CN112102311B - Thyroid nodule image processing method and device and computer equipment - Google Patents
Thyroid nodule image processing method and device and computer equipment Download PDFInfo
- Publication number
- CN112102311B CN112102311B CN202011035723.7A CN202011035723A CN112102311B CN 112102311 B CN112102311 B CN 112102311B CN 202011035723 A CN202011035723 A CN 202011035723A CN 112102311 B CN112102311 B CN 112102311B
- Authority
- CN
- China
- Prior art keywords
- nodule
- image
- area
- cystic
- region
- 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.)
- Active
Links
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
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20132—Image cropping
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Quality & Reliability (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Image Analysis (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
The application relates to artificial intelligence, is applicable to intelligent medical treatment, provides a processing method, a device and computer equipment of thyroid nodule images, and the method comprises the following steps: generating a nodule mask image from the ultrasound image; generating a nodule image corresponding to the minimum bounding rectangle of the nodule mask image; processing the knot image to generate a first knot image, and acquiring a sound and shadow candidate region; determining a sound shadow region from the first nodule image; cutting a sound image area in the nodule image to obtain a second nodule image, and processing the second nodule image to obtain a third nodule image; acquiring a cystic candidate region of the third nodule image; calculating a first area of the cystic candidate region, a second area of the nodule region and a third area related to the first area and the second area, and generating a rough classification result of thyroid nodule; determining a cystic area from the third nodule image; and generating a structure fine classification result of the thyroid nodule according to the area of the cystic area, the second area and the third area, and accurately identifying the structure of the thyroid nodule.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a thyroid nodule image processing method, a thyroid nodule image processing device and computer equipment.
Background
Thyroid nodule is a very common clinical condition of the thyroid gland, which is a lump existing in the thyroid gland and has benign and malignant characteristics. If a patient's thyroid nodule is diagnosed as malignant, then he has a high probability of developing thyroid cancer. The ultrasonic image is the preferred method for thyroid examination due to the advantages of low price, no risk, simple use and the like. In an ultrasonic image, thyroid nodules can be well found according to the intensity of echoes (namely gray values in the image) and the properties of the thyroid nodules can be judged. The composition of thyroid nodules is an important index of thyroid image reporting and data systems, which is divided into solidity, cystic, and cystic. When ultrasound images of thyroid nodules are generated, this would greatly aid in screening for signs of malignancy of thyroid nodules if the composition of the thyroid nodules can be quickly determined.
However, in the conventional method for determining the composition of a thyroid nodule, the gray value is generally used as a criterion, and whether the thyroid nodule is cystic or solid is primarily determined by determining the gray distribution of an ultrasound image. For example, the HI value may be used to make the determination (the difference between the standard deviation and the mean of the gray scale at all points within the thyroid nodule), and the greater the HI value, the higher the probability that the thyroid nodule is solid. However, this method of determining the composition uses only the gray value as a criterion for identifying the composition of the thyroid nodule, and does not consider other factors related to the composition of the thyroid nodule, so that a large identification error is likely to be caused, and the accuracy of identifying the composition of the thyroid nodule is low.
Disclosure of Invention
The main purpose of the application is to provide a thyroid nodule image processing method, a thyroid nodule image processing device, a thyroid nodule image processing computer device and a thyroid nodule image storage medium, and aims to solve the technical problem that an existing thyroid nodule composition judging method is low in recognition accuracy.
The application provides a thyroid nodule image processing method, which comprises the following steps:
acquiring an input ultrasonic image with thyroid nodules, marking a nodule area on the ultrasonic image, and generating a nodule mask image corresponding to the nodule area;
intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
performing binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
determining a sound and image area from the first nodule image according to the non-sound and image part in the sound and image candidate area and the nodule boundary of the nodule area;
cutting the sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
Acquiring a black partial region in the third nodule image, wherein the black partial region is positioned in the nodule region, and taking the black partial region as a cystic candidate region;
calculating a first area of the cystic candidate region, a second area of the nodule region, and a third area of the nodule region corresponding to an intersection of the sound shadow region;
generating a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area and the third area, wherein the rough classification result of the thyroid nodule is cystic or cystic, or the rough classification result of the thyroid nodule is real or real;
determining a cystic region from the third nodule image based on the gradient of the third nodule image and the cystic candidate region;
invoking a preset calculation rule corresponding to the composition rough classification result to calculate and generate a composition fine classification result of the thyroid nodule according to a fourth area, the second area and the third area of the cystic area, wherein the composition fine classification result comprises: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
Optionally, the step of determining a sound image area from the first nodule image according to the non-sound image portion and the nodule boundary in the sound image candidate area includes:
Acquiring a non-sound image part in the sound image candidate area;
performing elimination treatment on the non-sound image part to obtain a sound image candidate area after treatment;
acquiring a first communication domain passing through a nodule boundary of the nodule region in the processed sound and image candidate region;
and taking the first connected domain as the sound image area.
Optionally, the step of generating a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area and the third area includes:
calculating a first difference of the second area minus the third area;
calculating a quotient obtained by dividing the first difference by 2;
judging whether the first area is larger than the quotient value or not;
if the first area is larger than the quotient, judging that the rough classification result of the thyroid nodule is cystic or mainly cystic;
and if the first area is not larger than the quotient value, judging that the rough classification result of the thyroid nodule is true or true.
Optionally, the step of determining a cystic region from the third nodule image according to the gradient of the third nodule image and the cystic candidate region comprises:
Calculating the gradient of the third nodule image and generating a corresponding gradient map;
searching seed points in the gradient map, wherein the gray scale of the seed points is smaller than a first preset threshold value, the gradient of the seed points is smaller than a second preset threshold value, and the seed points are in the cystic candidate area;
performing region growing treatment on the seed points according to a second preset rule to generate corresponding appointed connected domains;
carrying out corrosion treatment on the appointed connected domain;
calculating the product of the second area and a specified value;
screening second connected domains with the connected domain area larger than the product from all the appointed connected domains;
and determining the second connected domain as the cystic region.
Optionally, the step of performing region growing processing on the seed points according to a third preset rule to generate corresponding specified connected domains includes:
acquiring a designated seed point, wherein the designated seed point is any seed point in all the seed points;
merging pixels adjacent to the specified seed points into the same set according to the composition conditions of the communication areas to generate a specified pixel set, wherein the composition conditions of the communication areas comprise the same pixel value and adjacent pixel values;
And taking the generated appointed set as a connected domain corresponding to the appointed seed point.
Optionally, after the step of calculating the gradient of the third nodule image and generating a corresponding gradient map, the method includes:
screening out abnormal gradients in the gradient map through a median filtering algorithm;
deleting the abnormal gradient from the gradient map.
Optionally, the step of calling a preset calculation rule corresponding to the composition coarse classification result to calculate and generate the composition fine classification result of the thyroid nodule according to a fourth area, the second area and the third area of the cystic area includes:
calculating a second difference of the second area minus the third area;
calculating a ratio between the fourth area and the second difference;
when the coarse classification result is that the cystic is dominant or cystic, judging whether the ratio is larger than a preset first ratio threshold;
if the ratio is larger than the first ratio threshold, judging that the structural fine classification result of the thyroid nodule is cystic, otherwise judging that the structural fine classification result of the thyroid nodule is cystic;
when the coarse classification result is that the reality is dominant or the reality is dominant, judging whether the ratio is smaller than a preset second ratio threshold;
And if the ratio is smaller than the second ratio threshold, judging that the structural fine classification result of the thyroid nodule is real, otherwise, judging that the structural fine classification result of the thyroid nodule is real.
The application also provides a processing apparatus of thyroid nodule image, include:
the first acquisition module is used for acquiring an input ultrasonic image with thyroid nodules, marking the nodule areas on the ultrasonic image and generating a nodule mask image corresponding to the nodule areas;
the first generation module is used for intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
the first processing module is used for carrying out binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
a first determining module, configured to determine a sound-image region from the first nodule image according to a non-sound-image portion in the sound-image candidate region and a nodule boundary of the nodule region;
the second processing module is used for cutting the sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
The second acquisition module is used for acquiring a black partial region which is positioned inside the nodule region in the third nodule image, and taking the black partial region as a cystic candidate region;
a calculation module for calculating a first area of the cystic candidate region, a second area of the nodule region, and a third area of the nodule region corresponding to an intersection of the sound shadow region;
the second generation module is used for generating a composition coarse classification result of the thyroid nodule according to the first area, the second area and the third area and a first preset rule, wherein the composition coarse classification result is mainly cystic or cystic, or the composition coarse classification result is mainly solid or solid;
a second determining module, configured to determine a cystic area from the third nodule image according to the gradient of the third nodule image and the cystic candidate area;
the third generating module is configured to invoke a preset calculation rule corresponding to the rough classification result to generate a fine classification result of the thyroid nodule according to a fourth area of the cystic area, the second area and the third area, where the fine classification result includes: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
The application also provides a computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the steps of the above method when executing the computer program.
The present application also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of the above method.
The thyroid nodule image processing method, the thyroid nodule image processing device, the computer equipment and the storage medium have the following beneficial effects:
the thyroid nodule image processing method, device, computer equipment and storage medium comprise a two-step recognition process from thick to thin for thyroid nodule composition, in the thick recognition process, firstly, sound image areas in the nodule image are extracted and removed by utilizing sound image physical characteristics during thyroid ultrasonic scanning, so that the situation that the sound image areas in the nodule image are misjudged to be cystic areas is avoided, and therefore, a coarse classification result of the composition of the thyroid nodule is accurately generated according to a first area of candidate cystic areas in the nodule image after the sound image areas are removed, a second area of the nodule areas and a third area corresponding to an intersection of the nodule areas and the sound image areas. And after the rough classification result is obtained, determining a cystic area in the nodule image by combining the gray level and gradient information of the nodule image, and calling a preset calculation rule corresponding to the rough classification result to quickly and accurately calculate and generate a fine classification result of the thyroid nodule according to a fourth area of the cystic area, the second area and the third area. The present invention can accurately identify the structure of a thyroid nodule, and on the one hand, an accurate analysis report concerning the structure of a thyroid nodule can be further generated, and on the other hand, the structure of a thyroid nodule can be input into a corresponding neural network model to determine whether the thyroid nodule is benign or malignant.
Drawings
FIG. 1 is a flow chart of a method of processing thyroid nodule images according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a processing device for thyroid nodule images according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of a computer device according to an embodiment of the present application.
The realization, functional characteristics and advantages of the present application will be further described with reference to the embodiments, referring to the attached drawings.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. The term "and/or" as used herein includes all or any element and all combination of one or more of the associated listed items.
It will be understood by those skilled in the art that all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs unless defined otherwise. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The digital medical treatment method and the digital medical treatment device can be applied to the digital medical treatment field in the smart city, so that the construction of the smart city is promoted.
Referring to fig. 1, a method for processing thyroid nodule images according to an embodiment of the present application includes:
s1: acquiring an input ultrasonic image with thyroid nodules, marking a nodule area on the ultrasonic image, and generating a nodule mask image corresponding to the nodule area;
s2: intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
s3: performing binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
S4: determining a sound and image area from the first nodule image according to the non-sound and image part in the sound and image candidate area and the nodule boundary of the nodule area;
s5: cutting the sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
s6: acquiring a black partial region in the third nodule image, wherein the black partial region is positioned in the nodule region, and taking the black partial region as a cystic candidate region;
s7: calculating a first area of the cystic candidate region, a second area of the nodule region, and a third area of the nodule region corresponding to an intersection of the sound shadow region;
s8: generating a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area and the third area, wherein the rough classification result of the thyroid nodule is cystic or cystic, or the rough classification result of the thyroid nodule is real or real;
s9: determining a cystic region from the third nodule image based on the gradient of the third nodule image and the cystic candidate region;
S10: invoking a preset calculation rule corresponding to the composition rough classification result to calculate and generate a composition fine classification result of the thyroid nodule according to a fourth area, the second area and the third area of the cystic area, wherein the composition fine classification result comprises: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
As described in steps S1 to S10, the execution subject of the embodiment of the method is a thyroid nodule image processing apparatus. In practical applications, the processing device for thyroid nodule images may be implemented by a virtual device, for example, a software code, or may be implemented by an entity device in which related execution codes are written or integrated, and may perform man-machine interaction with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device. The thyroid nodule image processing device in the embodiment can quickly and accurately identify the structure of a thyroid nodule in the ultrasonic image. Specifically, firstly, an input ultrasonic image with thyroid nodule is acquired, and the nodule is marked on the ultrasonic imageAnd generating a nodule mask image corresponding to the nodule region. And then intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image. After the nodule image is obtained, performing binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image. The OTSU segmentation algorithm is a self-adaptive method for calculating a threshold value, firstly, the original slice image is grayed, all pixel points of the image are divided into two different categories, namely a foreground, namely a target and a background according to the gray value of the image, the difference between the two categories is represented by the inter-category variance, the larger the difference between the foreground and the background is, the larger the inter-category variance of the two categories is, and the OTSU segmentation algorithm finds a proper threshold value by traversing all gray values, so that the inter-category variance of the two categories is maximum. Further, a first threshold value, which is optimal for binarization of an image corresponding to the first nodule image, may be determined by the OTSU segmentation algorithm, and the first nodule image may be obtained by performing binary segmentation processing on the first nodule image according to the first threshold value, and the sound image candidate region may be a foreground (white) portion in the first nodule image. Specifically, the formula of the OTSU segmentation algorithm is as follows: omega 0 =N 0 /(M*N),ω 1 =N 1 /(M*N),σ=ω 0 ω 1 (μ 0 -μ 1 ) 2 . M is the length of the nodule image, N is the width of the nodule image, M is the nodule area, N 0 N is the area of the foreground pixel point 1 The first nodule image can be divided into a foreground part and a background part by a certain threshold T for the area of the background pixel point, and the proportion of the foreground pixel point is omega 0 The proportion of background pixel points is omega 1 The gray average of the foreground and background is μ, respectively 0 、μ 1 And finally obtaining the inter-class variance sigma. By means of traversal, different thresholds T are calculated, and the threshold T capable of minimizing the inter-class variance sigma is found, which is the result of segmentation by the OTSU segmentation algorithm. After the sound and image candidate region is obtained, the non-sound and image part and the nodule region in the sound and image candidate region are used as the basisAnd determining a final sound image region from the first nodule image. The method may further comprise selecting a first connected domain passing through a nodule boundary of the nodule region in the sound and image candidate region after the elimination processing as the final sound and image region by selecting a non-sound and image portion from the sound and image candidate region according to physical characteristics of the sound and image. And then cutting out the final sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image. Wherein, since the sound image area in the second nodule image is already eliminated, that is, the situation that the sound image area in the nodule image is misjudged as the cystic area does not occur. The process of performing the binary segmentation of the second nodule image according to the OTSU segmentation algorithm may refer to the process of performing the binary segmentation of the nodule image according to the OTSU segmentation algorithm, and may be performed by changing the node area (m×n) used before to a predetermined nodule area, that is, subtracting the area of the nodule area from a third area (m×n- (m×n) ×n) a of the nodule area corresponding to the intersection of the final sound image areas, where a is the final sound image area. After the third nodule image is obtained, a black partial region (i.e., a background portion within the nodule region) within the nodule region in the third nodule image is acquired, and the black partial region is used as a cystic candidate region. Wherein, the black partial region inside the nodule region refers to the background partial region inside the nodule region. And calculating a first area of the candidate region, a second area of the nodule region, and a third area corresponding to an intersection of the nodule region and the final sound image region. And then generating a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area and the third area, wherein the rough classification result of the thyroid nodule is mainly cystic or cystic, or the rough classification result of the thyroid nodule is mainly solid or solid. In addition, the area of the sound image region included in the nodule can be obtained by subtracting the area of the sound image region from the area of the node from the area of the cystic candidate region And comparing the sizes of the half of the areas of the designated nodules, and generating a rough classification result corresponding to the structural composition of the thyroid nodule according to the corresponding size comparison result. And then determining a final cystic region from the third nodule image based on the gradient of the third nodule image and the cystic candidate region. The method comprises the steps of combining gray level and gradient of a nodule image to accurately identify a cystic region in the nodule image, generating a gradient map of a third nodule image, determining seed points meeting preset conditions from the gradient map, and obtaining the final cystic region according to a seed filling method. And finally, according to the fourth area, the second area and the third area of the final cystic area, calling a preset calculation rule corresponding to the composition rough classification result to calculate and generate a composition fine classification result of the thyroid nodule, wherein the composition fine classification result comprises the following steps: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity. In addition, for two different rough classification results of the thyroid nodule, two different calculation rules for generating a fine classification result are correspondingly preset, and the two different calculation rules are provided with two corresponding ratio thresholds, specifically, the ratio between the final cystic area and the designated nodule area obtained by subtracting the area of the sound shadow area included in the nodule from the node area can be calculated, and the ratio is compared with the preset ratio threshold, so that the fine classification result of the thyroid nodule can be quickly and accurately generated according to the corresponding size comparison result. In the two-step recognition process from thick to thin for thyroid nodule formation, firstly, sound image areas in a nodule image are extracted and eliminated by utilizing sound image physical characteristics during thyroid ultrasonic scanning in the thick recognition process, so that the situation that the sound image areas in the nodule image are misjudged as cystic areas does not occur, and thus, a rough classification result of the thyroid nodule formation is accurately generated according to a first area of candidate cystic areas in the nodule image after the sound image areas are eliminated, a second area of the nodule areas and a third area corresponding to an intersection of the nodule areas and the sound image areas. After the rough classification result is obtained, the method is carried out And determining a cystic area in the nodule image by combining the gray level and gradient information of the nodule image, and calling a preset calculation rule corresponding to the composition rough classification result to quickly and accurately calculate and generate a composition fine classification result of the thyroid nodule according to a fourth area of the cystic area, the second area and the third area. The present invention can accurately identify the structure of a thyroid nodule, and on the one hand, an accurate analysis report concerning the structure of a thyroid nodule can be further generated, and on the other hand, the structure of a thyroid nodule can be input into a corresponding neural network model to determine whether the thyroid nodule is benign or malignant. The method and the device are applicable to the fields of intelligent medical treatment and digital medical treatment, and medical image analysis is optimized.
Further, in an embodiment of the present application, the step S4 of determining the sound image area from the first nodule image according to the non-sound image portion and the nodule boundary in the sound image candidate area includes:
s400: acquiring a non-sound image part in the sound image candidate area;
s401: performing elimination treatment on the non-sound image part to obtain a sound image candidate area after treatment;
S402: acquiring a first communication domain passing through a nodule boundary of the nodule region in the processed sound and image candidate region;
s403: and taking the first connected domain as the sound image area.
As described in steps S400 to S403, the step of determining a sound image area from the first nodule image according to the non-sound image portion and the nodule boundary in the sound image candidate area may specifically include: the non-sound image portion in the sound image candidate area is first acquired. Wherein, because the sound image has a starting point which is always in the interior of the nodule, the sound image can be continued for a distance outside the boundary of the nodule, and the figure in the sound image does not have a large number of physical properties of faults. Thus, according to the physical characteristics, the non-sound image part in the sound image candidate region can be screened out by traversing each column in the first nodule image from top to bottom by the linear array probe. And then, performing elimination processing on the non-sound image part to obtain a sound image candidate area after processing. Wherein, by eliminating the above-mentioned non-sound image part in the sound image candidate region, the accuracy of the subsequently generated processed sound image candidate region can be improved. And then acquiring a first connected domain which passes through the nodule boundary of the nodule region in the processed sound and image candidate region. Before the first connected domain passing through the nodule boundary of the nodule region in the processed sound and image candidate region is obtained, the processed sound and image candidate region can be subjected to expansion operation in advance, so that tiny small points in the processed sound and image candidate region are eliminated, the influence of the tiny small points on the subsequently obtained first connected domain is avoided, and the accuracy of the obtained first connected domain is improved. And finally, when the first connected domain is obtained, the first connected domain is taken as the final sound image area so that the final sound image area can be cut out from the nodule image to obtain a corresponding second nodule image, and the sound image area in the second nodule image is eliminated, namely, the situation that the sound image area in the nodule image is misjudged as a cystic area is avoided, and the identification processing of the thyroid nodule structure can be accurately performed according to the second nodule image.
Further, in an embodiment of the present application, the step S8 of generating the rough classification result of the thyroid nodule according to the first preset rule according to the first area, the second area and the third area includes:
s800: calculating a first difference of the second area minus the third area;
s801: calculating a quotient obtained by dividing the first difference by 2;
s802: judging whether the first area is larger than the quotient value or not;
s803: if the first area is larger than the quotient, judging that the rough classification result of the thyroid nodule is cystic or mainly cystic;
s804: and if the first area is not larger than the quotient value, judging that the rough classification result of the thyroid nodule is true or true.
As described in steps S800 to S804, the step of generating the rough classification result of the thyroid nodule according to the first preset rule based on the first area, the second area and the third area may specifically include: first, a first difference is calculated by subtracting the third area from the second area. The second area is an area of a nodule region, the third area is an area corresponding to an intersection of the nodule region and the final sound image region, and the accurate designated nodule area can be obtained by calculating a first difference value obtained by subtracting the third area from the second area, that is, by subtracting the area of the sound image region included in the nodule from the nodule area. And then calculating a quotient obtained by dividing the first difference by 2. And judging whether the first area is larger than the quotient value. And if the first area is judged to be larger than the quotient, judging that the rough classification result of the thyroid nodule is cystic or cystic. Wherein the first area is an area of a cystic candidate region in the third nodule image, and when the area of the cystic candidate region is greater than one-half of the area of the designated nodule, the thyroid gland may be primarily determined to be cystic or cystic. And if it is determined that the first area is not larger than the quotient, determining that the result of the rough classification of the thyroid nodule is true or true. When the area of the candidate cystic region is not more than one-half of the area of the specific nodule, the thyroid gland may be primarily determined to be solid or solid. According to the method, the size comparison is carried out on the area of the cystic candidate region and one half of the area of the designated nodule obtained by subtracting the area of the sound image region included in the nodule from the area of the node, and then the structure rough classification result corresponding to the structure composition of the thyroid nodule is generated according to the corresponding size comparison result, so that the structure fine classification result of the thyroid nodule can be further generated according to the structure rough classification result.
Further, in an embodiment of the present application, the step S9 of determining the cystic area from the third nodule image according to the gradient of the third nodule image and the cystic candidate area includes:
s900: calculating the gradient of the third nodule image and generating a corresponding gradient map;
s901: searching seed points in the gradient map, wherein the gray scale of the seed points is smaller than a first preset threshold value, the gradient of the seed points is smaller than a second preset threshold value, and the seed points are in the cystic candidate area;
s902: performing region growing treatment on the seed points according to a second preset rule to generate corresponding appointed connected domains;
s903: carrying out corrosion treatment on the appointed connected domain;
s904: calculating the product of the second area and a specified value;
s905: screening second connected domains with the connected domain area larger than the product from all the appointed connected domains;
s906: and determining the second connected domain as the cystic region.
As described in the above steps S900 to S906, the present embodiment is not limited to using only the gradation information of the nodule image, but can accurately identify the cystic area in the nodule image in combination with the gradation and gradient of the nodule image. Specifically, the step of determining the cystic region from the third nodule image according to the gradient of the third nodule image and the cystic candidate region may include: first, the gradient of the third nodule image is calculated, and a corresponding gradient map is generated. Wherein, can be through the formula Calculating to generate a gradient map corresponding to the third nodule image g x And g is equal to y And respectively calculating the third nodule image in the horizontal direction and the vertical direction by using edge detection operators to obtain a gradient map g. And then searching seed points in the gradient map, wherein the gray scale is smaller than a first preset threshold value, the gradient is smaller than a second preset threshold value and the seed points are in the cystic candidate area. The first preset threshold and the second preset threshold are not particularly limited, and may be set according to actual requirements, for example, may be set to 2 and 4 respectively. After the seed points are obtained, according to the first stepAnd carrying out region growing treatment on the seed points by using two preset rules to generate corresponding appointed connected domains. The second preset rule may specifically be a seed filling method, where a flow of the seed filling method includes: and selecting a foreground pixel point as a seed point, combining foreground pixels adjacent to the seed into the same set according to the composition condition of the communication area, and finally obtaining a pixel set which is a communication area. After the specified connected domain is obtained, the specified connected domain is subjected to corrosion treatment. The specified connected domain is subjected to corrosion treatment, so that holes can be filled, and connecting parts of the specified connected domain and the non-nodular region can be removed, thereby realizing the refinement of the thyroid nodule shape. And then calculating the product of the second area and the specified value. The above-mentioned specific numerical values are not particularly limited, and may be set according to actual requirements, for example, may be set to 0.001. And screening out a second connected domain with the connected domain area larger than the product from all the specified connected domains. In order to avoid the influence of the nonsensical regions on the cystic region, the nonsensical regions need to be removed from the cystic candidate region. And finally, determining the second connected domain as the final cystic region so that the subsequent constructional fine classification result of the thyroid nodule can be quickly and accurately calculated and generated by calling a preset calculation rule corresponding to the constructional coarse classification result according to the obtained final fourth area of the cystic region, the second area and the third area.
Further, in an embodiment of the present application, the step S902 of performing the region growing process on the seed points according to the third preset rule to generate the corresponding specified connected domain includes:
s9020: acquiring a designated seed point, wherein the designated seed point is any seed point in all the seed points;
s9021: merging pixels adjacent to the specified seed points into the same set according to the composition conditions of the communication areas to generate a specified pixel set, wherein the composition conditions of the communication areas comprise the same pixel value and adjacent pixel values;
s9022: and taking the generated appointed set as a connected domain corresponding to the appointed seed point.
As described in the above steps S9020 to S9022, the step of performing the region growing process on the seed points according to the third preset rule to generate the corresponding specified connected domain may specifically include: first, a designated seed point is obtained, wherein the designated seed point is any seed point in all the seed points. Then, according to the construction conditions of the communication area, combining the pixels adjacent to the specified seed points into the same set to generate a specified pixel set; wherein the constituent conditions of the connected regions include that the pixel values are the same and the pixel values are adjacent. And finally, taking the generated designated set as a connected domain corresponding to the designated seed point. According to the construction conditions of the communication areas, the appointed communication areas of the seed points after the growth filling treatment can be obtained quickly and conveniently, so that the final cystic area in the third nodule image can be determined quickly and accurately according to the appointed communication areas, and the construction fine classification result of the thyroid nodule can be calculated and generated quickly and accurately by calling a preset calculation rule corresponding to the construction coarse classification result according to the fourth area of the final cystic area, the second area and the third area.
Further, in an embodiment of the present application, after the step S900 of calculating the gradient of the third nodule image and generating a corresponding gradient map, the method includes:
s9000: screening out abnormal gradients in the gradient map through a median filtering algorithm;
s9001: deleting the abnormal gradient from the gradient map.
After calculating the gradient of the third nodule image and generating a corresponding gradient map, the gradient map may be further median filtered to filter out unwanted noise data, as described in steps S9000 to S9001. Specifically, after the step of calculating the gradient of the third nodule image and generating a corresponding gradient map, the method includes: firstly, screening out abnormal gradients in the gradient map through a median filtering algorithm. After the abnormal gradient is obtained, the abnormal gradient is deleted from the gradient map. According to the embodiment, the abnormal gradient in the gradient map is removed by using a median filtering algorithm, so that the accuracy of the seed points searched later can be improved, and the accuracy of the final cystic region generated later can be improved.
Further, in an embodiment of the present application, the step S10 of generating the structure fine classification result of the thyroid nodule by calling a preset calculation rule corresponding to the structure coarse classification result according to the fourth area, the second area and the third area of the cystic area includes:
S1000: calculating a second difference of the second area minus the third area;
s1001: calculating a ratio between the fourth area and the second difference;
s1002: when the coarse classification result is that the capsulization is dominant or the capsulization is dominant, judging whether the ratio is larger than a preset first ratio threshold;
s1003: if the ratio is larger than the first ratio threshold, judging that the structural fine classification result of the thyroid nodule is cystic, otherwise judging that the structural fine classification result of the thyroid nodule is cystic;
s1004: when the rough classification result is that the reality is dominant or the reality is high, judging whether the ratio is smaller than a preset second ratio threshold;
s1005: and if the ratio is smaller than the second ratio threshold, judging that the structural fine classification result of the thyroid nodule is real, otherwise, judging that the structural fine classification result of the thyroid nodule is real.
As described in steps S1000 to S1005, two different calculation rules for generating the coarse classification result are set in advance for the thyroid nodule, and two corresponding ratio thresholds are set for the two different calculation rules. The step of calling a preset calculation rule corresponding to the composition coarse classification result to calculate and generate the composition fine classification result of the thyroid nodule according to the fourth area, the second area and the third area of the cystic area may specifically include: first, a second difference is calculated by subtracting the third area from the second area. And calculating a ratio between the fourth area and the second difference. The second area is an area of a nodule region, the third area is an area corresponding to an intersection of the nodule region and the final sound image region, and the accurate designated nodule area can be obtained by calculating a first difference value obtained by subtracting the third area from the second area, that is, by subtracting the area of the sound image region included in the nodule from the nodule area. In addition, the ratio between the fourth area and the second difference is the ratio between the area of the final cystic area and the area of the designated nodule. And when the rough classification result is mainly cystic or cystic, judging whether the ratio is larger than a preset first ratio threshold. The first ratio threshold is not specifically limited, and may be set according to actual requirements, for example, may be set to 0.95. And if the ratio is larger than the first ratio threshold, judging that the structural fine classification result of the thyroid nodule is cystic, otherwise, judging that the structural fine classification result of the thyroid nodule is mainly cystic. And when the rough classification result is mainly or practically, judging whether the ratio is smaller than a preset second ratio threshold. The second ratio threshold is not specifically limited, and may be set according to actual requirements, for example, may be set to 0.08. And if the ratio is smaller than the second ratio threshold, judging that the structural fine classification result of the thyroid nodule is real, otherwise, judging that the structural fine classification result of the thyroid nodule is dominant. After the rough classification result of the thyroid nodule is obtained, the specific nodule area obtained by subtracting the area of the sound shadow area included in the nodule from the final cystic area and the node area is calculated, and the specific nodule area is compared with the preset specific threshold value, so that the fine classification result of the thyroid nodule can be quickly and accurately generated according to the corresponding size comparison result. The present embodiment can accurately identify the structure of a thyroid nodule, and on the one hand, an accurate analysis report concerning the structure of a thyroid nodule can be further generated, and on the other hand, the structure of a thyroid nodule can be input into a corresponding neural network model to determine whether the thyroid nodule is benign or malignant.
The processing method of the thyroid nodule image in the embodiment of the application can also be applied to the field of blockchains, for example, data such as the structural fine classification result of the thyroid nodule is stored on the blockchain. By storing and managing the constituent classification results of the thyroid nodules using a blockchain, the safety and tamper resistance of the constituent classification results of the thyroid nodules can be effectively ensured.
The blockchain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. The Blockchain (Blockchain), which is essentially a decentralised database, is a string of data blocks that are generated by cryptographic means in association, each data block containing a batch of information of network transactions for verifying the validity of the information (anti-counterfeiting) and generating the next block. The blockchain may include a blockchain underlying platform, a platform product services layer, an application services layer, and the like.
The blockchain underlying platform may include processing modules for user management, basic services, smart contracts, operation monitoring, and the like. The user management module is responsible for identity information management of all blockchain participants, including maintenance of public and private key generation (account management), key management, maintenance of corresponding relation between the real identity of the user and the blockchain address (authority management) and the like, and under the condition of authorization, supervision and audit of transaction conditions of certain real identities, and provision of rule configuration (wind control audit) of risk control; the basic service module is deployed on all block chain node devices, is used for verifying the validity of a service request, recording the service request on a storage after the effective request is identified, for a new service request, the basic service firstly analyzes interface adaptation and authenticates the interface adaptation, encrypts service information (identification management) through an identification algorithm, and transmits the encrypted service information to a shared account book (network communication) in a complete and consistent manner, and records and stores the service information; the intelligent contract module is responsible for registering and issuing contracts, triggering contracts and executing contracts, a developer can define contract logic through a certain programming language, issue the contract logic to a blockchain (contract registering), invoke keys or other event triggering execution according to the logic of contract clauses to complete the contract logic, and simultaneously provide a function of registering contract upgrading; the operation monitoring module is mainly responsible for deployment in the product release process, modification of configuration, contract setting, cloud adaptation and visual output of real-time states in product operation, for example: alarms, monitoring network conditions, monitoring node device health status, etc.
Referring to fig. 2, in an embodiment of the present application, there is further provided a processing apparatus for thyroid nodule image, including:
a first acquisition module 1, configured to acquire an input ultrasonic image with thyroid nodule, mark a nodule region on the ultrasonic image, and generate a nodule mask image corresponding to the nodule region;
the first generation module 2 is used for intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
the first processing module 3 is used for performing binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
a first determining module 4, configured to determine a sound image area from the first nodule image according to a non-sound image portion in the sound image candidate area and a nodule boundary of the nodule area;
the second processing module 5 is configured to cut the sound and shadow area from the nodule image to obtain a corresponding second nodule image, and perform binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
A second acquiring module 6, configured to acquire a black partial area inside a nodule area in the third nodule image, and use the black partial area as a cystic candidate area;
a calculation module 7, configured to calculate a first area of the cystic candidate region, a second area of the nodule region, and a third area corresponding to an intersection of the nodule region and the sound shadow region;
a second generating module 8, configured to generate a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area, and the third area, where the rough classification result of the thyroid nodule is mainly cystic or cystic, or the rough classification result of the thyroid nodule is mainly solid or solid;
a second determining module 9, configured to determine a cystic area from the third nodule image according to the gradient of the third nodule image and the cystic candidate area;
a third generating module 10, configured to generate a composition fine classification result of the thyroid nodule according to a fourth area of the cystic area, the second area and the third area by calling a preset calculation rule corresponding to the composition coarse classification result, where the composition fine classification result includes: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
In this embodiment, the implementation process of the functions and roles of the first acquiring module, the first generating module, the first processing module, the first determining module, the second processing module, the second acquiring module, the calculating module, the second generating module, the second determining module and the third generating module in the processing device of the thyroid nodule image is specifically detailed in the implementation process corresponding to steps S1 to S10 in the processing method of the thyroid nodule image, and will not be described herein.
Further, in an embodiment of the present application, the first determining module includes:
a first acquisition sub-module for acquiring a non-sound image part in the sound image candidate region;
the first processing submodule is used for carrying out elimination processing on the non-sound image part to obtain a sound image candidate region after processing;
a second obtaining sub-module, configured to obtain a first connected domain in the processed sound and image candidate region, where the first connected domain passes through a nodule boundary of the nodule region;
and the first determining submodule is used for taking the first communication domain as the sound image area.
In this embodiment, the implementation process of the functions and actions of the first acquiring sub-module, the first processing sub-module, the second acquiring sub-module and the first determining sub-module in the thyroid nodule image processing apparatus is specifically described in the implementation process corresponding to steps S400 to S403 in the thyroid nodule image processing method, which is not described herein.
Further, in an embodiment of the present application, the second generating module includes:
a first calculation sub-module for calculating a first difference of the second area minus the third area;
the second calculation sub-module is used for calculating a quotient obtained by dividing the first difference value by 2;
the first judging submodule is used for judging whether the first area is larger than the quotient or not;
the first judging submodule is used for judging that the rough classification result of the thyroid nodule is cystic or mainly cystic if the first area is larger than the quotient value;
and the second judging submodule is used for judging that the formation rough classification result of the thyroid nodule is the reality or the reality if the first area is not larger than the quotient value.
In this embodiment, the implementation process of the functions and roles of the first calculation submodule, the second calculation submodule, the first determination submodule and the second determination submodule in the thyroid nodule image processing apparatus is specifically described in the implementation process corresponding to steps S800 to S804 in the thyroid nodule image processing method, and will not be described herein.
Further, in an embodiment of the present application, the second determining module includes:
A third computing sub-module for computing gradients of the third nodule image and generating a corresponding gradient map;
the searching sub-module is used for searching seed points which are in the cystic candidate area and have the gray scale smaller than a first preset threshold value and the gradient smaller than a second preset threshold value in the gradient map;
the first processing submodule is used for carrying out region growing processing on the seed points according to a second preset rule to generate corresponding appointed connected domains;
the second processing submodule is used for carrying out corrosion treatment on the appointed connected domain;
a fourth calculation sub-module for calculating the product of the second area and a specified value;
a screening submodule, configured to screen a second connected domain with a connected domain area larger than the product from all the specified connected domains;
a determining submodule for determining the second connected domain as the cystic region.
In this embodiment, the implementation processes of the functions and actions of the third calculation sub-module, the search sub-module, the first processing sub-module, the second processing sub-module, the fourth calculation sub-module, the screening sub-module and the determination sub-module in the thyroid nodule image processing apparatus are specifically detailed in the implementation processes corresponding to steps S900 to S906 in the thyroid nodule image processing method, and are not repeated here.
Further, in an embodiment of the present application, the first processing sub-module includes:
an obtaining unit, configured to obtain a specified seed point, where the specified seed point is any seed point in all the seed points;
a generation unit, configured to combine pixels adjacent to the specified seed point into the same set according to a composition condition of the connected region, and generate a specified pixel set, where the composition condition of the connected region includes the same pixel value and the pixel values are adjacent;
and the determining unit is used for taking the generated specified set as a connected domain corresponding to the specified seed point.
In this embodiment, the implementation process of the functions and actions of the acquiring unit, the generating unit and the determining unit in the thyroid nodule image processing apparatus is specifically described in the implementation process corresponding to steps S9020 to S9022 in the thyroid nodule image processing method, and will not be described herein.
Further, in an embodiment of the present application, the second determining module includes:
the screening unit is used for screening out abnormal gradients in the gradient map through a median filtering algorithm;
and the deleting unit is used for deleting the abnormal gradient from the gradient map.
In this embodiment, the implementation process of the functions and actions of the screening unit and the deleting unit in the thyroid nodule image processing apparatus is specifically described in the implementation process corresponding to steps S9000 to S9001 in the thyroid nodule image processing method, and will not be described herein.
Further, in an embodiment of the present application, the third generating module includes:
a fifth calculation sub-module for calculating a second difference of the second area minus the third area;
a sixth calculation sub-module for calculating a ratio between the fourth area and the second difference;
the second judging submodule is used for judging whether the ratio is larger than a preset first ratio threshold value or not when the coarse classification result is that the cystic is dominant or cystic;
a third judging submodule, configured to judge that the structural fine classification result of the thyroid nodule is cystic if the ratio is greater than the first ratio threshold, or judge that the structural fine classification result of the thyroid nodule is cystic;
the third judging sub-module is used for judging whether the ratio is smaller than a preset second ratio threshold value when the coarse classification result is that the reality is dominant or the reality is high;
And the fourth judging submodule is used for judging that the constitution fine classification result of the thyroid nodule is real if the ratio is smaller than the second ratio threshold value, otherwise judging that the constitution fine classification result of the thyroid nodule is real.
In this embodiment, the implementation process of the functions and actions of the fifth calculation submodule, the sixth calculation submodule, the second judgment submodule, the third judgment submodule and the fourth judgment submodule in the thyroid nodule image processing apparatus are specifically detailed in the implementation process corresponding to steps S1000 to S1005 in the thyroid nodule image processing method, and are not described herein again.
Referring to fig. 3, a computer device is further provided in the embodiment of the present application, where the computer device may be a server, and the internal structure of the computer device may be as shown in fig. 3. The computer device includes a processor, a memory, a network interface, a display screen, an input device, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used to store data such as the nodule mask image, the nodule image, the third nodule image, the final cystic region, and the resulting finely classified results of thyroid nodules. The network interface of the computer device is used for communicating with an external terminal through a network connection. The display screen of the computer equipment is an indispensable image-text output equipment in the computer and is used for converting digital signals into optical signals so that characters and graphics can be displayed on the screen of the display screen. The input device of the computer equipment is a main device for exchanging information between the computer and a user or other equipment, and is used for conveying data, instructions, certain sign information and the like into the computer. The computer program is executed by a processor to implement a method of processing thyroid nodule images.
The processor executes the steps of the thyroid nodule image processing method:
acquiring an input ultrasonic image with thyroid nodules, marking a nodule area on the ultrasonic image, and generating a nodule mask image corresponding to the nodule area;
intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
performing binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
determining a sound and image area from the first nodule image according to the non-sound and image part in the sound and image candidate area and the nodule boundary of the nodule area;
cutting the sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
acquiring a black partial region in the third nodule image, wherein the black partial region is positioned in the nodule region, and taking the black partial region as a cystic candidate region;
calculating a first area of the cystic candidate region, a second area of the nodule region, and a third area of the nodule region corresponding to an intersection of the sound shadow region;
Generating a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area and the third area, wherein the rough classification result of the thyroid nodule is cystic or cystic, or the rough classification result of the thyroid nodule is real or real;
determining a cystic region from the third nodule image based on the gradient of the third nodule image and the cystic candidate region;
invoking a preset calculation rule corresponding to the composition rough classification result to calculate and generate a composition fine classification result of the thyroid nodule according to a fourth area, the second area and the third area of the cystic area, wherein the composition fine classification result comprises: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
Those skilled in the art will appreciate that the structures shown in fig. 3 are only block diagrams of portions of structures that may be associated with the aspects of the present application and are not intended to limit the scope of the apparatus, or computer devices on which the aspects of the present application may be implemented.
An embodiment of the present application further provides a computer readable storage medium, on which a computer program is stored, where the computer program when executed by a processor implements a method for processing thyroid nodule images, specifically:
Acquiring an input ultrasonic image with thyroid nodules, marking a nodule area on the ultrasonic image, and generating a nodule mask image corresponding to the nodule area;
intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
performing binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
determining a sound and image area from the first nodule image according to the non-sound and image part in the sound and image candidate area and the nodule boundary of the nodule area;
cutting the sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
acquiring a black partial region in the third nodule image, wherein the black partial region is positioned in the nodule region, and taking the black partial region as a cystic candidate region;
calculating a first area of the cystic candidate region, a second area of the nodule region, and a third area of the nodule region corresponding to an intersection of the sound shadow region;
Generating a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area and the third area, wherein the rough classification result of the thyroid nodule is cystic or cystic, or the rough classification result of the thyroid nodule is real or real;
determining a cystic region from the third nodule image based on the gradient of the third nodule image and the cystic candidate region;
invoking a preset calculation rule corresponding to the composition rough classification result to calculate and generate a composition fine classification result of the thyroid nodule according to a fourth area, the second area and the third area of the cystic area, wherein the composition fine classification result comprises: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
The digital medical treatment method and the digital medical treatment device can be applied to the digital medical treatment field in the smart city, so that the construction of the smart city is promoted.
In summary, the processing method, the device, the computer equipment and the storage medium for the thyroid nodule image provided in the embodiments of the present application include a two-step process of identifying a thyroid nodule from thick to thin, in the course of coarse identification, firstly, sound image areas in the nodule image are extracted and removed by using sound image physical characteristics during thyroid ultrasound scanning, so that a situation that the sound image areas in the nodule image are misjudged as cystic areas cannot occur, and thus, a coarse classification result of the thyroid nodule is accurately generated according to a first area of candidate cystic areas in the nodule image after the sound image areas are removed, a second area of the nodule areas, and a third area corresponding to an intersection of the nodule areas and the sound image areas. And after the rough classification result is obtained, determining a cystic area in the nodule image by combining the gray level and gradient information of the nodule image, and calling a preset calculation rule corresponding to the rough classification result to quickly and accurately calculate and generate a fine classification result of the thyroid nodule according to a fourth area of the cystic area, the second area and the third area. By the method and the device, the structure of the thyroid nodule can be accurately identified, on one hand, an analysis report related to the structure of the thyroid nodule can be further generated, and on the other hand, the structure of the thyroid nodule can be input into a corresponding neural network model to judge whether the thyroid nodule is benign or malignant.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by hardware associated with a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium provided herein and used in embodiments may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, apparatus, article or method that comprises the element.
The foregoing description is only of the preferred embodiments of the present application, and is not intended to limit the scope of the claims, and all equivalent structures or equivalent processes using the descriptions and drawings of the present application, or direct or indirect application in other related technical fields are included in the scope of the claims of the present application.
Claims (10)
1. A method for processing an image of a thyroid nodule, comprising:
acquiring an input ultrasonic image with thyroid nodules, marking a nodule area on the ultrasonic image, and generating a nodule mask image corresponding to the nodule area;
Intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
performing binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
determining a sound and image area from the first nodule image according to the non-sound and image part in the sound and image candidate area and the nodule boundary of the nodule area;
cutting the sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
acquiring a black partial region in the third nodule image, wherein the black partial region is positioned in the nodule region, and taking the black partial region as a cystic candidate region;
calculating a first area of the cystic candidate region, a second area of the nodule region, and a third area of the nodule region corresponding to an intersection of the sound shadow region;
generating a rough classification result of the thyroid nodule according to a first preset rule according to the first area, the second area and the third area, wherein the rough classification result of the thyroid nodule is cystic or cystic, or the rough classification result of the thyroid nodule is real or real;
Determining a cystic region from the third nodule image based on the gradient of the third nodule image and the cystic candidate region;
invoking a preset calculation rule corresponding to the composition rough classification result to calculate and generate a composition fine classification result of the thyroid nodule according to a fourth area, the second area and the third area of the cystic area, wherein the composition fine classification result comprises: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
2. The method of claim 1, wherein the step of determining a sound image region from the first nodule image based on the non-sound image portion of the sound image candidate region and the nodule boundary comprises:
acquiring a non-sound image part in the sound image candidate area;
performing elimination treatment on the non-sound image part to obtain a sound image candidate area after treatment;
acquiring a first communication domain passing through a nodule boundary of the nodule region in the processed sound and image candidate region;
and taking the first connected domain as the sound image area.
3. The method according to claim 1, wherein the step of generating the rough classification result of the thyroid nodule according to the first preset rule based on the first area, the second area, and the third area comprises:
Calculating a first difference of the second area minus the third area;
calculating a quotient obtained by dividing the first difference by 2;
judging whether the first area is larger than the quotient value or not;
if the first area is larger than the quotient, judging that the rough classification result of the thyroid nodule is cystic or mainly cystic;
and if the first area is not larger than the quotient value, judging that the rough classification result of the thyroid nodule is true or true.
4. The method of processing thyroid nodule images of claim 1, wherein the step of determining a cystic region from the third nodule image from the gradient of the third nodule image and the cystic candidate region comprises:
calculating the gradient of the third nodule image and generating a corresponding gradient map;
searching seed points in the gradient map, wherein the gray scale of the seed points is smaller than a first preset threshold value, the gradient of the seed points is smaller than a second preset threshold value, and the seed points are in the cystic candidate area;
performing region growing treatment on the seed points according to a second preset rule to generate corresponding appointed connected domains;
carrying out corrosion treatment on the appointed connected domain;
calculating the product of the second area and a specified value;
Screening second connected domains with the connected domain area larger than the product from all the appointed connected domains;
and determining the second connected domain as the cystic region.
5. The method for processing a thyroid nodule image according to claim 4, wherein the step of performing region growing processing on the seed points according to a third preset rule to generate corresponding specified connected regions comprises:
acquiring a designated seed point, wherein the designated seed point is any seed point in all the seed points;
merging pixels adjacent to the specified seed points into the same set according to the composition conditions of the communication areas to generate a specified pixel set, wherein the composition conditions of the communication areas comprise the same pixel value and adjacent pixel values;
and taking the generated designated set as a connected domain corresponding to the designated seed point.
6. The method of processing thyroid nodule images of claim 4, wherein after the step of computing gradients of the third nodule image and generating a corresponding gradient map, comprising:
screening out abnormal gradients in the gradient map through a median filtering algorithm;
Deleting the abnormal gradient from the gradient map.
7. The method according to claim 1, wherein the step of calling a preset calculation rule corresponding to the coarse classification result to calculate and generate the fine classification result of the thyroid nodule according to a fourth area, the second area and the third area of the cystic area comprises:
calculating a second difference of the second area minus the third area;
calculating a ratio between the fourth area and the second difference;
when the coarse classification result is that the capsulization is dominant or the capsulization is dominant, judging whether the ratio is larger than a preset first ratio threshold;
if the ratio is larger than the first ratio threshold, judging that the structural fine classification result of the thyroid nodule is cystic, otherwise judging that the structural fine classification result of the thyroid nodule is cystic;
when the rough classification result is that the reality is dominant or the reality is high, judging whether the ratio is smaller than a preset second ratio threshold;
and if the ratio is smaller than the second ratio threshold, judging that the structural fine classification result of the thyroid nodule is real, otherwise, judging that the structural fine classification result of the thyroid nodule is real.
8. A processing apparatus for thyroid nodule images, comprising:
the first acquisition module is used for acquiring an input ultrasonic image with thyroid nodules, marking the nodule areas on the ultrasonic image and generating a nodule mask image corresponding to the nodule areas;
the first generation module is used for intercepting an image corresponding to the minimum circumscribed rectangle of the nodule mask image to obtain a nodule image;
the first processing module is used for carrying out binary segmentation processing on the nodule image according to a preset OTSU segmentation algorithm, generating a processed first nodule image, and acquiring a sound and image candidate region from the first nodule image;
a first determining module, configured to determine a sound-image region from the first nodule image according to a non-sound-image portion in the sound-image candidate region and a nodule boundary of the nodule region;
the second processing module is used for cutting the sound image area from the nodule image to obtain a corresponding second nodule image, and performing binary segmentation processing on the second nodule image according to the OTSU segmentation algorithm to obtain a processed third nodule image;
the second acquisition module is used for acquiring a black partial region which is positioned inside the nodule region in the third nodule image, and taking the black partial region as a cystic candidate region;
A calculation module for calculating a first area of the cystic candidate region, a second area of the nodule region, and a third area of the nodule region corresponding to an intersection of the sound shadow region;
the second generation module is used for generating a composition coarse classification result of the thyroid nodule according to the first area, the second area and the third area and a first preset rule, wherein the composition coarse classification result is mainly cystic or cystic, or the composition coarse classification result is mainly solid or solid;
a second determining module, configured to determine a cystic area from the third nodule image according to the gradient of the third nodule image and the cystic candidate area;
the third generating module is configured to invoke a preset calculation rule corresponding to the rough classification result to generate a fine classification result of the thyroid nodule according to a fourth area of the cystic area, the second area and the third area, where the fine classification result includes: the cystic nature is dominant, the cystic nature, the solidity is dominant or solidity.
9. A computer device comprising a memory and a processor, the memory having stored therein a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011035723.7A CN112102311B (en) | 2020-09-27 | 2020-09-27 | Thyroid nodule image processing method and device and computer equipment |
PCT/CN2020/132595 WO2021169452A1 (en) | 2020-09-27 | 2020-11-30 | Thyroid nodule image processing method and apparatus, and computer device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011035723.7A CN112102311B (en) | 2020-09-27 | 2020-09-27 | Thyroid nodule image processing method and device and computer equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112102311A CN112102311A (en) | 2020-12-18 |
CN112102311B true CN112102311B (en) | 2023-07-18 |
Family
ID=73782679
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011035723.7A Active CN112102311B (en) | 2020-09-27 | 2020-09-27 | Thyroid nodule image processing method and device and computer equipment |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN112102311B (en) |
WO (1) | WO2021169452A1 (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113940702A (en) * | 2021-10-11 | 2022-01-18 | 什维新智医疗科技(上海)有限公司 | Thyroid nodule echo analysis device |
CN114332003A (en) * | 2021-12-27 | 2022-04-12 | 什维新智医疗科技(上海)有限公司 | Thyroid nodule texture analysis device |
CN114708156A (en) * | 2022-03-15 | 2022-07-05 | 平安科技(深圳)有限公司 | Method, device and equipment for generating reference image of ultrasonic image and storage medium |
CN118115415B (en) * | 2024-04-29 | 2024-07-05 | 陕西省核工业二一五医院 | Ultrasonic image optimization processing method and system based on artificial intelligence |
CN118135340B (en) * | 2024-05-06 | 2024-07-23 | 天津市肿瘤医院(天津医科大学肿瘤医院) | Lung image focus pre-marking method, system and medium based on lung region segmentation |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2003105675A2 (en) * | 2002-06-18 | 2003-12-24 | Lifespan Biosciences, Inc. | Computerized image capture of structures of interest within a tissue sample |
WO2016085236A1 (en) * | 2014-11-25 | 2016-06-02 | 재단법인 아산사회복지재단 | Method and system for automatic determination of thyroid cancer |
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 |
CN111461202A (en) * | 2020-03-30 | 2020-07-28 | 上海尽星生物科技有限责任公司 | Real-time thyroid nodule ultrasonic image identification method and device |
CN111612752A (en) * | 2020-05-15 | 2020-09-01 | 江苏省人民医院(南京医科大学第一附属医院) | Ultrasonic image thyroid nodule intelligent detection system based on fast-RCNN |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030161513A1 (en) * | 2002-02-22 | 2003-08-28 | The University Of Chicago | Computerized schemes for detecting and/or diagnosing lesions on ultrasound images using analysis of lesion shadows |
EP3397773B1 (en) * | 2015-12-30 | 2021-04-28 | Technische Universität Dresden | Method and means for diagnosing tumors |
CN107680678B (en) * | 2017-10-18 | 2020-12-01 | 北京航空航天大学 | Thyroid ultrasound image nodule diagnosis system based on multi-scale convolution neural network |
CN108734694A (en) * | 2018-04-09 | 2018-11-02 | 华南农业大学 | Thyroid tumors ultrasonoscopy automatic identifying method based on faster r-cnn |
CN109003260B (en) * | 2018-06-28 | 2021-02-09 | 深圳视见医疗科技有限公司 | CT image pulmonary nodule detection method, device and equipment and readable storage medium |
CN111553919B (en) * | 2020-05-12 | 2022-12-30 | 上海深至信息科技有限公司 | Thyroid nodule analysis system based on elastic ultrasonic imaging |
-
2020
- 2020-09-27 CN CN202011035723.7A patent/CN112102311B/en active Active
- 2020-11-30 WO PCT/CN2020/132595 patent/WO2021169452A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2003105675A2 (en) * | 2002-06-18 | 2003-12-24 | Lifespan Biosciences, Inc. | Computerized image capture of structures of interest within a tissue sample |
WO2016085236A1 (en) * | 2014-11-25 | 2016-06-02 | 재단법인 아산사회복지재단 | Method and system for automatic determination of thyroid cancer |
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 |
CN111461202A (en) * | 2020-03-30 | 2020-07-28 | 上海尽星生物科技有限责任公司 | Real-time thyroid nodule ultrasonic image identification method and device |
CN111612752A (en) * | 2020-05-15 | 2020-09-01 | 江苏省人民医院(南京医科大学第一附属医院) | Ultrasonic image thyroid nodule intelligent detection system based on fast-RCNN |
Non-Patent Citations (1)
Title |
---|
声触诊组织成像鉴别甲状腺实性结节良恶性的价值研究;侯金香;王兴田;李曾;詹维伟;周伟;张丽;左云鹏;陈聪;;齐齐哈尔医学院学报(01);第73-75页 * |
Also Published As
Publication number | Publication date |
---|---|
WO2021169452A1 (en) | 2021-09-02 |
CN112102311A (en) | 2020-12-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112102311B (en) | Thyroid nodule image processing method and device and computer equipment | |
Kalfarisi et al. | Crack detection and segmentation using deep learning with 3D reality mesh model for quantitative assessment and integrated visualization | |
Hoang | An Artificial Intelligence Method for Asphalt Pavement Pothole Detection Using Least Squares Support Vector Machine and Neural Network with Steerable Filter‐Based Feature Extraction | |
Wu et al. | Improvement of crack-detection accuracy using a novel crack defragmentation technique in image-based road assessment | |
Feng et al. | Pavement crack detection and segmentation method based on improved deep learning fusion model | |
Chambon et al. | Automatic road pavement assessment with image processing: Review and comparison | |
Tsai et al. | Multiscale crack fundamental element model for real-world pavement crack classification | |
Huyan et al. | Illumination compensation model with k-means algorithm for detection of pavement surface cracks with shadow | |
Su et al. | Deep convolutional neural network–based pixel-wise landslide inventory mapping | |
CN111899246B (en) | Slide digital information quality detection method, device, equipment and medium | |
Shi et al. | A novel underwater dam crack detection and classification approach based on sonar images | |
Tsai et al. | Comprehensive, quantitative crack detection algorithm performance evaluation system | |
CN111415336B (en) | Image tampering identification method, device, server and storage medium | |
Hoang et al. | Fast local Laplacian‐based steerable and Sobel filters integrated with adaptive boosting classification tree for automatic recognition of asphalt pavement cracks | |
Huyan et al. | Pixelwise asphalt concrete pavement crack detection via deep learning‐based semantic segmentation method | |
Gao et al. | Real-time detecting one specific tampering operation in multiple operator chains | |
CN115239655A (en) | Thyroid ultrasonic image tumor segmentation and classification method and device | |
Kokila et al. | Hybrid Behrens-Fisher-and gray contrast–based feature point selection for building detection from satellite images | |
Al Qudah et al. | Using Artificial Intelligence applications for E-Government services as iris recognition | |
CN113793332A (en) | Experimental instrument defect identification and classification method and system | |
Wilm et al. | Fast whole-slide cartography in colon cancer histology using superpixels and CNN classification | |
CN112529888A (en) | Face image evaluation method, device, equipment and medium based on deep learning | |
CN113610971B (en) | Fine-grained three-dimensional model construction method and device and electronic equipment | |
Worms et al. | Record events attribution in climate studies | |
Kee et al. | Cracks identification using mask region-based denoised deformable convolutional network |
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 |