Skeletal age determination of the hand: a comparison of methods

ORIGINAL ARTICLE

Skeletal age determination of the hand: a comparison of methods

S. Schmidt & I. Nitz & S. Ribbecke & R. Schulz &

H. Pfeiffer & A. Schmeling

Received: 20 August 2012 /Accepted: 27 February 2013 /Published online: 13 March 2013# Springer-Verlag Berlin Heidelberg 2013

Abstract Until final completion of maturation processesat the age of approximately 18 years, determination ofthe skeletal age of the hand plays a central role inforensic age diagnostics in living persons in criminalproceedings. In this process, assessment of hand radio-graphs relies primarily on the stage of development ofthe epiphyseal nuclei, the increase in size of the indi-vidual bones and of the hand skeleton as a whole,changes in the shape of the various skeletal elementsand ossification of the epiphyseal plates. To achievethis, there are a variety of methodological approachesbased on two different fundamental principles. Themethods proposed by Greulich and Pyle, Thiemann etal. and Gilsanz and Ratib rank among the so-called atlastechniques, whilst the methods proposed by Tanner etal. and Roche et al. are classified as so-called bone-specific techniques. In order to be applicable in the fieldof criminal procedure, the methods of estimating theskeletal age of the hand developed with clinical aspectsin mind must satisfy the demands of a high degree ofestimate accuracy and good reproducibility of the esti-mated results. In the course of the present study, a studypopulation of 92 persons was used to compare the

above-mentioned atlas and bone-specific techniques fordetermining hand skeleton age in view of these qualita-tive criteria. Estimate accuracy was studied usingPearson’s correlation coefficients, and weighted kappacoefficients were determined for studying the intra-andinterobserver agreement of an estimate result. In theinter-method comparison, a basically good agreementwas shown between the skeletal ages and the chrono-logical age of the test persons on the one hand and theskeletal age diagnoses of one or of two examiners onthe other. No general advantage of the methodologicalapproach of the bone-specific technique was discerniblein the course of comparison; in the female gender,particularly, the RUS2 and RUS3 score of the methodof Tanner et al. proved unfavourable. For age estimationpractice in criminal proceedings, the atlas methods ofGreulich and Pyle and Thiemann et al. are particularlyrecommendable.

Keywords Forensic age estimation . Bone age . Handskeleton . Comparison of methods . Estimate accuracy .

Reproducibility

Introduction

Forensic age estimation is a current focus of research inter-est in the field of forensic medicine. Over the last few years,it has become increasingly important to determine, in par-ticular, the age of living persons [5, 19, 20, 32, 37, 38].

In the context of criminal proceedings, the StudyGroup on Forensic Age Diagnostics currently recom-mends that, in addition to a physical inspection and adental examination, skeletal age should be determined[23]. In principle, this is accomplished initially by

S. Schmidt (*) :R. Schulz :H. Pfeiffer :A. SchmelingInstitute of Legal Medicine, Münster University Hospital,Röntgenstraße 23, 48149 Münster, Germanye-mail: [email protected]

I. NitzBerlin, Germany

S. RibbeckeDr. Gladitz Statistical Service, Schwedter Straße 225, 10435Berlin, Germany

Int J Legal Med (2013) 127:691–698DOI 10.1007/s00414-013-0845-4

examining the degree of maturity of the skeletal ele-ments of the hand. From a forensic viewpoint, assess-ment of the hand skeleton is applied in questionablecases to help monitor whether any juristically relevantage limits have been exceeded. As the persons in ques-tion may, under certain circumstances, risk serious con-sequences, the result of the examination is coupled withthe demand for a particularly high level of validity. Tobe forensically applicable, the methods available musttherefore offer a high degree of estimate accuracy andfacilitate reliable diagnoses.

To date, a multiplicity of very varied techniques havebeen developed for determining skeletal age based on an X-ray image of the hand. All existing methods may essentiallybe classified according to two different basic principles:atlas and bone-specific techniques.

