WO2023070946A1 - 基于超声图像的测量设备、方法、介质及电子设备 - Google Patents

基于超声图像的测量设备、方法、介质及电子设备 Download PDF

Info

Publication number
WO2023070946A1
WO2023070946A1 PCT/CN2022/070646 CN2022070646W WO2023070946A1 WO 2023070946 A1 WO2023070946 A1 WO 2023070946A1 CN 2022070646 W CN2022070646 W CN 2022070646W WO 2023070946 A1 WO2023070946 A1 WO 2023070946A1
Authority
WO
WIPO (PCT)
Prior art keywords
scale
target
target object
ultrasound image
key points
Prior art date
Application number
PCT/CN2022/070646
Other languages
English (en)
French (fr)
Inventor
叶秋意
谷晓林
Original Assignee
上海杏脉信息科技有限公司
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 上海杏脉信息科技有限公司 filed Critical 上海杏脉信息科技有限公司
Publication of WO2023070946A1 publication Critical patent/WO2023070946A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/44—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device
    • A61B8/4411—Device being modular
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/5223—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for extracting a diagnostic or physiological parameter from medical diagnostic data

Definitions

  • the invention relates to a measuring device, in particular to a measuring device, method, medium and electronic device based on an ultrasonic image.
  • Ultrasound imaging is the use of ultrasound beams to scan the human body, and obtain images of internal organs by receiving and processing reflected signals.
  • ultrasound imaging methods are often used to determine the location, size, and shape of organs, and to determine the scope and physical properties of lesions. For example, in the scene of ultrasonic detection of thyroid, breast, liver and gallbladder, etc., if nodules or other lesions are found, the size of the lesions needs to be evaluated.
  • the ultrasound images displayed by ultrasound equipment from different manufacturers are of different sizes, and the size ratio between the image display and the real human organs is different. Medical personnel need to visually observe the proportional relationship between the ultrasound image and the real human body and identify the The size of the lesion was visually measured after the lesion. Even if the ultrasound equipment provided by some manufacturers can display the reference scale line while displaying the image, the display methods and positions of the ultrasound equipment of different manufacturers are different, and the scaling ratio of the real human body is different, which requires medical staff to use an ultrasound Learn to understand the image display characteristics of the device before installing the device, otherwise it may lead to inaccurate assessment of the patient's lesion size. This method is inefficient and prone to measurement errors caused by human subjective factors such as doctor experience.
  • the purpose of the present invention is to provide a measurement device, method, medium and electronic device based on ultrasonic images, which are used to solve the above-mentioned problems in the prior art.
  • the first aspect of the present invention provides a measurement device based on ultrasonic images
  • the device includes: an ultrasonic image acquisition module for acquiring target ultrasonic images; a scale key point acquisition module for Obtain the scale key points in the target ultrasonic image; the scale acquisition module is used to obtain the scale of the target ultrasonic image according to the distance on the map and the actual distance between the scale key points; the target object detection module is used to detect The target object in the target ultrasound image; the object size measurement module, used to obtain the actual size of the target object according to the size on the map of the target object and the scale; The key points of the scale, the scale bar, the detection result of the target object and/or the actual size of the target object are superimposed on the image.
  • the scale key point acquisition module uses a neural network model to process the target ultrasound image to obtain the scale key point.
  • the target object detection module includes: a target object acquisition unit configured to segment the target ultrasound image to obtain a model of the target object; a target parameter acquisition unit configured to The target parameter is obtained according to the model of the target object, wherein the target parameter is used to reflect the size of the target object on the map.
  • the target object detection module includes: a detection frame acquisition unit, configured to acquire a detection frame of the target object according to the target ultrasonic image; a target parameter acquisition unit, configured to obtain a detection frame of the target object according to the The detection frame of the target object acquires target parameters, wherein the target parameters are used to reflect the size of the target object on the map.
  • the scale acquisition module includes: a scale key point identification unit, which is used to identify the scale key point according to the ordinate of the scale key point; the distance acquisition unit on the map , for acquiring the distance on the map between the key points of the scale according to the identification of the key points of the scale; the scale acquisition unit is used for obtaining the target ultrasound according to the distance on the map and the actual distance between the key points of the scale The scale of the image.
  • the scale key point identification unit includes: a scale key point sorting subunit, configured to sort the scale key points according to the ordinate of the scale key point;
  • the point identification subunit is configured to sequentially identify the key points of the scale according to the sorting result of the key points of the scale.
  • the scale obtaining unit obtains the scale according to the distance on the map and the actual distance between two key points of the scale, or, the scale obtaining unit obtains the scale according to a plurality of The scale bar is obtained from the average on-map distance and the average actual distance between the scale key points.
  • a second aspect of the present invention provides a measurement method based on ultrasonic images, the method comprising: acquiring a target ultrasonic image; acquiring key points on a scale in the target ultrasonic image; Acquire the scale of the target ultrasound image from the actual distance; detect the target object in the target ultrasound image; acquire the actual size of the target object according to the size on the map of the target object and the scale; The key points of the scale, the scale bar, the detection result of the target object and/or the actual size of the target object are superimposed on the image.
  • a third aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the ultrasonic image-based measurement method described in the second aspect of the present invention is implemented.
  • a fourth aspect of the present invention provides an electronic device, which includes: a memory storing a computer program; a processor connected to the memory in communication, and executing the computer program based on the second aspect of the present invention when calling the computer program. Measurement methods of ultrasound images.
  • the ultrasonic image-based measurement device in one or more embodiments of the present invention has the following beneficial effects:
  • the ultrasonic image-based measurement device can obtain the scale of the target ultrasonic image based on the scale key points in the target ultrasonic image, and the actual size of the target object can be obtained based on the scale and the size of the target object on the map. This process basically does not require manual participation, has high efficiency, and can avoid measurement errors caused by human subjective reasons.
  • the measurement device based on ultrasonic images can be externally connected to different ultrasonic devices, as long as the ultrasonic images output by these ultrasonic devices can be obtained, they can be processed. Therefore, the measurement device based on ultrasonic images supports of Instant Connect.
  • FIG. 1A is a schematic structural diagram of an embodiment of the ultrasonic image-based measurement device of the present invention.
  • FIG. 1B is an example diagram of a target ultrasound image in a specific embodiment of the measurement device based on ultrasound images according to the present invention.
  • FIG. 2 is a flow chart of the training method of the first neural network model in a specific embodiment of the ultrasonic image-based measuring device of the present invention.
  • FIG. 3A is a schematic structural diagram of a target object detection module in a specific embodiment of the ultrasonic image-based measurement device of the present invention.
  • FIG. 3B is a flow chart of the training method of the second neural network model in a specific embodiment of the ultrasonic image-based measuring device of the present invention.
  • FIG. 3C is a schematic structural diagram of a target object detection module in another specific embodiment of the ultrasonic image-based measurement device of the present invention.
  • FIG. 4A is a schematic structural diagram of a scale acquisition module in a specific embodiment of the ultrasonic image-based measuring device of the present invention.
  • FIG. 4B is a schematic structural diagram of a scale key point marking unit in a specific embodiment of the ultrasonic image-based measuring device of the present invention.
  • Fig. 4C is a diagram showing an example of the marking results of key points on the scale in a specific embodiment of the ultrasonic image-based measuring device according to the present invention.
  • FIG. 5 is a flow chart of an embodiment of the ultrasonic image-based measurement method of the present invention.
  • FIG. 6 is a schematic structural diagram of an electronic device according to a specific embodiment of the present invention.
