WO2020062839A1 - 一种检测骨龄的方法及装置 - Google Patents

一种检测骨龄的方法及装置 Download PDF

Info

Publication number
WO2020062839A1
WO2020062839A1 PCT/CN2019/082681 CN2019082681W WO2020062839A1 WO 2020062839 A1 WO2020062839 A1 WO 2020062839A1 CN 2019082681 W CN2019082681 W CN 2019082681W WO 2020062839 A1 WO2020062839 A1 WO 2020062839A1
Authority
WO
WIPO (PCT)
Prior art keywords
bone age
epiphysis
target
key point
bone
Prior art date
Application number
PCT/CN2019/082681
Other languages
English (en)
French (fr)
Inventor
魏子昆
杨忠程
丁泽震
Original Assignee
杭州依图医疗技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 杭州依图医疗技术有限公司 filed Critical 杭州依图医疗技术有限公司
Priority to SG11202002139SA priority Critical patent/SG11202002139SA/en
Publication of WO2020062839A1 publication Critical patent/WO2020062839A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G06T7/75—Determining position or orientation of objects or cameras using feature-based methods involving models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10116—X-ray image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30008—Bone

Definitions

  • the invention relates to the technical field of machine learning, and in particular, to a method and a device for detecting bone age.
  • Bone age is the abbreviation of skeletal age. It is the age at which adolescents' bone development level is compared with bone development standards. It can more accurately reflect the maturity of the body than age, height, and weight, and more accurately reflect the individual. Growth level and maturity.
  • bone age films are mainly shot, and the bone age is estimated by manually viewing the bone age films.
  • This method relies on artificial estimation of the bone age based on the bone age films.
  • the accuracy is subject to human subjective factors and the image is large and the efficiency is low.
  • Embodiments of the present application provide a method and device for detecting bone age.
  • an embodiment of the present application provides a method for detecting bone age, including:
  • For each target epiphysis determine the bone age fraction of the target epiphysis according to the characteristics of the target epiphysis and the position of the target epiphysis;
  • the bone age of the user to be detected is determined according to the bone age score of each target epiphysis.
  • a positioning model is used to automatically determine the position of the target epiphysis, and then the characteristics of the target epiphysis and the target are combined.
  • the position of the epiphysis determines the bone age score, and the bone age of the user to be tested is determined based on the bone age score, without artificially judging the bone age based on the bone age film, thereby improving the accuracy of detecting the bone age and improving the efficiency of detecting the bone age.
  • the positioning model includes a coarse positioning module and a fine positioning module
  • the determining a position of each target epiphysis in the bone age film by using a positioning model includes:
  • For each coarse segmented region input the coarse segmented region into a corresponding fine positioning module to determine the coordinates of a second key point corresponding to each target epiphysis in the coarse segmented region;
  • the position of each target epiphysis is determined according to the coordinates of a second key point corresponding to each target epiphysis in the coarsely divided region.
  • the bone age piece is a hand bone piece
  • the first key point includes a first key point in the thumb region, a first key point in the middle finger region, a first key point in the little finger region, and a first key in the wrist region. point;
  • the determining one or more coarsely divided regions in the bone age slice according to the coordinates of a first key point corresponding to each target epiphysis includes:
  • a coarse division region is determined by expanding a preset distance outward according to the long axis direction and the short axis direction.
  • the method before inputting the coarse segmentation region to a corresponding fine positioning module, the method further includes:
  • the bone age slice of the coarsely divided region is adjusted to meet the requirements of the fine positioning module.
  • an embodiment of the present application provides a device for detecting bone age, including:
  • a positioning module configured to determine a position of each target epiphysis in the bone age slice by using a positioning model
  • a processing module configured to determine, for each target epiphysis, a bone age fraction of the target epiphysis according to a characteristic of the target epiphysis and a position of the target epiphysis;
  • a detection module is configured to determine a bone age of the user to be detected according to a bone age score of each target epiphysis.
  • the positioning model includes a coarse positioning module and a fine positioning module
  • the positioning module is specifically configured to:
  • For each coarse segmented region input the coarse segmented region into a corresponding fine positioning module to determine the coordinates of a second key point corresponding to each target epiphysis in the coarse segmented region;
  • the position of each target epiphysis is determined according to the coordinates of a second key point corresponding to each target epiphysis in the coarsely divided region.
  • the bone age piece is a hand bone piece
  • the first key point includes a first key point in the thumb region, a first key point in the middle finger region, a first key point in the little finger region, and a first key in the wrist region. point;
  • the positioning module is specifically configured to:
  • a coarse division region is determined by expanding a preset distance outward according to the long axis direction and the short axis direction.
  • the positioning module is further configured to:
  • the bone age slice of the coarse segmented region is adjusted to meet the requirements of the fine positioning module.
  • an embodiment of the present application provides a device for detecting bone age, including at least one processor and at least one memory, wherein the storage unit stores a computer program, and when the program is executed by the processor, So that the processor executes the steps of the method according to the first aspect.
  • an embodiment of the present application provides a computer-readable medium that stores a computer program executable by a device that detects bone age, and when the program is run on the device that detects bone age, the device that makes the device detect bone age Perform the steps of the method described in the first aspect.
  • an embodiment of the present invention provides a computer program product.
  • the computer program product includes a computing program stored on a non-transitory computer-readable storage medium.
  • the computer program includes program instructions. When executed by a computer, the computer is caused to execute the method according to the first aspect.
  • a positioning model is used to automatically determine the position of the target epiphysis, and then the characteristics of the target epiphysis and the target are combined.
  • the position of the epiphysis determines the bone age score, and the bone age of the user to be tested is determined based on the bone age score, without artificially judging the bone age based on the bone age film, thereby improving the accuracy of detecting the bone age and improving the efficiency of detecting the bone age.
  • the rough segmented region containing the target epiphysis is positioned first, and then the position of the target epiphysis is located from the coarse segmented region, thereby improving the accuracy of locating the target epiphysis.
  • FIG. 1 is a schematic flowchart of a method for detecting bone age according to an embodiment of the present application
  • FIG. 2 is a schematic diagram of a bone age slice provided by an embodiment of the present application.
  • FIG. 3 is a schematic diagram of an epiphysis provided by an embodiment of the present application.
  • FIG. 4 is a schematic flowchart of a method for adjusting a bone age piece according to an embodiment of the present application
  • FIG. 5 is a schematic structural diagram of a deep residual network according to an embodiment of the present application.
  • FIG. 6 is a schematic structural diagram of a convolutional feature extraction block provided by an embodiment of the present application.
  • FIG. 7 is a schematic flowchart of a method for positioning a target epiphysis according to an embodiment of the present application.
  • FIG. 8 is a schematic diagram of a coarsely divided region according to an embodiment of the present application.
  • FIG. 9 is a schematic diagram of a thumb area according to an embodiment of the present application.
  • FIG. 10 is a schematic diagram of a thumb area according to an embodiment of the present application.
  • FIG. 11 is a schematic structural diagram of a device for detecting bone age according to an embodiment of the present application.
  • FIG. 12 is a schematic structural diagram of a device for detecting bone age according to an embodiment of the present application.
  • Bone age short for bone age, which needs to be determined with the help of specific images of bones in X-ray photography. Usually, an X-ray film of the wrist of the left hand of a person is taken. The doctor observes the development of the ossification center of the left palmar phalanx, wrist and radius ulna to determine the bone age.
  • the technical solution for detecting the bone age in the embodiments of the present application is suitable for detecting the bone age of adolescents in a hospital, providing a reference for judging adolescents' growth and development, and helping doctors diagnose diseases such as chromosomal abnormalities, hereditary diseases, and endocrine diseases.
  • FIG. 1 exemplarily illustrates a schematic flowchart of a method for detecting bone age provided by an embodiment of the present application.
  • the process may be performed by a device for detecting bone age, and specifically includes the following steps:
  • Step S101 Obtain a bone age slice of a user to be detected.
  • the bone age film refers to a specific image taken by X-rays, and an X-ray film of a person's left hand is usually taken as a bone age film. For example, as shown in FIG. 2.
  • the pre-processing process mainly includes gland segmentation and image normalization.
  • Gland segmentation includes the following steps: First, Gaussian filtering is applied to the entire bone age slice, and the filtered result is binarized.
  • the binarization threshold is obtained by obtaining the maximum class spacing method of the image gray histogram.
  • the binarized result is inflated to obtain individual area blocks by the flood method, and the area of each area block is counted. The largest area block is retained and the segmented hand bone image. Paste the segmented hand bone image onto a pure black image that matches the length and width of the hand bone image.
  • the image normalization includes the following steps:
  • the bone age slice is an image in the dicom format. First, according to the dicom information, a window width and a window level are selected, and the bone age slice image is converted into a png format. By adding black borders on the top or both sides of the bone age film image, the bone age film image aspect ratio is adjusted to 1: 1, and finally the bone age film image is scaled to a size of 512 * 512.
  • step S102 the position of each target epiphysis in the bone age slice is determined using a positioning model.
  • the localization model is a trained deep residual network.
  • the epiphysis is the ossification point that appears in the cartilage at both ends of the bone during the development of the bone. It is usually found at the ends of long bones, margins of flat bones, nodules, tuberosities and protrusions. After most of the epiphyseal ossification, there is only one layer of cartilage plate in the adjacent part of the diaphysis, that is, the epiphyseal cartilage. Through the proliferation and ossification of chondrocytes, bones are continuously lengthened.
  • the bone age piece of the hand bone shown in FIG. 3 includes phalangeal callus, ulna callus, and radial callus. It should be noted that, in the bone age piece shown in FIG. 3, the bone callus identified by the circle is only an example identification. Part of the epiphysis, the epiphysis in the hand bone is not limited to these.
  • Step S103 For each target epiphysis, determine the bone age fraction of the target epiphysis according to the characteristics of the target epiphysis and the position of the target epiphysis.
  • the bone age score is determined by querying the bone age scoring criteria according to the characteristics of the target epiphysis and the location of the target epiphysis. Different positions and different features of the epiphysis correspond to different bone age fractions. When scoring, the bone age scores determined by different bone age score standards will also be different.
  • the bone age score standards include, but are not limited to, the TW3 bone age score standard and G-P (atlas method).
  • Step S104 Determine the bone age of the user to be detected according to the bone age score of each target epiphysis.
  • a positioning model is used to automatically determine the position of the target epiphysis, and then the characteristics of the target epiphysis and the target are combined.
  • the position of the epiphysis determines the bone age score, and the bone age of the user to be tested is determined based on the bone age score, without artificially judging the bone age based on the bone age film, thereby improving the accuracy of detecting the bone age and improving the efficiency of detecting the bone age.
  • each bone in the reference position includes the following steps, as shown in Figure 4:
  • Step S401 Acquire the coordinates of a preset reference point.
  • the coordinates of the preset reference point may be preset coordinates of a part of bones, which are used to indicate the reference position of some bones.
  • the coordinates of the preset reference point may be preset coordinates of a point related to the middle finger, which is used to indicate the reference of the middle finger.
  • the coordinates of the position and the preset reference point may also be preset coordinates of a point related to the little finger, which is used to indicate the reference position of the little finger.
  • step S402 the coordinates of key points in the bone age film are determined by using an adjustment model.
  • the adjustment model is determined by training the deep residual network after multiple bone age films in which the coordinates of the key points are marked in advance as training samples.
  • the coordinates of the key point and the coordinates of the preset reference point are coordinates in the same coordinate system.
  • multiple bone age slices are obtained, and then key points around the little finger in the bone age slices are manually labeled, and then the bone age slices marked with the key points are input to a deep residual network for training.
  • the objective function of the deep residual network meets a preset condition .
  • Determine the adjustment model When the bone age piece of the user to be detected is obtained, the bone age piece is input into the adjustment model to determine the key points around the little finger in the bone age piece.
  • Step S403 Determine the correspondence between the current position of each bone of the bone age slice and the reference position according to the coordinates of the preset reference point and the coordinates of the key points.
  • step S404 each bone in the bone age piece is adjusted to a reference position according to the corresponding relationship.
  • the current position of the little finger in the bone age film and the reference position of the little finger in the bone age film can be determined according to the coordinates of the preset reference point and the key point coordinates.
  • Corresponding relationship can be further obtained between the current position and the reference position of other bones in the bone age film, wherein the corresponding relationship includes translation relationship and rotation relationship. Then, each bone in the bone age film is adjusted to the reference position according to the corresponding relationship.
  • the positioning model is obtained by training the bone age piece at the reference position as a training sample, before positioning the target epiphysis in the bone age piece, first adjust each bone in the bone age piece of the user to be tested to the reference position, thereby improving The accuracy of positioning target epiphysis was improved.