A large number of X-ray atlases are in existence,depicting the normal ossification of the human hand [10,12, 17, 21, 24, 33, 39, 40]. These make it possible toestimate the chronological age of a child or adolescent bycomparing the overall maturation pattern of a given X-rayimage of the hand with a collection of standard images(Fig. 1). Amongst these atlas methods, the standard worksby Greulich and Pyle [12] and Thiemann et al. [40], inparticular, are applied in criminal age estimation practice.In recent years, the digital skeletal maturation atlas ofGilsanz and Ratib [10] has provided idealised standards ofnormal hand skeleton development classified specifically bygender and age.

Although the different ossification centres of the handskeleton occur in a certain regular order, and changes totheir size and shape as well as closure of the growthplates take place more or less following set principles,hand radiographs with inhomogeneous maturation patternsalso have to be assessed occasionally. This phenomenon istaken more strongly into consideration in another method-ological approach. Bone-specific techniques make it pos-sible, by means of specific assessment systems, toestimate individual chronological age on the basis of thedegree of maturity manifested by selected skeletal ele-ments in an X-ray image of the hand (Fig. 2). Amongstthe techniques most used in forensic work are the bone-specific methods of Tanner et al. [11, 34, 36]. The methodof Roche et al. (FELS method [22]) has been establishedfor a long time, primarily in clinical use.

To determine the most suitable method for use in criminalforensic work, the present study aims to compare the esti-mate accuracy of the atlas methods of Greulich and Pyle,Thiemann et al. and Gilsanz and Ratib, and the bone-specific methods of Tanner et al. (here, the RUS2 andRUS3 scores) and Roche et al. Furthermore, all thesemethods will also be studied comparatively in relation tothe reproducibility of skeletal age diagnoses.

Test persons and methods

Radiographs of the left hand of a total of 48 male and44 female children and adolescents aged between 12and 16 years were evaluated retrospectively. Table 1shows case group figures subdivided by gender andchronological age.

The X-ray images of the hand were created in theperiod from 1986 to 2002 in a specialised osteologypractice in Papenburg, Germany. The evaluation includ-ed exclusively hand radiographs of children and adoles-cents whose physical development was appropriate totheir age. If any indications of a disease which mightinfluence skeletal maturation emerged, this led to exclu-sion from the study.

All the X-ray images were on a scale of 1:1. They hadpreviously been digitalized using an X-ray scanner andanonymised by randomly distributed numerical data names.It was not possible to derive any person-related informationfrom the images themselves. Only the head of the study,who was not involved in the collection of data, knew the ageand gender of the persons under investigation in the form ofa key file.

With the use of the program Synedra view personal 3version 3.2.0.0, all the X-ray images were assessed in thecourse of a blind study by two examiners with much expe-rience in the evaluation of hand radiographs, independentlyof one another. One of the examiners repeated the evaluationfollowing an interval of at least 3 months. In the case of eachindividual X-ray image, skeletal age was determined usingthe methods of Greulich and Pyle, Thiemann et al., Gilsanzand Ratib, Tanner et al. (both RUS2 and RUS3) as well asRoche et al.

It is only possible to determine the age of a skeletonaccording to Roche et al. by using special software(FELShw). The chronological age of a person that is as-sumed to be known is also needed for the calculations. Sinceit is not possible to provide appropriate information underthe conditions of the forensic age estimation, we used theskeletal age previously ascertained according to Greulichand Pyle as an example in accordance with a recommendedalternative approach [6].

The statistical interpretation of the data ensued with theaid of IBM SPSS STATISTICS 19 software. All calculationswere undertaken divided by gender.

To study the criterion of estimate accuracy as acorrelation between the skeletal age ascertained in de-pendence on the different methods on the one hand, andthe chronological age of the test persons on the other,we applied the respective Pearson’s correlation coeffi-cients (ρ). To analyse the criterion of reproducibility asa correlation between the method-dependent diagnosesof skeletal age made by one and the same examiner

692 Int J Legal Med (2013) 127:691–698

(intraobserver agreement) and by the two examiners(interobserver agreement), the weighted kappa coeffi-cients (κ) were calculated.

Results

As the result of the studies on estimate accuracy,Tables 2 and 3 show the correlation coefficients calcu-lated for the observed method spectrum, divided bygender. These are taken here as a basic measure of thelinear correlation between the characteristics of method-dependent skeletal age and the chronological age of thetest persons.