  • the measurement device 1 based on ultrasonic images includes an ultrasonic image acquisition module 10, a scale key point acquisition module 11, a scale An acquisition module 12 , a target object detection module 13 , an object size measurement module 14 and an overlay display module 15 .
  • the ultrasound image acquisition module 10 is used to acquire target ultrasound images. Specifically, the ultrasound image acquiring module 10 may be connected to the ultrasound equipment in communication, and acquire the ultrasound image output by the ultrasound equipment as the target ultrasound image.
  • the ultrasound image acquisition module 10 is preferably able to be connected to multiple different types of ultrasound equipment, and can acquire ultrasound images output by these ultrasound equipment in different display modes, different display positions and/or different scaling ratios as the target ultrasound image .
  • the scale key point acquisition module 11 is connected to the ultrasound image acquisition module 10 for acquiring scale key points in the target ultrasound image.
  • FIG. 1B is an example diagram of a target ultrasonic image involved in this embodiment, wherein the target ultrasonic image includes an image body and scale key points, and the scale key points refer to the Points on the indicated scale, usually automatically generated by ultrasound scanning equipment.
  • only one scale key point is marked in FIG. 1B , and it can be understood that the white point below the scale key point in the figure is also a scale key point.
  • the scale acquisition module 12 is connected to the scale key point acquisition module 11, and is used to acquire the scale of the target ultrasound image according to the distance on the map and the actual distance between the scale key points.
  • the scale is, for example, the ratio of the distance on the map between the key points of the scale and the actual distance.
  • the target object detection module 13 is used to detect the target object in the target ultrasound image, more specifically, to detect the size of the target object in the target ultrasound image, wherein the target object is, for example, For thyroid nodules, the size includes but not limited to length, circumference, area, volume, etc.
  • the object size measurement module 14 is connected to the scale acquisition module 12 and the target object detection module 13, and is used for obtaining the actual size of the target object according to the size on the map of the target object and the scale. For example, when the minimum detection frame is used to realize the detection of the target object, and the minimum detection frame is a rectangle, the object size measurement module 14 can acquire the The actual length and actual width of the minimum detection frame are obtained to obtain the actual radial size of the target object.
  • the superimposed display module 15 is connected with the ultrasonic image acquisition module 10, the scale key point acquisition module 11, the scale acquisition module 12, the target object detection module 13 and the object size measurement module 14, for The key point of the scale, the scale bar, the detection result of the target object (such as the minimum detection frame, the outline of the lesion) and/or the actual size of the target object are superimposed and displayed in the target ultrasound image.
  • the connection manners between the superimposed display module 15 and the above other modules include direct connection and indirect connection through other modules.
  • the overlay display module 15 may be directly connected to the object size measurement module 14 , or indirectly connected to the scale acquisition module 12 through the object size measurement module 14 .
  • the superimposed display module 15 is also used to determine the display position and/or display mode of the key point of the scale, the scale, the detection result of the target object and/or the actual size of the target object, so as to Avoid occlusion between displayed content.
  • the overlay display module 15 may display the scale key points, the scale bar and the actual size of the target object on one side of the target ultrasound image.
  • the ultrasonic image-based measurement device in this embodiment can obtain the scale of the target ultrasonic image based on the scale key points in the target ultrasonic image, and based on the scale and the size of the target object on the map, it can be obtained The actual size of the target object.
  • This process basically does not require manual participation, has high efficiency, and can avoid measurement errors caused by human subjective reasons.
  • the ultrasonic image measuring device 1 further includes a housing, and the hardware structures corresponding to the above-mentioned modules of the ultrasonic image measuring device 1 are arranged in the housing.
  • the housing has a communication interface, through which the ultrasonic image acquisition module 10 is communicatively connected with the ultrasound equipment, so as to acquire the target ultrasound image from the ultrasound equipment.
  • the ultrasonic image measuring device 1 can be connected and used immediately with the ultrasonic device, that is, as long as the ultrasonic image measuring device 1 is externally connected to the ultrasonic device, the ultrasonic image measuring device 1 can
  • the ultrasonic image output by the ultrasonic device is processed, so as to obtain and display the corresponding key points of the scale, the scale, the detection result of the target object and/or the actual size of the target object.
  • the scale key point acquisition module 11 uses a first neural network model to process the target ultrasound image to obtain the scale key point, wherein the first neural network model is A trained neural network model.
  • the training method of the first neural network model described in this embodiment includes:
  • the initial model of the neural network in this embodiment is used to extract the key points of the scale in the ultrasound image.
  • a deep learning network model such as U-Net or V-Net can be used.
  • a structure of the target object detection module 13 includes a target object acquisition unit 131 and a target parameter acquisition unit 132 .
  • the target object acquisition unit 131 is configured to segment the target ultrasound image to obtain a model of the target object.
  • the target object obtaining unit 131 uses a second neural network model to process the target ultrasound image to realize segmentation of the target ultrasound image, so as to obtain a model of the target object.
  • the implementation method for training the second neural network model in this embodiment includes:
  • the initial neural network model in this embodiment is used to implement a segmentation algorithm for the target object.
  • a deep learning model such as U-Net or V-Net may be used.
  • the acquisition of the target object model by the target object acquisition unit 131 using the second neural network model is only a feasible manner of this embodiment, but the present invention is not limited thereto.
  • the target parameter acquisition unit 132 is connected to the target object acquisition unit 131 and configured to acquire target parameters according to the target object model, wherein the target parameter is used to reflect the size of the target object on the map.
  • the target parameters include, but are not limited to, length, perimeter, area, volume, and the like.
  • the target parameter obtaining unit 132 may obtain a manually marked parameter value as the target parameter.
  • the target parameter acquisition unit 132 may use traditional algorithms such as key point detection to acquire the target parameters.
  • the target parameter acquiring unit 132 may acquire the maximum diameter and the minimum diameter of the model of the target object obtained after image segmentation as the target parameters.
  • another structure of the target object detection module 13 includes a detection frame acquisition unit 133 and a target parameter acquisition unit 134 .
  • the detection frame obtaining unit 133 is configured to obtain a detection frame of the target object according to the target ultrasound image. Specifically, the detection frame obtaining unit 133 can use a trained deep learning model, a neural network model, etc. to process the target ultrasound image to obtain the detection frame of the target object, or obtain the detection frame of the target object according to the labeling result input by the user. To obtain the detection frame of the target object, the boundary of the target object can also be obtained according to the model of the target object, and then the detection frame of the target object can be obtained according to the pixels on the boundary, but the present invention is not limited to the above-mentioned several In practical applications, other techniques may also be used to obtain the detection frame of the target object.
  • the detection frame of the target object may be of any shape, and the detection frame of the target object completely includes the target object.
  • the detection frame of the target object is the minimum detection frame (or called the minimum bounding frame) of the target object, the minimum detection frame completely contains the target object, and the minimum detection frame is preferably a polygon, further Preferably rectangular.
  • the target parameter acquisition unit 134 is connected to the detection frame acquisition unit 133 and configured to acquire target parameters according to the detection frame of the target object, wherein the target parameter is used to reflect the size of the target object on the map.
  • the target parameter acquisition unit 134 may acquire geometric dimensions such as width, height, area, and perimeter of the detection frame as the target parameter.
  • the two schemes for obtaining the target parameters listed above are only optional schemes of the present invention, but the present invention is not limited thereto. In practical applications, other methods can also be used to obtain the target parameters, such as , the target parameters can be obtained by drawing outlines, displaying meridians, and the like.
