Artificial Intelligence Theory and Applications, vol.2, pp.308-318, 2021 (Peer-Reviewed Journal)
Bone age is an effective indicator for diagnosing various diseases and to
determine bone ages of livings. The earliest well-known studies belong to the
Greulich-Pyle and Tanner-Whitehouse, as a result bone age development
atlases were published using hand and wrist radiography. Atlases works well
for the younger ages between 0-18, while they have deviations at elder ages.
Kazuro Anhara and Takao Suzuki emphasized the importance of changes in
pubic symphysis of pubic bones belonging to 20 to 40 years old cases who
were not alive for further ages. All this researches focuses on the hand intensive
works. However, automation of bone age detection using artificial intelligence
techniques such as image processing of radiological images is important in
order to prevent human side-effects on the evaluation, they are called
automated methods. Some examples are automatic bone age estimation fully
automatic with carpal bone segmentation using fuzzy classification, fuzzybased radius model for bone age estimation including image preprocessing,
and neural network applications mostly seen on the literature. It is obvious,
artificial intelligence promises faster bone age estimation and to minimize
different evaluations between experts. However, new studies are needed for
applying new techniques (deep learning) efficiently and discovering new bones
to estimate elder ages accurately in the field of forensic informatics especially.