The coefficients ascertained for the study populationare characterised across the methods, in both genders,by values which quite predominantly lie close together.In detail, the coefficients determined for female testpersons, in particular, using the atlas methods ofGreulich and Pyle as well as Thiemann et al., but alsousing the bone-specific method of Roche et al., indicatea comparatively high linear correlation measure betweenthe skeletal ages determined in each case and the rele-vant chronological age (0.766≤ρ≤0.802). Most notably,the skeletal ages determined in accordance with thebone-specific method of Tanner et al. lead, in thiscontext, to significantly poorer results (0.615≤ρ≤0.717).

Particular good estimate accuracies can also be achievedin the male gender using the atlas methods according toGreulich and Pyle or Thiemann et al. and according to thebone-specific method of Roche et al. (0.745≤ρ≤0.791).There are no similar restrictions on the method accordingto Tanner et al. either in the RUS2 or in the RUS3 score(0.760≤ρ≤0.793).

In Tables 4 and 5, the results of the study on thereproducibility of the method-dependent skeletal agediagnoses for one or both examiners, respectively, areshown with the gender-specific weighted kappa coeffi-cients. The coefficients for the study population assumealmost exclusively high values of over 0.9 in bothgenders, regardless of the method applied. More differ-entiated observation shows that, besides the bone-specific method of Roche et al., each of the atlasmethods studied produces a comparatively high degreeof intraobserver agreement (0.959≤κ≤0.980) amongstthe female test persons. By contrast, the correspondingcoefficients for the RUS2 and the RUS3 score of thebone-specific method of Tanner et al. display a signifi-cantly lower value (κ=0.913). In the interobserver com-parison, the weighted kappa coefficients calculated forthe atlas method of Thiemann et al. and the bone-specific method of Roche et al. represent a very highdegree of agreement of results amongst the female test

persons (0.933≤κ≤0.953). Here too, the coefficients ofthe two tested versions of the bone-specific method ofTanner et al. are lower (0.849≤κ≤0.917). In the malegender, an overall tendency towards higher weightedkappa coefficients can be ascertained, and at the sametime, methodologically induced discrepancies are lesssignificant. Here, particularly high linear correlation mea-sures emerge for the atlas method of Greulich and Pyle, inparticular (intraobserver agreement, κ=0.985; interobserveragreements, κ1=0.949, κ2=0.966) as well as the bone-specific method of Roche et al. (intraobserver agreement,κ=0.992; interobserver agreements, κ1=0.986, κ2=0.984).

Discussion

Only recently could some of the today more or lessestablished clinical methods to determine skeletal age onthe basis of an X-ray image of the hand also be utilised forforensic age diagnostics [25, 26, 28–31]. The methodsavailable can only be used in forensic age estimation if theysatisfy the criteria of a high degree of accuracy and repro-ducibility of the diagnoses.

To evaluate the criterion of the estimate accuracy of themethods analysed, use was made of Pearson’s correlationcoefficient. This indicates the strength of the linear correla-tion between two metric variables and can thus be used as ameasure of the linear correlation between the characteristicsof method-dependent skeletal age on the one hand and of thechronological age of the test persons on the other [2]. As thecorrelation does not take into consideration whether thereare differences in the mean values of the two distributionsand whether the individual value ranges which result differ,the method-specific estimate accuracy can be evaluatedwithout the additionally superimposing influence of thesecular degree of acceleration in the different referencepopulations.

The correlation coefficients submitted with this studymake it clear that, overall, a fundamentally high degreeof agreement exists between the characteristics studied.They thus express the generally good estimate accuracyof all the assessment methods observed. This becomesparticularly apparent not only in the case of the atlasmethods of Greulich and Pyle and Thiemann et al., butalso in the case of the bone-specific method of Roche etal. By contrast, the RUS2 and RUS3 versions of thebone-specific method of Tanner et al. stand out becausethey manifest the lowest correlation coefficients in thefemale gender.