  • the scale acquisition module 12 includes a scale key point identification unit 121 , a distance acquisition unit 122 on the map, and a scale acquisition unit 123 .
  • the scale key point identification unit 121 is connected to the scale key point acquisition module 11, and is used to identify the scale key point according to the ordinate of the scale key point, wherein the identification of the scale key point is used for
  • the vertical relative positions between the scale key points are marked, and the distance on the longitudinal map between different scale key points can be obtained according to the marks. For example, if the labels of two scale key points are a and b respectively, the distance between them on the graph is a-b, where a and b are preferably positive integers.
  • the on-map distance acquisition unit 122 is connected to the scale key point identification unit 121, and is configured to acquire the distance on the longitudinal map between the scale key points according to the identification of the scale key points.
  • the scale acquiring unit 123 is connected to the on-map distance acquiring unit 122, and is configured to acquire the scale of the target ultrasound image according to the longitudinal on-map distance and the longitudinal actual distance between the key points of the scale.
  • the longitudinal actual distance of the key points of the scale can be obtained according to the position, number and/or coordinates of the key points of the scale.
  • longitudinal direction in this embodiment refers to the direction substantially parallel to the line connecting the key points of the scale.
  • the scale acquisition module in this embodiment further includes a preprocessing unit, configured to preprocess the scale key points to delete abnormal points therein.
  • the abnormal point is, for example, a key point with an abnormal abscissa.
  • the scale key point can be used as an abnormal point and delete.
  • the scale key point identification unit 121 is connected to the scale key point acquisition module 11 through the preprocessing unit, and the scale acquisition unit 123 is connected to the scale key point acquisition module 11 through the preprocessing unit. In this way, it can be ensured that the scale acquisition unit acquires the scale of the target image only according to the distance on the map and the actual distance between the key points of the normal scale, which is beneficial to improve the accuracy of the acquired scale.
  • the scale key point identification unit 121 includes a scale key point sorting subunit 1211 and a scale key point identification subunit 1212 .
  • the scale key point sorting subunit 1211 is connected to the scale key point acquisition module 11, and is used to sort the scale key points according to the ordinate of the scale key point, for example, according to the ordinate from large to small Or sort the scale key points in ascending order.
  • the scale key point identification subunit 1212 is connected to the scale key point sorting subunit 1211, and is used to sequentially identify the scale key points according to the sorting of the scale key points. For example, when the key point sorting subunit 1211 sorts the scale key points in descending order of the vertical coordinates, the scale key point identification subunit 1212 may sequentially identify the scale key points as 0 , 1, ..., N, wherein, N is a positive integer.
  • N is a positive integer.
  • the scale acquiring unit 124 can acquire the scale according to the distance on the map and the actual distance between two key points of the scale, or the scale acquiring unit 124 can also acquire the scale according to multiple The scale bar is obtained by the average on-map distance and the average actual distance between the scale key points.
  • the scale acquiring unit 124 may also acquire the scale by other methods, which is not limited in the present invention.
  • the object size measurement module determines the scale of the target ultrasound image according to the following formula:
  • L is the scale bar of the target ultrasound image
  • l i, j is the longitudinal actual distance between the i-th scale key point and the j-th scale key point, that is, the i-th scale key point and the j-th scale key point
  • the actual distance of the scale key point in the vertical direction, the distance value can be obtained according to the vertical coordinates of the two scale key points
  • l m, n is the vertical actual distance between the mth scale key point and the nth scale key point, That is, the actual distance between the mth scale key point and the nth scale key point in the vertical direction, the distance value can be obtained according to the vertical coordinate
  • the object size measurement module in this embodiment can be based on any two scale key point pairs (i, j) and (m, n) Get a scale value. Based on this, the object size measurement module preferably acquires values of multiple scales according to different scale key point pairs, and takes an average value of these values as the scale of the target ultrasound image.
  • the method provided by this embodiment can avoid the problem that the acquired scale is not accurate enough due to the omission of the key point of the scale, and is beneficial to further improve the accuracy of the result.
  • the target object is a thyroid nodule
  • the ultrasonic image-based measurement device can obtain its upper and lower diameter, width diameter and height diameter according to the ultrasound image of the thyroid nodule, and then obtain its aspect ratio , where the aspect ratio is height diameter/width diameter. Based on the upper and lower diameter, width diameter, height diameter and aspect ratio of thyroid nodules, the malignancy can be obtained. This process basically does not require manual participation, and the efficiency is high and there will be no errors caused by human subjective factors.
  • the present invention also provides a measurement method based on ultrasonic images.
  • the measurement method based on an ultrasonic image in an embodiment of the present invention includes:
  • steps S51-S56 in this embodiment correspond one-to-one to the corresponding modules in the ultrasonic image-based measuring device 1 shown in FIG. 1A .
  • step S54 can be executed first and then Steps S52 and S53 may be executed, and steps S54 and S52 may also be executed at the same time, and the present invention does not limit the relative execution order of these steps.
  • the present invention also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the ultrasonic image-based measuring method shown in FIG. 5 is realized.
  • the present invention also provides an electronic device.
  • the electronic device 600 includes a memory 610 and a processor 620, wherein the memory 610 stores a computer program, and the processor 620 is connected to the memory 610 in communication , implementing the measurement method based on the ultrasonic image shown in FIG. 5 when the computer program is called.
  • the electronic device 600 further includes a display 630, which is connected in communication with the memory 610 and the processor 620, and is configured to display a related GUI interactive interface of the measurement method based on an ultrasound image.
  • a display 630 which is connected in communication with the memory 610 and the processor 620, and is configured to display a related GUI interactive interface of the measurement method based on an ultrasound image.
  • the present invention also provides a measurement device based on ultrasonic images, which can implement the measurement method based on ultrasonic images in the present invention, but the implementation hardware of the measurement method based on ultrasonic images in the present invention Including but not limited to the structure of the ultrasonic image-based measuring equipment listed in this embodiment, all structural deformations and replacements in the prior art based on the principle of the present invention are included in the scope of protection of the present invention.
  • the ultrasonic image-based measuring device of the present invention can obtain the scale of the target ultrasonic image based on the key points of the scale in the target ultrasonic image, and the actual size of the target object can be obtained based on the scale and the minimum detection frame of the target object. size. This process basically does not require manual participation, has high efficiency, and can avoid measurement errors caused by human subjective reasons.
  • the measurement equipment based on ultrasonic images can also use the preprocessing unit to delete the key points of the scale. Existing abnormal points, so as to avoid errors caused by misidentification.
  • the scale acquisition module can pass This formula is used to determine the scale, so as to avoid errors caused by missing key points of the scale. Therefore, the ultrasonic image-based measuring device of the present invention has good accuracy.
  • the measurement equipment based on ultrasonic images can be externally connected to different ultrasonic equipment, as long as the ultrasonic images output by these ultrasonic equipment can be obtained, they can be processed. Therefore, the measurement equipment based on ultrasonic images supports Connect-and-play devices.
  • the present invention effectively overcomes various shortcomings in the prior art and has high industrial application value.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Medical Informatics (AREA)
  • Veterinary Medicine (AREA)
  • Biophysics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Public Health (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Physics & Mathematics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Physiology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

