate and time of irradiation should be calibrating first , then we can decided

I think that it depends on what you compare. If the comparison is related to commissioning data, you should look to point dose > 10%. There is no direct answer. You can look at the recent paper of Low on Medical Physics

I agree with most previous replies but would like to add a few comments/caveats:

(1) The full dose range should be evaluated to assay the target as well as organs at risk. Local Gamma discrepancies must be placed in the context of the underlying structures (e.g Gamma maps within contoured regions).

(2) A poor Gamma score is a nice 'alarm' but it does not identify the cause of a discrepancy ...was the alarm triggered by a local dose error (delta-Dose,D) or large distance to a nearby matching point (delta_distance, d)? This may need a followup investigation by manipulating the threshold values to isolate the dominant term in the Gamma(D,d) vector.

(3) Gamma scores can be evaluated in 2D (in-plane measure of distance "d") or (true distance "d" searched in the nearby 3D volume).

Make sure you understand what your code is doing!

(4) The Gamma alarm set points should be tied back to the acceptable DVHs - the clinical decision-making criteria. e.g. What gamma threshold values and score are neededto avoid 'rounding' the DVH curve in the target volume ?

Similarly, a different set of gamma thresholds may be needed to score organs at risk . Unfortunately, this is NOT a simple linear transform!

(5) The "% dose deviation" can be misinterpreted - Is it a local % difference OR or a % -point difference, normalized to a reference point (like the maximum dose). For example, in a low dose region (e.g. 40 % of isocentric dose), would a dose value of 45% be scored as a 5% deviation OR a 5/40 = 12.5 % deviation?

Bottom line - understand your code, its limitations, and test it with a known dose distribution before clinical use ! It could lead to you to a false sense of security...

J2B

I appreciate all the comments of Dr.Battista. Thanks

Thank you for the comments of Dr.Battista.

I think that the Dr Battista answer is comprehnsive

Just some general thoughts added: As said above the gamma index is not a fixed equation. Vendors "adapt" the gamma pass rate as their devices need it. There are many ways to name a gamma index. So you can not even compare 2 gamma pass rate without lots of background information. Sayings like "we just treat with pass rates above 98%" don't give you any information, it just sounds nice.

We are working on common "baselines" so medical physicists can compare worldwide exactly for this reason.