  • the positioning model includes a coarse positioning module and a fine positioning module.
  • Both the coarse positioning module and the fine positioning module are deep residual networks.
  • the coarse positioning module is based on multiple bone age slices labeled with key points in advance as training samples , Determined after training the deep residual network.
  • the fine positioning module uses bone age slices of multiple coarsely segmented regions marked with key points in advance as training samples and is determined after training the deep residual network.
  • the number of fine positioning modules is determined according to the number of coarse segmented regions output by the coarse positioning module.
  • multiple bone age slices are obtained as training samples. For each bone age piece, pretreat the bone age piece, and then adjust each bone in the bone age piece to the reference position.
  • the process of adjusting the position of the bone age piece and pretreating the bone age piece has been described in the foregoing. I will not repeat them here.
  • the key points are marked by each labeler in each bone age film.
  • the key points are the points near the epiphysis in the bone age film, and each key point corresponds to a serial number.
  • the data augmentation methods include, but are not limited to:
  • the training samples are then input to the deep residual network for training.
  • the loss function is calculated according to the coordinates of the labeled key points and the coordinates of the key points predicted by the network.
  • the loss function is trained by back propagation.
  • the training optimization algorithm uses the sgd algorithm with momentum and step attenuation.
  • the structure of the above-mentioned deep residual network is shown in FIG. 5 and includes N consecutive convolutional feature extraction blocks and a fully connected layer.
  • N convolutional feature extraction blocks For any two consecutive first volumes of the N convolutional feature extraction blocks, Convolution feature extraction block and second convolution feature extraction block, the second image feature output by the second convolution feature extraction block and the first image feature output by the first convolution feature extraction block are added as a third convolution feature extraction block Input or the output of N consecutive convolutional feature extraction blocks.
  • the third convolutional feature extraction block is a convolutional feature extraction block that is located after the second convolutional feature extraction block and is continuous with the second convolutional feature extraction block.
  • the convolution feature extraction block includes L convolution modules, where L is greater than 0. Any convolution module includes a convolution layer, a BN layer, and an excitation layer, as shown in FIG. 6.
  • the training process of the fine positioning module is specifically introduced below: multiple bone age slices are obtained as training samples. For each bone age piece, adjust each bone in the bone age piece to the reference position, and then preprocess the bone age piece. According to a preset rule, a plurality of coarse segmentation regions are segmented from the bone age film, for example, four coarse segmentation regions are divided into a thumb region, a middle finger region, a little finger region, and a wrist region. A fine positioning module is trained for each coarse segmentation region. For any fine positioning module, the key point is marked by the labeler in each bone age slice of the coarse segmentation region, or the key points already marked in the training sample corresponding to the coarse segmentation module can be directly used. The key point is the epiphysis in the coarse segmentation region.
  • each key point corresponds to a serial number.
  • the bone age slice of the coarsely divided region is adjusted to a normal image.
  • the thumb area may be an inclined area in the same direction as the thumb direction. After the thumb area is intercepted, the thumb area is rotated by a certain angle and adjusted to a normal image. Then perform data augmentation on the training samples to increase the data amount to 10 times the original data amount.
  • the data augmentation methods include, but are not limited to:
  • the training samples are then input to the deep residual network for training.
  • the loss function is calculated according to the coordinates of the labeled key points and the coordinates of the key points predicted by the network.
  • the loss function is trained by back propagation.
  • the training optimization algorithm uses the sgd algorithm with momentum and step attenuation.
  • the deep residual network includes M consecutive convolutional feature extraction blocks and a fully connected layer. The convolutional feature extraction block is shown in Figure 6.
  • the positioning model is used to determine the position of each target epiphysis in the bone age slice, as shown in FIG. 7, which specifically includes the following steps:
  • step S701 the bone age piece is input into the coarse positioning module, and the coordinates of the first key point corresponding to each target epiphysis in the bone age piece are determined.
  • the bone age piece is a hand bone piece
  • the first key point may be a key point of each epiphyseal attachment in the hand bone, and each first key point corresponds to a serial number.
  • Step S702 Determine one or more coarse segmented regions in the bone age slice according to the coordinates of the first key point corresponding to each target epiphysis, and the coarse segmented regions include one or more target epiphyses.
  • the coarse segmentation region may be a thumb region, an index finger region, a middle finger region, a ring finger region, a little finger region, a wrist region, and the like.
  • the specific position of the coarse segmentation region and the target epiphysis contained in the coarse segmentation region may be set in advance.
  • Step S703 For each coarse segmented area, input the coarse segmented area into a corresponding fine positioning module, and determine the coordinates of the second key point corresponding to each target epiphysis in the coarse segmented area.
  • Step S704 Determine the position of each target epiphysis according to the coordinates of the second key point corresponding to each target epiphysis in the coarse segmentation region.
  • the positioning model locates the position of each epiphysis, first use the coarse positioning module to segment the coarse segmentation region from the bone age film, reduce the positioning range of the target epiphysis, and then locate the position of each target epiphysis from the coarse segmentation region, thereby improving the positioning. The accuracy of the target epiphysis.
  • the coarsely divided region is a thumb region, a middle finger region, a little finger region, and a wrist region
  • the first key point includes the first key point of the thumb region, the first key point of the middle finger region, and the little finger region
  • the long axis direction and the short axis direction are determined according to the coordinates of the first key point of the region, and the center point of the region is determined.
  • the center point is taken as the center, and the preset distance is extended outward according to the long axis direction and the short axis direction to determine the coarse division region, as shown in FIG. 8.
  • principal component analysis may be performed on the coordinates of the first key point to determine a long axis direction and a short axis direction.
  • the coarsely divided region may be a rectangle having a ratio of the major axis direction to the minor axis direction of 1: 3.
  • step S703 before the coarse segmented area is input to the corresponding fine positioning module, the bone age slice of the coarse segmented area is adjusted to meet the requirements of the fine positioning module.
  • a bone age piece in the thumb area is taken from the bone age piece shown in FIG. 8, and the bone age piece in the thumb area is shown in FIG. 9.
  • a normal image is obtained, as shown in FIG. 10. Then input the normal image of the thumb area into the fine positioning module.
  • step S704 after determining the coordinates of the second key point corresponding to each target epiphysis in the coarse segmentation region, determine that each target epiphysis corresponds to the second key point according to the serial number of the second key point, and then according to the coordinates of the second key point Calculate the rectangular box of the target epiphysis.
  • the features of the target epiphysis are identified using a recognition model, and the recognition model is determined by training a deep residual network by using multiple epiphyseal images labeled with the target epiphyseal feature type in advance. Because there are differences in the characteristics of each target epiphysis, a recognition model can be trained for each target epiphysis, thereby improving the accuracy of identifying the characteristics of the target epiphysis.
  • the device 1100 includes:
  • An acquisition module 1101 configured to acquire a bone age slice of a user to be detected
  • a positioning module 1102 configured to determine a position of each target epiphysis in the bone age slice by using a positioning model
  • a processing module 1103, configured to determine, for each target epiphysis, a bone age fraction of the target epiphysis according to a characteristic of the target epiphysis and a position of the target epiphysis;
  • the detection module 1104 is configured to determine a bone age of the user to be detected according to a bone age score of each target epiphysis.
  • the positioning model includes a coarse positioning module and a fine positioning module
  • the positioning module 1102 is specifically configured to:
  • For each coarse segmented region input the coarse segmented region into a corresponding fine positioning module to determine the coordinates of a second key point corresponding to each target epiphysis in the coarse segmented region;
  • the position of each target epiphysis is determined according to the coordinates of a second key point corresponding to each target epiphysis in the coarsely divided region.
  • the bone age piece is a hand bone piece
  • the first key point includes a first key point in the thumb region, a first key point in the middle finger region, a first key point in the little finger region, and a first key in the wrist region. point;
  • the positioning module 1102 is specifically configured to:
  • a coarse division region is determined by expanding a preset distance outward according to the long axis direction and the short axis direction.
  • the positioning module 1102 is further configured to:
  • the bone age slice of the coarse segmented region is adjusted to meet the requirements of the fine positioning module.
  • the embodiment of the present application provides a device for detecting bone age.
  • the device includes at least one processor 1201 and a memory 1202 connected to the at least one processor, which is not limited in the embodiment of the present application.
  • the specific connection medium between the processor 1201 and the memory 1202, and the connection between the processor 1201 and the memory 1202 through a bus in FIG. 12 is taken as an example.
  • the bus can be divided into an address bus, a data bus, a control bus, and the like.
  • the memory 1202 stores instructions that can be executed by at least one processor 1201.
  • the at least one processor 1201 can execute the steps included in the foregoing method for detecting bone age by executing the instructions stored in the memory 1202.
  • the processor 1201 is a control center of a device for detecting bone age, and can use various interfaces and lines to connect various parts of the device for detecting bone age.
  • the processor 1201 may include one or more processing units, and the processor 1201 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface, and an application program, etc.
  • the tuning processor mainly handles wireless communication. It can be understood that the foregoing modem processor may not be integrated into the processor 1201.
  • the processor 1201 and the memory 1202 may be implemented on the same chip, and in some embodiments, they may also be implemented separately on separate chips.
  • the processor 1201 may be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application specific integrated circuit (ASIC), a field programmable gate array or other programmable logic device, a discrete gate, or a transistor
  • the logic device and discrete hardware components can implement or execute the methods, steps and logic block diagrams disclosed in the embodiments of the present application.
  • a general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in combination with the embodiments of the present application may be directly implemented by a hardware processor, or may be performed by a combination of hardware and software modules in the processor.
  • the memory 1202 is a non-volatile computer-readable storage medium and can be used to store non-volatile software programs, non-volatile computer executable programs, and modules.
  • the memory 1202 may include at least one type of storage medium, for example, it may include a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (RAM), a static random access memory (Static Random Access Memory, SRAM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Magnetic Memory, Disk , CDs and more.
  • the memory 1202 is any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.
  • the memory 1202 in the embodiment of the present application may also be a circuit or any other device capable of implementing a storage function, and is configured to store program instructions and / or data.
  • an embodiment of the present application further provides a computer-readable medium that stores a computer program executable by a device that detects bone age, and when the program is run on the device that detects bone age, the detected bone age is made The device performs the steps of the method of detecting bone age.
  • an embodiment of the present application further provides a computer program product.
  • the computer program product includes a computing program stored on a non-transitory computer-readable storage medium.
  • the computer program includes program instructions. When the program instructions are executed by a computer, the computer is caused to execute a method for detecting bone age.
  • the embodiments of the present invention may be provided as a method, a system, or a computer program product. Therefore, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
  • computer-usable storage media including, but not limited to, disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to work in a particular manner such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, the instructions
  • the device implements the functions specified in one or more flowcharts and / or one or more blocks of the block diagram.
  • These computer program instructions can also be loaded on a computer or other programmable data processing device, so that a series of steps can be performed on the computer or other programmable device to produce a computer-implemented process, which can be executed on the computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more flowcharts and / or one or more blocks of the block diagrams.
  • the embodiments of the present invention may be provided as a method, a system, or a computer program product. Therefore, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
  • computer-usable storage media including, but not limited to, disk storage, CD-ROM, optical storage, etc.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to work in a particular manner such that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, the instructions
  • the device implements the functions specified in one or more flowcharts and / or one or more blocks of the block diagram.
  • These computer program instructions can also be loaded on a computer or other programmable data processing device, so that a series of steps can be performed on the computer or other programmable device to produce a computer-implemented process, which can be executed on the computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more of the flowcharts and / or one or more of the blocks of the block diagram.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Medical Informatics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