One possible cause for the differences ascertainedbetween the methods in relation to estimate accuracywhich should be discussed is, first of all, the limitedrepresentativity of the standard criteria selected. In

Int J Legal Med (2013) 127:691–698 693

accordance with the present results, a particular advan-tage of the atlas methods according to Greulich and Pyleand to Thiemann et al. seems to be that they combinehand radiograms that are representative in terms of agewith the verbal definition of various age-specific maturitycriteria for the assessment of the skeletal maturity pattern.In contrast, the atlas method according to Gilsanz andRatib dispenses with the specification of standard criteriathat must be checked as an obligatory measure; thisseems to limit estimate accuracy particularly where aninexperienced examiner is concerned. Compared with theatlas methods, the bone-specific methods of Roche et al.and Tanner et al., with their varying emphases on differ-ent designated standard criteria of the state of maturity ofindividual elements of the hand skeleton, offer a signif-icant difference. In particular, the extremely comprehen-sive and complex system proposed by Roche et al., withits 13 skeletal proportions and 98 maturity indicators,which are divided into 232 stages and entered into theoverall result in a differentiated manner, is evidently wellable to represent the process of maturation of the handskeleton. By contrast, the principle of skeletal age esti-mation of the hand introduced by Tanner et al. in itsapplication of the clinically most significant RUS2 andRUS3 scores to the female test persons in the studypopulation leads to a noticeable decrease in estimateaccuracy. In this context, the identical stage differentia-tion of the radius and ulna used in both systems must beevaluated as deficiently inaccurate in view of the majorinfluence of the state of maturity of the distal epiphysesof the forearm on overall skeletal age, particularly in theadvanced phase of development [29, 41]. Thus, whencomparing the genders, the earlier completion of matura-tion of the female hand skeleton increasingly leads to alack of options for differentiation, particularly at a higherage, thus entailing the risk of incorrect estimates. As thecarpalia have also already achieved their mature state inthe corresponding age segment, this problem can also notbe overcome by including them in the more comprehen-sive TW20 score of the method of Tanner et al. In fact,the only conceivable solution would be to additionallytake into consideration other characteristics of the devel-opment of the radius and ulna, such as, for example, theforensically significant regression of the epiphyseal scar[3, 27]. Finally, a further general circumstance whichmay potentially compromise the estimate accuracy ofthe method of Tanner et al. must also be seen in thecompletely arbitrary assignment of point values to certainmanifestations of characteristics [9]. In contrast to themethod of Roche et al., the two tested scores do nottake into account the probabilities of the interindividuallyvarying presence of maturity indicators in the referencepopulation [18].

The limitation of the estimate accuracy of a methodfor determining the skeletal age of the hand is alsoconceivable as an expression of the inhomogeneity ofthe reference population in question. Amongst the atlastechniques analysed, the method of Greulich and Pyle isbased on a more or less clearly defined population of1,000 Americans of mainly North European origin, bornin the USA, and primarily from economically well-situated families. They were studied between 1931 and1942 at the age of 0 to 18 years in the course of theBrush Foundation longitudinal study. The method ofThiemann et al. is based on a random sample of 5,200children and adolescents aged from 0 to 18 years, select-ed in a standardised manner and analysed in 1977 to1979 in the former GDR. However, even the bone-specific method of Roche et al. is based on a studypopulation, examined from 1932 to 1977 in the courseof the FELS longitudinal study, of 355 male and 322female US-American children and adolescents with arelatively coherent national social structure. By contrast,according to the current recommendations of the StudyGroup on Forensic Age Diagnostics (AGFAD) [23], theatlas method of Gilsanz and Ratib does not meet thecriteria for a reference study applicable for purposes ofcriminal proceedings, as no details at all are given on thereference population used. Thus, amongst other factors,potential inhomogeneities in the composition of this ref-erence population cannot be excluded as a cause of theoverwhelmingly most unfavourable estimate accuraciesamongst all the atlas methods studied. The random sam-ple in the TW2 bone-specific method of Tanner et al. iscomposed of some 3,000 British children from the mid-dle and lower classes. The relevant X-ray images wereevaluated between 1946 and 1970. With the TW3 meth-od of Tanner et al., which did not appear until 2001, thisreference population was adapted to the secular trend.The updated reference values are now additionally basedon studies from Belgium [4], Spain [14] and USA [35].A further potential source of blurring of the linear com-bination of characteristics of skeletal age and the chro-nological age of the test persons, particularly in thefemale gender, may be identified in the mixture of testpersons of differing socioeconomic status, different geo-graphic origin and heterogeneous secular degree of accel-eration which characterises the random samples used inthe TW methods.

