Critical Review of Bone Age Assessment Methods

Globally, age has become the benchmark for many social events such as alcohol consumption, marriage and employment. And with it, entails legal responsibilities and ramifications. With the rise in illegal immigrants from war-torn and impoverished nations into developing countries, an individual’s age becomes one of the most important factors in determining their subsequent treatment. An illegal immigrant coming into the country is without birth records, as a result of poorly maintained birth records or undocumented births from their country of origin (Mohammed, et al., 2015). As such, skeletal age is used instead to determine the chronological age, more specifically the biological maturity, of the individual.

The assessment of skeletal age is generally achieved by assessing the radiographs of hands and wrists because they contain many bones - 8 carpals, 5 metacarpals and 14 phalanges. It is the individual growth and degree of maturation of ossification centres in each bone, appearing at particular ages, that serves as the basis for skeletal age (De Sanctis, et al., 2014). Due to the dominance of right-handed individuals, the risk of injury to the right hand and wrist are increased. Therefore, the left hand and wrist are typically radiographed and assessed (Satoh, 2015). The Greulich-Pyle and Tanner-Whitehouse methods are commonly utilised for skeletal age assessment, internationally. However, these methods are not without their respective shortcomings. Hence, the development of other methods such as ultrasonography, MRI and automated computerised assessments, in recent years (Satoh, 2015).

This literature review will be evaluating each method.

Greulich-Pyle Method

Greulich-Pyle (GP) method is considered to be the standard when it comes to the estimation of age in modern children and adolescents around the world, due to its simpleness. Radiographs of the left hand and wrist are taken and compared to standardised radiographs within the GP atlas, to estimate an individual’s chronological age. The atlas is a chronological compilation of anthropometric data and radiographs of 1000 healthy Ohio children progressing through childhood. The longitudinal study collected data at intervals of 3 to 12 months, between 1931 and 1942. However, it was the 1959 edition that was separated into two series: one following female development and the other following male development (Tsehay, Afework, & Mesifin, 2017).

In spite of the commonality of utilising GP assessments, there has been a lot of controversy over the accuracy and reliability of its predictions due to methodology, mainly because it compares radiographs of other ethnic population samples to the reference population. In one study for medicolegal purposes, the chronological age of 4.5 to 9.4-year-old Pakistani children was found to be significantly underestimated, with the average difference between the skeletal and chronological ages for males and females being 15.78 and 6.65 months, respectively. The study contained 139 males and 81 females from Pakistan, all of which had their skeletal ages calculated according to the GP atlas and compared to their respective chronological ages (Mughal, Hassan, & Ahmed, 2014). Compared to the former study, the GP method also accurately predicted the chronological age of 535 children from an Italian sample population of another study, in particular, those aged between 7 to 9 and 10.4 to 11.5 years (Santoro, et al., 2012). For the latter study, this means that despite ethnic differences between the reference and contemporary populations, the analysis could be considered to be reliable. But this prediction could be explained by both populations being of Caucasian descent.

There are also many other issues that the GP method overlooks, such as factors that can influence the development of bone in the contemporary sample population, which are not reflected in the reference population. Factors such as the variation in bone development between subsequent generations due to differences in socioeconomic status (Hsieh, Liu, Tiu, Chen, & Jong, 2011) and whether individuals are in possession of endocrine diseases (Xing, Cheng, Wergedal, & Mohan, 2014).

Save your time!
We can take care of your essay

• Proper editing and formatting
• Free revision, title page, and bibliography
• Flexible prices and money-back guarantee

Place Order

Tanner-Whitehouse Method

Unlike the Greulich-Pyle method, there are two assessment systems in the Tanner-Whitehouse (TW) method: “RUS” (radio, ulna and selected metacarpals and phalanges) and ‘Carpal’, which evaluates all carpal bones except pisiform. The Tanner-Whitehouse 2 (TW2) method is an earlier version of the TW method which analysed the level of maturity of up to 20 regions of interest in specific bones. Radiographs of the contemporary population are taken, and a numerical score is assigned to particular stages of development for each bone. The sum of all scores produces a total maturity score that correlates to a skeletal age – separate for males and females. The reference population of the TW2 method were UK children possessing average socioeconomic status (Satoh, 2015).