本发明提供一种基于超声图像的测量设备、方法、介质及电子设备。所述设备包括：超声图像获取模块，用于获取目标超声图像；刻度关键点获取模块，用于获取目标超声图像中的刻度关键点；比例尺获取模块，用于根据刻度关键点之间的图上距离与实际距离获取目标超声图像的比例尺；目标对象检测模块，用于检测目标超声图像中的目标对象；对象尺寸测量模块，用于根据目标对象的图上尺寸和比例尺获取目标对象的实际尺寸；叠加显示模块，用于在目标超声图像中叠加显示刻度关键点、比例尺、目标对象的检测结果和/或目标对象的实际尺寸。所述设备能够自动获取目标对象的实际尺寸，该过程基本无需人工参与，效率较高。

Description

基于超声图像的测量设备、方法、介质及电子设备 技术领域
本发明涉及一种测量设备，特别是涉及一种基于超声图像的测量设备、方法、介质及电子设备。
背景技术
超声成像是利用超声声束扫描人体，通过对反射信号的接收、处理以获得体内器官的图象。实际应用中，超声成像方法常用来判断脏器的位置、大小、形态，确定病灶的范围和物理性质等。例如，在对甲状腺、乳腺、肝胆等部位进行超声检测的场景中，如发现结节或者其他病灶，需要评估病灶的尺寸。
现有技术中，不同厂商的超声设备所显示的超声图像的大小尺寸不同，图像显示与人体真实器官的大小比例不同，医务人员需要凭经验目测超声图像和真实人体之间的比例关系、识别出病灶后目测病灶的大小尺寸。即使有的厂商提供的超声设备可以在显示图像的同时显示参考刻度线，但不同厂商的的超声设备显示方式和显示位置不同、相对真实人体的缩放比例不同，这就要求医务人员使用一款超声设备前先学习了解该设备图像显示特点，否则可能导致对患者病灶尺寸的评估不够准确。此种方式效率较低，且容易出现由于医生经验等人为主观因素所导致的测量误差。
发明内容
鉴于以上所述现有技术的缺点，本发明的目的在于提供一种基于超声图像的测量设备、方法、介质及电子设备，用于解决现有技术中存在的上述问题。
为实现上述目的及其他相关目的，本发明的第一方面提供一种基于超声图像的测量设备，所述设备包括：超声图像获取模块，用于获取目标超声图像；刻度关键点获取模块，用于获取所述目标超声图像中的刻度关键点；比例尺获取模块，用于根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺；目标对象检测模块，用于检测所述目标超声图像中的目标对象；对象尺寸测量模块，用于根据所述目标对象的图上尺寸和所述比例尺获取所述目标对象的实际尺寸；叠加显示模块，用于在所述目标超声图像中叠加显示所述刻度关键点、所述比例尺、所述目标对象的检测结果和/或所述目标对象的实际尺寸。
于所述第一方面的一实施例中，所述刻度关键点获取模块利用一神经网络模型对所述目 标超声图像进行处理，以得到所述刻度关键点。
于所述第一方面的一实施例中，所述目标对象检测模块包括：目标对象获取单元，用于对所述目标超声图像进行分割以得到所述目标对象的模型；目标参数获取单元，用于根据所述目标对象的模型获取目标参数，其中，所述目标参数用于反映所述目标对象的图上尺寸。
于所述第一方面的一实施例中，所述目标对象检测模块包括：检测框获取单元，用于根据所述目标超声图像获取所述目标对象的检测框；目标参数获取单元，用于根据所述目标对象的检测框获取目标参数，其中，所述目标参数用于反映所述目标对象的图上尺寸。
于所述第一方面的一实施例中，所述比例尺获取模块包括：刻度关键点标识单元，用于根据所述刻度关键点的纵坐标对所述刻度关键点进行标识；图上距离获取单元，用于根据所述刻度关键点的标识获取所述刻度关键点之间的图上距离；比例尺获取单元，用于根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺。
于所述第一方面的一实施例中，所述刻度关键点标识单元包括：刻度关键点排序子单元，用于根据所述刻度关键点的纵坐标对所述刻度关键点进行排序；刻度关键点标识子单元，用于根据所述刻度关键点的排序结果对所述刻度关键点依次进行标识。
于所述第一方面的一实施例中，所述比例尺获取单元根据两个所述刻度关键点之间的图上距离与实际距离获取所述比例尺，或者，所述比例尺获取单元根据多个所述刻度关键点之间的平均图上距离与平均实际距离获取所述比例尺。
本发明的第二方面提供一种基于超声图像的测量方法，所述方法包括：获取目标超声图像；获取所述目标超声图像中的刻度关键点；根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺；检测所述目标超声图像中的目标对象；根据所述目标对象的图上尺寸和所述比例尺获取所述目标对象的实际尺寸；在所述目标超声图像中叠加显示所述刻度关键点、所述比例尺、所述目标对象的检测结果和/或所述目标对象的实际尺寸。
本发明的第三方面提供一种计算机可读存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现本发明第二方面所述基于超声图像的测量方法。
本发明的第四方面提供一种电子设备，所述电子设备包括：存储器，存储有一计算机程序；处理器，与所述存储器通信相连，调用所述计算机程序时执行本发明第二方面所述基于超声图像的测量方法。
如上所述，本发明一个或多个实施例中所述基于超声图像的测量设备具有以下有益效果：
所述基于超声图像的测量设备能够基于目标超声图像中的刻度关键点来获取所述目标超声图像的比例尺，基于所述比例尺以及目标对象的图上尺寸即可得到所述目标对象的实际尺 寸。该过程基本无需人工参与，效率较高，且能够避免由于人为主观原因所导致的测量误差。
此外，所述基于超声图像的测量设备能够外接在不同的超声设备上，只要能够获取这些超声设备输出的超声图像即可对其进行处理，因此，所述基于超声图像的测量设备支持与超声设备的即连即用。
附图说明
图1A显示为本发明所述基于超声图像的测量设备于一具体实施例中的结构示意图。
图1B显示为本发明所述基于超声图像的测量设备于一具体实施例中目标超声图像的示例图。
图2显示为本发明所述基于超声图像的测量设备于一具体实施例中第一神经网络模型的训练方法流程图。
图3A显示为本发明所述基于超声图像的测量设备于一具体实施例中目标对象检测模块的结构示意图。
图3B显示为本发明所述基于超声图像的测量设备于一具体实施例中第二神经网络模型的训练方法流程图。
图3C显示为本发明所述基于超声图像的测量设备于另一具体实施例中目标对象检测模块的结构示意图。
图4A显示为本发明所述基于超声图像的测量设备于一具体实施例中比例尺获取模块的结构示意图。
图4B显示为本发明所述基于超声图像的测量设备于一具体实施例中刻度关键点标识单元的结构示意图。
图4C显示为本发明所述基于超声图像的测量设备于一具体实施例中刻度关键点的标识结果示例图。
图5显示为本发明所述基于超声图像的测量方法于一具体实施例中的流程图。
图6显示为本发明所述电子设备于一具体实施例中的结构示意图。
元件标号说明
1        基于超声图像的测量设备
10       超声图像获取模块
11       刻度关键点获取模块
12       比例尺获取模块
121      刻度关键点标识单元
1211     刻度关键点排序子单元
1212     刻度关键点标识子单元
122      图上距离获取单元
123      比例尺获取单元
13       目标对象检测模块
131       目标对象获取单元
132      目标参数获取单元
133      检测框获取单元
134      目标参数获取单元
14       对象尺寸测量模块
15       叠加显示模块
600      电子设备
610      存储器
620      处理器
630      显示器
S21～S24步骤
S31～S34步骤
S51～S56步骤
具体实施方式
以下通过特定的具体实例说明本发明的实施方式，本领域技术人员可由本说明书所揭露的内容轻易地了解本发明的其他优点与功效。本发明还可以通过另外不同的具体实施方式加以实施或应用，本说明书中的各项细节也可以基于不同观点与应用，在没有背离本发明的精神下进行各种修饰或改变。需说明的是，在不冲突的情况下，以下实施例及实施例中的特征可以相互组合。
需要说明的是，以下实施例中所提供的图示仅以示意方式说明本发明的基本构想，图示中仅显示与本发明中有关的组件而非按照实际实施时的组件数目、形状及尺寸绘制，其实际实施时各组件的型态、数量及比例可为一种随意的改变，且其组件布局型态也可能更为复杂。 此外，在本文中，诸如“第一”、“第二”等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。