In evaluating the criterion of the reproducibility ofskeletal age diagnoses, the weighted kappa coefficientwas applied as a measure of observer agreement be-tween the method-dependent diagnoses of one and thesame examiner and of the two examiners [42].

The weighted kappa coefficients identified in thecourse of the present study attest to the overall excellent

694 Int J Legal Med (2013) 127:691–698

agreement between the diagnostic results reached by oneor both examiners in the case of all the methods stud-ied. For evaluators with good experience, the achievableskeletal age diagnoses can therefore be considered reli-able and, for the most part, independent of the observer.

The inter-method differences in the reproducibility ofskeletal age diagnoses ascertained here may be explained,for the most part, by the margin of interpretation confrontingthe examiner, for example, as a result of the insufficientlyexact description of methodological procedure and of therelevant estimate criteria, but also as a result of ambiguousreference images and illustrative graphics. In the overallview, the existing results initially reveal a generally lowerpotential for discrimination between the methods studied inthe case of the male gender. This effect, in turn, may bederived as a consequence of a later completion of maturationaccompanied by a longer-term conservation of maturitycriteria which can be differentiated unequivocally. In thegroup of bone-specific methods studied, in particular, thespecification of a few selected indicators of the developmentof the hand skeleton appears to have a limiting effectamongst female test persons. Thus, the lowest concordancesof estimate results in the intra- and interobserver comparisonwere achieved with the method of Tanner et al. in the groupof female study participants. Moreover, it proves problem-atic here too, independent of gender, that many examinersfind it very difficult to understand the texts describing thestages, which promotes a superficial orientation by the ac-companying illustrations [8]. However, the primary use ofgraphics reduced to the essential characteristics is a potentialsource of incorrect diagnoses [13]. Amongst the atlasmethods analysed, the lowest rate of intra- and interobserveragreement is to be established for the method of Gilsanz andRatib. In this connection too, the complete lack of a descrip-tion and/or schematic characterisation of each of the mostimportant age-relevant morphological criteria must be notedcritically.

Much discussion has taken place on the potential ad-vantages of the principle of the bone-specific technique ascompared with the atlas method in age estimation practice.In particular, the weighting of characteristics, which isachievable using the relevant methods, is of outstandinginterest for the clinician in the event of a dissociatedskeletal maturation. Forensic age estimations are, however,contraindicated in the case of such maturation patterns,which is why questions need to be asked about the ben-efits of the bone-specific methods in this area of applica-tion in view of the greater difficulty of learning thisprocedure and the considerable additional expenditure oftime in some cases. The data presented here confirm theassertion of various earlier studies that the greater expen-diture of time does not stand in any acceptable relation-ship to the potential increase in estimate accuracy as

compared with established atlas methods [1, 7, 15, 16,41]. The results also make it clear that bone-specificmethods likewise make little contribution to an improve-ment in the reproducibility of skeletal age diagnoses. Thisbecomes particularly apparent in the case of the clinicallyworld-renowned second and third versions of the methodof Tanner et al. In the overall comparison, both of theseare least able to fulfil the qualitative criteria that aredecisive for forensic application and should therefore notbe favoured, especially in the context of criminal pro-ceedings. By contrast, the results of the present studyconfirm the scientifically high significance of the methodof Roche et al. Moreover, in this case, the computer-supported calculation of the skeletal age is based on theperson’s clinically known chronological age; this was anobstacle to the use of the method in forensic age estima-tion. The recommendation implemented in this study toreplace the chronological age by the skeletal age deter-mined for the test persons according to Greulich and Pylecannot be applied to the forensic use of the methodaccording to Roche et al. in view of the incalculable riskof a systematic assessment error. Here, more far-reachingstudies would be indispensable.

