CN108876772B - Lung cancer image diagnosis system and method based on big data - Google Patents

Lung cancer image diagnosis system and method based on big data Download PDF

Info

Publication number
CN108876772B
CN108876772B CN201810567040.2A CN201810567040A CN108876772B CN 108876772 B CN108876772 B CN 108876772B CN 201810567040 A CN201810567040 A CN 201810567040A CN 108876772 B CN108876772 B CN 108876772B
Authority
CN
China
Prior art keywords
image
module
contour
data
lung cancer
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
Application number
CN201810567040.2A
Other languages
Chinese (zh)
Other versions
CN108876772A (en
Inventor
奉水东
凌宏艳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of South China
Original Assignee
University of South China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of South China filed Critical University of South China
Priority to CN201810567040.2A priority Critical patent/CN108876772B/en
Publication of CN108876772A publication Critical patent/CN108876772A/en
Application granted granted Critical
Publication of CN108876772B publication Critical patent/CN108876772B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/194—Segmentation; Edge detection involving foreground-background 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30061—Lung
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30096—Tumor; Lesion

Landscapes

  • Engineering & Computer Science (AREA)
  • 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)
  • Apparatus For Radiation Diagnosis (AREA)
  • Image Processing (AREA)

Abstract

The invention belongs to the technical field of medical diagnosis, and discloses a lung cancer image diagnosis system and method based on big data.A image acquisition module performs image segmentation by adopting a gray threshold method, performs defect repair on a binary image by using a mathematical morphology method, tracks and stores contour information through chain codes, performs image contour extraction with a single-pixel edge, and obtains user image data information; collecting gas exhaled by a user through an exhalation collecting module; the body temperature data of the user is detected through the body temperature detection module. The lung cancer image diagnosis method based on big data has strong operability and is easy to carry out. The method is widely applied in practice, and can be applied to image data processing in other fields. The image diagnosis method has strong robustness; the accuracy of predicting early cell lung cancer can be improved; meanwhile, the cloud service module can greatly improve the analysis and calculation speed of the diagnosis data and improve the diagnosis efficiency.

Description

Lung cancer image diagnosis system and method based on big data
Technical Field
The invention belongs to the technical field of medical diagnosis, and particularly relates to a lung cancer image diagnosis system and method based on big data.
Background
Currently, the current state of the art commonly used in the industry is such that:
lung cancer is one of the most rapidly growing malignancies that threaten human health and life. In many countries, the incidence and mortality of lung cancer have been reported to be significantly higher in recent 50 years, with lung cancer incidence and mortality in men accounting for the first of all malignancies, in women accounting for the second, and mortality accounting for the second. The etiology of lung cancer is not completely clear up to now, and a large amount of data show that a large amount of smoking for a long time has a very close relationship with the occurrence of lung cancer. Existing studies have demonstrated that: the probability of lung cancer of a large number of smokers in a long term is 10-20 times that of non-smokers, and the smaller the smoking starting age is, the higher the probability of lung cancer is. In addition, smoking not only directly affects the health of the user, but also has adverse effects on the health of surrounding people, so that the lung cancer prevalence of passive smokers is obviously increased. The incidence of lung cancer in urban residents is higher than that in rural areas, which may be related to urban atmospheric pollution and carcinogens contained in smoke dust. So no smoking should be advocated and the urban sanitation work should be enhanced. However, the existing lung cancer diagnosis image data is inaccurate; meanwhile, the lung cancer diagnosis data processing speed is low, and the diagnosis efficiency is low.
The image processing technology based on computer vision is based on optics and integrates various comprehensive technologies of modern science, such as optoelectronics, computer technology, image processing technology and the like. The geometric characteristic parameters of the acquired object are obtained by processing the image of the acquired object, and the extraction of the target object outline becomes a main factor influencing the measurement accuracy. Therefore, in a computer vision-based measurement system, it is very important to select a suitable contour extraction method for ensuring the measurement accuracy.
The basic method of contour extraction is the edge detection method, i.e. by means of spatial differential operators, by convolving the template with the image.
The classical edge detection method is a local operator method, such as a gradient operator, a Sobel operator, a Roberts operator, a Canny operator and the like, and has the advantages of simple realization, high operation speed and the like, but has the following defects: the detected edge cannot be guaranteed to be single-pixel wide, and isolated or only small-segment continuous edges often appear. Therefore, thinning is required while trying to connect the intermittent edge pixels so that contour extraction can be accomplished. Obviously, the processing procedure is too complex, and the precision of the extracted contour cannot be guaranteed. In some cases, it is not even possible to extract the contour of the image due to the influence of noise.
In summary, the problems of the prior art are as follows:
the existing lung cancer diagnosis image data is inaccurate; meanwhile, the lung cancer diagnosis data processing speed is low, and the diagnosis efficiency is low.
The existing lung cancer image diagnosis method based on big data has poor operability and complex image diagnosis method. The application is limited in practice.
The traditional edge detection method has the characteristics of weak anti-interference performance, low precision and the like, and cannot meet the actual requirements of medical engineering image technology.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a lung cancer image diagnosis system and method based on big data.
The present invention is achieved in such a way that a big-data-based lung cancer image diagnosis method includes the steps of:
the image acquisition module is used for carrying out image segmentation by adopting a gray threshold method, repairing defects of a binary image by adopting a mathematical morphology method, tracking and storing contour information through chain codes, and extracting an image contour with a single pixel edge to obtain user image data information; collecting gas exhaled by a user through an exhalation collecting module; detecting body temperature data of a user through a body temperature detection module;
the central control module dispatches the gas component detection module to pump the user gas collected by the air bag into the reaction gas chamber, so that the gas can react with the porphyrin sensor in advance in the reaction gas chamber to detect the exhaled harmful gas component;
the screening module detects the content of YKL-40 and NSE in the blood sample, and analyzes and judges the risk of image lesion; performing pathological analysis on the detected image data through a pathological analysis module;
detection data signal s (t) is carried out by centralizing big data computing resources through cloud service module
Figure BDA0001684801970000031
A non-linear transformation process in which
Figure BDA0001684801970000032
A represents the amplitude of the signal, a (m) represents the symbol sign of the signal, p (t) represents the shaping function, fcWhich is indicative of the carrier frequency of the signal,
Figure BDA0001684801970000033
representing the phase of the signal; and displaying the detection data information through the display module.
Further, in the method of dividing an image by the gray threshold method, if the gray scale interval of the image f (x, y) is [ Zmin, Zmax ]]Within this interval, a threshold value Z is settAnd Z ismin<Zt<ZmaxMaking all gray values in the image less than or equal to ZtAll of the new gray values of pixels of (2) are 0 and greater than ZtThe new gray values of the pixels of (a) are all 1, and an output binary image ft (x, y) is constructed through the threshold segmentation:
Figure BDA0001684801970000034
wherein the gray threshold is Zt；
The method specifically comprises the following steps:
1) determining the maximum gray-scale value Z of the imagemaxAnd the minimum gray value ZminLet a threshold value ZtHas an initial value of (Z)max+Zmin)/2；
2) According to a threshold value ZtAn image is divided into a target and a background, and an average gradation value Z of the target and the background is obtainedOAnd ZB；
3) Finding a new threshold value Zt+1＝(ZO+ZB)/2；
4) If Z ist＝Zt+1If so, obtaining the threshold value; otherwise, order Zt＝Zt+1Turning to the step 2) to continue iterative computation.
Further, the method for defect repair of the binary image by using the mathematical morphology method comprises the following steps:
setting A as an input binary image, and performing morphological processing on the image by adopting a structural element B;
a is expanded by B, denoted as a ≦ B ≦ x ≦ a + B for some a ∈ a and B ∈ B;
a is eroded by B, denoted as a Θ B, as a Θ B ═ { x | (x + B) ∈ a, for each B ∈ B };
a is calculated by B, denoted as a σ B, as: a θ B is (a Θ B) · B;
a is closed by B, and expressed as a · B, where a · B is ═ a ∞ B.
Further, the method for storing the contour information by chain code tracking comprises the following steps:
firstly, obtaining an initial contour point by adopting a line scanning technology, recording coordinates of the initial contour point as Start-X and Start-Y, taking the initial contour point as a current point, and turning to the second step; if the contour point can not be obtained after scanning, turning to the fourth step;
secondly, scanning 8 neighborhoods adjacent to the current point according to the sequence of the direction chain codes, if the contour point is encountered, immediately stopping tracking the contour, recording the tracked direction chain code value, and turning to the third step; if no contour point is encountered, setting a mark for ending the contour tracking, setting the scanning starting points as Start-X and Start-Y, and turning to the first step;
filling the scanned contour points with ground color, setting the current points as the tracked contour points, and turning to the second step;
fourthly, using a mark to mark all the contours to finish tracking; obtaining chain code sequences List of all the outline information; the format of the chain code sequence comprises the coordinates of a starting point and the values of direction chain codes, and different outlines in the sequence are separated by outline ending marks; the recording method comprises that for a closed contour, firstly, the starting point of the contour coordinate, then the direction chain code sequence and then the contour ending mark are included; because the self-closed tracking of the contour is adopted, the values in the sequence are recorded according to the sequence of the contour lines, and convenience is provided for the subsequent processing of the contour. The chain code tracking method can realize one-time tracking to obtain all the contours;
the method for extracting the image contour with the single-pixel edge comprises the following steps:
1) and (3) calculating the predicted value of the target contour by using a Sobel operator or a color space clustering method:
Figure BDA0001684801970000041
2) treating the image object outline as a set of N unit line segments { dl }i}1，2，...，NFor i ═ 1, 2, …, N;
3) at C0Is found iniCorresponding position according to C0Tangent line of corresponding position in the middle is dliGenerating an initial particle set according to the sampling reference value;
4) guiding the particles to gather towards the known optimal solution direction continuously according to the state transition model, avoiding the degradation method in the standard particle filtering process to realize the state transition of the particles, and calculating the contour point set corresponding to each particle;
5) calculating the weight of the particles according to the established observation model;
6) calculating parameter dl obtained in the current iteration by weighted average of particle setsj (i)＝(kj (i)，bj (i))；
7) If it is
Figure BDA0001684801970000051
Where ε is taken to be 0.5, the weighted average of the particle sets is obtained as dliEstimating parameters, otherwise, turning to step 4).
Further, the screening module screening method comprises the following steps:
firstly, detecting the content of YKL-40 and NSE in a blood sample;
then, if the NSE content is in the critical value range and the YKL-40 content is larger than the critical value range, preliminarily judging that the blood sample has larger cell lesion image risk;
and if the content of the NSE is larger than the critical value range of the NSE and the content of the YKL-40 is equal to or smaller than the critical value range of the NSE, preliminarily judging that the blood sample has less risk of cytopathic image.
Another object of the present invention is to provide a computer program for implementing the big-data based lung cancer image diagnosis method.
Another object of the present invention is to provide an information data processing terminal for implementing the big-data-based lung cancer image diagnosis method.
Another object of the present invention is to provide a computer-readable storage medium including instructions which, when executed on a computer, cause the computer to perform the big-data based lung cancer image diagnosis method.
Another object of the present invention is to provide a big-data-based lung cancer image diagnosis system for implementing the big-data-based lung cancer image diagnosis method, including:
the device comprises an image acquisition module, an expiration acquisition module, a body temperature detection module, a central control module, a gas component detection module, a screening module, a pathological analysis module, a cloud service module and a display module;
the image acquisition module is connected with the central control module and is used for acquiring user image data information through the camera;
the breath collecting module is connected with the central control module and is used for collecting the gas exhaled by the user through the air bag;
the body temperature detection module is connected with the central control module and used for detecting body temperature data of a user through the temperature sensor;
the central control module is connected with the image acquisition module, the breath acquisition module, the body temperature detection module, the gas component detection module, the screening module, the pathological analysis module, the cloud service module and the display module and is used for controlling the modules to normally work;
the gas component detection module is connected with the central control module and used for pumping the user gas collected by the air bag into the reaction gas chamber, so that the gas can react with the porphyrin sensor in advance in the reaction gas chamber to detect the exhaled harmful gas components;
the screening module is connected with the central control module and is used for extracting a blood sample of the user to analyze and judge the risk of the pathological changes;
the pathological analysis module is connected with the central control module and is used for carrying out pathological analysis on the detected data;
the cloud service module is connected with the central control module and used for processing the detection data by centralizing big data computing resources through a cloud server;
and the display module is connected with the central control module and is used for displaying the detection data information.
Another object of the present invention is to provide a lung cancer image diagnosis apparatus equipped with the big-data based lung cancer image diagnosis system.
The invention has the advantages and positive effects that:
the screening module is used for screening the early cell lung cancer by joint detection of the keratanase protein (YKL-40) and the neuron-specific enolase (NSE). Because the sensitivity of the keratanase protein (YKL-40) is higher than that of neuron-specific enolase (NSE), the accuracy of predicting early cell lung cancer can be improved by the method; meanwhile, the cloud service module can greatly improve the analysis and calculation speed of the diagnosis data and improve the diagnosis efficiency.
The lung cancer image diagnosis method based on big data has strong operability and is easy to carry out. The method is widely applied in practice, and can be applied to image data processing in other fields. The image diagnosis method has strong robustness.
The invention provides a practical outline extraction method by combining the characteristics of a computer vision measurement technology. The method adopts a gray threshold value method to carry out image segmentation, adopts a mathematical morphology method to carry out defect repair on a binary image, and realizes the extraction of the image contour with a single pixel edge by tracking and storing contour information through chain codes. The principle and the implementation method of the key technology are provided. Experiments show that compared with the classical edge detection method, the method has the characteristics of strong anti-interference performance, high precision and the like, and can meet the actual requirements of medical engineering.
The contour extracted by the method has the characteristics of continuity, high precision, wide single pixel and the like, and compared with a classical edge detection method, the contour extraction method provided by the invention has the advantages of strong anti-interference performance and good robustness, and can meet the requirements of a computer vision image measurement technology.
The binary image obtained by threshold segmentation of the invention may have defects such as broken lines, pits, burrs and the like, and particularly, if the image with the defects is directly used for contour extraction under the condition that the noise exists in the image, the contour with the defects such as broken points, burrs and the like can be obtained, which causes difficulty in further processing the contour and even influences the measurement precision, so measures must be taken to eliminate the defects.
Mathematical morphology is a nonlinear filtering method in which the expansion and erosion operations have a very intuitive geometric background, allowing the image to be thickened or thinned in a direction that depends on the selected structural elements. Based on the characteristic, the invention adopts a mathematical morphology method to carry out defect repair on the binary image so as to eliminate defects and noise in the image.
The invention concentrates big data computing resources to detect data signals s (t) through the cloud service module
Figure BDA0001684801970000071
And through nonlinear transformation processing, accurate detection data can be obtained, and conditions are provided for later-stage requirements.
Drawings
FIG. 1 is a flow chart of a big data-based lung cancer diagnosis method provided by the practice of the present invention.
Fig. 2 is a block diagram of a lung cancer image diagnosis system based on big data according to an embodiment of the present invention.
In the figure: 1. an image acquisition module; 2. an expired air collection module; 3. a body temperature detection module; 4. a central control module; 5. a gas component detection module; 6. a screening module; 7. a pathology analysis module; 8. a cloud service module; 9. and a display module.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, the big data-based lung cancer diagnosis method provided by the embodiment of the present invention includes the following steps:
s101, acquiring user image data information through an image acquisition module; collecting gas exhaled by a user through an exhalation collecting module; detecting body temperature data of a user through a body temperature detection module;
s102, the central control module dispatches the gas component detection module to pump the user gas collected by the air bag into the reaction gas chamber, so that the gas can react with the porphyrin sensor in the reaction gas chamber to detect the exhaled harmful gas components;
s103, extracting a blood sample of the user through a screening module, analyzing and judging the risk of the lesion; performing pathological analysis on the detected data through a pathological analysis module;
s104, processing the detection data by centralizing big data computing resources through a cloud service module; and displaying the detection data information through the display module.
As shown in fig. 2, the lung cancer image diagnosis system based on big data according to the embodiment of the present invention includes: the device comprises an image acquisition module 1, an expiration acquisition module 2, a body temperature detection module 3, a central control module 4, a gas component detection module 5, a screening module 6, a pathology analysis module 7, a cloud service module 8 and a display module 9.
The image acquisition module 1 is connected with the central control module 4 and is used for acquiring user image data information through a camera;
the breath collecting module 2 is connected with the central control module 4 and is used for collecting the gas exhaled by the user through the air bag;
the body temperature detection module 3 is connected with the central control module 4 and is used for detecting body temperature data of a user through a temperature sensor;
the central control module 4 is connected with the image acquisition module 1, the breath acquisition module 2, the body temperature detection module 3, the gas component detection module 5, the screening module 6, the pathology analysis module 7, the cloud service module 8 and the display module 9 and is used for controlling the normal work of each module;
the gas component detection module 5 is connected with the central control module 4 and used for pumping the user gas collected by the air bag into the reaction gas chamber, so that the gas can react with the porphyrin sensor in advance in the reaction gas chamber to detect the exhaled harmful gas components;
the screening module 6 is connected with the central control module 4 and is used for extracting a blood sample of the user to analyze and judge the lesion risk;
the pathological analysis module 7 is connected with the central control module 4 and is used for carrying out pathological analysis on the detected data;
the cloud service module 8 is connected with the central control module 4 and used for processing the detection data through the cloud server centralized big data computing resources;
and the display module 9 is connected with the central control module 4 and is used for displaying the detection data information.
The invention is further described with reference to specific examples.
The lung cancer image diagnosis method based on big data provided by the embodiment of the invention comprises the following steps:
the image acquisition module is used for carrying out image segmentation by adopting a gray threshold method, repairing defects of a binary image by adopting a mathematical morphology method, tracking and storing contour information through chain codes, and extracting an image contour with a single pixel edge to obtain user image data information; collecting gas exhaled by a user through an exhalation collecting module; detecting body temperature data of a user through a body temperature detection module;
the central control module dispatches the gas component detection module to pump the user gas collected by the air bag into the reaction gas chamber, so that the gas can react with the porphyrin sensor in advance in the reaction gas chamber to detect the exhaled harmful gas component;
the screening module detects the content of YKL-40 and NSE in the blood sample, and analyzes and judges the risk of image lesion; performing pathological analysis on the detected image data through a pathological analysis module;
detection data signal s (t) is carried out by centralizing big data computing resources through cloud service module
Figure BDA0001684801970000091
A non-linear transformation process in which
Figure BDA0001684801970000092
A represents the amplitude of the signal, a (m) represents the symbol sign of the signal, p (t) represents the shaping function, fcWhich is indicative of the carrier frequency of the signal,
Figure BDA0001684801970000101
representing the phase of the signal; and displaying the detection data information through the display module.
In the method of image segmentation by gray threshold method, if the gray scale interval of the image f (x, y) is [ Zmin, Zmax]Within this interval, a threshold value Z is settAnd Z ismin<Zt<ZmaxMaking all gray values in the image less than or equal to ZtAll of the new gray values of pixels of (2) are 0 and greater than ZtThe new gray values of the pixels of (a) are all 1, and an output binary image ft (x, y) is constructed through the threshold segmentation:
Figure BDA0001684801970000102
wherein the gray threshold is Zt；
The method specifically comprises the following steps:
1) determining the maximum gray-scale value Z of the imagemaxAnd the minimum gray value ZminLet a threshold value ZtHas an initial value of (Z)max+Zmin)/2；
2) According to a threshold value ZtAn image is divided into a target and a background, and an average gradation value Z of the target and the background is obtainedOAnd ZB；
3) Finding a new threshold value Zt+1＝(ZO+ZB)/2；
4) If Z ist＝Zt+1If so, obtaining the threshold value; otherwise, order Zt＝Zt+1Turning to the step 2) to continue iterative computation.
Further, the method for defect repair of the binary image by using the mathematical morphology method comprises the following steps:
setting A as an input binary image, and performing morphological processing on the image by adopting a structural element B;
a is expanded by B, denoted as a ≦ B ≦ x ≦ a + B for some a ∈ a and B ∈ B;
a is eroded by B, denoted as a Θ B, as a Θ B ═ { x | (x + B) ∈ a, for each B ∈ B };
a is calculated by B, denoted as a σ B, as: a θ B is (a Θ B) · B;
a is closed by B, and expressed as a · B, where a · B is ═ a ∞ B.
Further, the method for storing the contour information by chain code tracking comprises the following steps:
firstly, obtaining an initial contour point by adopting a line scanning technology, recording coordinates of the initial contour point as Start-X and Start-Y, taking the initial contour point as a current point, and turning to the second step; if the contour point can not be obtained after scanning, turning to the fourth step;
secondly, scanning 8 neighborhoods adjacent to the current point according to the sequence of the direction chain codes, if the contour point is encountered, immediately stopping tracking the contour, recording the tracked direction chain code value, and turning to the third step; if no contour point is encountered, setting a mark for ending the contour tracking, setting the scanning starting points as Start-X and Start-Y, and turning to the first step;
filling the scanned contour points with ground color, setting the current points as the tracked contour points, and turning to the second step;
fourthly, using a mark to mark all the contours to finish tracking; obtaining chain code sequences List of all the outline information; the format of the chain code sequence comprises the coordinates of a starting point and the values of direction chain codes, and different outlines in the sequence are separated by outline ending marks; the recording method comprises that for a closed contour, firstly, the starting point of the contour coordinate, then the direction chain code sequence and then the contour ending mark are included; because the self-closed tracking of the contour is adopted, the values in the sequence are recorded according to the sequence of the contour lines, and convenience is provided for the subsequent processing of the contour. The chain code tracking method can realize one-time tracking to obtain all the contours;
the method for extracting the image contour with the single-pixel edge comprises the following steps:
1) and (3) calculating the predicted value of the target contour by using a Sobel operator or a color space clustering method:
Figure BDA0001684801970000111
2) treating the image object outline as a set of N unit line segments { dl }i}1，2，...，NFor i ═ 1, 2, …, N;
3) at C0Is found iniCorresponding position according to C0Tangent line of corresponding position in the middle is dliGenerating an initial particle set according to the sampling reference value;
4) guiding the particles to gather towards the known optimal solution direction continuously according to the state transition model, avoiding the degradation method in the standard particle filtering process to realize the state transition of the particles, and calculating the contour point set corresponding to each particle;
5) calculating the weight of the particles according to the established observation model;
6) calculating parameter dl obtained in the current iteration by weighted average of particle setsj (i)＝(kj (i)，bj (i))；
7) If it is
Figure BDA0001684801970000112
Where ε is taken to be 0.5, the weighted average of the particle sets is obtained as dliEstimating parameters, otherwise, turning to step 4).
The screening module screening method comprises the following steps:
firstly, detecting the content of YKL-40 and NSE in a blood sample;
then, if the NSE content is in the critical value range and the YKL-40 content is larger than the critical value range, preliminarily judging that the blood sample has larger cell lesion image risk;
and if the content of the NSE is larger than the critical value range of the NSE and the content of the YKL-40 is equal to or smaller than the critical value range of the NSE, preliminarily judging that the blood sample has less risk of cytopathic image.
The sample provided by the invention is a blood sample.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When used in whole or in part, can be implemented in a computer program product that includes one or more computer instructions. When loaded or executed on a computer, cause the flow or functions according to embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL), or wireless (e.g., infrared, wireless, microwave, etc.)). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (ssd)), among others.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (6)

1. A big data-based lung cancer image diagnosis method is characterized by comprising the following steps:
the image acquisition module is used for carrying out image segmentation by adopting a gray threshold method, repairing defects of a binary image by adopting a mathematical morphology method, tracking and storing contour information through chain codes, and extracting an image contour with a single pixel edge to obtain user image data information; collecting gas exhaled by a user through an exhalation collecting module; detecting body temperature data of a user through a body temperature detection module;
the central control module dispatches the gas component detection module to pump the user gas collected by the air bag into the reaction gas chamber, so that the gas can react with the porphyrin sensor in advance in the reaction gas chamber to detect the exhaled harmful gas component;
the screening module detects the content of YKL-40 and NSE in the blood sample, and analyzes and judges the risk of image lesion; performing pathological analysis on the detected image data through a pathological analysis module;
centralizing large data computing resources for detecting data signals through cloud service modules (t) carrying out
Figure FDA0003210443720000011
A non-linear transformation process in which
Figure FDA0003210443720000012
A represents the amplitude of the signal, a (m) represents the symbol sign of the signal, p (t) represents the shaping function, fcWhich is indicative of the carrier frequency of the signal,
Figure FDA0003210443720000013
representing the phase of the signal; displaying the detection data information through a display module;
in the method of image segmentation by gray threshold method, if the gray scale interval of the image f (x, y) is [ Zmin, Zmax]Within this interval, a threshold value Z is settAnd Z ismin<Zt<ZmaxMaking all gray values in the image less than or equal to ZtAll of the new gray values of pixels of (2) are 0 and greater than ZtThe new gray values of the pixels of (a) are all 1, and an output binary image ft (x, y) is constructed through the threshold segmentation:
Figure FDA0003210443720000014
wherein the gray threshold is Zt；
The method specifically comprises the following steps:
1) determining the maximum gray-scale value Z of the imagemaxAnd the minimum gray value ZminLet a threshold value ZtHas an initial value of (Z)max+Zmin)/2；
2) According to a threshold value ZtAn image is divided into a target and a background, and an average gradation value Z of the target and the background is obtainedOAnd ZB；
3) Finding a new threshold value Zt+1＝(ZO+ZB)/2；
4) If Z ist＝Zt+1If so, obtaining the threshold value; otherwise, order Zt＝Zt+1Turning to the step 2) to continue iterative computation; the method for repairing the defect of the binary image by using the mathematical morphology method comprises the following steps:
setting A as an input binary image, and performing morphological processing on the image by adopting a structural element B;
a is dilated by B, denoted as a ≦ B, as: a ≦ B ≦ { x | x ≦ a + B, for some a ∈ a and B ∈ B };
a is eroded by B, denoted A Θ B, as: a Θ B ═ { x | (x + B) ∈ a, for each B ∈ B };
a is calculated by B, denoted as a σ B, as: a θ B is (a Θ B) · B;
a is closed by B, expressed as A.B, and is: a · B ═ (a ≦ B) Θ B;
the method for storing the contour information by chain code tracking comprises the following steps:
firstly, obtaining an initial contour point by adopting a line scanning technology, recording coordinates of the initial contour point as Start-X and Start-Y, taking the initial contour point as a current point, and turning to the second step; if the contour point can not be obtained after scanning, turning to the fourth step;
secondly, scanning 8 neighborhoods adjacent to the current point according to the sequence of the direction chain codes, if the contour point is encountered, immediately stopping tracking the contour, recording the tracked direction chain code value, and turning to the third step; if no contour point is encountered, setting a mark for ending the contour tracking, setting the scanning starting points as Start-X and Start-Y, and turning to the first step;
filling the scanned contour points with ground color, setting the current points as the tracked contour points, and turning to the second step;
fourthly, using a mark to mark all the contours to finish tracking; obtaining chain code sequences List of all the outline information; the format of the chain code sequence comprises the coordinates of a starting point and the values of direction chain codes, and different outlines in the sequence are separated by outline ending marks; the recording mode comprises the following steps: for a closed contour, firstly, the starting point of a contour coordinate, then a direction chain code sequence and then a contour ending mark are included;
the method for extracting the image contour with the single-pixel edge comprises the following steps:
1) and (3) calculating the predicted value of the target contour by using a Sobel operator or a color space clustering method:
Figure FDA0003210443720000031
2) treating the image object outline as a set of N unit line segments { dl }i}1，2，…，NFor i ═ 1, 2, …, N;
3) at C0Is found iniCorresponding position according to C0Tangent line of corresponding position in the middle is dliGenerating an initial particle set according to the sampling reference value;
4) guiding the particles to gather towards the known optimal solution direction continuously according to the state transition model, avoiding the degradation method in the standard particle filtering process to realize the state transition of the particles, and calculating the contour point set corresponding to each particle;
5) calculating the weight of the particles according to the established observation model;
6) calculating parameter dl obtained in the current iteration by weighted average of particle setsj (i)＝(kj (i)，bj (i))；
7) If it is
Figure FDA0003210443720000032
Where ε is taken to be 0.5, the weighted average of the particle sets is obtained as dliEstimating parameters, otherwise, turning to step 4).
2. The big-data based lung cancer image diagnosis method as claimed in claim 1, wherein the screening module screening method comprises:
firstly, detecting the content of YKL-40 and NSE in a blood sample;
then, if the NSE content is in the critical value range and the YKL-40 content is larger than the critical value range, preliminarily judging that the blood sample has larger cell lesion image risk;
and if the content of the NSE is larger than the critical value range of the NSE and the content of the YKL-40 is equal to or smaller than the critical value range of the NSE, preliminarily judging that the blood sample has less risk of cytopathic image.
3. An information data processing terminal for implementing the big data based lung cancer image diagnosis method according to any one of claims 1 to 2.
4. A computer-readable storage medium comprising instructions which, when executed on a computer, cause the computer to perform the big-data based lung cancer image diagnosis method according to any one of claims 1 to 2.
5. A big-data-based lung cancer image diagnosis system implementing the big-data-based lung cancer image diagnosis method according to claim 1, comprising:
the image acquisition module is connected with the central control module and is used for acquiring user image data information through the camera;
the breath collecting module is connected with the central control module and is used for collecting the gas exhaled by the user through the air bag;
the body temperature detection module is connected with the central control module and used for detecting body temperature data of a user through the temperature sensor;
the central control module is connected with the image acquisition module, the breath acquisition module, the body temperature detection module, the gas component detection module, the screening module, the pathological analysis module, the cloud service module and the display module and is used for controlling the modules to normally work;
the gas component detection module is connected with the central control module and used for pumping the user gas collected by the air bag into the reaction gas chamber, so that the gas can react with the porphyrin sensor in advance in the reaction gas chamber to detect the exhaled harmful gas components;
the screening module is connected with the central control module and is used for extracting a blood sample of the user to analyze and judge the risk of the pathological changes;
the pathological analysis module is connected with the central control module and is used for carrying out pathological analysis on the detected data;
the cloud service module is connected with the central control module and used for processing the detection data by centralizing big data computing resources through a cloud server;
and the display module is connected with the central control module and is used for displaying the detection data information.
6. A lung cancer image diagnosis apparatus equipped with the big-data based lung cancer image diagnosis system according to claim 5.
CN201810567040.2A 2018-06-05 2018-06-05 Lung cancer image diagnosis system and method based on big data Active CN108876772B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810567040.2A CN108876772B (en) 2018-06-05 2018-06-05 Lung cancer image diagnosis system and method based on big data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810567040.2A CN108876772B (en) 2018-06-05 2018-06-05 Lung cancer image diagnosis system and method based on big data

Publications (2)

Publication Number Publication Date
CN108876772A CN108876772A (en) 2018-11-23
CN108876772B true CN108876772B (en) 2021-10-12

Family

ID=64336173

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810567040.2A Active CN108876772B (en) 2018-06-05 2018-06-05 Lung cancer image diagnosis system and method based on big data

Country Status (1)

Country Link
CN (1) CN108876772B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109724689B (en) * 2019-02-21 2021-11-30 哈尔滨工程大学 Device and method for measuring acoustic characteristics of single bubble in water
CN110246584A (en) * 2019-06-13 2019-09-17 吉林大学第一医院 A kind of prediction and evaluation method and system of lung cancer sample
CN110533658A (en) * 2019-09-02 2019-12-03 山东大学齐鲁医院 Intelligent pulmonary emphysema diagnostic message processing system and method, information data processing terminal
CN114266724A (en) * 2021-11-16 2022-04-01 中国航空工业集团公司雷华电子技术研究所 High-voltage line detection method based on radar infrared visible light image fusion
CN114821114B (en) * 2022-03-28 2024-04-30 南京业恒达智能系统有限公司 Groove cutting robot image processing method based on vision system

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106461664A (en) * 2013-12-30 2017-02-22 斯克利普斯研究所 Circulating tumor cell diagnostics for lung cancer
CN107767362A (en) * 2017-09-01 2018-03-06 苏州侠洛信息科技有限公司 A kind of early screening of lung cancer device based on deep learning

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101420970A (en) * 2003-11-20 2009-04-29 血管技术国际股份公司 Implantable sensors and implantable pumps and anti-scarring agents
EP1892671A3 (en) * 2006-08-23 2009-07-29 Medison Co., Ltd. System and method for determining the volume of an object by image processing
US9775844B2 (en) * 2014-03-19 2017-10-03 Infinity Pharmaceuticals, Inc. Heterocyclic compounds and uses thereof
CN103942803B (en) * 2014-05-05 2017-05-17 北京理工大学 SAR (Synthetic Aperture Radar) image based automatic water area detection method
CN107765012B (en) * 2016-08-16 2020-10-27 华明康生物科技（深圳）有限公司 Early non-small cell lung cancer screening method and kit
CN106372390B (en) * 2016-08-25 2019-04-02 汤一平 A kind of self-service healthy cloud service system of prevention lung cancer based on depth convolutional neural networks
CN107016665B (en) * 2017-02-16 2021-05-04 浙江大学 CT pulmonary nodule detection method based on deep convolutional neural network
CN107180426B (en) * 2017-06-06 2020-12-08 西北工业大学 Migratable multi-model integration-based computer-aided lung nodule classification device

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106461664A (en) * 2013-12-30 2017-02-22 斯克利普斯研究所 Circulating tumor cell diagnostics for lung cancer
CN107767362A (en) * 2017-09-01 2018-03-06 苏州侠洛信息科技有限公司 A kind of early screening of lung cancer device based on deep learning

Also Published As

Publication number Publication date
CN108876772A (en) 2018-11-23

Similar Documents

Publication Publication Date Title
CN108876772B (en) Lung cancer image diagnosis system and method based on big data
JP5281826B2 (en) Image processing apparatus, image processing program, and image processing method
CN108464840B (en) Automatic detection method and system for breast lumps
CN110772286B (en) System for discernment liver focal lesion based on ultrasonic contrast
CN110555868A (en) method for detecting small moving target under complex ground background
CN111462201B (en) Follow-up analysis system and method based on novel coronavirus pneumonia CT image
CN110263662B (en) Human body contour key point and key part identification method based on grading
Osma-Ruiz et al. Segmentation of the glottal space from laryngeal images using the watershed transform
CN110084830A (en) A kind of detection of video frequency motion target and tracking
US7203349B2 (en) Bronchial wall thickening recognition for reduced false-positives in pulmonary nodule detection
CN109543498A (en) A kind of method for detecting lane lines based on multitask network
Yao et al. Position-based anchor optimization for point supervised dense nuclei detection
CN110580697B (en) Video image processing method and system for measuring thickness of fetal nape transparency from ultrasonic video image
KR101690050B1 (en) Intelligent video security system
CN108364289B (en) IVOCT image vulnerable plaque automatic detection method
CN109215059A (en) Local data&#39;s correlating method of moving vehicle tracking in a kind of video of taking photo by plane
CN111401102A (en) Deep learning model training method and device, electronic equipment and storage medium
Jiang et al. Blood vessel tracking in retinal images
Gao et al. Intelligent crack damage detection system in shield tunnel using combination of retinanet and optimal adaptive selection
CN114419061A (en) Method and system for segmenting pulmonary artery and vein blood vessels
Yadav et al. Edge detection of images using Prewitt algorithm comparing with Sobel algorithm to improve accuracy
Fan et al. Aging aircraft rivet site inspection using magneto-optic imaging: Automation and real-time image processing
Meejaroen et al. Detection of fibrosis in liver biopsy images by using Bayesian classifier
Acharya et al. Segmentation of pap smear images to diagnose cervical cancer types and stages
Li et al. (Retracted) Automatic reading recognition of pointer barometer based on machine vision

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