一种检测骨龄的方法及装置，涉及机器学习技术领域，该方法包括：获取待检测用户的骨龄片，采用定位模型确定骨龄片中各目标骨骺的位置。针对每一个目标骨骺，根据目标骨骺的特征以及目标骨骺的位置确定目标骨骺的骨龄分数，根据各目标骨骺的骨龄分数确定待检测用户的骨龄。在获取待检测用户的骨龄片后，调整骨龄片中的各骨头至基准位置，具体包括：获取预设基准点的坐标（401）；采用调整模型确定骨龄片中关键点的坐标，调整模型是以预先标记关键点的坐标的多张骨龄片为训练样本，对深度残差网络进行训练后确定的（402）；根据预设基准点的坐标和关键点的坐标确定骨龄片各骨头的当前位置与基准位置之间的对应关系（403）；根据对应关系将骨龄片中的各骨头调整至基准位置（404）。由于骨骺的不同形态代表着不同的骨龄阶段，不同位置的骨骺的形态也会存在一定的区别，因此采用定位模型自动确定目标骨骺的位置，然后结合目标骨骺的特征以及目标骨骺的位置确定骨龄分数，基于骨龄分数确定待检测用户的骨龄，而不需要人工主观根据骨龄片判断骨龄，从而一方面提高了检测骨龄的精度，另一方面提高了检测骨龄的效率。

