Skeletal age, referred to as bone age, that is, the growth cycle of the bones, many people's physiological age and the age of the bones are different differences.
But the age of the bones can not be identified with the naked eye, can only be used to determine the age of the bone age meter film. The wrist is the best place to look, as it contains a large number of long, short, and round bones, reflecting the growth and maturity of the entire body, and is easy to use and accurate.
Introduction of bone age assessment method
In China, bone age assessment was introduced into the sports world after 1979 with the gradual development of scientific selection of athletes. In the beginning, the American G-P slice method was used (Greulich-Greulich and Pyle-Pyle, professors of anatomy at Stanford Medical University. They developed a standardized atlas of bone age films by taking 2-21 x-rays each of more than 1000 children from upper-class, affluent families in the United States under continuous observation from 1931 to 1942). Evaluation of bone age using G-P standardized films is a qualitative method, i.e., comparing several standardized films to determine the age, which is subject to error by the inexperienced because it involves looking at the degree of development of more than 20 bones in a single film.
Later, the British TW2 method (Tanner-Whitehouse) was introduced, which is a method of bone age assessment based on a sample of middle-class children in the United Kingdom. The degree of development of 20 wrist bones was evaluated, and then the bone age was calculated by scoring. It is a quantitative method. In 1985, when China and Japan collaborated on a study of the physical fitness of children and adolescents, the bone age of adolescents and adolescents in Tokyo and Beijing was assessed by this method.
In 1979, Prof. Li Guozhen of China's medical profession developed the first Chinese bone age percentage counting method based on the hand X-rays of 1,936 children and adolescents aged 0-18 years in Beijing.
In 1987-88, China's National Sports Commission organized a study of Chinese bone age assessment standards, in each of the country's six regions, a city (Harbin, Shijiazhuang, Xi'an, Changsha, Chongqing, Fuzhou)*** more than 20,000 children and adolescents aged 0-19 years of age, wrist bone age film, and on the basis of the formulation of the "Chinese wrist bone development standards CHH Chinese wrist bone development standard CHN method". 1992 was approved as the Chinese people's *** and the national industry standard officially announced and implemented.
CHN method
Wrist bone X-ray filming: left hand five fingers apart, palm down close to the X-ray box, hand straight and forearm into a straight line, X-ray lens on the third metacarpal bone, from the back of the hand of about 90 centimeters. X-rays in addition to all the hand, the wrist bone, should also include the distal end of the radius bone 3-4 centimeters (CHN method stipulates that the shooting of the left hand, the United States has studied the 450 left and right hand of the In the United States, 450 people were studied for left and right hand bone age, of which only 13 people, or 2.9%, did not have the same bone age).
Assessment: The CHN method*** assesses the development of 14 bones. They were: radius, 1,3,5 metacarpals, 1,3,5 proximal phalanges, 3,5 middle phalanges, 1,3,5 distal phalanges, capitellum, and hook bone. The radius is divided into 10 grades, the capitate into 7 grades, and the other 12 bones into 8 grades. Each bone has a different score for each grade. The age of the bones can be determined by adding up the scores of the 14 bones and checking the table. Now, by using the computerized evaluation software, the age of the bones can be evaluated automatically by entering the grade of each bone. After systematic training and repeated practice, and after accurately mastering the criteria for distinguishing the grades of each bone, the error in the evaluation of films by the CHN method is not large.
Bone development process: (take radius as an example)
In the newborn, the epiphyseal part is still cartilage, then with the age, the center of ossification begins to appear, and the center of ossification grows gradually, and develops to different stages, showing different shapes, and finally, the epiphyseal part is completely ossified, and fuses with the diaphysis as a whole. The different shapes of the ossification centers are used as the basis for grading bone development in the assessment of bone age.
Before ossification, the epiphysis is transparent on X-ray.
The process includes the following four steps:
1. Image pre-processing: filtering the image in order to improve the speed and accuracy of image recognition.
The image pre-processing process is a filtering process of the image, to eliminate interference to retain the parts that need to be processed, and filter out the unwanted parts. As the imaging device reads in the hand bone image will expand the field of view, which contains a lot of invalid pixel points, which will affect the later processing, increasing the amount of computation. Therefore, the first task is to remove as accurately as possible the portion of the image that contains the object being measured from the read-in image. The main difficulty of this process is to determine whether the image belongs to the part that needs to be retained, and must ensure its accuracy, otherwise it will have a serious impact on the next step of image analysis and judgment.
This process is mainly based on the thresholding method, because the x-light source is a parallel beam of light obtained by focusing through a lens, the center part of the circle is strong, and the circumference part of the light is weak, or even very weak, which brings us a lot of trouble in judging the threshold. The solution is to divide the circle into a number of concentric circles, take the threshold value on different arcs to filter the background, thus obtaining a pre-processed image.
In the preprocessing process, we also judged the quality of the image, such as judging whether the image is too blurred, whether the framing range of the image is too far, whether the bones to be judged are complete.
2. Localization of the image: Determine the approximate location or extent of the bones that are important for determining bone age.
Because the bone development is mainly affected by the epiphyses between the phalanges and the finger bones and between the metacarpal bones, such as the flexor epiphyses, ulnar epiphyses, metacarpal epiphyses, and the epiphyses of each finger, etc., therefore, the judgment of the age of the test subject is mainly based on the analysis and judgment of the specific bones on the palm of the hand, therefore, the first and foremost task is to determine the exact location of the bones of the hand to be analyzed after the image preprocessing.
For the image whose main content is the part of the finger bones, considering that in the image of the x-ray film, the grayscale of each finger bone position is always darker relative to the color of the other positions, therefore, after weighted average and median filter smoothing, we get the values of the coordinates of the valley points of several groups of darker points (i.e., the grayscale value) of a smaller value, and then, based on the shape of the finger, we use the equation of a straight line or a curve fitted to it, so that the straight and curved lines pass roughly through the center of each finger.
Then, because in each finger, the location between the phalanges and the phalanges, where the phalangeal epiphyses are located, is brighter relative to the rest of the finger, i.e., the gray value is higher, the gray value is taken in turn on the fitted straight line or curve, and after smoothing, the peaks with the larger values are found as the locations where the several phalangeal epiphyses are located. If the number of epiphyseal points is insufficient, the approximate positions of the missed epiphyseal points can be calculated from the positions of the other epiphyseal points according to the approximate ratio of the length of the knuckle, which was tested to be not more than 5% deviation.
The localization of metacarpal and metacarpal epiphyses is therefore more difficult than the localization of phalangeal epiphyses. After analyzing a large number of images, we found a pattern. Based on the flexor and ulna, which occupy the main part of the lower part of the image, we can determine the upper, lower, left and right edges of the flexor and ulna epiphyses to be analyzed, and thus determine the location and approximate range of the flexor and ulna epiphyses. The location of the hook bone, the capitellum, and the metacarpal epiphysis of the thumb can be determined from the extension of the fitting curves of the individual phalanges, respectively. The centers of these known metacarpal bones are then extended outward in a certain angular range, and the locations of the metacarpal bones such as the triangular bone, the lunate bone, the large and small polygonal bones, and the navicular bone are then determined.
3. Characterization of the image: the bones that have been localized to extract their various shape features.
Based on the actual needs of image analysis, we extract the features and analyze the main aspects of these bones are the presence or absence of relative size and ratio, shape, area comparison, the degree of healing and so on.
Each phalangeal epiphysis was first analyzed. After determining its position, the smaller individual epiphyses in the original image are first enlarged by a zoom-in algorithm, with some rotation of the more oblique bones. Then in the enlarged and rotated image, the sobel edge detection operator is used to separate the epiphysis from the diaphysis, the width ratio of the epiphysis to the diaphysis is calculated, as well as the algorithm of contour line tracing is utilized to obtain an array of data on the shape contours of some of the finger epiphyses that are emphasized in the analysis.
For the analysis of the metacarpal part, we mainly determine whether some metacarpal bones exist or not. Then after segmenting and enlarging the metacarpal region of some key analyzed metacarpals, we get their size ratio, contour shape, etc., and match them with the standard template to get the closest results. For the metacarpal bone images with large bone age, since some bones have begun or completely healed, it is necessary to analyze and judge some residual traces at the healing place, so as to get the accurate bone development.
4. Comprehensive judgment: Based on the shape and characteristics of each bone, the age of the subject is judged comprehensively.
First of all, the ratio and healing of the epiphysis of the finger bone are used to determine the approximate range of the subject's bone age. Then a comprehensive judgment is made based on the metacarpal bones.
It will cost a few thousand dollars