Register for Updates | Search | Contacts | Site Map | Member Login

ICRP: Free the Annals!

View Comment

Submitted by Albrecht Kellerer, ICRP C1
   Commenting as an individual
Document 2005 ICRP Recommendation
 
PART 2 of 3
Intricacies of measurement:
The computation of E requires first the quantification of the particle types and energies and also of the direc-tional distribution of the incident field. With this information it is then necessary to derive the internal radia-tion field and the resulting organ absorbed doses, DT,R in the exposed body or its mathematical phantom. For comparatively simple incident radiation fields of known composition this is not too complicated.
However, for complex incident fields, and especially in unknown fields, the computation can be difficult. Measurements in a phantom are then more attractive, because the exposure of a phantom provides the inter-nal radiation field ready made. But there is a strange problem. With the current definition even perfect meas-urements of the radiation field at all points in the phantom are insufficient to determine the effective dose or the radiation weighted organ doses. The average absorbed dose, DT, to an organ can be measured, but no measurement within the organ can provide the subdivision into the doses DT,R. There must be something amiss in the definition of the effective dose or the organ weighted doses, when these quantities can not be determined even from the full knowledge of the radiation field in the exposed body or phantom.
The former definition according to Eqs(1,2), i.e. the use of an LET-dependent weighting factor, did permit a measurement – or an approximate measurement – of effective dose equivalent by bringing into the radiation field a phantom that corresponds to the exposed individual, and then to determine the equivalent doses at the relevant organ sites. Such measurements can – at least in principle – be performed with small microdosimet-ric detectors.
The current restrictive definition in terms of Eq(3,4) has unnecessarily removed this possibility, because it has abandoned the local radiation weighting factor that corresponds to the overall weighting factor wR. The quality factor Q(L) still exists, but – contrary to the original intention – it is not compatible with the radiation weighting factor wR and can, therefore, not be used as closely equivalent alternative to wR.
Since the direct approach is ruled out, a measurement of E would require prior spectroscopy of the external radiation. Separate exposures of the phantom to the various components of the incident field (i.e. the particle types and energy bins) would then have to be performed. In each of these exposures the organ absorbed doses would have to be determined, which can, in principle, be done with small dosimeters. It is obvious that in a high energy field with various components such a procedure would be prohibitive. The problem is not due to inherent “immeasurability”, it is due to the strange mixture of external (wR) and internal (DT,R) ingre-dients in the current definition of the effective dose or the organ weighted doses.
The same considerations applies, of course, to less ambitious, approximate measurements which sub-stitute the measurement at different locations in a realistic phantom by measurements at several depths in a spherical phantom or in some other simplified geometry. In all cases there remains the striking consequence of the exclusive use of wR with no alternative formulation in terms of a corresponding lo-cal, LET-dependent radiation factor: Even if the radiation field is completely known at all posi-tions in the phantom, the information is insufficient to determine the effective dose or the radiation weighted organ doses. As has been pointed out, the definition in line with Eq(4) requires, in addition, the extraneous information on the composition of the incident radiation.
Conclusion: The measurement of the effective dose is possible in terms of suitable phantoms and small do-simeters. However Eq(4) complicates the measurement unnecessarily, because it requires spectroscopy of the incident field in addition to measurements in a phantom for each of the radiation components. Under the anticipation that effective dose can not be measured – which would be a strange attribute of a physical quan-tity – a restrictive definition has been adopted in ICRP 60 that goes a long way towards a nearly unmeasur-able quantity. Specifying – in line with the proposal in ICRP 92 – a local, LET-dependent weighting factor that is equivalent to wR will remove the difficulty. Eqs(1,2) and (3,4) are then essentially equivalent. The effective dose E can thus – where it is convenient – be expressed in terms of the radiation weighting factor wR, i.e. in terms of Eq.(4). Alternatively – in measurements and particularly in complex, high energy fields – Eq(2) can be used.


The considerations in the preceding sections can be summarised:
Due to computational problems at the time, values of wR for neutrons were given in ICRP 60 that dif-fered substantially from representative values of the effective quality factor. Perhaps in order to neutralise this problem, the weighting factor wR was separated from Q(L). However, measurements in unknown radia-tion fields require the use of a local radiation weighting factor that depends – in the absence of a better alter-native – on linear energy transfer L. There are different options to achieve this:

Option 1:
The local weighting factor is set equal to the current Q(L):
= 1 for L <=10keV/m
w(L) = Q(L) = 0.32 L  2.2 for 10keV/m <= L <=100keV/m
= 300/  L for 100keV/m <= L

With a standard anthropomorphic phantom (average overADAM and EVE) and for an isotropic field the resulting radiation weighting factor is well described by the empirical equation :
wR = 2 + 10 exp[ln(En))2 /4] + 2.5 exp[ln(En/20))2 /12]
where En is the energy of the incident neutrons in MeV.


Fig.1: Option 1 to provide equivalent radiation weighting factors wR and w(L). The current convention for the quality factor is used for w(L). The radiation weighting factor for neutrons then needs to be substantially reduced to attain equivalence.

Thus, if Q(L) is used as local LET-dependent weighting factor w(L), and the values wR for neutrons are changed accordingly, the resulting values of effective dose will be considerably smaller than those obtained in terms of the current wR (see Fig.1). Since the current values of effective dose have already been incorpo-rated into official regulations, such as the European Radiation Protection Directive, it is doubtful whether the radical reduction of the effective dose values will be deemed acceptable.

Option 2:
Faced with the current imbalance between Q(L) and wR ICRP 92 proposed to modify wR moderately to make it compatible with an LET-dependent weighting factor w(L) = 1.6 Q(L) – 0.6 that is patterned after Q(L) but has an enlarged excess over unity (Fig.2).
= 1 for L <=10keV/m
w(L) = 0.5L  4 for 10keV/m <= L <=100keV/m
= 420/  L for 100keV/m <= L
With a standard anthropomorphic phantom (average of ADAM and EVE) the resulting over-all weighting factor is well described by the empirical equation 3:
wR = 2.6 + 16 exp[ln(En))2 /4] + 4 exp[ln(En/20))2 /12]




Fig.2: Option 2 to provide equivalent radiation weighting factors wR and Q(L). The LET-dependent local weighting factor is patterned after the current convention Q(L), but the excess over unity is increased by the factor 1.6 in order to avoid a large reduction of the current values wR for neutrons.

The resulting values are closely in line with the proposal in ICRP 92 (Fig.4.4 and formula in foot-note 9) except for the somewhat different and simpler numerical approximation which provides correct values even at extremely high neutron energies. Likewise, Eq(8) (or curve C of Fig.1) in the Draft for Consultation is close to the ICRP 92 proposal, but it lacks the close agreement with a specified LET-dependent weighting factor w(L). Also – as does the proposal in ICRP 92 – it gives the unduly high value wR=5 for very high neutron energies.

