CN114881925A - Good malignant decision maker of node based on elasticity ultrasonic image - Google Patents

Good malignant decision maker of node based on elasticity ultrasonic image Download PDF

Info

Publication number
CN114881925A
CN114881925A CN202210323631.1A CN202210323631A CN114881925A CN 114881925 A CN114881925 A CN 114881925A CN 202210323631 A CN202210323631 A CN 202210323631A CN 114881925 A CN114881925 A CN 114881925A
Authority
CN
China
Prior art keywords
nodule
malignancy
elastic
texture
calcification
Prior art date
Legal status (The legal status 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 status listed.)
Granted
Application number
CN202210323631.1A
Other languages
Chinese (zh)
Other versions
CN114881925B (en
Inventor
韩东旭
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shiwei Xinzhi Medical Technology Shanghai Co ltd
Original Assignee
Shiwei Xinzhi Medical Technology Shanghai Co ltd
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 Shiwei Xinzhi Medical Technology Shanghai Co ltd filed Critical Shiwei Xinzhi Medical Technology Shanghai Co ltd
Priority to CN202210323631.1A priority Critical patent/CN114881925B/en
Publication of CN114881925A publication Critical patent/CN114881925A/en
Application granted granted Critical
Publication of CN114881925B publication Critical patent/CN114881925B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • 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
    • G06T2207/10132—Ultrasound 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30068—Mammography; Breast
    • 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
    • G06T2207/30096—Tumor; Lesion

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)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Quality & Reliability (AREA)
  • Radiology & Medical Imaging (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Multimedia (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

The present invention relates to a nodule quality and malignancy determination device based on an elastic ultrasound image, including: an image acquisition module: for acquiring an ultrasound image with a nodule; an elastic signal acquisition module: the elastic signal is used for extracting the elastic signal of the ultrasonic image to obtain an elastic ultrasonic image; a nodule region extraction module: the elastic ultrasonic image processing device is used for intercepting the nodule boundary of the elastic ultrasonic image to obtain a nodule area; a feature acquisition module: for obtaining echo, texture and calcification features of the nodule region; a nodule quality and malignancy determination module: for determining the malignancy and/or benign of a nodule based on echogenic, textural and calcification features of the nodule region. The method can effectively judge the malignancy and the goodness of the nodule in the elastic ultrasonic image.