Description

一种检测骨龄的方法及装置
本申请要求在2018年9月30日提交中国专利局、申请号为201811163000.8、发明名称为“一种检测骨龄的方法及装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及机器学习技术领域，尤其涉及一种检测骨龄的方法及装置。
背景技术
“骨龄”是骨骼年龄的简称，是青少年儿童骨骼发育水平同骨发育标准比较而得出的发育年龄，它比年龄、身高、体重更能精确的反映出身体的成熟程度，更加准确地反映个体的生长发育水平和成熟程度。
临床上通过检测骨龄来判读儿童的生物学年龄，通过生物年龄与日历年龄的差异来评估儿童发育状况，了解儿童性成熟的趋势，预测儿童的成年身高等，并广泛用于影响儿童生长发育疾病的治疗监测，对一些儿科内分泌疾病的诊断有很大帮助。
现有技术中主要通过拍摄骨龄片，由人工查看骨龄片估计骨龄。该方法依赖人工根据骨龄片估计骨龄，精度受人的主观因素影像大，效率较低。
发明内容
由于现有技术中依赖人工根据骨龄片估计骨龄，精度受人的主观因素影像大，效率较低的问题，本申请实施例提供了一种检测骨龄的方法及装置。
第一方面，本申请实施例提供了一种检测骨龄的方法，包括：
获取待检测用户的骨龄片；
采用定位模型确定所述骨龄片中各目标骨骺的位置；
针对每一个目标骨骺，根据所述目标骨骺的特征以及所述目标骨骺的位置确定所述目标骨骺的骨龄分数；
根据各目标骨骺的骨龄分数确定所述待检测用户的骨龄。
由于骨骺的不同形态代表着不同的骨龄阶段，不同位置的骨骺的形态也会存下一定的区别，因此本申请实施例中采用定位模型自动确定目标骨骺的位置，然后结合目标骨骺的特征以及目标骨骺的位置确定骨龄分数，基于骨龄分数确定待检测用户的骨龄，而不需要人工主观根据骨龄片判断骨龄，从而一方面提高了检测骨龄的精度，另一方面提高了检测骨龄的效率。
可选地，所述定位模型包括粗定位模块和细定位模块；
所述采用定位模型确定所述骨龄片中各目标骨骺的位置，包括：
将所述骨龄片输入所述粗定位模块，确定所述骨龄片中各目标骨骺对应的第一关键点的坐标；
根据所述各目标骨骺对应的第一关键点的坐标确定所述骨龄片中的一个或多个粗分割区域，所述粗分割区域包括一个或多个目标骨骺；
针对每一个粗分割区域，将所述粗分割区域输入对应的细定位模块，确定所述粗分割区域中各目标骨骺对应的第二关键点的坐标；
根据所述粗分割区域中各目标骨骺对应的第二关键点的坐标确定所述各目标骨骺的位置。
可选地，所述骨龄片为手骨片；所述第一关键点包括拇指区域的第一关键点、中指区域的第一关键点、小指区域的第一关键点及手腕区域的第一关键点；
所述根据所述各目标骨骺对应的第一关键点的坐标确定所述骨龄片中的一个或多个粗分割区域，包括：
针对每个区域的第一关键点，根据所述区域的第一关键点的坐标确定长轴方向和短轴方向；
确定所述区域的中心点；
以所述中心点为中心，根据所述长轴方向和所述短轴方向向外扩展预设距离，确定粗分割区域。
可选地，将所述粗分割区域输入对应的细定位模块之前，还包括：
调整所述粗分割区域的骨龄片以满足所述细定位模块的要求。
第二方面，本申请实施例提供了一种检测骨龄的装置，包括：
获取模块，用于获取待检测用户的骨龄片；
定位模块，用于采用定位模型确定所述骨龄片中各目标骨骺的位置；
处理模块，用于针对每一个目标骨骺，根据所述目标骨骺的特征以及所述目标骨骺的位置确定所述目标骨骺的骨龄分数；
检测模块，用于根据各目标骨骺的骨龄分数确定所述待检测用户的骨龄。
可选地，所述定位模型包括粗定位模块和细定位模块；
所述定位模块具体用于：
将所述骨龄片输入所述粗定位模块，确定所述骨龄片中各目标骨骺对应的第一关键点的坐标；
根据所述各目标骨骺对应的第一关键点的坐标确定所述骨龄片中的一个或多个粗分割区域，所述粗分割区域包括一个或多个目标骨骺；
针对每一个粗分割区域，将所述粗分割区域输入对应的细定位模块，确定所述粗分割区域中各目标骨骺对应的第二关键点的坐标；
根据所述粗分割区域中各目标骨骺对应的第二关键点的坐标确定所述各目标骨骺的位置。
可选地，所述骨龄片为手骨片；所述第一关键点包括拇指区域的第一关键点、中指区域的第一关键点、小指区域的第一关键点及手腕区域的第一关键点；
所述定位模块具体用于：
针对每个区域的第一关键点，根据所述区域的第一关键点的坐标确定长轴方向和短轴方向；
确定所述区域的中心点；
以所述中心点为中心，根据所述长轴方向和所述短轴方向向外扩展预设距离，确定粗分割区域。
可选地，所述定位模块还用于：
将所述粗分割区域输入对应的细定位模块之前，调整所述粗分割区域的骨龄片以满足所述细定位模块的要求。
第三方面，本申请实施例提供了一种检测骨龄的设备，包括至少一个处理器、以及至少一个存储器，其中，所述存储单元存储有计算机程序，当所述程序被所述处理器执行时，使得所述处理器执行第一方面所述方法的步骤。
第四方面，本申请实施例提供了一种计算机可读介质，其存储有可由检测骨龄的设备执行的计算机程序，当所述程序在检测骨龄的设备上运行时，使得所述检测骨龄的设备执行第一方面所述方法的步骤。
第五方面，本发明实施例提供一种计算机程序产品，所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算程序，所述计算机程序包括程序指令，当所述程序指令被计算机执行时，使所述计算机执行上述第一方面所述的方法。
由于骨骺的不同形态代表着不同的骨龄阶段，不同位置的骨骺的形态也会存下一定的区别，因此本申请实施例中采用定位模型自动确定目标骨骺的位置，然后结合目标骨骺的特征以及目标骨骺的位置确定骨龄分数，基于骨龄分数确定待检测用户的骨龄，而不需要人工主观根据骨龄片判断骨龄，从而一方面提高了检测骨龄的精度，另一方面提高了检测骨龄的效率。由于在定位骨龄片中的目标骨骺的位置时，先定位包含目标骨骺的粗分割区域，然后从粗分割区域中定位目标骨骺的位置，从而提高了定位目标骨骺的精度。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简要介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域的普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的一种检测骨龄的方法的流程示意图；
图2为本申请实施例提供的一种骨龄片的示意图；
图3为本申请实施例提供的一种骨骺的示意图；
图4为本申请实施例提供的一种调整骨龄片的方法的流程示意图；
图5为本申请实施例提供的一种深度残差网络的结构示意图；
图6为本申请实施例提供的一种卷积特征提取块的结构示意图；
图7为本申请实施例提供的一种定位目标骨骺的方法的流程示意图；
图8为本申请实施例提供的一种粗分割区域的示意图；
图9为本申请实施例提供的一种拇指区域的示意图；
图10为本申请实施例提供的一种拇指区域的示意图；
图11为本申请实施例提供的一种检测骨龄的装置的结构示意图；
图12为本申请实施例提供的一种检测骨龄的设备的结构示意图。
具体实施方式
为了使本发明的目的、技术方案和优点更加清楚，下面将结合附图对本发明作进一步地详细描述，显然，所描述的实施例仅仅是本发明一部份实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其它实施例，都属于本发明保护的范围。
为了方便理解，下面对本发明实施例中涉及的名词进行解释。
骨龄：骨骼年龄的简称，需要借助于骨骼在X光摄像中的特定图像来确定。通常要拍摄人左手手腕部的X光片，医生通过X光片观察左手掌指骨、腕骨及桡尺骨下端的骨化中心的发育程度，来确定骨龄。
本申请实施例中检测骨龄的技术方案适用于医院检测青少年的骨龄，为医生判定青少年生长发育情况提供参考，同时帮助医生诊断染色体异常、遗传性疾病、内分泌疾病等疾病。
图1示例性示出了本申请实施例提供的一种检测骨龄的方法的流程示意图，该流程可以由检测骨龄的装置执行，具体包括以下步骤：
步骤S101，获取待检测用户的骨龄片。
骨龄片指采用X光拍摄的特定图像，通常拍摄人左手的X光片作为骨龄片，示例性地，如图2所示。
获取待检测用户的骨龄片之后，对骨龄片进行预处理，预处理过程主要包括腺体分割以及图像归一化。
