A method and device for assessing adolescent elbow bone age using lateral elbow X-ray images and an artificial intelligence model

Abstract

The present invention is a bone age assessment method executed by a bone age assessment device, which includes: (a) acquiring an X-ray image of the lateral side of the elbow; (b) dividing the olecranon region from the X-ray image; (c) analyzing the morphological characteristics of the olecranon ossification center in the divided olecranon region; and (d) determining bone age based on the analyzed morphological characteristics.

Description

[Technical Field]

[0001] The present invention relates to a bone age assessment technique, and more particularly to a bone age assessment method and apparatus using lateral X-ray images of the elbow of adolescent children. [Background technology]

[0002] Bone age assessment in children and adolescents is important for accurately assessing a patient's skeletal maturity, which may differ from the patient's chronological age. The most widely used method for bone age assessment is hand and carpal age determination, but this method has drawbacks: it is inaccurate during adolescence, has a large age gap, and its clinical applicability is limited. Therefore, elbow radiographs are widely used to assess bone age in adolescents.

[0003] The first two years of adolescence are the acceleration phase, during which rapid growth spurts occur. Peak height velocity (PHV), the period of greatest height growth during adolescence, and sexual maturity at Tanner stage 2 are indicators of this phase. During this period, the ossification process of the olecranon epiphyseal plate undergoes very distinctive morphological changes. As the ossification process of the olecranon epiphyseal plate is completed, the acceleration phase of early adolescence ends and the deceleration phase of late adolescence begins. Therefore, assessing bone age based on the rapidly changing ossification process of the olecranon epiphyseal plate allows for accurate assessment of the rapid growth period of early adolescence.

[0004] Existing methods for assessing elbow bone age that utilize the ossification process of the olecranon epiphyseal line include the Sauvegrain et al. method and the Dimeglio et al. method. The Sauvegrain et al. method is the most traditionally used, but it requires two X-ray images, one for the anterior-posterior and one for the lateral side of the elbow, resulting in high radiation exposure. It also has drawbacks, such as a complex interpretation method, as bone age is determined by assessing the morphology of each part of the elbow and adding up the scores. The Dimeglio et al. method requires only one X-ray image for the lateral side of the elbow, but it cannot reflect the diversity of the ossification process of the olecranon epiphyseal line, resulting in limitations and inaccuracy in its application in actual clinical practice. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Korean Patent Publication No. 10-2019-0142234 Summary of the Invention [Problem to be solved by the invention]

[0006] The objective of the present invention is to provide a method and device for assessing bone age that morphologically subdivides the ossification center formation and fusion stages observed during the maturation process of the olecranon epiphyseal line, which can be confirmed by lateral elbow X-rays during adolescence, and reflects the diverse and distinctive change processes.

[0007] The problems to be solved by the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description. [Means for solving the problem]

[0008] A first aspect of the present invention for achieving the above object is a bone age assessment method executed by a bone age assessment device, which includes: (a) acquiring an X-ray image of the lateral side of the elbow; (b) dividing the olecranon region in the X-ray image; (c) analyzing the morphological characteristics of the olecranon ossification center in the divided olecranon region; and (d) determining bone age based on the analyzed morphological characteristics.

[0009] Preferably, step (b) may include predicting a bone region in the X-ray image on a pixel-by-pixel basis, and segmenting a region corresponding to the olecranon from the X-ray image based on the result of the pixel-by-pixel prediction.

[0010] Preferably, step (c) may include determining an area for an olecranon ossification center in the olecranon region and determining morphological characteristics of the olecranon ossification center.

[0011] Preferably, step (d) may include classifying the olecranon bone age according to predetermined age-specific olecranon characteristic criteria based on the morphological characteristics of the olecranon ossification center.

[0012] Preferably, the pre-set olecranon characteristic standards for each age may be set in six-month increments from 9.5 to 13 years old for girls, and from 11.5 to 15 years old for boys.

[0013] Preferably, the pre-set olecranon characteristic standard by age may further include olecranon characteristic standards for 11.25 years old for girls and 13.25 years old for boys.

[0014] Preferably, the pre-set characteristic criteria of the olecranon by age may be classified into a primary ossification process and a secondary ossification process according to the order of fusion of the ossification center of the olecranon and the secondary ossification center.

[0015] Preferably, the first ossification process may comprise a small olecranon ossification center stage where the height of the olecranon body is 20% or more and less than 50%, an enlarged olecranon ossification center stage where the height of the olecranon body is 50% or more, an appearance stage of a secondary ossification center, a fusion stage of the olecranon ossification center and the secondary ossification center, an expansion stage of the fused ossification center, a partial fusion stage of less than 50% of the olecranon epiphysis, a partial fusion stage of more than 50% of the olecranon epiphysis, and a complete fusion stage of the olecranon epiphysis.

[0016] Preferably, the bone age assessment method may further include a stage in which another secondary ossification center appears after the fusion of the olecranon ossification center and the secondary ossification center between the stage in which the olecranon ossification center and the secondary ossification center fuse and the stage in which the fused ossification center expands.

[0017] Preferably, the second ossification process may comprise a small olecranon ossification center stage in which the height of the olecranon body is 20% or more and less than 50%, an enlarged olecranon ossification center stage in which the height of the olecranon body is 50% or more, an appearance stage of a secondary ossification center, a fusion stage of the secondary ossification center and the olecranon body, an expansion stage of the olecranon ossification center that is fused with the olecranon body and not fused, a partial fusion stage of the secondary ossification center that is fused with the olecranon ossification center and less than 50% of the olecranon body, a partial fusion stage of 50% or more of the olecranon epiphysis, and a complete fusion stage of the olecranon epiphysis.

[0018] Preferably, the bone age assessment method may further include a stage of appearance of another secondary ossification center after fusion of the olecranon ossification center and the secondary ossification center between the stage of fusion of the secondary ossification center and the body of the olecranon and the stage of expansion of the fused secondary ossification center-olecranon body and the unfused olecranon ossification center.

[0019] A second aspect of the present invention for achieving the above object is a bone age evaluation device that includes an image acquisition unit that acquires an X-ray image of the lateral side of the elbow, a region division unit that divides the olecranon region in the X-ray image, a morphological analysis unit that analyzes the morphological characteristics of the olecranon ossification center in the divided olecranon region, and a bone age determination unit that determines bone age based on the analyzed morphological characteristics.

[0020] Preferably, the region dividing unit may predict a bone region in the X-ray image on a pixel-by-pixel basis, and divide a region corresponding to the olecranon in the X-ray image based on the result of the pixel-by-pixel prediction.

[0021] Preferably, the morphological analysis unit may determine a region for an olecranon ossification center in the olecranon region and determine morphological characteristics of the olecranon ossification center.

[0022] Preferably, the bone age determining unit may classify the olecranon bone age according to a characteristic standard of the olecranon for each age that is preset based on the morphological characteristics of the olecranon ossification center.

[0023] A third aspect of the present invention to achieve the above object is characterized in that a bone age assessment method is performed in a computer program stored on a computer-readable recording medium, when instructions of the computer program are executed. [Effects of the Invention]

[0024] As described above, according to the present invention, since a single X-ray image corresponding to the lateral side of the elbow is used, it is possible to reduce radiation exposure during the examination and to broaden and subdivide the range of ages that can be determined.

[0025] In addition, the classification reflects the characteristic morphological changes in the olecranon epiphyseal line that can be observed in lateral elbow X-ray images, which has the effect of enabling more accurate bone age assessment in various clinical cases when determining the bone age of adolescent children and young people.

[0026] In addition, accurate pubertal bone age assessment can be used to evaluate the maturity of the patient's musculoskeletal system, which can play a key role in determining the timing of diagnosis and treatment for various endocrine disorders, short stature, fractures, scoliosis, or leg length inequality.

[0027] It also improves the work process for radiologists, making diagnosis easier and more intuitive, and allows for the use of AI programs to assist radiologists, pediatricians, or orthopedic surgeons with their work and provide faster interpretation. [Brief explanation of the drawings]

[0028] [Figure 1] 1 is a block diagram of a bone age assessment method according to a preferred embodiment of the present invention. [Figure 2] 1 is a flowchart illustrating a bone age assessment method according to one embodiment. [Figure 3] 1 is an exemplary diagram illustrating a bone age assessment method according to an embodiment; [Figure 4] FIG. 10 is an exemplary diagram illustrating division of an olecranon region according to one embodiment. [Figure 5] FIG. 10 is an exemplary diagram illustrating the analysis of morphological features of ossification centers according to one embodiment. [Figure 6] FIG. 1 is an exemplary diagram illustrating a deep learning algorithm according to an embodiment. [Figure 7] FIG. 1 is an exemplary diagram illustrating a deep learning algorithm according to an embodiment. [Figure 8] FIG. 1 is an exemplary diagram for explaining determination of bone age according to one embodiment. [Figure 9] FIG. 1 is an exemplary diagram for explaining determination of bone age according to one embodiment. [Figure 10] FIG. 1 is a diagram for explaining the performance of a bone age assessment method according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0029] The advantages and features of the present invention, as well as methods for achieving them, will become clearer with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, and can be realized in various different forms. The present embodiments are provided merely to complete the disclosure of the present invention and fully convey the scope of the invention to those skilled in the art to which the present invention pertains. The present invention is defined only by the scope of the claims. The same reference symbols refer to the same elements throughout the specification. The term "and / or" includes each and every combination of one or more of the referenced items.

[0030] Even if terms such as "first," "second," etc. are used to describe various elements, components, and / or sections, it is understood that these elements, components, and / or sections are not limited by these terms. These terms are used merely to distinguish one element, component, or section from another element, component, or section. Therefore, it is understood that a first element, first component, or first section referred to below may also be a second element, second component, or second section within the technical spirit of the present invention.

[0031] Furthermore, in each step, identification symbols (e.g., a, b, c, etc.) are used for convenience of explanation, and the identification symbols do not describe the order of each step, and each step may be performed in a different order from that specified unless a specific order is clearly stated in the context. That is, each step may be performed in the same order as specified, may be performed substantially simultaneously, or may be performed in the opposite order.

[0032] The terms used in this specification are for the purpose of describing embodiments and are not intended to limit the present invention. In this specification, the singular form includes the plural form unless the context clearly dictates otherwise. When used in this specification, the words "comprises" and / or "comprising" do not exclude the presence or addition of one or more other components, steps, operations and / or elements to a referenced configuration, step, operation and / or element.

[0033] Unless otherwise defined, all terms (including technical and scientific terms) used herein are used in the sense that they can be commonly understood by those skilled in the art to which the present invention pertains. Furthermore, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless they are clearly and specifically defined.

[0034] Furthermore, when describing embodiments of the present invention, if it is determined that a detailed description of a known function or configuration may unnecessarily obscure the gist of the present invention, the detailed description will be omitted. Furthermore, the terms used below are defined in consideration of the functions in the embodiments of the present invention, and may vary depending on the intentions or practices of a user or operator. Therefore, the definitions should be based on the contents of this specification as a whole.

[0035] FIG. 1 is a block diagram showing a bone age evaluation device according to a preferred embodiment of the present invention.

[0036] The bone age assessment device 100 is a device for performing the bone age assessment method according to the present invention, and assesses bone age by receiving a partial lateral X-ray image of an elbow and analyzing the shape of the bone. That is, the bone age assessment device 100 assesses the bone age of adolescent children and young people by utilizing the characteristics of the olecranon ossification process, which is characteristic of the acceleration phase, which corresponds to the first two years of adolescence.

[0037] Preferably, the bone age evaluation device 100 is a computer on which an application or program for executing the bone age evaluation method can be installed and executed, and is equipped with a user interface and capable of controlling data input and output. Here, the term "computer" refers to any type of hardware device including at least one processor, and may also encompass software configurations operating on corresponding hardware devices according to embodiments. For example, the term "computer" may be understood to include, but is not limited to, smartphones, tablet PCs, desktops, laptops, and user clients and applications running on each device.

[0038] 1, the bone age evaluation device 100 includes an image acquisition unit 110, an area division unit 120, a morphological analysis unit 130, a bone age determination unit 140, and a control unit 150. The control unit 150 controls the operations and data flow of the image acquisition unit 110, the area division unit 120, the morphological analysis unit 130, and the bone age determination unit 140.

