Bone Age-automatic Evaluation System Based On Active Contour Model Of The Edge Of The Wrist Bone Extract

The value of the maturity of bones growth has been wildly used in the area of defending medicine, clinical medicine, physical training, and justice, etc. It has become a novel problem in the medicine area. The assessment of skeletal age is drawing data of skeletal age through the observing the hand-wrist photography. The commonly used method in our country is CHN. The standard of CHN method is one which has been putted forward based on the assessment of Chinese hand-wrist skeletal in 1992. It draws a conclusion through the observation the maturity of the fourteen bones of enfant hand-wrist.Today many experts have been devoted into this research area, associating the computer vision with the assessment of skeletal age, and realizing the automation of the assessment of skeletal age, to make the process more effective and the result more reliable. The computer-aided skeletal age score system is a system disposing hand-wrist photography, namely, to analysis a hand-wrist photography based on the CHN method. To build a whole system we face many troubles, so we made slow progress, and have many problems need to be solved. The segmentation of bones and the extraction of hand-wrist features are the greatest obstacles in the development of the automation assessment system. The CHN method deal with fourteen bones of hand-wrist, so firstly we need to segment each bone from whole image, namely to locate the bones. In CHN method, classification mainly depends on the contour of bones, so before we could do the next work, we should get the contour of each bones accurately.The extraction of bones is related to the segmentation of bones. Image segmentation is one of the most important technology in the image processing. It has been researched for decades, and thousands kinds of algorithms hadbeen putted forward, but there is not still a general theory to treat with this problem. Active Contour Model is a method interactive between human and machine brought forward by Kass in 1987, it's a contour extract model from top to bottom. This model takes advantage of the advance recognition ability of human, and focus on the neighborhood of the contour, so it elevates the reliability and accuracy of the extraction of contour. Compared with the traditional method, it has obvious advantages. For decades, many experts devoted into this area, and brought forward many optimization methods from aspects of the problem, such as the expression of model, the energy function, and the optimization methods, so, this method has become more and more better, and it has been applied in wide fields.There are many questions in hand-wrist photography, such as the lower contrast, unevenly lighting, the effect of muscle and parenchyma, the irregular of contour, etc. The traditional edge detecting method can't get accurate edge of bones, for this reason, this paper use the ASM to detect the edge of bones, and make some improvements, it get ideal result in our tests.This paper has made some discussion and attempt on the question of how to apply the ASM to the automation assessment of skeletal age system, and made a good foundation for the next work.