As the present results show, the demands made byage estimation practice in criminal proceedings on amethod of skeletal age estimation of the hand are par-ticularly well met by the atlas methods analysed. How-ever, in relation to the method of Gilsanz and Ratib,besides the methodological weaknesses highlighted here,it must be kept in mind that in the case of the 14-plusage groups relevant to criminal proceedings, there is arisk of overestimating chronological age in the femalegender by up to 7.2 months [30]. For this reason, it isonly of very limited use for the purposes of forensicage estimation. With an extremely strongly acceleratedreference population in the case of the methods of bothGreulich and Pyle and Thiemann et al., the risk ofoverestimating chronological age and thus of causingprejudice to a defendant in criminal proceedings isminimised. Both methods can be recommended uncon-ditionally for use in the area of criminal proceedings.Compared with the method of Greulich and Pyle, whichis today one of the most established worldwide, themethod of Thiemann et al. has so far been appliedmainly in the German-speaking world. Its predestinatingqualities are primarily to be seen in the more up-to-datereference population and an improved scientific studyconcept involving defined inclusion criteria.

Acknowledgments The present study was supported by a grant fromthe German Research Foundation (SCHM 1609/5-1). The authorswould like to thank Dr Karl Minas MD for his kind loan of theevaluated hand radiographs.

Int J Legal Med (2013) 127:691–698 695

Appendix

Greulich & Pyle atlasstandard 25 (male)

skeletal age: 14 years

Greulich & Pyle atlasstandard 26 (male)

skeletal age: 15 years

radiograph to be evaluated (male)

chronological age: 14 years

Fig. 1 Atlas technique(Greulich and Pyle method). Thehand radiograph to be assessed iscompared with age- and gender-specific reference images. Itfulfils the maturity criteria ofGreulich–Pyle standard 25(skeletal age, 14 years), but notyet the maturity criteria ofGreulich–Pyle standard 26(skeletal age, 15 years)

Fig. 2 Bone-specific technique(RUS score of the method ofTanner et al.). For each of theepiphyses incorporated in thescore, a stage of maturity is firstdetermined to which a specificnumber of points of maturity areassigned. (In the example, theepiphysis of the phalanx medialisof the third digit observedmanifests characteristics ofmaturity stage G, to which 32maturity points are assigned.)The sum of all points of maturityidentified in this way finallyyields skeletal age

696 Int J Legal Med (2013) 127:691–698

References

1. Andersen E (1971) Comparison of Tanner-Whitehouse andGreulich-Pyle methods in a large scale Danish survey. Am JPhys Anthropol 35:373–376

2. Backhaus K, Erichson B, Plinke W, Weiber R (2006) MultivariateAnalysemethoden–Eine anwendungsorientierte Einführung.Springer, Berlin

3. Baumann U, Schulz R, Heinecke A, Schmeling A, Schmidt S (2008)Reference study on the time frame for ossification of the distal radiusand ulnar epiphysis on the hand radiograph. Forensic Sci Int 191:15–18

4. Beunen G, Lefevre J, Ostyn M, Renson R, Simons J, vanGernen D (1990) Skeletal maturity in Belgian youths assessed

Table 1 Case groupfigures Age (in years) Male Female

12 10 10

13 10 10

14 10 10

15 10 10

16 8 4

Σ 48 44

Table 2 Pearson correlation coefficients (ρ; female)

A1 A2 B

Greulich and Pyle 0.792 0.785 0.766

Thiemann et al. 0.783 0.770 0.773

Gilsanz and Ratib 0.762 0.792 0.750

Tanner et al. (RUS2) 0.717 0.664 0.644

Tanner et al. (RUS3) 0.697 0.639 0.615

Roche et al. 0.778 0.802 0.784

Agreement between the diagnosis of method-dependent skeletal age bythe first (left columns) and the second examiner (right column) andchronological age of test persons

A1 first assessment by examiner A, A2 second assessment by examinerA, B assessment by examiner B

Table 3 Pearson correlation coefficients (ρ; male)

A1 A2 B


Thiemann et al. 0.771 0.753 0.775


Tanner et al. (RUS2) 0.760 0.764 0.793

Tanner et al. (RUS3) 0.762 0.763 0.775

Roche et al. 0.751 0.745 0.766

Agreement between the diagnosis of method-dependent skeletal age bythe first (left columns) and the second examiner (right column) andchronological age of test persons