[0039] The image acquiring unit 110 acquires an X-ray image. Preferably, the X-ray image may be captured by a separate imaging device connected to the bone age evaluation device 100 by wire or wirelessly and transmitted to the bone age evaluation device 100, or may be captured by an imaging device included in the bone age evaluation device 100 and transmitted to the image acquiring unit 110.

[0040] The region dividing unit 120 divides the X-ray image acquired by the image acquiring unit 110 into an olecranon region, which is a region necessary for determining bone age. Preferably, various deep learning algorithms may be applied to the method of dividing the olecranon region.

[0041] The morphological analysis unit 130 analyzes the morphological characteristics of the bones in the olecranon region. Preferably, the morphological analysis unit 130 may use a deep learning algorithm to determine the region for the olecranon ossification center and analyze the morphological characteristics of each bone that constitutes the olecranon ossification center.

[0042] The bone age determining unit 140 determines the bone age based on the morphological characteristics of the bone. Preferably, the bone age determining unit 140 may determine the bone age based on a bone age determination criterion that matches the morphological characteristics of the bone with the bone age set in advance.

[0043] The operations performed by each component of the bone age assessment device 100 shown in Fig. 1 will be described in detail below with reference to Fig. 2. Although each step described with reference to Fig. 2 is described as being performed by different components, this is not limiting, and at least some of the steps may be the same or may be performed by different components depending on the embodiment.

[0044] FIG. 2 is a flowchart illustrating a bone age assessment method according to one embodiment.

[0045] Referring to Fig. 2, the image acquisition unit 110 acquires an X-ray image of the lateral side of the elbow (step S210). Preferably, referring to Fig. 3, the image acquisition unit 110 may preprocess the X-ray image. That is, the image acquisition unit 110 may remove unnecessary areas from the X-ray image, recognize and select a lateral view, and perform normalization.

[0046] The region segmentation unit 120 segments the olecranon region from the X-ray image (step S220). Preferably, the region segmentation unit 120 may apply a deep learning algorithm to the X-ray image to predict a bone region on a pixel-by-pixel basis and segment the X-ray image into a region corresponding to the olecranon based on the pixel-by-pixel prediction result. More specifically, referring to FIG. 3, the region segmentation unit 120 may apply a deep learning algorithm, YOLOv5, to the X-ray image, which is the input image, to detect a bone portion corresponding to the olecranon. The region segmentation unit 120 may apply a YOLOv5 network to automatically crop the olecranon region from the X-ray image and set the image of the cropped olecranon region as a region of interest (ROI). For example, referring to FIG. 4, (a) in FIG. 4 corresponds to the original X-ray image, and (b) corresponds to the image segmented for the olecranon region. The region segmentation unit 120 may apply YOLOv5, the first step of the cascade model, to an X-ray image such as (a) to detect the bone portion corresponding to the olecranon, and then segment only the region corresponding to the olecranon to obtain an image such as (b). That is, the region segmentation unit 120 may crop only the olecranon region from the X-ray image via the YOLOv5 network and extract the cropped olecranon region as a region of interest. Here, step S230, described below, may be performed based on the extracted region of interest.

[0047] In one embodiment, the region segmentation unit 120 may apply YOLOv5 to an X-ray image, which is an input image, to detect a region corresponding to the olecranon, and may segment the region corresponding to the olecranon using EfficientDet. Here, EfficientDet is a deep learning algorithm improved based on the backbone of the EfficientNet model, and may have an additional segmentation layer, which may be used by the region segmentation unit 120 to segment the olecranon region.

[0048] The morphological analysis unit 130 analyzes the morphological features of the olecranon ossification center in the segmented olecranon region (step S230). Preferably, the morphological analysis unit 130 may apply a deep learning algorithm to the olecranon region to determine the region for the olecranon ossification center and determine the morphological features of the olecranon ossification center. Referring to FIG. 3, the morphological analysis unit 130 may classify the morphological features of the olecranon ossification center using EfficientDet, a cascaded deep learning algorithm. That is, the morphological analysis unit 130 may apply EfficientDet to the segmented olecranon region to classify the olecranon region and the bony prominence. Here, EfficientDet is a network that achieves high efficiency and accuracy using complex scaling that simultaneously adjusts depth, width, and resolution, and is advantageous for accurately identifying objects for elbow shapes that vary from person to person. For example, referring to Figure 5, (a) of Figure 5 corresponds to the olecranon region acquired through the region dividing unit 120, and (b) of Figure 5 corresponds to the region relative to the olecranon ossification center, which serves as a reference for determining the morphological characteristics of the olecranon ossification center. The morphological analysis unit 130 applies the deep learning algorithm EfficientDet to the olecranon region as shown in (a) to determine the region relative to the olecranon ossification center of the olecranon region as shown in (b) and determine the morphological characteristics of the region relative to the olecranon ossification center. That is, the deep learning algorithm EfficientDet analyzes the morphological characteristics based on the highlighted region shown in (b) of Figure 5.

[0049] More specifically, EfficientDet, a deep learning algorithm applied to the region segmentation unit 120, the morphological analysis unit 130, and the bone age determination unit 140 described below, is configured to suit the present invention by incorporating the backbone of an EfficientNet model. As shown in FIG. 6, the Efficient-Net located on the left side is used as the backbone, and the backbone EfficientNet is applied as b4 (i.e., EfficientNet-b4). In the final stage, four Bi-directional Feature Pyramid Network (BiFPN) layers are applied as feature networks. The box network is removed from the class network that classifies object classes and the box network that generates bounding boxes that determine the position and size of objects, and a segmentation layer is added. Here, b4 of EfficientNet-b4 is assigned a number depending on the layer depth, such as b1, b2, or b4, and the layer depth may be applied or changed in various ways. This deep learning algorithm adds segmentation prediction logit, which predicts X-ray images pixel by pixel, and segmentation prediction logit can predict bone areas from X-ray images pixel by pixel. Using this improved EfficientDet has the advantage of lowering the error rate and allowing for faster calculations because it eliminates the need to check each box.

[0050] Referring to Figure 7, Efficient-Net, which is the backbone structure of EfficientDet and is used as the backbone network, consists of multiple convolutional layers (Convolutional (Conv) layers), MBconv blocks, conv1x1 blocks, and pooling and fully connected (FC) layers. Conv3x3 stacks one convolutional layer with a 3x3 kernel using 32 channels, followed by one MBConv1 block with a 3x3 kernel and 16 channels, two MBConv6 blocks with a 3x3 kernel and 24 channels, two MBConv6 blocks with a 5x5 kernel and 40 channels, and three MBConv6 blocks with a 3x3 kernel and 80 channels. Here, MBConv6 further performs depth-related batch normalization and a swish process in MBconv. Finally, a fully connected layer consisting of a convolutional layer using a 1x1 kernel, a pooling layer, and a dense layer is stacked. Because the data input to the deep learning algorithm is a 2D X-ray image rather than a 3D image with z-axis depth, layers may be stacked in four units, such as Input → P1 / 2 → P2 / 4 → P3 / 8 → P4 / 16 → P5 / 32, and various scales of features may be integrated through a BiFPN layer to improve object detection accuracy. Segmentation logits are applied to layers passing through the BiFPN layer, corresponding to a network structure in which segmentation of the elbow region and classification of that region are performed in the final stage. Here, layers may be stacked deeper as needed. Here, segmentation corresponds to identifying the exact boundary of the olecranon region through the region segmentation unit 120 and separating the corresponding region from other regions. The region segmentation unit 120 may segment the olecranon region using EfficientDet, and classification corresponds to bone age determination performed through the bone age determination unit 140, described below.

[0051] The bone age determination unit 140 determines the bone age based on the morphological characteristics of the olecranon ossification center analyzed in step S230 (step S240). Preferably, the bone age determination unit 140 may apply a deep learning algorithm, EfficientDet, to classify the olecranon bone age according to pre-set olecranon characteristic criteria for each age based on the morphological characteristics of the olecranon ossification center. Here, the deep learning algorithm is the same as the algorithm applied in step S230.

[0052] Preferably, the preset olecranon characteristic standards for each age may be set in six-month increments from 9.5 to 13 years old for girls and from 11.5 to 15 years old for boys. Here, olecranon characteristic standards for 11.25 years old, which is the interval between 11 and 11.5 years old for girls, and for 13.25 years old, which is the interval between 13 and 13.5 years old for boys, may be further set. Referring to FIG. 8 , a preset age-specific olecranon characteristic standard table shows the characteristics and shapes of the olecranon for girls (F) and boys (M) by age. For 11-year-old girls and 13-year-old boys, 11.5-year-old girls and 13.5-year-old boys, and 12-year-old girls and 14-year-old boys, two shapes corresponding to stages D and D', L and L', and G and G' may appear, respectively. These shapes correspond to the order of fusion in the ossification process.

[0053] More specifically, the pre-set characteristic criteria for the olecranon by age may be classified into the first ossification process and the second ossification process according to the order of fusion between the ossification center of the olecranon and the secondary ossification center. That is, if the ossification center of the olecranon and the secondary ossification center fuse first after the appearance of the secondary ossification center in the ossification process, the first ossification process can be achieved, and if the secondary ossification center and the body of the olecranon fuse first, the second ossification process can be achieved.

[0054] Referring to FIG. 9, the first ossification process is as follows: - Stage O2, which corresponds to 9.5 years old girls and 11.5 years old boys (A small ossification center (20-50% height of the olecranon body)). - Stage O3 (An enlarged olecranon ossification center (>50% height of olecranon body)) for girls aged 10 years and boys aged 12 years. - Stage B (appearance of an accessory ossification center (two ossification centers)) for girls aged 10.5 years and boys aged 12.5 years - Stage D (fusion of two ossification centers (a half-moon shape)) for girls aged 11 and boys aged 13. - Stage L (Enlargement of fused ossification centers (a rectangular shape)), which corresponds to 11.5 years of age for girls and 13.5 years of age for boys. - Stage G (partial fusion (<50%) of the olecranon apophysis) for girls aged 12 years and boys aged 14 years. Stage H (partial fusion (>50%) of the olecranon apophysis) for girls 12.5 years of age and boys 14.5 years of age, and -It can consist of stage F (Complete fusion of the olecranon apophysis), which corresponds to 13 years of age for girls and 15 years of age for boys.

[0055] Next, referring to FIG. 9, the second ossification process is as follows: - Stage O2, which corresponds to 9.5 years old girls and 11.5 years old boys (A small ossification center (20-50% height of the olecranon body)). - Stage O3 (enlarged olecranon ossification center (>50% height of olecranon body)) for girls aged 10 years and boys aged 12 years. - Stage B (appearance of an accessory ossification center (two ossification centers)) for girls aged 10.5 years and boys aged 12.5 years. -D' stage (fusion of the accessory ossification center and the olecranon body) for girls aged 11 and boys aged 13. - L' stage (Enlargement of unfused olecranon ossification center and fused accessory ossification center-olecranon body) for girls aged 11.5 years and boys aged 13.5 years. - Stage G', which applies to 12-year-old girls and 14-year-old boys (Partial fusion (<50%) of fused accessory ossification center - olecranon body and olecranon ossification center). *- Stage H (partial fusion (>50%) of the olecranon apophysis), which corresponds to 12.5 years of age for girls and 14.5 years of age for boys, and -It can consist of stage F (Complete fusion of the olecranon apophysis), which corresponds to 13 years of age for girls and 15 years of age for boys.

[0056] Here, between the D and L stages of the first ossification process and the D' and L' stages of the second ossification process, there is a further stage, BD, which corresponds to 11.25 years of age in girls and 13.25 years of age in boys [appearance of another accessory ossification center after fusion of previous ossification centers; a half-moon shape + additional accessory ossification center].

[0057] That is, the bone age determination unit 140 determines the bone age based on the age corresponding to the stage to which the morphological characteristics of the olecranon ossification center acquired through the morphological analysis unit 130 match among the stages of the olecranon characteristic standards for each age set in advance.

[0058] FIG. 10 is a diagram for explaining the performance of a bone age evaluation method according to an embodiment.