于本发明的一实施例中提供一种基于超声图像的测量设备，具体地，请参阅图1A，所述基于超声图像的测量设备1包括超声图像获取模块10、刻度关键点获取模块11、比例尺获取模块12、目标对象检测模块13、对象尺寸测量模块14以及叠加显示模块15。
所述超声图像获取模块10用于获取目标超声图像。具体地，所述超声图像获取模块10可以与超声设备通信相连，并获取超声设备输出的超声图像作为所述目标超声图像。所述超声图像获取模块10优选为能够与多种不同类型的超声设备相连，并能够获取这些超声设备输出的不同显示方式、不同显示位置和/或不同缩放比例的超声图像作为所述目标超声图像。
所述刻度关键点获取模块11与所述超声图像获取模块10相连，用于获取所述目标超声图像中的刻度关键点。例如，请参阅图1B，显示为本实施例中涉及的一目标超声图像的示例图，其中，该目标超声图像包括图像本体和刻度关键点，所述刻度关键点是指该目标超声图像中用于指示刻度的点，通常由超声扫描设备自动生成。此外，出于显示清楚的目的，图1B中仅标识了一个刻度关键点，可以理解的是，图中该刻度关键点下方的白色点也为刻度关键点。
所述比例尺获取模块12与所述刻度关键点获取模块11相连，用于根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺。其中，所述比例尺例如为所述刻度关键点之间的图上距离与实际距离的比值。
所述目标对象检测模块13用于检测所述目标超声图像中的目标对象，更具体地说，用于检测所述目标超声图像中所述目标对象的图上尺寸，其中，所述目标对象例如为甲状腺结节，该尺寸包括但不限于长度、周长、面积、体积等。
所述对象尺寸测量模块14与所述比例尺获取模块12和所述目标对象检测模块13相连，用于根据所述目标对象的图上尺寸和所述比例尺获取所述目标对象的实际尺寸。例如，当采用最小检测框来实现所述目标对象的检测、且所述最小检测框为矩形时，所述对象尺寸测量模块14可以根据所述最小检测框的图上长度和图上宽度获取所述最小检测框的实际长度和实际宽度，进而得到所述目标对象的实际径向尺寸。
所述叠加显示模块15与所述超声图像获取模块10、所述刻度关键点获取模块11、所述比例尺获取模块12、所述目标对象检测模块13以及所述对象尺寸测量模块14相连，用于在所述目标超声图像中叠加显示所述刻度关键点、所述比例尺、所述目标对象的检测结果(例 如最小检测框、病灶轮廓)和/或所述目标对象的实际尺寸。其中，所述叠加显示模块15与上述其余模块之间的连接方式包括直接相连以及通过其他模块间接相连。例如，所述叠加显示模块15可以与所述对象尺寸测量模块14直接相连，也可以通过所述对象尺寸测量模块14与所述比例尺获取模块12间接相连。
优选地，所述叠加显示模块15还用于确定所述刻度关键点、所述比例尺、所述目标对象的检测结果和/或所述目标对象的实际尺寸的显示位置和/或显示方式，以避免显示内容之间的遮挡。例如，所述叠加显示模块15可以在所述目标超声图像的一侧显示所述刻度关键点、所述比例尺以及所述目标对象的实际尺寸。
根据以上描述可知，本实施例所述基于超声图像的测量设备能够基于目标超声图像中的刻度关键点来获取所述目标超声图像的比例尺，基于所述比例尺以及目标对象的图上尺寸即可得到所述目标对象的实际尺寸。该过程基本无需人工参与，效率较高，且能够避免由于人为主观原因所导致的测量误差。
于本发明的一实施例中，所述超声图像测量设备1还包括一外壳，所述超声图像测量设备1的上述各模块对应的硬件结构设置于所述外壳中。此外，所述外壳具有通信接口，所述超声图像获取模块10通过该通信接口与超声设备通信相连，以从超声设备获取所述目标超声图像。通过此种方式，所述超声图像测量设备1能够实现与超声设备的即连即用，即，只要将所述超声图像测量设备1外接在超声设备上，所述超声图像测量设备1即可对超声设备输出的超声图像进行处理，从而获取相应的刻度关键点、比例尺、目标对象的检测结果和/或目标对象的实际尺寸并进行显示。
于本发明的一实施例中，所述刻度关键点获取模块11利用第一神经网络模型对所述目标超声图像进行处理，以得到所述刻度关键点，其中，所述第一神经网络模型为一训练好的神经网络模型。
可选地，请参阅图2，本实施例中所述第一神经网络模型的训练方法包括：
S21，构建一神经网络初始模型，本实施例中所述神经网络初始模型用于提取超声图像中的刻度关键点，例如可采用U-Net或V-Net等深度学习网络模型。
S22，获取第一训练数据，其中，所述第一训练数据是标注有刻度关键点的超声图像，对所述刻度关键点的标注可以采用人工等方式实现，本发明对此不做限制。
S23，使用所述第一训练数据对所述神经网络初始模型进行训练，以得到所述第一神经网络模型。
S24，对所述第一神经网络模型进行测试。
应当理解的是，上述通过第一神经网络模型来获取所述刻度关键点的方案仅为本发明的一种优选方案，实际应用中还可以采用其他算法或者人工标记等方式来获得所述刻度关键点，本发明对此不做限制。
请参阅图3A，于本发明的一实施例中，所述目标对象检测模块13的一种结构包括目标对象获取单元131和目标参数获取单元132。
所述目标对象获取单元131用于对所述目标超声图像进行分割以得到所述目标对象的模型。
可选地，所述目标对象获取单元131利用一第二神经网络模型对所述目标超声图像进行处理以实现对所述目标超声图像的分割，从而得到所述目标对象的模型。
请参阅图3B，本实施例中对所述第二神经网络模型进行训练的实现方法包括：
S31，构建一神经网络初始模型，本实施例中所述神经网络初始模型用于实现对所述目标对象的分割算法，例如，可以采用U-Net或V-Net等深度学习模型。
S32，获取第二训练数据，其中，所述第二训练数据是标注有所述目标对象二元分割掩膜的超声图像，对所述二元分割掩膜的标注可以采用人工等方式实现，本发明对此不做限制。
S33，利用所述第二训练数据对所述神经网络初始模型进行训练，以得到所述第二神经网络模型。
S34，对所述第二神经网络模型进行测试。
需要说明的是，所述目标对象获取单元131利用所述第二神经网络模型获取所述目标对象的模型仅为本实施例的一种可行方式，但本发明并不以此为限。
所述目标参数获取单元132与所述目标对象获取单元131相连，用于根据所述目标对象的模型获取目标参数，其中，所述目标参数用于反映所述目标对象的图上尺寸。所述目标参数包括但不限于长度、周长、面积、体积等。
可选地，所述目标参数获取单元132可以获取人工标记的参数值作为所述目标参数。
可选地，所述目标参数获取单元132可以采用关键点检测等传统算法来获取所述目标参数。
可选地，所述目标参数获取单元132可以获取图像分割以后得到的所述目标对象的模型的最大径和最小径作为所述目标参数。
请参阅图3C，于本发明的一实施例中，所述目标对象检测模块13的另一种结构包括检测框获取单元133以及目标参数获取单元134。
所述检测框获取单元133用于根据所述目标超声图像获取所述目标对象的检测框。具体 地，所述检测框获取单元133可以利用训练好的深度学习模型、神经网络模型等对所述目标超声图像进行处理来得到所述目标对象的检测框，也可以根据用户输入的标注结果来获取所述目标对象的检测框，还可以根据所述目标对象的模型获取所述目标对象的边界进而根据该边界上的像素点获取所述目标对象的检测框，但本发明并不限于上述几种方式，实际应用中也可以采用其他技术来得到所述目标对象的检测框。
此外，所述目标对象的检测框可以为任意形状，且所述目标对象的检测框全包含所述目标对象。优选地，所述目标对象的检测框为所述目标对象的最小检测框(或称为最小包围框)，该最小检测框全包含所述目标对象，且所述最小检测框优选为多边形，进一步优选为矩形。
所述目标参数获取单元134与所述检测框获取单元133相连，用于根据所述目标对象的检测框获取目标参数，其中，所述目标参数用于反映所述目标对象的图上尺寸。例如，所述目标参数获取单元134可以获取所述检测框的宽度、高度、面积、周长等几何尺寸作为所述目标参数。
需要说明的是，以上列举的获取所述目标参数的两种方案仅为本发明的可选方案，但本发明并不限于此，实际应用中也可以采用其他方法来获取所述目标参数，例如，可以采用轮廓勾勒、显示经线等方式来获取所述目标参数。
请参阅图4A，于本发明的一实施例中，所述比例尺获取模块12包括刻度关键点标识单元121、图上距离获取单元122和比例尺获取单元123。
所述刻度关键点标识单元121与所述刻度关键点获取模块11相连，用于根据所述刻度关键点的纵坐标对所述刻度关键点进行标识，其中，所述刻度关键点的标识用于标记所述刻度关键点之间的纵向相对位置，不同刻度关键点之间的纵向图上距离可以根据其标识得到。例如，若两个刻度关键点的标识分别为a和b，则二者之间的图上距离为a-b，其中，a和b优选为正整数。