In 2001, publication of the Tanner-Whitehouse 3 (TW3) method was introduced. While the new methodology maintained the two systems of assessment as in TW2, the reference population and radiographs for the systems was changed to those belonging to children from North America. Thereby, improving on the accuracy and reproducibility of the original TW and TW2 method (Ortega, et al., 2006). There are also reports of standardised TW methods, where longitudinal studies for different regions and populations are conducted and utilised as the new reference population, in place of the original TW reference population. And by doing so, most of the issues that the GP method overlooks is addressed, as the relationship between the total maturity score and the bone age is changed to become more suitable for that particular group. However, while this makes it more favourable for accuracy compared to the GP method, the TW does require more time to process and reproduce the complex data into a chronological age estimate (Zhang, et al., 2013). According to a 1994 study, the average time to perform a GP assessment was approximately six times quicker than a TW2 assessment (King, et al., 1994).

Ultrasonography

One the of major disadvantages present in both GP and TW methods is that individuals are exposed to ionising radiation for radiographs. Ultrasonography or ultrasound (US) is a less harmful alternative in its initial stages of clinical application. A 2009 study assessed the accuracy of skeletal age predications from the wrists of 100 children via US, and it was found that the linear correlation between the predictions by GP, TW and US methods varied greatly. The strongest correlation between the three methods was found in the ‘normal’ bone age group (80.0% to 86.1%). However, the US method overestimated delayed bone ages and underestimated advanced bone ages. Thereby, producing weaker correlations of 77.1% to 86.9% in ‘delayed’ bone age group and 62.2% to 81.1% in ‘advanced’ bone age group (Khan, Miller, Hoggard, Somani, & Sarafoglou, 2009). Therefore, despite sufficiently providing an assessment for skeletal age estimation, the correlations between the US and GP and TW methods are not accurate or valid enough to become a reliable alternative.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) is another method that does not require exposure to ionising radiation. A 2014 preliminary study found that a strong linear correlation existed between MRI predications and the chronological ages of 179 individuals. The methodology of the study consisted of T1-weighted MRI scans of the hand and wrist being collected and evaluated by two blinded radiologist. Depending on various factors such as cartilage appearance and vacuolisation, provisional calcification, progression of and completion of ossification, a correlating skeletal age prediction would be produced (Tomei, et al., 2014). The only problem with MRI assessments, despite strong correlational results, would be that MRI equipment are not readily accessible and is relative more expensive than radiographic methods.

Automated Computerised Assessments

Automated skeletal age assessments by a software called BoneXpert is recently being adopted by many hospitals. One of the major problems that automated assessments are able to solve is the variability in manual skeletal age predictions, as it highly dependent on the individual performing the rating. However, for this method, the radiographs of abnormal bone morphologies are rejected. The BoneXpert method first reconstructs the borders of 15 bones from radiographs of the patient’s hand, of which 13 bones are automatically assigned with intrinsic bones ages. The intrinsic bone ages are not only dependent on various scores derived from preliminary analysis, but also the consensus bone age concept that defines the best bone age estimation for each bone. This is possible due to the software being capable of containing a large database of hand radiographs from children of varying ages. Thereby producing for a common skeletal age model – separate for males and females. The intrinsic bones ages are then converted into GP and TW bone ages (Thodberg, Kreiborg, Juul, & Pedersen, 2009).

As the BoneXpert method is dependent on its database, it is also possible to eliminate the issues with ethnic differences, which are present in the GP method. In a 2010 study, a large database of 1380 hand radiographs of children with varying ethnic backgrounds were analysed using BoneXpert. It was found that the root-mean-square (RMS) deviation between two manually-rated bone ages were approximately 7.5 months, and the RMS deviation between the automated bone age and the average of the manually-rated bone ages was 7.3 months (Thodberg & Sävendahl, Validation and reference values of automated bone age determination for four ethnicities., 2010). These results suggest that utilisation of BoneXpert is a valid and reliable alternative.