[0059] Referring to FIG. 10, the inter-observer agreement between two pediatric radiologists is shown, and it can be seen that when the bone age assessment method according to the present invention (Novel classification and Olecranon Bone Age (Olecranon BA)) is applied, it is higher than the existing olecranon bone age assessment methods, Elbow BA_Dimeglio and Elbow BA_Sauvegrain. Here, Novel classification shows the results of an analysis based on the stages expressed as O2 to F, and Olecranon BA shows the results of an analysis based on age. Stages D and D', L and L', and G and G' have the same bone age but are classified as different stages, so they were analyzed separately.

[0060] Furthermore, referring to Table 1 below, it can be seen that the bone age evaluation method according to the present invention exhibits high reliability when compared with existing elbow bone age evaluation methods, such as Sauvegrain (Elbow BA_Sauvegrain) and Dimeglio (Elbow BA_Dimeglio), and existing hand bone age evaluation methods, such as Greulich-Pyle and Tanner-Whitehouse hybrid (Hand BA_GP / TW3) and Korean pediatric standard bone age (Hand BA_KS).

[0061] [Table 1] [Table 1]

[0062] In addition, the AI ​​model, a deep learning algorithm, applied to the bone age assessment method of the present invention exhibits high accuracy and low error rate. Specifically, for the internal Novel Classification dataset, it exhibited an accuracy of 0.96, a sensitivity of 0.80, and a specificity of 0.98, while for the external dataset, it exhibited an accuracy of 0.87, a sensitivity of 0.76, and a specificity of 0.91. Furthermore, for olecranon bone age, internal validation showed an MSE of 0.088 and an RMSE of 0.296, while external validation showed an MSE of 0.125 and an RMSE of 0.398. Here, internal validation was used to evaluate how well the algorithmic model or classification method applied to perform the bone age assessment method fit the training data, and external validation was used to evaluate how well the algorithmic model or classification method generalized to new data.

[0063] Meanwhile, the steps of a method or algorithm described in connection with embodiments of the present invention may be implemented directly in hardware, as a software module executed by hardware, or a combination thereof. The software module may reside in Random Access Memory (RAM), Read Only Memory (ROM), Erasable Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), Flash Memory, a hard disk, a removable disk, a CD-ROM, or any other form of computer-readable storage medium well known in the art to which the present invention pertains.

[0064] The components of the present invention may be implemented as a program (or application) stored on a medium for execution in combination with a computer, which is hardware. The components of the present invention may be implemented as software programs or software elements. Similarly, embodiments include various algorithms implemented as a combination of data structures, processes, routines, or other programming constructs, and may be implemented in programming or scripting languages ​​such as C, C++, Java, assembler, etc. Functional aspects may be implemented as algorithms executed on one or more processors.

[0065] Although the preferred embodiments of the bone age assessment method and device according to the present invention have been described, the present invention is not limited thereto, and various modifications can be made within the scope of the claims, the detailed description of the invention, and the accompanying drawings, which also belong to the present invention. [Explanation of symbols]

[0066] 100: Bone age evaluation device, 110: Image acquisition unit, 120: Region division unit, 130: Morphological analysis unit, 140: Bone age determination unit, 150: Control unit

Claims

1. In a bone age assessment method performed by a bone age assessment device, (a) obtaining a lateral x-ray image of the elbow; (b) segmenting an olecranon region from the X-ray image; (c) analyzing the morphological characteristics of the olecranon ossification center in the segmented olecranon region; and (d) determining bone age based on the analyzed morphological characteristics.

2. The (b) is predicting a bone region in the X-ray image pixel by pixel; and The bone age evaluation method according to claim 1 , further comprising: dividing the X-ray image into regions corresponding to the olecranon based on the pixel-by-pixel prediction results.

3. The (c) is determining a region for an olecranon ossification center in the olecranon region; and The bone age assessment method of claim 2, further comprising determining morphological characteristics of the olecranon ossification center.

4. The (d) is The bone age evaluation method according to claim 3, further comprising classifying the olecranon bone age according to predetermined age-specific olecranon characteristic standards based on the morphological characteristics of the olecranon ossification center.

5. The preset age-specific olecranon characteristic standards are:

5. The bone age evaluation method according to claim 4, wherein the age range is set in six-month increments from 9.5 to 13 years for girls and from 11.5 to 15 years for boys.

6. In the preset age-specific olecranon characteristic standards, The bone age evaluation method according to claim 5, further comprising setting olecranon characteristic standards for the ages of 11.25 for girls and 13.25 for boys.

7. The preset age-specific olecranon characteristic standards are: The bone age evaluation method according to claim 4, wherein the bone age is classified into a primary ossification process and a secondary ossification process according to the order of fusion of the olecranon ossification center and the secondary ossification center.

8. The first ossification process is 8. The bone age evaluation method according to claim 7, characterized in that it comprises the following stages: a small olecranon ossification center stage in which the height of the olecranon body is 20% or more and less than 50%, an enlarged olecranon ossification center stage in which the height of the olecranon body is 50% or more, an appearance stage of a secondary ossification center, a fusion stage of the olecranon ossification center and the secondary ossification center, an expansion stage of the fused ossification center, a partial fusion stage of less than 50% of the olecranon epiphysis, a partial fusion stage of 50% or more of the olecranon epiphysis, and a complete fusion stage of the olecranon epiphysis.

9. The bone age evaluation method according to claim 8, further comprising a step of appearance of another secondary ossification center after fusion of the olecranon ossification center and the secondary ossification center between the fusion step of the olecranon ossification center and the secondary ossification center and the expansion step of the fused ossification center.

10. The second ossification process is 8. The bone age evaluation method according to claim 7, characterized in that it comprises the following stages: a small olecranon ossification center stage in which the height of the olecranon body is 20% or more and less than 50%; an enlarged olecranon ossification center stage in which the height of the olecranon body is 50% or more; an appearance stage of a secondary ossification center; a fusion stage of the secondary ossification center and the olecranon body; an expansion stage of the olecranon ossification center where the fused secondary ossification center is not fused with the olecranon body; a partial fusion stage of the secondary ossification center is fused with the olecranon ossification center and less than 50% of the olecranon body; a partial fusion stage of 50% or more of the olecranon epiphysis; and a complete fusion stage of the olecranon epiphysis.

11. The bone age evaluation method according to claim 10, further comprising a stage of appearance of another secondary ossification center after fusion of the olecranon ossification center and the secondary ossification center between the stage of fusion of the secondary ossification center and the body of the olecranon and the stage of expansion of the fused secondary ossification center - the body of the olecranon and the olecranon ossification center that is not fused.

12. an image acquisition unit for acquiring an X-ray image of the lateral side of the elbow; a region dividing unit that divides an olecranon region from the X-ray image; a morphological analysis unit that analyzes morphological characteristics of the olecranon ossification center in the divided olecranon region; a bone age determining unit that determines bone age based on the analyzed morphological characteristics.

13. The region dividing unit The bone age evaluation device according to claim 12, wherein a bone region in the X-ray image is predicted in pixel units, and a region corresponding to the olecranon in the X-ray image is divided based on the result of the pixel-by-pixel prediction.

14. The morphological analysis unit The bone age evaluation device according to claim 13, wherein a region for an olecranon ossification center is determined in the olecranon region, and morphological characteristics of the olecranon ossification center are determined.

15. The bone age determination unit The bone age evaluation device according to claim 14, wherein the olecranon bone age is classified according to a predetermined characteristic standard of the olecranon for each age based on the morphological characteristics of the olecranon ossification center.

16. The preset age-specific olecranon characteristic standards are:

16. The bone age evaluation device according to claim 15, wherein the age is set in six-month increments from 9.5 to 13 years for girls and from 11.5 to 15 years for boys.

17. In the preset age-specific olecranon characteristic standards, The bone age evaluation device according to claim 16, wherein olecranon characteristic standards are further set for the ages of 11.25 for girls and 13.25 for boys.

18. The preset age-specific olecranon characteristic standards are: The bone age evaluation device according to claim 15, wherein the bone age is classified into a primary ossification process and a secondary ossification process according to the order of fusion of the olecranon ossification center and the secondary ossification center.

19. A computer program stored on a computer-readable recording medium, A computer program stored on a computer-readable recording medium, characterized in that the instructions of the computer program, when executed, cause a method according to any one of claims 1 to 11 to be performed.

Images