Description

Good malignant decision maker of node based on elasticity ultrasonic image
Technical Field
The present invention relates to the field of medical image processing technologies, and in particular, to a nodule malignancy and malignancy determination device based on an elastic ultrasound image.
Background
Nowadays, with the increasing demand for rapid and accurate diagnosis and the shortage of clinical personnel, computer analysis methods have been increasingly applied to assist routine clinical diagnosis and show good effects. The ultrasonic imaging is a noninvasive, non-radiative and low-cost tumor diagnosis technology, and can well assist clinicians in judging benign and malignant symptoms of cancers. However, due to the low quality of the ultrasound images, the analysis of tumor nodules in the ultrasound images is a challenging task and is highly susceptible to subjective factors of physicians, resulting in misdiagnosis. As one of potential solutions, the corresponding computer analysis method can be used for well and accurately analyzing nodules of different pathological parts so as to assist clinicians in making more accurate judgment and reduce misdiagnosis.
As an important basis for diagnosing various cancers, the analysis of texture signs in ultrasonic images plays a crucial role in clinical diagnosis. Hard nodules are more prone to damage to surrounding normal tissue and exhibit stronger malignant characteristics; in contrast, soft-textured nodules are less likely to cause damage to surrounding normal tissue, showing more benign features. Unfortunately, due to the limitation of the self-imaging principle and the influence of surrounding hyperplastic tissues and various types of noise, the conventional ultrasonic means is difficult to analyze the texture of the nodule, and is very likely to cause misdiagnosis. Elastic ultrasound is used as a novel ultrasonic imaging technology, and the acquired original ultrasonic signals are processed, so that the hardness degree of a scanned part can be represented on an ultrasonic image in a color mode, and the defect that the display effect of the traditional conventional ultrasound on tissue texture is poor is overcome. However, limited by the current technology, the conventional examination method can only perform subjective or semi-subjective analysis on the texture information in the elastic ultrasound image, which limits the wider application of the elastic ultrasound technology.
Disclosure of Invention
The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a nodule quality and malignancy determination device based on an elastic ultrasound image, which can effectively determine the quality and malignancy of a nodule in the elastic ultrasound image.
The technical scheme adopted by the invention for solving the technical problems is as follows: provided is a nodule malignancy and well-malignancy determination device based on an elastic ultrasound image, including:
an image acquisition module: for acquiring an ultrasound image with a nodule;
an elastic signal acquisition module: the elastic signal is used for extracting the elastic signal of the ultrasonic image to obtain an elastic ultrasonic image;
a nodule region extraction module: the elastic ultrasonic image processing device is used for intercepting the nodule boundary of the elastic ultrasonic image to obtain a nodule area;
a feature acquisition module: for obtaining echo, texture and calcification features of the nodule region;
a nodule quality and malignancy determination module: for determining the malignancy and/or benign of a nodule based on echogenic, textural and calcification features of the nodule region.
The module for judging whether the nodule is good or bad passes
Figure BDA0003572666040000021
To determine the malignancy and malignancy of the nodule, wherein R BM Indicates the percent malignancy, t base Representing the number of benign and malignant read bases, R low Indicating the proportion of hypoechos in the nodule, t compo Indicates the benign or malignant coefficient of texture, omega compo Represents the macroscopic texture classification of nodules, cytic represents the macroscopic cystic property, solid represents the macroscopic solidity, t calc Indicating the benign or malignant coefficient of calcification.
The coefficient of malignancy t compo Satisfies the following conditions:
Figure BDA0003572666040000022
wherein, ω is compo Represents the nodule macro texture classification, the cysic the macro cystic and the solid the macro solidity.
The index t represents the benign or malignant coefficient of calcification calc Satisfies the following conditions:
Figure BDA0003572666040000023
wherein ^ is a logical AND operation, omega calc Indicating classification of nodule calcification, calcium indicating calcification, none indicating non-calcification, R S Is the mean elasticity measure of the hyperechoic within the nodule region and
Figure BDA0003572666040000024
CheckEcho (x, y) is the type of echo within the nodule region,
Figure BDA0003572666040000025
1 in (1) * To indicate a function, return 1 if CheckEcho (x, y) belongs to string, otherwise return 0; s e (x, y) is a measure of elasticity information and
Figure BDA0003572666040000026
s e (x, y) is the color saturation of the elastic signal and
Figure BDA0003572666040000027
R x,y representing the luminance value, G, of the pixel (x, y) in the red channel x,y Representing the luminance value of the pixel (x, y) in the green channel, B x,y Representing the luminance value of the pixel (x, y) in the blue channel,
Figure BDA0003572666040000028
is an elastic weighting constant and
Figure BDA0003572666040000031
Figure BDA0003572666040000032
is an elastic weighting coefficient and
Figure BDA0003572666040000033
ω e for the texture classification of the elasticity signal, soft means that the texture is soft, medium means that the texture is moderate, and hard means that the texture is hard; and lambda is a positive and negative operation coefficient.
A nodule benign and malignant grading module is also included: for ranking the benign or malignant nature of the nodules according to the percent malignancy.
The grade module of the goodness and malignancy of the nodule passes
Figure BDA0003572666040000034
Grading the benign or malignant nature of the nodule, wherein R BM Indicates the percent malignancy, normal for the nodule, abnormal for the nodule, lesknormal for the presence of abnormal nodules, t negative Is a normal index threshold, t positive Is an abnormality index threshold.
The normal index threshold value t negative 0.12, the abnormality index threshold t positive ＝0.62。
Advantageous effects
Due to the adoption of the technical scheme, compared with the prior art, the invention has the following advantages and positive effects: the method can accurately and quantitatively analyze the malignancy degree of the nodule, further grade the malignancy degree of the nodule and prompt subsequent coping means, greatly facilitates the daily work of an ultrasonic doctor and improves the consistency and the accuracy of diagnosis.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention;
FIG. 2 is a schematic diagram of elastic signal region extraction according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a potential echo reference region of an embodiment of the present invention;
FIG. 4 is a schematic diagram of an echo reference region in accordance with an embodiment of the present invention;
fig. 5 is a schematic diagram of an extraction result of the elasticity information measure value in the embodiment of the present invention.
Detailed Description
The invention will be further illustrated with reference to the following specific examples. It should be understood that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention. Further, it should be understood that various changes or modifications of the present invention may be made by those skilled in the art after reading the teaching of the present invention, and such equivalents may fall within the scope of the present invention as defined in the appended claims.
An embodiment of the present invention relates to a nodule malignancy and malignancy determination apparatus based on an elastic ultrasound image, and with reference to fig. 1, the apparatus includes:
an image acquisition module: for acquiring an ultrasound image with a nodule;
an elastic signal acquisition module: the elastic signal is used for extracting the elastic signal of the ultrasonic image to obtain an elastic ultrasonic image;
a nodule region extraction module: the elastic ultrasonic image processing device is used for intercepting the nodule boundary of the elastic ultrasonic image to obtain a nodule area;
a feature acquisition module: for obtaining echo, texture and calcification features of the nodule region;
a nodule quality and malignancy determination module: for determining the malignancy and/or benign of a nodule based on echogenic, textural and calcification features of the nodule region.
The present embodiment will be described in detail below:
further, the elastic signal acquisition module includes:
a detection unit: for extracting a color region I in the ultrasound image by a threshold determination method based on RGB channels c And using eight neighborhoods as radii to the color area I c Searching the color pixels to obtain a plurality of independent connected domains, selecting the connected domain with the largest area in all the connected domains, and fitting the minimum boundary frame of the connected domain with the largest area to obtain an elastic signal region I e ；
An extraction unit: for extracting the elastic signal region I by RGB color channel luminance comparison e The elasticity signal in (1).
Further, the formula of the threshold judgment method in the detection unit is as follows:
Figure BDA0003572666040000041
wherein, color (x, y) is the recognition result of the ultrasound at the pixel (x, y), true indicates that the recognition result is color, false indicates that the recognition result is black and white, and G (x,y) Is the brightness value, R, of the pixel (x, y) on the green channel of the ultrasound image (x,y) Is the brightness value of pixel (x, y) on the red channel of the ultrasound image, B (x,y) The brightness value of the pixel (x, y) on the blue channel of the ultrasound image is thr, which is a preset threshold value.
Further, the extraction unit is provided with
Figure BDA0003572666040000051
To extract the elastic signal region I e Wherein, the CheckInfo (x, y) is the elastic signal extraction result of the elastic ultrasound image at the pixel (x, y), and the pixel (x, y) is located in the elastic signal area I e Inner, D e Representing a channel decision equation, c (x,y) Representing an elastic signal region I e Luminance value at (x, y) pixel in channel C, (C \ C) (x,y) Representing an elastic signal region I e Set of luminance values at (x, y) pixels in two other channels than channel c, L c Is the elastic conversion coefficient; when blue is used to indicate soft texture and red is used to indicate hard texture, { L R ＝250,L G ＝150,L B 50 }; when red is used to indicate soft texture and blue is used to indicate hard texture, { L R ＝50,L G ＝150,L B ＝250}。
Referring to fig. 2, (a) is a breast ultrasound image with elastic signals; (b) colored region I extracted for crude extraction c (ii) a (c) For thin identified elastic signal regions I e 。
The embodiment obtains the extracted elastic information classification image, and the image can well mark the elastic information in the nodule. In addition, the embodiment also measures the ratio of each type of elastic signal in the nodule area, and the elastic signals are divided into three types, namely hard, soft and medium. If by ω e E { soft, medium, hard } represents the category to be solved, the calculation formula can be specifically expressed as:
Figure BDA0003572666040000052
wherein, X e Representing an elastic signal region I e Set of all pixel points, X, shown internally ROI Representing the set of all pixel points within the nodule region,
Figure BDA0003572666040000053
1 in (1) * To indicate the function, when the elastic signal extracts the return value of the equation CheckInfo (x, y) belongs to
Figure BDA0003572666040000054
If yes, returning to 1, otherwise, returning to 0;
Figure BDA0003572666040000055
1 in (1) * To indicate the function, when the pixel (X, y) is at X ROI When in the marked area, 1 is returned, otherwise 0 is returned,
Figure BDA0003572666040000056
is a lookup value for the elasticity class sought and is based on the elastic transformation coefficient L c Is defined as t soft ＝50,t medium ＝150,t hard 250 }. A higher hard texture ratio measured using this method of measurement indicates more hard components in the nodule region, whereas a higher soft texture ratio measured indicates more soft components in the nodule region.
The feature acquisition module comprises an echo extraction unit, a texture extraction unit and a calcification extraction unit, and is described in detail as follows:
(1) echo extraction unit
If with X ROI Representing the set of all pixel points in the nodule region, the binary image I about the nodule region can be extracted ROI Making it satisfy that the region inside all nodules is 1 (true) and the region outside all nodules is 0 (false), which can be expressed by the following formula:
Figure BDA0003572666040000061
wherein,
Figure BDA0003572666040000062
1 in (1) * To indicate the function, when the pixel I in the ultrasound image is (x,y) At X ROI And returning to 1 when the marked area is in the area, and otherwise, returning to 0.
Further, after the binary image of the nodule region is extracted, morphological dilation operation is performed on the nodule region, and the dilated image is subtracted from the image before dilation to finally obtain a potential echo reference extraction region, wherein a specific formula is as follows:
Figure BDA0003572666040000063
wherein, I ref In order to be a potential echo reference region,
Figure BDA0003572666040000064
for the dilation operator, S is a 25-pixel wide disk-shaped structuring element.
Referring to fig. 3, the area marked by the small irregular white circles is a nodule, and the area between the small irregular white circles and the large irregular white circles is a potential echo reference area.
Determines a potential echo reference extraction area I ref Then, the present embodiment classifies it into ω with the extracted elasticity information e (soft, hard and moderate) combination, wherein the area judged to be soft or hard texture is regarded as non-reference tissue elimination, only the area judged to be soft or hard texture is reserved, and finally the echo reference area I is determined ref' The specific formula can be expressed as:
Figure BDA0003572666040000065
wherein, I ref' (x, y) is an echo reference region,
Figure BDA0003572666040000067
is I ref The value at the pixel (x, y), Λ is the logical AND operation, CheckInfo (x, y) is the elastic signal extraction equation, the output result is soft, hard or moderate, t medium It means that the texture is moderate,
Figure BDA0003572666040000066
1 in (1) * To indicate the function, the tag returned when the elastic signal extraction equation CheckInfo (x, y) belongs to t medium If yes, then return to 1 (true), if notThen 0 is returned (false).
Referring to fig. 4, the region identified by numeral 1 is a nodule, and the region identified by numeral 2 is a reference region of the nodule echo after final optimization.
Furthermore, the embodiment further includes an echo type determination module, which determines the type of the echo by comparing the type of the echo with the type of the echo type ref' Averaging the pixels shown by the internal standard to obtain the pixel brightness expression reference value M of the equal echo ref The specific calculation formula is as follows:
Figure BDA0003572666040000071
wherein, X ref Is I ref' Set of all pixel points shown by internal standards, I x,y Is the gray scale brightness value of the ultrasound image at pixel (x, y).
After the brightness representation of the iso-echo is determined, the present embodiment uses M ref The reference values of other echo types are further determined for reference, and are divided into four types, namely strong, equal/high, low and no echo, and the judgment process of each type of echo can be expressed by the following formula:
Figure BDA0003572666040000072
where no represents anechoic, low represents hypoechoic, medhigh represents iso/hyperechoic, strong represents hyperechoic, t n Determining a threshold value, t, for a first echo l Determining a threshold value, t, for the second echo mh The threshold is determined for the third echo, defined experimentally as { t } n ＝0.2,t l ＝1,t mh ＝1.75}。
Obtaining the occupation ratio of various echoes in the nodule region according to the determined echo classification, and if the occupation ratio is omega, obtaining the occupation ratio of various echoes in the nodule region E E { no, low, medhiigh, strong } represents the echo category, the proportion of each echo can be calculated by the following formula:
Figure BDA0003572666040000073
wherein, X ROI The method is a set of all pixel points in a nodule region, no is no echo, low is low echo, medhigh is equal echo, strong echo is strong echo, CheckEcho (x, y) is an echo label returned by an echo judgment equation at a pixel (x, y),
Figure BDA0003572666040000074
1 in (1) * To indicate the function, when CheckEcho (x, y) belongs to ω E If so, 1 is returned, otherwise, 0 is returned.
(2) Texture extraction unit
In examinations using elastic ultrasound, the sonographer often uses color changes as a basis for observation to determine the texture of the scanned region. However, due to the lack and difference of human eyes' sensitivity to various colors, it is difficult to accurately and consistently interpret various colors and their variations, which leads to an increase in misdiagnosis rate. Therefore, the elastic ultrasonic image is analyzed based on the principle of color saturation analysis, so that the purity of various signals in the elastic ultrasonic image is accurately evaluated, and further the quantitative analysis of the elastic information is realized. In the signal purity analysis module, the characteristics of each pixel in the RGB color space are specifically utilized for comparison, and if R is used, R is used x,y 、G x,y And B x,y Respectively, the brightness values in the red, green and blue channels at the target pixel (x, y) position, the color saturation s of the elastic signal e The calculation process of (a) can be embodied by the following formula:
Figure BDA0003572666040000081
where max () denotes taking the maximum value, min () denotes taking the minimum value, s e ∈[0,1]. The embodiment further comprises a first judging module: when the color saturation s of the elastic signal e If the saturation is larger than the preset saturation threshold, the elastic signal is clear (pure); color saturation s of the elasticity signal e If the saturation level is less than the preset saturation threshold, the elastic signal is fuzzy (impure).
Further, the specific calculation formula of the elastic information measure value is as follows:
Figure BDA0003572666040000082
wherein S is e (x, y) is a measure of elasticity,
Figure BDA0003572666040000083
for the elastic weighting constant, it is defined as:
Figure BDA0003572666040000084
Figure BDA0003572666040000085
for the elastic weighting coefficients, it is defined according to different elastic classifications as:
Figure BDA0003572666040000086
wherein soft means soft texture, medium means moderate texture, hard means hard texture, lambda is positive and negative operation coefficient, when the texture classification omega of the elastic signal e Medium, if the elastic signal reading of the hard texture at (x, y) is greater than the elastic signal reading of the soft texture, λ ═ 1; if the elastic signal reading of the hard texture at (x, y) is less than the elastic signal reading of the soft texture, λ is-1. It is to be understood here that each pixel (x, y) has three channel readings, when texture ω is measured e For medium, the readings of the channels corresponding to soft and hard need to be compared again, so as to determine the assignment of λ. The embodiment further comprises a second judging module: elastic information measuring value S e (x, y) is between 0 and 1 when the elasticity information measure value S is e (x, y) is greater than the first texture threshold, indicating that the texture of the ultrasound image at pixel (x, y) is hard; the bulletMeasuring value S of sexual information e (x, y) is less than the second texture threshold, indicating that the texture of the ultrasound image at pixel (x, y) is soft; the elastic information measure value S e (x, y) is between the first texture threshold and the second texture threshold, indicating that the ultrasound image is moderately textured at pixel (x, y).
Referring to fig. 5, (a) is a breast ultrasound image with elastic signals, and the white frame is the elastic signal area; (b) for the extracted elastic information metric values, it is easy to find that the higher the elastic information metric value is, the brighter the corresponding pixel is.
In the examination of thyroid nodules or breast nodules, the nodules, which are mostly composed of fluid, are called cystic nodules, which are less likely to cause malignant invasion of normal tissues around the nodules due to the nature of the fluid itself. In contrast, the malignant character of a nodule is mostly present on the solid part of the nodule. Therefore, the analysis of the texture components of nodules is an essential part for judging the quality and malignancy of the nodules.
Assuming that the judgment result CheckContent (x, y) composed of the cystic property and the real property at each pixel point (x, y) in the nodule region can be calculated by the following formula:
Figure BDA0003572666040000091
wherein mass represents the real property, cys represents the cystic property,
Figure BDA0003572666040000092
representing the corresponding elastic information measure, t, at pixel point (x, y) mc The threshold for cystic solidity segmentation was experimentally set to 0.26. Finally, all the pixel points marked as cystic can be summarized as the cystic component of the nodule.
The present embodiment lists the distribution of different textures of the nodes in the form of percentage values for the purpose of description and quantification thereof. Percent of cystic composition of the nodule R cyst Percentage of substantial composition R mass It can be expressed as:
Figure BDA0003572666040000101
and R is known cyst +R mass ＝1，X ROI Is a collection of pixel points within the nodule region,
Figure BDA0003572666040000102
1 in (1) * To indicate the function, when the tag returned by the CheckContent (x, y) belongs to the cyst, 1 is returned, otherwise 0 is returned;
Figure BDA0003572666040000103
1 in (1) * To indicate a function, a 1 is returned when the tag returned by the CheckContent (x, y) belongs to mass, and a 0 is returned otherwise.
In addition, the present embodiment further defines a set of thresholds t cyct ,t mass And the properties of the nodules are classified into cystic property, real property or cystic-solid mixed property according to the components respectively, and the detailed classification rules are as follows:
Figure BDA0003572666040000104
wherein, cysteine represents macroscopic cystic property, solid represents macroscopic property, texture represents macroscopic mixing property, and threshold { t } cyct ,t mass Experimentally determined as: { t cyct ＝0.85,t mass ＝0.9}。
(3) Calcification extraction unit
As an important basis for diagnosing various types of cancers, calcification plays an important role in the analysis of ultrasound images. Calcifications, especially microcalcifications, are mostly bright punctate hyperechoic areas. However, due to the limitation of the self-imaging principle and the influence of surrounding hyperplastic tissues and various noises, the accuracy of the conventional ultrasonic means for judging the calcification of the focus part is not ideal, and misdiagnosis is easily caused. As a conventional ultrasound image enhancement technique, elastic ultrasound additionally provides information about the texture of the scanned region in addition to the conventional echo intensity characteristics. Because the calcified area is mostly composed of hard tissues, the short board with insufficient diagnostic evidence expressed according to the echo can be well filled in the analysis of the texture of the scanned part, and the defect that the diagnosis accuracy of the traditional conventional ultrasound on calcification (especially microcalcification with stronger malignant characteristics) is poor is overcome.
Specifically, in the present embodiment, a hyperechoic region within a nodule is determined by analyzing the features of echoes within the nodule. Based on the idea that the calcified part is composed of hard tissue, the present embodiment calculates the average elasticity measurement value R in the strong echo region S As a first feature of calcification analysis, a specific calculation method thereof is as follows:
Figure BDA0003572666040000111
wherein R is S Is the average elasticity measure of the hyperechos in the nodule region, CheckEcho (x, y) is the type of echo in the nodule region,
Figure BDA0003572666040000112
1 in (1) * To indicate a function, return 1 if CheckEcho (x, y) belongs to string, otherwise return 0; s e (x, y) is a measure of the elasticity information at pixel (x, y).
Secondly, calcification should by definition grow on solid tissue, so this embodiment also calculates the solid texture ratio around the hyperechoic area as a second feature of the calcification analysis. Specifically, the present embodiment first performs a binary image I composed of hyperechoic regions in the nodule region S Performing morphological dilation operation, and subtracting the dilated image from the image before dilation to obtain analysis region I around the hyperechoic echo O This process can be embodied by the formula:
Figure BDA0003572666040000113
wherein,
Figure BDA0003572666040000114
for the dilation operator, S is a 25-pixel wide disk-shaped structuring element. Next, in the present embodiment, the substantial proportion in the analysis area is calculated, and a specific calculation process can be expressed by the following formula:
Figure BDA0003572666040000115
wherein, X O Indicates the analysis region I O The CheckContent (x, y) is a judgment result formed by the cystic property or the real property of each pixel point (x, y) in the nodule area,
Figure BDA0003572666040000116
1 in (1) * To indicate a function, a 1 is returned when the return value of CheckContent (x, y) belongs to mass, otherwise a 0 is returned.
Finally, the two extracted features are combined in the embodiment, and finally, judgment of whether each pixel point (x, y) belongs to calcification is obtained, and the specific classification rule can be expressed by a formula as follows:
Figure BDA0003572666040000117
wherein calceification represents calcification, none represents non-calcification, and t S For calcification elasticity threshold, set at 0.7, t O The threshold for calcification texture is set to 0.9 through experiments, and Λ is a logical and operation.
Further, the nodule malignancy and well-malignancy determination module is described in detail below:
as a conventional method for diagnosing various cancers, the analysis of the characteristics of the echo, texture and calcification in the nodule plays an important role in the diagnosis process of the ultrasonic image. In particular, hypoechoic nodules often have a higher likelihood of malignancy than other types of echoes; the predominantly solid nodules often have a higher likelihood of malignancy than the predominantly cystic nodules; nodules with calcifications often have a higher malignancy potential than non-calcified nodules. Sonographers often integrate these three different features to make a global judgment of the malignancy and benign of the nodule. However, this type of judgment is often influenced by the experience and qualification of the physician, and it is difficult to produce an objective and consistent judgment.
This embodiment is based on the percent malignancy R BM To determine the malignancy and well-being of the nodule, R BM The greater the value, the more malignant the nodule may be; conversely, the smaller the number, the weaker the malignancy of the nodule, R BM The formula of (1) is:
Figure BDA0003572666040000121
wherein, t base Indicates a base number of good or malignant reads and t base ＝6.72，R low Indicating the proportion of hypoechos in the nodule, t compo The benign and malignant coefficient of the texture is represented, and is set according to different texture classifications as follows:
Figure BDA0003572666040000122
wherein, ω is compo Represents the nodule macro texture classification, cysic represents the macro cystic property, and solid represents the macro solidity;
t calc the benign and malignant coefficient of calcification is shown, and is set according to different calcification performances:
Figure BDA0003572666040000123
wherein, ω is calc Indicating the classification of calcifications in tumors, calcium indicating calcifications, none indicating non-calcifications, R S The mean elasticity measure of the hyperechoic within the nodule region.
The embodiment further comprises a nodule benign and malignant grading module: in terms of calculated percent malignancy R BM On the basis, a corresponding grading diagnosis rule is established, and the grading specific rule is as follows:
Figure BDA0003572666040000124
wherein, normal shows that the tumor is normal and does not indicate possible malignancy; abnormal indicates abnormal tumor expression, suggesting possible malignancy; less normal indicates that there is a tumor with a heterosis but no apparent abnormality, suggesting further analysis in combination with other features; t is t negative Is a normal index threshold value, defined as 0.12 through experiments; t is t positive The anomaly index threshold was defined experimentally as 0.62.
Therefore, the method can accurately and quantitatively analyze the malignancy degree of the nodule, further grade the malignancy degree of the nodule and prompt a subsequent coping means, greatly facilitates the daily work of an ultrasonic doctor and improves the diagnosis consistency and accuracy.
The foregoing descriptions of specific exemplary embodiments of the present invention have been presented for purposes of illustration and description. It is not intended to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain certain principles of the invention and its practical application to enable one skilled in the art to make and use various exemplary embodiments of the invention and various alternatives and modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims and their equivalents.

Claims (7)

1. A nodule malignancy and well-malignancy determination device based on an elastic ultrasound image includes:
an image acquisition module: for acquiring an ultrasound image with a nodule;
an elastic signal acquisition module: the elastic signal is used for extracting the elastic signal of the ultrasonic image to obtain an elastic ultrasonic image;
a nodule region extraction module: the elastic ultrasonic image processing device is used for intercepting the nodule boundary of the elastic ultrasonic image to obtain a nodule area;
a feature acquisition module: for obtaining echo, texture and calcification characteristics of the nodule region;
a nodule quality and malignancy determination module: for determining the malignancy and/or benign of a nodule based on echogenic, textural and calcification features of the nodule region.
2. The apparatus according to claim 1, wherein the nodule malignancy and well-being determination module determines the malignancy and well-being of the nodule by
Figure FDA0003572666030000011
To determine the malignancy and malignancy of the nodule, wherein R BM Indicates the percent malignancy, t base Representing the number of benign and malignant read bases, R low Indicating the proportion of hypoechos in the nodule, t compo Indicates the benign or malignant coefficient of texture, omega compo Represents the macroscopic texture classification of nodules, cytic represents the macroscopic cystic property, solid represents the macroscopic solidity, t calc Indicating the benign or malignant coefficient of calcification.
3. The apparatus according to claim 2, wherein the texture benign-malignant coefficient t is a coefficient of malignancy of the nodule compo Satisfies the following conditions:
Figure FDA0003572666030000012
wherein, ω is compo Represents the nodule macro texture classification, the cysic the macro cystic and the solid the macro solidity.
4. The apparatus according to claim 2, wherein the coefficient t indicating the malignancy and malignancy of the calcification is the same as or different from the coefficient t calc Satisfies the following conditions:
Figure FDA0003572666030000013
wherein ^ is a logical AND operation, omega calc Indicating classification of nodule calcification, calcium indicating calcification, none indicating non-calcification, R S Is the mean elasticity measure of the hyperechoic within the nodule region and
Figure FDA0003572666030000014
CheckEcho (x, y) is the type of echo within the nodule region,
Figure FDA0003572666030000015
1 in (1) * To indicate a function, return 1 if CheckEcho (x, y) belongs to string, otherwise return 0; s. the e (x, y) is a measure of elasticity information and
Figure FDA0003572666030000016
s e (x, y) is the color saturation of the elastic signal and
Figure FDA0003572666030000021
R x,y representing the luminance value, G, of the pixel (x, y) in the red channel x,y Representing the luminance value of the pixel (x, y) in the green channel, B x,y Representing the luminance value of the pixel (x, y) in the blue channel,
Figure FDA0003572666030000022
is an elastic weighting constant and
Figure FDA0003572666030000023
Figure FDA0003572666030000024
is an elastic weighting coefficient and
Figure FDA0003572666030000025
ω e for the texture classification of the elasticity signal, soft means that the texture is soft, medium means that the texture is moderate, and hard means that the texture is hard; and lambda is a positive and negative operation coefficient.
5. The apparatus according to claim 2, further comprising a nodule malignancy and benign grading module that: for ranking the benign or malignant nature of the nodules according to the percent malignancy.
6. The apparatus according to claim 5, wherein the nodule malignancy and well-being grading module is configured to grade the nodule malignancy and well-being by
Figure FDA0003572666030000026
Grading the benign or malignant nature of the nodule, wherein R BM Indicates the percent malignancy, normal for the nodule, abnormal for the nodule, lesknormal for the presence of abnormal nodules, t negative Is a normal index threshold, t positive Is an abnormality index threshold.
7. The apparatus for determining the malignancy and malignancy of a nodule according to claim 6, wherein the normality index threshold t is a threshold value t negative 0.12, the abnormality index threshold t positive ＝0.62。
CN202210323631.1A 2022-03-30 2022-03-30 Elastic ultrasonic image-based benign and malignant nodule judgment device Active CN114881925B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210323631.1A CN114881925B (en) 2022-03-30 2022-03-30 Elastic ultrasonic image-based benign and malignant nodule judgment device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210323631.1A CN114881925B (en) 2022-03-30 2022-03-30 Elastic ultrasonic image-based benign and malignant nodule judgment device

Publications (2)

Publication Number Publication Date
CN114881925A true CN114881925A (en) 2022-08-09
CN114881925B CN114881925B (en) 2024-07-23

Family

ID=82669433

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210323631.1A Active CN114881925B (en) 2022-03-30 2022-03-30 Elastic ultrasonic image-based benign and malignant nodule judgment device

Country Status (1)

Country Link
CN (1) CN114881925B (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108257135A (en) * 2018-02-01 2018-07-06 浙江德尚韵兴图像科技有限公司 The assistant diagnosis system of medical image features is understood based on deep learning method
US20180333140A1 (en) * 2015-11-10 2018-11-22 Exact Imaging, Inc. System comprising indicator features in high-resolution micro-ultrasound images
CN110838110A (en) * 2019-11-05 2020-02-25 张峰 System for identifying benign and malignant tumor based on ultrasonic imaging
CN111243042A (en) * 2020-02-28 2020-06-05 浙江德尚韵兴医疗科技有限公司 Ultrasonic thyroid nodule benign and malignant characteristic visualization method based on deep learning
CN111528914A (en) * 2020-06-02 2020-08-14 深圳市人民医院 Image processing method and device
CN112820399A (en) * 2021-01-26 2021-05-18 北京小白世纪网络科技有限公司 Method and device for automatically diagnosing benign and malignant thyroid nodules
CN112998748A (en) * 2019-12-19 2021-06-22 通用电气精准医疗有限责任公司 Method and system for strain automatic measurement and strain ratio calculation for ultrasonic elastography
US20210219944A1 (en) * 2018-05-31 2021-07-22 Mayo Foundation For Medical Education And Research Systems and Media for Automatically Diagnosing Thyroid Nodules
CN113450320A (en) * 2021-06-17 2021-09-28 浙江德尚韵兴医疗科技有限公司 Ultrasonic nodule grading and benign and malignant prediction method based on deeper network structure
CN113768546A (en) * 2021-09-14 2021-12-10 上海市第六人民医院 Ultrasound elastic image generation and processing system and method
CN114119458A (en) * 2021-09-14 2022-03-01 福州大学 Thyroid medullary cancer ultrasonic image identification method based on clinical priori knowledge guidance

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180333140A1 (en) * 2015-11-10 2018-11-22 Exact Imaging, Inc. System comprising indicator features in high-resolution micro-ultrasound images
CN108257135A (en) * 2018-02-01 2018-07-06 浙江德尚韵兴图像科技有限公司 The assistant diagnosis system of medical image features is understood based on deep learning method
US20210219944A1 (en) * 2018-05-31 2021-07-22 Mayo Foundation For Medical Education And Research Systems and Media for Automatically Diagnosing Thyroid Nodules
CN110838110A (en) * 2019-11-05 2020-02-25 张峰 System for identifying benign and malignant tumor based on ultrasonic imaging
CN112998748A (en) * 2019-12-19 2021-06-22 通用电气精准医疗有限责任公司 Method and system for strain automatic measurement and strain ratio calculation for ultrasonic elastography
CN111243042A (en) * 2020-02-28 2020-06-05 浙江德尚韵兴医疗科技有限公司 Ultrasonic thyroid nodule benign and malignant characteristic visualization method based on deep learning
CN111528914A (en) * 2020-06-02 2020-08-14 深圳市人民医院 Image processing method and device
CN112820399A (en) * 2021-01-26 2021-05-18 北京小白世纪网络科技有限公司 Method and device for automatically diagnosing benign and malignant thyroid nodules
CN113450320A (en) * 2021-06-17 2021-09-28 浙江德尚韵兴医疗科技有限公司 Ultrasonic nodule grading and benign and malignant prediction method based on deeper network structure
CN113768546A (en) * 2021-09-14 2021-12-10 上海市第六人民医院 Ultrasound elastic image generation and processing system and method
CN114119458A (en) * 2021-09-14 2022-03-01 福州大学 Thyroid medullary cancer ultrasonic image identification method based on clinical priori knowledge guidance

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
J WANG ET AL.: "Ultrasonic elastography and conventional ultrasound in the diagnosis of thyroid micro-nodules", 《PAK J MED SCI.》, vol. 35, no. 6, 31 December 2019 (2019-12-31), pages 1526 - 1531 *
薛淑恒: "超声声触弹性成像技术对乳腺良恶肿瘤诊断价值", 《中国卫生工程学》, vol. 19, no. 3, 30 June 2020 (2020-06-30), pages 426 - 428 *

Also Published As

Publication number Publication date
CN114881925B (en) 2024-07-23

Similar Documents

Publication Publication Date Title
CN107798679B (en) Breast region segmentation and calcification detection method for mammary gland molybdenum target image
US20060120608A1 (en) Detecting and classifying lesions in ultrasound images
JP2012512672A (en) Method and system for automatically detecting lesions in medical images
CN113450327B (en) Blood vessel distribution information recognition device of thyroid nodule
CN113450325B (en) Thyroid nodule benign and malignant recognition device
CN113269225A (en) Non-invasive liver epithelium sample vascular smooth muscle lipoma image classification device based on image omics
CN116630680B (en) Dual-mode image classification method and system combining X-ray photography and ultrasound
CN113689424B (en) Ultrasonic inspection system capable of automatically identifying image features and identification method
CN111528914A (en) Image processing method and device
CN112200815A (en) Thyroid nodule ultrasonic image segmentation method based on semantic segmentation network PSPNet
CN114881925B (en) Elastic ultrasonic image-based benign and malignant nodule judgment device
Habibollahi et al. Hyperechoic renal masses: differentiation of angiomyolipomas from renal cell carcinomas using tumor size and ultrasound radiomics
Sim et al. Radiomics Analysis of Magnetic Resonance Proton Density Fat Fraction for the diagnosis of hepatic steatosis in patients with suspected non-alcoholic fatty liver disease
CN114881922B (en) Elastic signal calcification analysis device based on ultrasonic image
CN113940702B (en) Thyroid nodule echo analysis device
CN114881924B (en) Quantitative analysis device based on ultrasonic image elastic signals
CN114926389B (en) Elastic signal extraction device for ultrasonic image
CN102124471A (en) Methods for enhancing vascular patterns in cervical imagery
CN113940704A (en) Thyroid-based muscle and fascia detection device
CN114881923A (en) Ultrasonic image elastic signal-based nodule echo analysis device
CN115018761A (en) Mammary nodule rear echo classification device
CN112734749A (en) Vocal leukoplakia auxiliary diagnosis system based on convolutional neural network model
CN114332003A (en) Thyroid nodule texture analysis device
Angulo-Rodrguez et al. Automated system for quantifying the level of preparation in colonoscopy
CN114022431A (en) Mammary nodule roughness measuring device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant