Keywords:
Bones, Paediatric, CAD, Digital radiography, Diagnostic procedure, Computer Applications-Detection, diagnosis, Metabolic disorders
Authors:
J. Song1, X. Fang1, Z. Yin2, Z. Xing2, P. Gong2, X. Li2, Y. Yu2, C. Gao1; 1Wuxi/CN, 2Beijing/CN
DOI:
10.26044/ecr2019/C-2335
Aims and objectives
Radiographic bone age assessment was commonly used in the evaluation of pediatric endocrinology problems,
children’s growth and genetic disorders[1], in which the patient's chronologic age was compared with their level of skeletal maturity based on a standardized reference.
At present,
either the atlas matching method of Greulich and Pyle [2–4] or the scoring assigning method of Tanner–Whitehouse (TW) [5,6] are the typical approaches used for bone age assessment in clinics,
by comparing a radiograph of the hand and wrist to an age-based atlas,
or determining age based on scoring specific radiographic features.
In order to be more suitable for contemporary Chinese children's bone age assessment,Shaoyan Zhang et al[7],in accordance with TW3 method[6],using the bone developmental level score table of the TW3 method and increasing several new skeletal maturity indications,
without changing the start and stop timeline of the TW3 bone development level,established the China 05 method,
including TW3-Chinese RUS (TW3-C RUS) and TW3-Chinese Capral (TW3-C Capral) and RUS-CHN,
based on a sample of contemporary Chinese children.
Compared with the atlas matching method,
the scoring assigning methods such as TW3 method and the China 05 method,
are subjective and qualitative. However,
the scoring assignment methods showed several limitations,
such as the complicated evaluation system,
much longer learning cycle,
highly skilled characteristics,
and more time-consumption,
which made them hard to widely apply in clinics.
To settle these problems,
in this study,
we have built an automatic artificial intelligence bone age assessment system based on the scoring assignment method of China 05.
The system adopted the latest deep learning technology and had the advantage of excellent accuracy and high-efficiency,
which made it an assistant and convenient new tool of children’s bone age assessment for pediatricians.