B. Specific Comments on Draft for Consultation
2005 Recommendations of the ICRP
Comments on Summary
(S14), line 6-8:
If a new special name for a unit or unit combination of the International System (SI) is to be introduced officially by the ICRP, the proposal needs to be sanctioned by the Bureau Interna-tional des Poids et Mesures through the ICRU.
Since numerous quantities share the same unit, the SI requires that a quantity be identified by its name, and not by the name of its unit. But this rule is frequently ignored. To reduce the resulting confusion, certain special names of units (for use with particular quantities) have been accepted. An example is Hz (hertz) as special name of the reciprocal second (s-1) when used with frequency which relates to periodic processes. Activity being a different quantity, the additional special name Bq (bequerel) has been adopted.
Another concession was made, because the quantities absorbed dose and dose equivalent are fre-quently not identified, but are simply termed dose. In view of this undesirable usage the ICRU rec-ommended – and the SI accepted – the two special names Gy (gray) and Sv (sievert) for the unit J/kg when used with absorbed dose and equivalent dose.
However, the ICRU and the SI will be reluctant to introduce yet another special name for J/kg when used with radiation weighted organ dose (the former equivalent dose to an organ). The reason is that this might open the floodgates too wide. If Sv (sievert) were to be used with effective dose, and – for example – Ta (taylor) with radiation weighted dose to an organ, then what symbol shall be used with weighted dose at a point? Li (lindell) ? Ultimately when every dose quantity has acquired its own special name for the unit, nothing will be gained, except that two names are linked to each quantity (and one of the major advantages of using coherent units, the easy check of equations for correct dimensionality is inhibited).
There are related issues and problems. Examples:
1. Certain special symbols and names for units of the International System of Units (SI) have been adopted and used inofficially. An example has been ret (´rad equivalent therapy´) for 0.01 J/kg when used with an absorbed dose weighted for effectiveness in therapy.
2. ICRP 92 (paragraph (297)) contains the debatable proposal to use the special symbol Gy-Eq for J/kg when an absorbed dose is multiplied by a weighting factor for “deterministic” radiation effects. This is not a new request, and certainly not – in my judgement – a convention that deserves official status. At best, it can serve as a notation that is used in certain specialised studies or reports, but whenever it is used it needs to be accompanied by a clearly spelled out ad hoc definition.
3. Over the years there has been an unresolved discussion at UNSCEAR, whether the particular weighted dose used at RERF (absorbed dose from -rays plus 10 times the absorbed dose from neutron recoils) can be expressed in Sv. Here, too, the reasonable conclusion is, to adopt whatever ad hoc convention seems practical, and to define and explain it whenever it is used. No general convention can be adopted for such specific issues.

These considerations are not meant to imply that there is not a problem that deserves serious atten-tion. Thus, a dose from radon is often quoted simply as “dose of x mSv”, without specification

whether effective dose is meant or radiation weighted dose to the lung. But such unacceptably loose terminology ought to be avoided by emphasising the need to identify a quantity by its name, not by its unit.
This may seem a lengthy discussion of a single technical point. But it is a point with major implica-tions. The Main Commission will be well advised to consider the issue with caution and – at any rate – in co-ordination with ICRU.

Added remark to this sentence: In view of the problems associated with the issue, the formulation needs to be pre-cise. Therefore the current text (if it were retained) ought to read “The Commission is considering a new special name for the unit of radiation weighted dose ”

(S14), line 5: “radiation weighted dose“ is not a bad name and it is adopted in the present comments. But – to stay with the above example – even “equivalent dose to the lung” is too often abbreviated. The longer new name “radiation weighted dose to the lung” might make the correct use of the name of the quantity even less common. Perhaps one should either stay with equivalent dose to the organ or find a shorter version of the proposed new name.

(S15), last sentence:
Why the reference to “a dosimetric model”? The statement seems to imply that the value of E is meaningful only if accompanied by additional dosimetric information (type and position of phan-tom?).
The original definition of the effective dose (ICRP 60, (A18)) appears to refer to a particular person and a particular exposure situation. Although the value of E may be difficult to determine with much accuracy, its value is then well defined. In principle one can re-create the exposure situation, substitute a tailor made phantom for the exposed person, and then perform the necessary measure-ments in the field and in the organs of the phantom. In this sense a value of E can be given without specification of “a dosimetric model”.
Assume, for instance, that a particular worker has performed a particular task at a particular work-ing place and that a careful analysis – in order to assess whether the effective dose limit has been exceeded – has provided the estimate E= (3710) mSv. The statement of this value makes sense, even if it is not accompanied by a specified dosimetric model (although the details of the assess-ment may need to be specified for validation).
I suspect, that the last sentence in (S15) refers to the – more common – case where an effective dose is specified – in a general way – for the potential exposure of “a person”, rather than a spe-cific individual. A dosimetric model (type of phantom and orientation) must then, indeed, be speci-fied to make the result meaningful.
Would it help to add at the end of the sentence: “ ..., whenever it is not related to a particular person and a particular exposure geometry” ?
Some related notions – for example the statement that E is “not measurable” – can be so readily misread that an extra effort towards clarity is desirable.
COMMENT CONTINUES IN A 3RD AND FINAL PART