Han K. Huang DSC, Presenter: Nothing to Disclose

James Sayre PhD, Abstract Co-Author: Nothing to Disclose

Vicente Gilsanz MD, PhD, Abstract Co-Author: Nothing to Disclose

Ewa Pietka PhD, Abstract Co-Author: Nothing to Disclose

Bone age assessment (BAA) is a frequent procedure performed on pediatric patients to evaluate growth and is key for growth disease diagnosis/management. Most common method for BAA is matching by a left wrist radiograph against a small reference set atlas of normal standards developed by Greulich & Pyle (G&P) in the 1950s from data collected by the Brush Foundation Study. Due to increasing racial diversity and changing nutrition/behavior, the G&P method may not accurately reflect skeletal development for today's children. This study compares the difference in BAA using G&P method with a large set of reacially diverse normal hand images of present-day children.

We have developed a digital hand atlas with 1,080 images collected from boys and girls of European, African, Hispanic, and Asian descent at the Childrens Hospital of Los Angeles; 5 images per age group of pre-pubertal children and 10 images during puberty of uniform distribution. All relevant patient information have also been collected and stored along with the digital images. The digital atlas contains features of 3 fingers including bone growth regions and phalange geometric parameters.

Two pediatric Radiologists performed BAA. There are two significant results: 1) The collected normal data was validated by three existing standards. 2) A difference of approximately 10 months lower in bone age was noted between both the digital atlas mean and the G&P mean over the 5-14 year old children as compared to the children of the Brush Foundation Studies. Plots and tables of various comparisons among ethnic origin and gender will be exhibited.

This discovery of the differing bone age assessments is puzzling since better nutrition and living environments today should produce the opposite effect. Possible explanations for these findings are related to at least one of four factors: true population change and ethnicity; differences in patient recruitment methods; quality of image presentation; and quality/training of the readers. This study will be further continued by enlarging the sample sizes and increasing statistical power and considering regional patient distribution effect which may have skewed analysis.

Huang, H, Sayre, J, Gilsanz, V, Pietka, E, How Accurate is the Greulich and Pyle Atlas for Bone Age Assessment of Children Today?.  Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL. http://archive.rsna.org/2004/4412039.html