Table 4 Kappa coefficients (κ; female)

Method A1/A2 A1/B A2/B


Thiemann et al. 0.959 0.933 0.948


Tanner et al. (RUS2) 0.913 0.865 0.917

Tanner et al. (RUS3) 0.913 0.849 0.913

Roche et al. 0.980 0.953 0.951

Agreement of the diagnoses of one and the same examiner (leftcolumn) and the diagnoses of both examiners (right columns)


Table 5 Kappa coefficients (κ; male)

Method A1/A2 A1/B A2/B


Thiemann et al. 0.936 0.953 0.915


Tanner et al. (RUS2) 0.961 0.934 0.954

Tanner et al. (RUS3) 0.951 0.922 0.955

Roche et al. 0.992 0.986 0.984

Agreement of the diagnoses of one and the same examiner (leftcolumn) and the diagnoses of both examiners (right columns)


Int J Legal Med (2013) 127:691–698 697

by the Tanner-Whitehouse method (TW2). Ann Hum Biol17:355–376

5. Cameriere R, De Luca S, De Angelis D, Merelli V, Giuliodori A,Cingolani M, Cattaneo C, Ferrante L (2012) Reliability ofSchmeling’s stages of ossification of medial clavicular epiphysesand its validity to assess 18 years of age in living subjects. Int JLegal Med 126:923–932

6. Chumlea C (2011) Persönliche Mitteilung7. Cole AJL, Webb L, Cole TJ (1988) Bone age estimation: a com-

parison of methods. Br J Radiol 61:683–6868. Cox LA (1996) Tanner-Whitehouse method of assessing skeletal

maturity: problems and common errors. HormRes 45(Suppl 2):53–559. Fende l H (1976) Die Method ik der rad io log ischen

Skeletalterbestimmung. Radiologe 16:370–38010. Gilsanz V, Ratib O (2005) Hand bone age: a digital atlas of skeletal

maturity. Springer, Berlin11. Graham CB (1972) Assessment of bone maturation—methods and

pitfalls. Radiol Clin N Amer 10:185–20212. Greulich WW, Pyle SI (1959) Radiographic atlas of skeletal de-

velopment of the hand and wrist. Stanford University Press,Stanford

13. Haavikko K, Kilpinen E (1973) Skeletal development of Finnishchildren in the light of hand-wrist roentgenograms. Proc Finn DentSoc 69:182–190

14. Hernández M, Sánchez E, Sobradillo B, Rincón JM (1991)Maduracion osea y prediction de la talla. Diaz de Santos, Madrid

15. Kemperdick H (1986) Die Skelettalterbestimmung beim Kind.Radiologe 26:216–221

16. King DG, Steventon DM, O’Sullivan MP, Cook AM, Hornsby VP,Jefferson IG (1994) Reproducibility of bone ages when performedby radiology registrars: an audit of Tanner and Whitehouse IIversus Greulich and Pyle methods. Br J Radiol 67:848–851

17. Kopczynska-Sikorska J (1969) Atlas radiologiczny rozwojukoscca dloni i nadgastka. PZWL, Warszawa

18. van Lenthe FJ, Kemper HC, van Mechelen W (1998) Skeletalmaturation in adolescence: a comparison between the Tanner-Whitehouse II and the Fels method. Eur J Pediatr 157:257–261

19. Olze A, Hertel J, Schulz R, Wierer T, Schmeling A (2012)Radiographic evaluation of Gustafson’s criteria for the purposeof forensic age diagnostics. Int J Legal Med 126:615–621

20. Olze A, van Niekerk P, Schulz R, Ribbecke S, Schmeling A (2012)The influence of impaction on the rate of third molar mineralisationin male black Africans. Int J Legal Med 126:869–874

21. Poland J (1898) Skiagraphic atlas showing the development of thebones of the wrist and hand, for the use of students and others.Smith, Elder, & Co, London

22. Roche AF, Chumlea WC, Thissen D (1988) Assessing the skeletalmaturity of the hand-wrist: Fels method. C.C. Thomas, Springfield

23. Schmeling A, Grundmann C, Fuhrmann A, Kaatsch HJ, Knell B,Ramsthaler F, Reisinger W, Riepert T, Ritz-Timme S, Rösing FW,Rötzscher K, Geserick G (2008) Criteria for age estimation inliving individuals. Int J Legal Med 122:457–460

24. Schmid F, Moll H (1960) Atlas der normalen und pathologischenHandskelettentwicklung. Springer, Berlin

25. Schmidt S, Koch B, Mühler M, Reisinger W, Schmeling A (2007)Optimizing the Thiemann-Nitz method for skeletal age determina-tion for forensic age diagnostics in live subjects. Scand J ForensicSci 13:5–7

26. Schmidt S, Koch B, Schulz R, Reisinger W, Schmeling A (2007)Comparative analysis of the applicability of the skeletal age deter-mination methods of Greulich-Pyle and Thiemann-Nitz for foren-sic age estimation in living subjects. Int J Legal Med 121:293–296

27. Schmidt S, Baumann U, Schulz R, Reisinger W, Schmeling A(2008) Study of age dependence of epiphyseal ossification of thehand skeleton. Int J Legal Med 122:51–54

28. Schmidt S, Koch B, Schulz R, Reisinger W, Schmeling A (2008)Studies in use of the Greulich-Pyle skeletal age method to assesscriminal liability. Leg Med (Tokyo) 10:190–195

29. Schmidt S, Nitz I, Schulz R, Schmeling A (2008) Applicability ofthe skeletal age determination method of Tanner and Whitehousefor forensic age diagnostics. Int J Legal Med 122:309–314

30. Schmidt S, Nitz I, Schulz R, Tsokos M, Schmeling A (2009) Thedigital atlas of skeletal maturity by Gilsanz and Ratib: a suitablealternative for age estimation of living individuals in criminalproceedings? Int J Legal Med 123:489–494

31. Schmidt S, Fracasso T, Pfeiffer H, Schmeling A (2010)Skelettaltersbestimmung der Hand. Rechtsmedizin 20:475–482

32. Schmidt S, Schmeling A, Zwiesigk P, Pfeiffer H, Schulz R (2011)Sonographic evaluation of apophyseal ossification of the iliac crestin forensic age diagnostics in living individuals. Int J Legal Med125:271–276

33. Siegert F (1935) Atlas der normalen Ossifikation der menschlichenHand. Georg Thieme, Leipzig

34. Tanner JM, Whitehouse RH, Marshall WA, Healy MJR, GoldsteinH (1983) Assessment of skeletal maturity and prediction of adultheight (TW2 method), 2nd edn. Academic, London

35. Tanner JM, Oshman D, Bahhage F, Healy M (1997) Tanner-Whithouse bone age reference values for North American children.J Pediatr 131:34–40

36. Tanner JM, Healy MJR, Goldstein H, Cameron N (2001)Assessment of skeletal maturity and prediction of adult height(TW3 method). W.B. Saunders, London

37. Thevissen PW, Galiti D, Willems G (2012) Human dental ageestimation combining third molar(s) development and tooth mor-phological age predictors. Int J Legal Med 126:883–887

38. Tisè M, Mazzarini L, Fabrizzi G, Ferrante L, Giorgetti R,Tagliabracci A (2011) Applicability of Greulich and Pyle methodfor age assessment in forensic practice on an Italian sample. Int JLegal Med 125:411–416

39. Todd TW (1937) Atlas of skeletal maturation: 1 The hand. C.V.Mosby Comp, St. Louis

40. Thiemann H-H, Nitz I, Schmeling A (2006) Röntgenatlas dernormalen Hand im Kindesalter. Thieme, Stuttgart

41. Weber R (1978). Genauigkeit der Skelettalterbestimmungen undGrößenprognosen nach den Methoden von Greulich & Pyle sowieTanner & Whitehouse. Diss, Freie Universität, Berlin

42. Wirtz M, Caspar F (2002) Beurteilerübereinstimmung undBeurteilerreliabilität. Hogrefe-Verlag, Göttingen, pp 157–235

698 Int J Legal Med (2013) 127:691–698

Documents

Skeletal age determination of the hand: a comparison of methods