腺体分割包括以下步骤：首先对整张骨龄片使用高斯滤波，将滤波的结果二值化，二值化的阈值通过求图像灰度直方图的最大类间距方法获得。然后将二值化的结果膨胀后通过漫水法(flood fill)获得一个个独立的区域块，统计每个区域块的面积。将面积最大的区域块保留，分割出来的手骨图像。将分割出的手骨图像粘贴到一张和手骨图像的长宽相符的纯黑图像上。
图像归一化包括以下步骤：骨龄片为dicom格式的图像，先根据dicom信息，选取窗宽窗位，转为png格式的骨龄片图像。通过在骨龄片图像上侧或两侧添加黑边，将骨龄片图像长宽比调整为1:1，最后将骨龄片图像缩放到512*512大小。
步骤S102，采用定位模型确定骨龄片中各目标骨骺的位置。
定位模型为训练好的深度残差网络。骨骺(epiphysis)是骨在发育过程中，骨两端的软骨中出现的骨化点。一般见于长骨两端、扁骨缘、结节、粗隆及突起等处。骨骺大部分骨化后，只在与骨干相邻部位留有一层软骨板，即骺软骨。通过软骨细胞的分裂增殖、骨化，使骨不断加长。示例性地，如图3所示的手骨的骨龄片中，包括指骨骨骺、尺骨骨骺、桡骨骨骺，需要说明的是，图3所示的骨龄片中，圆圈标识出的骨骺只是作为示例标识出的部分骨骺，手骨中的骨骺并不仅限于这些。
步骤S103，针对每一个目标骨骺，根据目标骨骺的特征以及目标骨骺的位置确定目标骨骺的骨龄分数。
根据目标骨骺的特征以及目标骨骺的位置查询骨龄评分标准确定骨龄分数。不同位置以及不同特征的骨骺对应不同的骨龄分数。在评分时，不同的骨龄评分标准确定的骨龄分数也会不一样，骨龄评分标准包括但不限于TW3骨龄评分标准、G-P(图谱法)。
步骤S104，根据各目标骨骺的骨龄分数确定待检测用户的骨龄。
由于骨骺的不同形态代表着不同的骨龄阶段，不同位置的骨骺的形态也会存下一定的区别，因此本申请实施例中采用定位模型自动确定目标骨骺的位置，然后结合目标骨骺的特征以及目标骨骺的位置确定骨龄分数，基于骨龄分数确定待检测用户的骨龄，而不需要人工主观根据骨龄片判断骨龄，从而一方面提高了检测骨龄的精度，另一方面提高了检测骨龄的效率。
可选地，在上述步骤S101中，由于获取的骨龄片中骨头的位置有时候并不是基准位置，由此影响后续定位目标骨骺的位置，故在获取待检测用户的骨龄片后，调整骨龄片中的各骨头至基准位置，具体包括以下步骤，如图4所示：
步骤S401，获取预设基准点的坐标。
预设基准点的坐标可以是预设的部分骨头的坐标，用于表示部分骨头的基准位置，比如预设基准点的坐标可以是与中指相关的点的预设坐标，用于表示中指的基准位置，预设基准点的坐标也可以是与小指相关的点的预设坐标，用于表示小指的基准位置。
步骤S402，采用调整模型确定骨龄片中关键点的坐标，调整模型是以预先标记关键点的坐标的多张骨龄片为训练样本，对深度残差网络进行训练后确定的。
关键点的坐标和预设基准点的坐标为同一坐标系中的坐标。
示例性地，获取多张骨龄片，然后人工标记骨龄片中小指周围的关键点，然后将标记关键点的骨龄片输入深度残差网络进行训练，当深度残差网络的目标函数满足预设条件时，确定调整模型。当获取待检测用户的骨龄片时，将骨龄片输入调整模型，确定骨龄片中小指周围的关键点。
步骤S403，根据预设基准点的坐标和关键点的坐标确定骨龄片各骨头的当前位置与基准位置之间的对应关系。
步骤S404，根据对应关系将骨龄片中的各骨头调整至基准位置。
当预设基准点和关键点都是与小指相关的点时，根据预设基准点的坐标 和关键点的坐标可以确定小指在骨龄片中的当前位置与小指在骨龄片中的基准位置之间的对应关系，进一步也可以得到骨龄片中其他骨头的当前位置与基准位置之间的对应关系，其中对应关系包括平移关系以及旋转关系。然后根据对应关系将骨龄片中的各骨头调整至基准位置。由于定位模型是以处于基准位置的骨龄片为训练样本训练得到的，故在定位骨龄片中各目标骨骺的位置之前，先将待检测用户的骨龄片中的各骨头调整至基准位置，从而提高了定位目标骨骺的精度。
在上述步骤S102中，定位模型包括粗定位模块和细定位模块，粗定位模块和细定位模块均为深度残差网络，其中，粗定位模块是以预先标记关键点的多张骨龄片为训练样本，对深度残差网络进行训练后确定的。细定位模块以预先标记关键点的多张粗分割区域的骨龄片为训练样本，对深度残差网络进行训练后确定的，细定位模块的数量根据粗定位模块输出的粗分割区域的数量确定。
下面具体介绍粗定位模块的训练过程：获取多张骨龄片作为训练样本。针对每一张骨龄片，对骨龄片进行预处理，再将骨龄片中的各骨头调整至基准位置，其中，调整骨龄片的位置以及对骨龄片进行预处理的过程已在前文中描述，此处不再赘述。由标注人员在每张骨龄片中标注关键点，关键点为骨龄片中骨骺附近的点，每一个关键点对应一个序号。然后对训练样本进行数据增强，将数据量增强至原来数据量的10倍，数据增强的方法包括但不限于：
1、随机旋转一定角度。
2、随机上下左右平移0～30像素。
3、随机缩放0.85～1.15倍。
4、对图像对比度和亮度进行少量抖动。
之后再将训练样本输入深度残差网络进行训练。训练时根据标注的关键点的坐标和网络预测的关键点的坐标计算loss函数，通过反向传播的方法训练，训练的优化算法使用带有动量和阶梯衰减的sgd算法。
可选地，上述深度残差网络的结构如图5所示，包括N个连续的卷积特征提取块以及一个全连接层，针对N个卷积特征提取块中任意两个连续的第一卷积特征提取块和第二卷积特征提取块，第二卷积特征提取块输出的第二图像特征与第一卷积特征提取块输出的第一图像特征相加后作为第三积特征提取块的输入或者N个连续的卷积特征提取块的输出。第三卷积特征提取块为位于第二卷积特征提取块之后且与第二卷积特征提取块连续的卷积特征提取块。卷积特征提取块包括L个卷积模块，L大于0，任意一个卷积模块中包括卷积层、BN层及激励层，具体如图6所示。
下面具体介绍细定位模块的训练过程：获取多张骨龄片作为训练样本。针对每一张骨龄片，将骨龄片中的各骨头调整至基准位置，再对骨龄片进行预处理。按照预设规则从骨龄片中分割出多个粗分割区域，比如分割出拇指区域、中指区域、小指区域以及手腕区域这四个粗分割区域。针对每个粗分割区域训练一个细定位模块。针对任意一个细定位模块，由标注人员在每张粗分割区域的骨龄片中标注关键点，也可以直接使用粗分割模块对应的训练样本中已经标注的关键点，关键点为粗分割区域中骨骺附近的点，每一个关键点对应一个序号。然后将粗分割区域的骨龄片调整为正位图像。比如，拇指区域可能为与拇指方向相同的倾斜的区域，则在截取拇指区域后，将拇指区域旋转一定的角度，调整为正位图像。然后对训练样本进行数据增强，将数据量增强至原来数据量的10倍，数据增强的方法包括但不限于：
1、随机旋转一定角度。
2、随机上下左右平移0～30像素。
3、随机缩放0.85～1.15倍。
4、对图像对比度和亮度进行少量抖动。
之后再将训练样本输入深度残差网络进行训练。训练时根据标注的关键点的坐标和网络预测的关键点的坐标计算loss函数，通过反向传播的方法训练，训练的优化算法使用带有动量和阶梯衰减的sgd算法。深度残差网络包括M个连续的卷积特征提取块以及一个全连接层，卷积特征提取块如图6所示。
采用上述方法确定定位模型中的粗定位模块和细定位模块之后，采用定位模型确定骨龄片中各目标骨骺的位置，如图7所示，具体包括以下步骤：
步骤S701，将骨龄片输入所粗定位模块，确定骨龄片中各目标骨骺对应的第一关键点的坐标。
可选地，骨龄片为手骨片，第一关键点可以是手骨中各个骨骺附件的关键点，每一个第一关键点对应一个序号。
步骤S702，根据各目标骨骺对应的第一关键点的坐标确定骨龄片中的一个或多个粗分割区域，粗分割区域包括一个或多个目标骨骺。
粗分割区域可以是拇指区域、食指区域、中指区域、无名指区域、小指区域、手腕区域等等，粗分割区域的具体位置以及粗分割区域包含的目标骨骺可以预先设置。
步骤S703，针对每一个粗分割区域，将粗分割区域输入对应的细定位模块，确定粗分割区域中各目标骨骺对应的第二关键点的坐标。
步骤S704，根据粗分割区域中各目标骨骺对应的第二关键点的坐标确定各目标骨骺的位置。
由于定位模型在定位各骨骺的位置时，先采用粗定位模块从骨龄片中分割出粗分割区域，缩小目标骨骺的定位范围，然后再从粗分割区域中定位各目标骨骺的位置，从而提高定位目标骨骺的精度。
可选地，在上述步骤S702中，粗分割区域为拇指区域、中指区域、小指区域、手腕区域，则第一关键点包括拇指区域的第一关键点、中指区域的第一关键点、小指区域的第一关键点及手腕区域的第一关键点。针对每个区域的第一关键点，根据区域的第一关键点的坐标确定长轴方向和短轴方向，确定区域的中心点。然后以中心点为中心，根据长轴方向和短轴方向向外扩展预设距离，确定粗分割区域，具体如图8所示。具体实施中，可以对第一关键点的坐标进行主成分分析(principal components analysis，pca)，确定长轴方向和短轴方向。粗分割区域可以是长轴方向与短轴方向的比例为1:3的矩形。
可选地，在上述步骤S703中，将粗分割区域输入对应的细定位模块之前， 调整粗分割区域的骨龄片以满足细定位模块的要求。示例性地，从图8所示骨龄片中截取拇指区域的骨龄片，拇指区域的骨龄片如图9所示。将拇指区域的骨龄片向左旋转一定角度后得到正位图像，如图10所示。然后将拇指区域的正位图像输入细定位模块。
在上述步骤S704中，确定粗分割区域中各目标骨骺对应的第二关键点的坐标之后，根据第二关键点的序号确定每一个目标骨骺对应第二关键点，然后根据第二关键点的坐标计算目标骨骺的矩形框。
在上述步骤S103中，目标骨骺的特征采用识别模型进行识别，识别模型是以预先标记目标骨骺特征类型的多张骨骺图像为训练样本，对深度残差网络进行训练后确定的。由于各个目标骨骺的特征存在差异，因此，可以针对每一个目标骨骺训练一个识别模型，从而提高识别目标骨骺的特征的精度。
基于相同的技术构思，本申请实施例提供了一种检测骨龄的装置，如图11所示，该装置1100包括：
获取模块1101，用于获取待检测用户的骨龄片；
定位模块1102，用于采用定位模型确定所述骨龄片中各目标骨骺的位置；
处理模块1103，用于针对每一个目标骨骺，根据所述目标骨骺的特征以及所述目标骨骺的位置确定所述目标骨骺的骨龄分数；
检测模块1104，用于根据各目标骨骺的骨龄分数确定所述待检测用户的骨龄。
可选地，所述定位模型包括粗定位模块和细定位模块；
所述定位模块1102具体用于：
将所述骨龄片输入所述粗定位模块，确定所述骨龄片中各目标骨骺对应的第一关键点的坐标；
根据所述各目标骨骺对应的第一关键点的坐标确定所述骨龄片中的一个或多个粗分割区域，所述粗分割区域包括一个或多个目标骨骺；
针对每一个粗分割区域，将所述粗分割区域输入对应的细定位模块，确定所述粗分割区域中各目标骨骺对应的第二关键点的坐标；
根据所述粗分割区域中各目标骨骺对应的第二关键点的坐标确定所述各目标骨骺的位置。
可选地，所述骨龄片为手骨片；所述第一关键点包括拇指区域的第一关键点、中指区域的第一关键点、小指区域的第一关键点及手腕区域的第一关键点；
所述定位模块1102具体用于：
针对每个区域的第一关键点，根据所述区域的第一关键点的坐标确定长轴方向和短轴方向；
确定所述区域的中心点；
以所述中心点为中心，根据所述长轴方向和所述短轴方向向外扩展预设距离，确定粗分割区域。
可选地，所述定位模块1102还用于：
将所述粗分割区域输入对应的细定位模块之前，调整所述粗分割区域的骨龄片以满足所述细定位模块的要求。
基于相同的技术构思，本申请实施例提供了一种检测骨龄的设备，如图12所示，包括至少一个处理器1201，以及与至少一个处理器连接的存储器1202，本申请实施例中不限定处理器1201与存储器1202之间的具体连接介质，图12中处理器1201和存储器1202之间通过总线连接为例。总线可以分为地址总线、数据总线、控制总线等。
在本申请实施例中，存储器1202存储有可被至少一个处理器1201执行的指令，至少一个处理器1201通过执行存储器1202存储的指令，可以执行前述的检测骨龄的方法中所包括的步骤。
其中，处理器1201是检测骨龄的设备的控制中心，可以利用各种接口和线路连接检测骨龄的设备的各个部分，通过运行或执行存储在存储器1202内的指令以及调用存储在存储器1202内的数据，从而实现检测骨龄。可选的，处理器1201可包括一个或多个处理单元，处理器1201可集成应用处理器和调制解调处理器，其中，应用处理器主要处理操作系统、用户界面和应用程 序等，调制解调处理器主要处理无线通信。可以理解的是，上述调制解调处理器也可以不集成到处理器1201中。在一些实施例中，处理器1201和存储器1202可以在同一芯片上实现，在一些实施例中，它们也可以在独立的芯片上分别实现。
处理器1201可以是通用处理器，例如中央处理器(CPU)、数字信号处理器、专用集成电路(Application Specific Integrated Circuit，ASIC)、现场可编程门阵列或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件，可以实现或者执行本申请实施例中公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者任何常规的处理器等。结合本申请实施例所公开的方法的步骤可以直接体现为硬件处理器执行完成，或者用处理器中的硬件及软件模块组合执行完成。
存储器1202作为一种非易失性计算机可读存储介质，可用于存储非易失性软件程序、非易失性计算机可执行程序以及模块。存储器1202可以包括至少一种类型的存储介质，例如可以包括闪存、硬盘、多媒体卡、卡型存储器、随机访问存储器(Random Access Memory，RAM)、静态随机访问存储器(Static Random Access Memory，SRAM)、可编程只读存储器(Programmable Read Only Memory，PROM)、只读存储器(Read Only Memory，ROM)、带电可擦除可编程只读存储器(Electrically Erasable Programmable Read-Only Memory，EEPROM)、磁性存储器、磁盘、光盘等等。存储器1202是能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质，但不限于此。本申请实施例中的存储器1202还可以是电路或者其它任意能够实现存储功能的装置，用于存储程序指令和/或数据。
基于同一发明构思，本申请实施例还提供了一种计算机可读介质，其存储有可由检测骨龄的设备执行的计算机程序，当所述程序在检测骨龄的设备上运行时，使得所述检测骨龄的设备执行检测骨龄的方法的步骤。
基于同一发明构思，本申请实施例还提供了一种计算机程序产品，所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算程序，所述 计算机程序包括程序指令，当所述程序指令被计算机执行时，使所述计算机执行检测骨龄的方法。
本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
尽管已描述了本发明的优选实施例，但本领域内的技术人员一旦得知了基本创造性概念，则可对这些实施例作出另外的变更和修改。所以，所附权利要求意欲解释为包括优选实施例以及落入本发明范围的所有变更和修改。
显然，本领域的技术人员可以对本发明进行各种改动和变型而不脱离本 发明的精神和范围。这样，倘若本发明的这些修改和变型属于本发明权利要求及其等同技术的范围之内，则本发明也意图包含这些改动和变型在内。
本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器，使得通过该计算机或其他可编程数据处理设备的处理器执行的指令可实现流程图中的一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图的一个流程或多个流程和/或方框图的一个方框或多个方框中指定的功能的步骤。
尽管已描述了本发明的优选实施例，但本领域内的技术人员一旦得知了基本创造性概念，则可对这些实施例作出另外的变更和修改。所以，所附权利要求意欲解释为包括优选实施例以及落入本发明范围的所有变更和修改。
显然，本领域的技术人员可以对本发明进行各种改动和变型而不脱离本 发明的精神和范围。这样，倘若本发明的这些修改和变型属于本发明权利要求及其等同技术的范围之内，则本发明也意图包含这些改动和变型在内。

Claims (11)

  1. 一种检测骨龄的方法，其特征在于，包括：
    获取待检测用户的骨龄片；
    采用定位模型确定所述骨龄片中各目标骨骺的位置；
    针对每一个目标骨骺，根据所述目标骨骺的特征以及所述目标骨骺的位置确定所述目标骨骺的骨龄分数；
    根据各目标骨骺的骨龄分数确定所述待检测用户的骨龄。
  2. 如权利要求1所述的方法，其特征在于，所述定位模型包括粗定位模块和细定位模块；
    所述采用定位模型确定所述骨龄片中各目标骨骺的位置，包括：
    将所述骨龄片输入所述粗定位模块，确定所述骨龄片中各目标骨骺对应的第一关键点的坐标；
    根据所述各目标骨骺对应的第一关键点的坐标确定所述骨龄片中的一个或多个粗分割区域，所述粗分割区域包括一个或多个目标骨骺；
    针对每一个粗分割区域，将所述粗分割区域输入对应的细定位模块，确定所述粗分割区域中各目标骨骺对应的第二关键点的坐标；
    根据所述粗分割区域中各目标骨骺对应的第二关键点的坐标确定所述各目标骨骺的位置。
  3. 如权利要求2所述的方法，其特征在于，所述骨龄片为手骨片；所述第一关键点包括拇指区域的第一关键点、中指区域的第一关键点、小指区域的第一关键点及手腕区域的第一关键点；
    所述根据所述各目标骨骺对应的第一关键点的坐标确定所述骨龄片中的一个或多个粗分割区域，包括：
    针对每个区域的第一关键点，根据所述区域的第一关键点的坐标确定长轴方向和短轴方向；
    确定所述区域的中心点；
    以所述中心点为中心，根据所述长轴方向和所述短轴方向向外扩展预设距离，确定粗分割区域。
  4. 如权利要求3所述的方法，其特征在于，将所述粗分割区域输入对应的细定位模块之前，还包括：
    调整所述粗分割区域的骨龄片以满足所述细定位模块的要求。
  5. 一种检测骨龄的装置，其特征在于，包括：
    获取模块，用于获取待检测用户的骨龄片；
    定位模块，用于采用定位模型确定所述骨龄片中各目标骨骺的位置；
    处理模块，用于针对每一个目标骨骺，根据所述目标骨骺的特征以及所述目标骨骺的位置确定所述目标骨骺的骨龄分数；
    检测模块，用于根据各目标骨骺的骨龄分数确定所述待检测用户的骨龄。
  6. 如权利要求5所述的装置，其特征在于，所述定位模型包括粗定位模块和细定位模块；
    所述定位模块具体用于：
    将所述骨龄片输入所述粗定位模块，确定所述骨龄片中各目标骨骺对应的第一关键点的坐标；
    根据所述各目标骨骺对应的第一关键点的坐标确定所述骨龄片中的一个或多个粗分割区域，所述粗分割区域包括一个或多个目标骨骺；
    针对每一个粗分割区域，将所述粗分割区域输入对应的细定位模块，确定所述粗分割区域中各目标骨骺对应的第二关键点的坐标；
    根据所述粗分割区域中各目标骨骺对应的第二关键点的坐标确定所述各目标骨骺的位置。
  7. 如权利要求6所述的装置，其特征在于，所述骨龄片为手骨片；所述第一关键点包括拇指区域的第一关键点、中指区域的第一关键点、小指区域的第一关键点及手腕区域的第一关键点；
    所述定位模块具体用于：
    针对每个区域的第一关键点，根据所述区域的第一关键点的坐标确定长 轴方向和短轴方向；
    确定所述区域的中心点；
    以所述中心点为中心，根据所述长轴方向和所述短轴方向向外扩展预设距离，确定粗分割区域。
  8. 如权利要求7所述的装置，其特征在于，所述定位模块还用于：
    将所述粗分割区域输入对应的细定位模块之前，调整所述粗分割区域的骨龄片以满足所述细定位模块的要求。
  9. 一种检测骨龄的设备，其特征在于，包括至少一个处理器、以及至少一个存储器，其中，所述存储单元存储有计算机程序，当所述程序被所述处理器执行时，使得所述处理器执行权利要求1～4任一权利要求所述方法的步骤。
  10. 一种计算机可读介质，其特征在于，其存储有可由检测骨龄的设备执行的计算机程序，当所述程序在检测骨龄的设备上运行时，使得所述检测骨龄的设备执行权利要求1～4任一所述方法的步骤。
  11. 一种计算机程序产品，其特征在于，所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算程序，所述计算机程序包括程序指令，当所述程序指令被计算机执行时，使所述计算机执行权利要求1～4任一所述方法。
PCT/CN2019/082681 2018-09-30 2019-04-15 一种检测骨龄的方法及装置 WO2020062839A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
SG11202002139SA SG11202002139SA (en) 2018-09-30 2019-04-15 Method and device for detecting bone age

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811163000.8 2018-09-30
CN201811163000.8A CN109377484B (zh) 2018-09-30 2018-09-30 一种检测骨龄的方法及装置

Publications (1)

Publication Number Publication Date
WO2020062839A1 true WO2020062839A1 (zh) 2020-04-02

Family

ID=65403564

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/082681 WO2020062839A1 (zh) 2018-09-30 2019-04-15 一种检测骨龄的方法及装置

Country Status (3)

Country Link
CN (1) CN109377484B (zh)
SG (1) SG11202002139SA (zh)
WO (1) WO2020062839A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111507953A (zh) * 2020-04-13 2020-08-07 武汉华晨酷神智能科技有限公司 一种ai骨龄快速识别方法
CN114601483A (zh) * 2022-05-11 2022-06-10 山东第一医科大学第一附属医院(山东省千佛山医院) 一种基于图像处理的骨龄分析方法及系统

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109377484B (zh) * 2018-09-30 2022-04-22 杭州依图医疗技术有限公司 一种检测骨龄的方法及装置
CN109994172A (zh) * 2019-03-06 2019-07-09 杭州津禾生物科技有限公司 骨龄在线数字化测算及身高数据库管理系统
CN113100791A (zh) * 2020-01-13 2021-07-13 上海西门子医疗器械有限公司 X-射线医学成像方法和系统以及计算机存储介质
CN111325734B (zh) * 2020-02-25 2023-06-27 杭州电子科技大学 一种基于视觉模型的骨龄预测方法及装置
CN111563874B (zh) * 2020-03-05 2023-04-28 北京深睿博联科技有限责任公司 骨龄评测方法和装置
CN111882517A (zh) * 2020-06-08 2020-11-03 杭州深睿博联科技有限公司 一种基于图卷积神经网络的骨龄评测方法、系统、终端及存储介质
CN113298842B (zh) * 2021-05-28 2023-07-14 平安科技（深圳）有限公司 骨龄预测方法、装置、设备及计算机可读存储介质
CN117252881B (zh) * 2023-11-20 2024-01-26 四川大学 基于手部x光图像的骨龄预测方法、系统、设备及介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102945545A (zh) * 2012-10-18 2013-02-27 重庆医科大学 一种鲁棒的骨龄评定图像预处理及骨关键点定位方法
CN107895367A (zh) * 2017-11-14 2018-04-10 中国科学院深圳先进技术研究院 一种骨龄识别方法、系统及电子设备
CN108334899A (zh) * 2018-01-28 2018-07-27 浙江大学 基于手骨x光片骨骼和关节量化信息集成的骨龄评估方法
CN109377484A (zh) * 2018-09-30 2019-02-22 杭州依图医疗技术有限公司 一种检测骨龄的方法及装置

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5928748B2 (ja) * 2014-07-31 2016-06-01 インターナショナル・ビジネス・マシーンズ・コーポレーションＩｎｔｅｒｎａｔｉｏｎａｌ Ｂｕｓｉｎｅｓｓ Ｍａｃｈｉｎｅｓ Ｃｏｒｐｏｒａｔｉｏｎ 同一種類の複数の認識対象物体が検索対象画像中に存在する場合に、それぞれの認識対象物体の位置および向きを精度良く求める手法
CN111671448A (zh) * 2017-12-08 2020-09-18 浙江大学 一种基于深度学习的骨龄检测方法和装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102945545A (zh) * 2012-10-18 2013-02-27 重庆医科大学 一种鲁棒的骨龄评定图像预处理及骨关键点定位方法
CN107895367A (zh) * 2017-11-14 2018-04-10 中国科学院深圳先进技术研究院 一种骨龄识别方法、系统及电子设备
CN108334899A (zh) * 2018-01-28 2018-07-27 浙江大学 基于手骨x光片骨骼和关节量化信息集成的骨龄评估方法
CN109377484A (zh) * 2018-09-30 2019-02-22 杭州依图医疗技术有限公司 一种检测骨龄的方法及装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111507953A (zh) * 2020-04-13 2020-08-07 武汉华晨酷神智能科技有限公司 一种ai骨龄快速识别方法
CN114601483A (zh) * 2022-05-11 2022-06-10 山东第一医科大学第一附属医院(山东省千佛山医院) 一种基于图像处理的骨龄分析方法及系统

Also Published As

Publication number Publication date
CN109377484A (zh) 2019-02-22
CN109377484B (zh) 2022-04-22
SG11202002139SA (en) 2020-05-28

Similar Documents

Publication Publication Date Title
WO2020062839A1 (zh) 一种检测骨龄的方法及装置
WO2020062840A1 (zh) 一种检测骨龄的方法及装置
CN109272002B (zh) 一种骨龄片的分类方法及装置
US11842487B2 (en) Detection model training method and apparatus, computer device and storage medium
CN110874594B (zh) 基于语义分割网络的人体外表损伤检测方法及相关设备
CN109741309B (zh) 一种基于深度回归网络的骨龄预测方法及装置
CN109509177B (zh) 一种脑部影像识别的方法及装置
CN111291825B (zh) 病灶分类模型训练方法、装置、计算机设备和存储介质
CN109255786B (zh) 一种检测骨龄的方法及装置
CN110533639B (zh) 一种关键点定位方法及装置
CN109376756B (zh) 基于深度学习的上腹部转移淋巴结节自动识别系统、计算机设备、存储介质
Almazroa et al. Optic disc segmentation for glaucoma screening system using fundus images
CN110689551B (zh) 四肢骨分割方法、装置、电子设备及可读存储介质
CN111915609A (zh) 病灶检测分析方法、装置、电子设备及计算机存储介质
CN110796659B (zh) 一种目标检测结果的鉴别方法、装置、设备及存储介质
CN116542966B (zh) 用于儿童内分泌异常检测的骨龄智能分析方法
Lee et al. Deep learning based cephalometric landmark identification using landmark-dependent multi-scale patches
CN110555856A (zh) 一种基于深度神经网络的黄斑水肿病变区域分割方法
CN112529900A (zh) 匹配乳腺图像中roi的方法、装置、终端与存储介质
CN116258671B (zh) 一种基于mr图像智能勾画方法、系统、设备及存储介质
Dickson et al. A Dual Channel Multiscale Convolution U-Net Methodfor Liver Tumor Segmentation from Abdomen CT Images
CN115880358A (zh) 定位模型的构建方法、影像标志点的定位方法及电子设备
WO2023047117A1 (en) A computer-implemented method, data processing apparatus, and computer program for active learning for computer vision in digital images
CN112529918B (zh) 一种脑部ct图像中脑室区域分割的方法、装置及设备
CN110516715B (zh) 一种手骨分类方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19867624

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19867624

Country of ref document: EP

Kind code of ref document: A1