所述图上距离获取单元122与所述刻度关键点标识单元121相连，用于根据所述刻度关键点的标识获取所述刻度关键点之间的纵向图上距离。
所述比例尺获取单元123与所述图上距离获取单元122相连，用于根据所述刻度关键点之间的纵向图上距离与纵向实际距离获取所述目标超声图像的比例尺。其中，所述刻度关键点的纵向实际距离可以根据所述刻度关键点的位置、编号和/或坐标得到。
需要说明的是，本实施例中所述的纵向是指实质上平行于刻度关键点连线的方向。
可选地，本实施例中所述比例尺获取模块还包括预处理单元，所述预处理单元用于对所述刻度关键点进行预处理，以删除其中的异常点。其中，所述异常点例如为横坐标异常的关 键点，例如，若某一刻度关键点的横坐标与其余刻度关键点横坐标的均值相差较大，则可以将该刻度关键点作为异常点并删除。所述刻度关键点标识单元121通过所述预处理单元与所述刻度关键点获取模块11相连，所述比例尺获取单元123通过所述预处理单元与所述刻度关键点获取模块11相连。通过此种方式，能够保证所述比例尺获取单元仅根据正常刻度关键点之间的图上距离与实际距离来获取所述目标图像的比例尺，有利于提升获取到的比例尺的准确度。
可选地，请参阅图4B，所述刻度关键点标识单元121包括刻度关键点排序子单元1211和刻度关键点标识子单元1212。
所述刻度关键点排序子单元1211与所述刻度关键点获取模块11相连，用于根据所述刻度关键点的纵坐标对所述刻度关键点进行排序，例如，可以按照纵坐标从大到小或者从小到大的顺序对所述刻度关键点进行排序。
所述刻度关键点标识子单元1212与所述刻度关键点排序子单元1211相连，用于根据所述刻度关键点的排序对所述刻度关键点进行依次标识。例如，当所述关键点排序子单元1211按照纵坐标从大到小的顺序对所述刻度关键点进行排序时，所述刻度关键点标识子单元1212可以依次将所述刻度关键点标识为0、1、…、N，其中，N为正整数。本实施例中对所述刻度关键点的标识结果例如图4C所示。
于本发明的一实施例中，所述比例尺获取单元124可以根据两个所述刻度关键点之间的图上距离与实际距离获取所述比例尺，或者，所述比例尺获取单元124也可以根据多个所述刻度关键点之间的平均图上距离与平均实际距离获取所述比例尺。此外，所述比例尺获取单元124也可以采用其他方法获取所述比例尺，本发明对此不做限制。
在实际应用中，获取刻度关键点的过程中容易出现遗漏，这会导致所述刻度关键点获取模块无法获取到所有的刻度关键点，进而会导致获取到的比例尺出现误差。针对这一问题，于本发明的一实施例中，所述对象尺寸测量模块根据以下公式确定所述目标超声图像的比例尺：
Figure PCTCN2022070646-appb-000001
其中，L为所述目标超声图像的比例尺，l i,j为第i个刻度关键点和第j个刻度关键点之间的纵向实际距离，也即，第i个刻度关键点和第j个刻度关键点在纵向的实际距离，该距离值可以根据这两个刻度关键点的纵坐标得到，l m,n为第m个刻度关键点和第n个刻度关键点之间的纵向实际距离，也即，第m个刻度关键点和第n个刻度关键点在纵向的实际距离，该距离值可以根据这两个刻度关键点的纵坐标得到，i＝j±1和/或m＝n±1，并且，
Figure PCTCN2022070646-appb-000002
f为取整函数，其可以为四舍五入取整函数、向上取整函数、向下取整函数等，max(l m,n,l i,j)为l m,n和l i,j之间的较大值，min(l m,n,l i,j)为l m,n和l i,j之间的较小值。例如，对于图4C所示的标识结果，若i＝2，j＝m＝3，n＝4，则
Figure PCTCN2022070646-appb-000003
此时，所述比例尺为
Figure PCTCN2022070646-appb-000004
又例如，对于图4C所示的标识结果，若i＝1，j＝2，m＝3，n＝4，则
Figure PCTCN2022070646-appb-000005
此时，所述比例尺为
Figure PCTCN2022070646-appb-000006
如上所述，本实施例中所述对象尺寸测量模块可以根据满足i＝j±1和/或m＝n±1这一条件的任意两个刻度关键点对(i,j)和(m,n)得到一个比例尺的值。基于此，所述对象尺寸测量模块优选为根据不同的刻度关键点对获取多个比例尺的值，并对这些值取平均值作为所述目标超声图像的比例尺。
根据以上描述可知，本实施例提供的方法能够避免因为刻度关键点遗漏而导致获取到的比例尺不够准确的问题，有利于进一步提升结果的准确度。
于本发明的一实施例中，所述目标对象为甲状腺结节，所述基于超声图像的测量设备能够根据甲状腺结节的超声图像获取其上下径、宽径和高径，进而获取其纵横比，其中，纵横比为高径/宽径。基于甲状腺结节的上下径、宽径、高径和纵横比能够得到其恶性程度，该过程基本无需人工参与，效率较高且不会出现人为主观因素所导致的误差。
基于以上对所述测量设备的描述，本发明还提供一种基于超声图像的测量方法。具体地，请参阅图5，于本发明的一实施例中所述基于超声图像的测量方法包括：
S51，获取目标超声图像。
S52，获取所述目标超声图像中的刻度关键点。
S53，根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺。
S54，检测所述目标超声图像中的目标对象。
S55，根据所述目标对象的图上尺寸和所述比例尺获取所述目标对象的实际尺寸。
S56，在所述目标超声图像中叠加显示所述刻度关键点、所述比例尺、所述目标对象的检测结果和/或所述目标对象的实际尺寸。
需要说明的是，本实施例中上述步骤S51～S56与图1A所示基于超声图像的测量设备1 中的相应模块一一对应，为节省说明书篇幅，此处不做过多赘述。
此外，上述标号S51～S54仅用于标识不同的步骤，而非用于限制这些步骤之间的执行顺序，实际应用中可以根据具体需求调整上述步骤的执行顺序，例如，可以先执行步骤S54再执行步骤S52和S53，也可以同时执行步骤S54与步骤S52，本发明对这些步骤之间的相对执行顺序不做限制。
基于以上对所述测量设备的描述，本发明还提供一种计算机可读存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现图5所示基于超声图像的测量方法。
基于以上对所述测量设备的描述，本发明还提供一种电子设备。请参阅图6，于本发明的一实施例中，所述电子设备600包括存储器610和处理器620，其中，所述存储器610存储有计算机程序，所述处理器620与所述存储器610通信相连，调用所述计算机程序时实现图5所示基于超声图像的测量方法。
可选地，所述电子设备600还包括显示器630，所述显示器630与所述存储器610和所述处理器620通信相连，用于显示所述基于超声图像的测量方法的相关GUI交互界面。
本发明所述基于超声图像的测量方法的保护范围不限于本实施例列举的步骤执行顺序，凡是根据本发明的原理所做的现有技术的步骤增减、步骤替换所实现的方案都包括在本发明的保护范围内。
本发明还提供一种基于超声图像的测量设备，所述基于超声图像的测量设备可以实现本发明所述的基于超声图像的测量方法，但本发明所述的基于超声图像的测量方法的实现硬件包括但不限于本实施例列举的基于超声图像的测量设备的结构，凡是根据本发明的原理所做的现有技术的结构变形和替换，都包括在本发明的保护范围内。
本发明所述基于超声图像的测量设备能够基于目标超声图像中的刻度关键点来获取所述目标超声图像的比例尺，基于所述比例尺以及目标对象的最小检测框即可得到所述目标对象的实际尺寸。该过程基本无需人工参与，效率较高，且能够避免由于人为主观原因所导致的测量误差。
此外，考虑到在测量过程中获取到的刻度关键点可能存在遗漏或者误识别等问题而导致获得的比例尺不够准确，所述基于超声图像的测量设备还可以利用预处理单元来删除刻度关键点中存在的异常点，从而避免误识别所导致的误差。并且，所述比例尺获取模块可以通过
Figure PCTCN2022070646-appb-000007
这一公式来确定比例尺，从而避免因为遗漏刻度关键点所导致的误差。因此，本发明所述基于超声图像的测量设备具有良好的准确度。
再者，所述基于超声图像的测量设备能够外接在不同的超声设备上，只要能够获取这些超声设备输出的超声图像即可对其进行处理，因此，所述基于超声图像的测量设备支持与超声设备的即连即用。
综上所述，本发明有效克服了现有技术中的种种缺点而具高度产业利用价值。
上述实施例仅例示性说明本发明的原理及其功效，而非用于限制本发明。任何熟悉此技术的人士皆可在不违背本发明的精神及范畴下，对上述实施例进行修饰或改变。因此，举凡所属技术领域中具有通常知识者在未脱离本发明所揭示的精神与技术思想下所完成的一切等效修饰或改变，仍应由本发明的权利要求所涵盖。

Claims (10)

  1. 一种基于超声图像的测量设备，其特征在于，所述设备包括：
    超声图像获取模块，用于获取目标超声图像；
    刻度关键点获取模块，用于获取所述目标超声图像中的刻度关键点；
    比例尺获取模块，用于根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺；
    目标对象检测模块，用于检测所述目标超声图像中的目标对象；
    对象尺寸测量模块，用于根据所述目标对象的图上尺寸和所述比例尺获取所述目标对象的实际尺寸；
    叠加显示模块，用于在所述目标超声图像中叠加显示所述刻度关键点、所述比例尺、所述目标对象的检测结果和/或所述目标对象的实际尺寸。
  2. 根据权利要求1所述的设备，其特征在于：所述刻度关键点获取模块利用一神经网络模型对所述目标超声图像进行处理，以得到所述刻度关键点。
  3. 根据权利要求1所述的设备，其特征在于，所述目标对象检测模块包括：
    目标对象获取单元，用于对所述目标超声图像进行分割以得到所述目标对象的模型；
    目标参数获取单元，用于根据所述目标对象的模型获取目标参数，其中，所述目标参数用于反映所述目标对象的图上尺寸。
  4. 根据权利要求1所述的设备，其特征在于，所述目标对象检测模块包括：
    检测框获取单元，用于根据所述目标超声图像获取所述目标对象的检测框；
    目标参数获取单元，用于根据所述目标对象的检测框获取目标参数，其中，所述目标参数用于反映所述目标对象的图上尺寸。
  5. 根据权利要求1所述的设备，其特征在于，所述比例尺获取模块包括：
    刻度关键点标识单元，用于根据所述刻度关键点的纵坐标对所述刻度关键点进行标识；
    图上距离获取单元，用于根据所述刻度关键点的标识获取所述刻度关键点之间的图上距离；
    比例尺获取单元，用于根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺。
  6. 根据权利要求5所述的设备，其特征在于，所述刻度关键点标识单元包括：
    刻度关键点排序子单元，用于根据所述刻度关键点的纵坐标对所述刻度关键点进行排序；
    刻度关键点标识子单元，用于根据所述刻度关键点的排序结果对所述刻度关键点依次进行标识。
  7. 根据权利要求5或6所述的设备，其特征在于：所述比例尺获取单元根据两个所述刻度关键点之间的图上距离与实际距离获取所述比例尺，或者，所述比例尺获取单元根据多个所述刻度关键点之间的平均图上距离与平均实际距离获取所述比例尺。
  8. 一种基于超声图像的测量方法，其特征在于，所述方法包括：
    获取目标超声图像；
    获取所述目标超声图像中的刻度关键点；
    根据所述刻度关键点之间的图上距离与实际距离获取所述目标超声图像的比例尺；
    检测所述目标超声图像中的目标对象；
    根据所述目标对象的图上尺寸和所述比例尺获取所述目标对象的实际尺寸；
    在所述目标超声图像中叠加显示所述刻度关键点、所述比例尺、所述目标对象的检测结果和/或所述目标对象的实际尺寸。
  9. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于：该计算机程序被处理器执行时实现权利要求8所述基于超声图像的测量方法。
  10. 一种电子设备，其特征在于，所述电子设备包括：
    存储器，存储有一计算机程序；
    处理器，与所述存储器通信相连，调用所述计算机程序时执行权利要求8所述基于超声图像的测量方法。
PCT/CN2022/070646 2021-10-25 2022-01-07 基于超声图像的测量设备、方法、介质及电子设备 WO2023070946A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111242249.XA CN113995435A (zh) 2021-10-25 2021-10-25 基于超声图像的测量设备、方法、介质及电子设备
CN202111242249.X 2021-10-25

Publications (1)

Publication Number Publication Date
WO2023070946A1 true WO2023070946A1 (zh) 2023-05-04

Family

ID=79923872

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/070646 WO2023070946A1 (zh) 2021-10-25 2022-01-07 基于超声图像的测量设备、方法、介质及电子设备

Country Status (2)

Country Link
CN (1) CN113995435A (zh)
WO (1) WO2023070946A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115937219B (zh) * 2023-03-14 2023-05-12 合肥合滨智能机器人有限公司 一种基于视频分类的超声图像部位识别方法及系统

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0739546A (ja) * 1993-07-27 1995-02-10 Ge Yokogawa Medical Syst Ltd 超音波診断装置
JP2004275362A (ja) * 2003-03-14 2004-10-07 Fuji Photo Film Co Ltd 画像実寸算出方法、画像計測方法および画像出力方法、ならびにそれらの装置
US20120302885A1 (en) * 2011-05-27 2012-11-29 Samsung Medison Co., Ltd. Providing a measuring item candidate group for measuring size of a target object in an ultrasound system
CN106023269A (zh) * 2016-05-16 2016-10-12 北京大学第医院 一种估算伤口面积的方法及装置
CN106419821A (zh) * 2016-08-31 2017-02-22 北京大学第三医院 一种关节镜下测量关节内结构的方法和装置
CN108062749A (zh) * 2017-12-12 2018-05-22 深圳大学 肛提肌裂孔的识别方法、装置和电子设备
CN109165645A (zh) * 2018-08-01 2019-01-08 腾讯科技（深圳）有限公司 一种图像处理方法、装置以及相关设备
CN113066093A (zh) * 2021-04-08 2021-07-02 南华大学附属第一医院 一种基于子宫内膜超声图像的自动分割分型及厚度测量方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0739546A (ja) * 1993-07-27 1995-02-10 Ge Yokogawa Medical Syst Ltd 超音波診断装置
JP2004275362A (ja) * 2003-03-14 2004-10-07 Fuji Photo Film Co Ltd 画像実寸算出方法、画像計測方法および画像出力方法、ならびにそれらの装置
US20120302885A1 (en) * 2011-05-27 2012-11-29 Samsung Medison Co., Ltd. Providing a measuring item candidate group for measuring size of a target object in an ultrasound system
CN106023269A (zh) * 2016-05-16 2016-10-12 北京大学第医院 一种估算伤口面积的方法及装置
CN106419821A (zh) * 2016-08-31 2017-02-22 北京大学第三医院 一种关节镜下测量关节内结构的方法和装置
CN108062749A (zh) * 2017-12-12 2018-05-22 深圳大学 肛提肌裂孔的识别方法、装置和电子设备
CN109165645A (zh) * 2018-08-01 2019-01-08 腾讯科技（深圳）有限公司 一种图像处理方法、装置以及相关设备
CN113066093A (zh) * 2021-04-08 2021-07-02 南华大学附属第一医院 一种基于子宫内膜超声图像的自动分割分型及厚度测量方法

Also Published As

Publication number Publication date
CN113995435A (zh) 2022-02-01

Similar Documents

Publication Publication Date Title
CN108764257B (zh) 一种多视角的指针式仪表识别方法
WO2020103417A1 (zh) 一种bmi评测方法、装置及计算机可读存储介质
CN108186051B (zh) 一种从超声图像中自动测量胎儿双顶径长度的图像处理方法及处理系统
CN104240180B (zh) 一种实现图像自动调整的方法及装置
CN108378869B (zh) 一种从超声图像中自动测量胎儿头围长度的图像处理方法及处理系统
CN111091538A (zh) 一种管道焊缝自动识别、缺陷检测方法及装置
CN105147311A (zh) 用于ct系统中的可视化设备辅助扫描定位方法和系统
CN108052909B (zh) 一种基于心血管oct影像的薄纤维帽斑块自动检测方法和装置
CN109035227A (zh) 对ct图像进行肺部肿瘤检测与诊断的系统
WO2020215485A1 (zh) 胎儿生长参数测量方法、系统及超声设备
CN113688817A (zh) 一种自动巡检的仪表识别方法及识别系统
WO2023070946A1 (zh) 基于超声图像的测量设备、方法、介质及电子设备
CN105371769B (zh) 一种测量动态裂纹长度的方法及装置
CN113749646A (zh) 基于单目视觉的人体身高测量方法、装置及电子设备
TWI675646B (zh) 乳房影像分析方法、乳房影像分析系統及非暫態電腦可讀取媒體
CN110051384B (zh) 结合医学统计信息的胎儿颈背透明物位置检测方法及系统
CN109978859B (zh) 一种基于可见失真池化的图像显示适应质量评估方法
WO2019117472A1 (ko) 아날로그 계기판의 측정값 인식 시스템 및 방법
KR100825960B1 (ko) 두부계측 방사선사진 상에 계측점을 지정하는 방법
CN111189826B (zh) 用于pH试纸测量待测溶液pH值的智能评分实验设备及方法
JP2001314374A (ja) 角膜内皮細胞測定装置
TW514515B (en) Method for obtaining boundary data, contour and boundary category of target image and positioning and guidance of image
CN114092399A (zh) 病灶标记方法、装置、电子设备及可读存储介质
Cui et al. Cobb Angle Measurement Method of Scoliosis Based on U-net Network
CN112884694B (zh) 一种平面显示面板的缺陷检测方法、装置、设备及介质

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: 22884866

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE