CN114782352A - Medical image recognition system and method based on artificial intelligence - Google Patents
Medical image recognition system and method based on artificial intelligence Download PDFInfo
- Publication number
- CN114782352A CN114782352A CN202210396421.5A CN202210396421A CN114782352A CN 114782352 A CN114782352 A CN 114782352A CN 202210396421 A CN202210396421 A CN 202210396421A CN 114782352 A CN114782352 A CN 114782352A
- Authority
- CN
- China
- Prior art keywords
- image
- lesion
- coefficient
- module
- sample image
- 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.)
- Pending
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
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- 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
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- 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
Abstract
The invention belongs to the field of medical images, relates to an image identification technology, and is used for solving the problem of low identification precision caused by the fact that the existing medical image identification system adopts a unified standard for identification aiming at medical images obtained by scanning different parameters, in particular to a medical image identification system and an identification method based on artificial intelligence, wherein the medical image identification system comprises an image acquisition module, an image detection module, a normalization processing module and a treatment evaluation module; the image acquisition module, the image detection module, the normalization processing module and the treatment evaluation module are sequentially connected; according to the invention, the image detection module is used for carrying out resolution detection on the acquired medical image to obtain the display coefficient, and the scanning effect of the CT scanner is judged through the numerical value of the display coefficient, so that the image with unqualified resolution is removed from a batch of sample images, the identification process of the medical image with unqualified resolution is saved, and the rescanning is carried out on the image with unqualified resolution.
Description
Technical Field
The invention belongs to the field of medical images, relates to an image recognition technology, and particularly relates to a medical image recognition system and a medical image recognition method based on artificial intelligence.
Background
In medical imaging, artificial intelligence has proven its ability to improve the efficiency of image analysis by rapidly and accurately labeling specific abnormal structures for reference by radiologists. In 2011, researchers at LangoneHealth, university of new york, found that this type of automated analysis could be 62% to 97% faster than radiologists in finding and matching specific lung nodules. Research results show that the efficiency of image analysis by this artificial intelligence allows radiologists to spend more time focusing on content review that requires more interpretation or judgment, resulting in a saving of $ 30 billion per year. Recent studies have also explored the search for artificial intelligence in pharmacy, molecular structure and biological proteins, and these exciting studies are all proving the ability of AI and expanding the ambition of AI.
The CT scanning device needs to set a large number of parameters and protocols when scanning images, and for the scanning of different patients, because parameters and protocols need to be set in a targeted manner, the image quality of the scanned images has great difference, and therefore, the accuracy of the recognition result when automatically recognizing such medical images according to the unified standard cannot meet the requirement.
The invention patent with publication number CN108229584A discloses a method and a device for multi-modal medical image recognition based on deep learning, which performs lesion information acquisition and image processing of a lesion region of a patient through a multi-modal medical image acquisition module and a multi-modal medical image display module, performs lesion information analysis and detection through a multi-modal medical image detection module, and finally determines a final lesion result through automatic recognition of the multi-modal medical image recognition module; however, parameters of scanning devices of the multi-modal medical image acquisition module are different when the multi-modal medical image acquisition module acquires images for different patients, so that the display effect of medical images of the same batch on the multi-modal medical image display module is very different, and medical images which cannot be identified exist.
Disclosure of Invention
The invention aims to provide a medical image recognition system and a recognition method based on artificial intelligence, which are used for solving the problem of low recognition precision caused by the fact that the existing medical image recognition system adopts a unified standard to recognize medical images obtained by scanning different parameters;
the technical problems to be solved by the invention are as follows: how to provide a medical image identification system and method which can carry out unified identification on medical images obtained by scanning different parameters.
The purpose of the invention can be realized by the following technical scheme:
a medical image recognition method based on artificial intelligence comprises the following steps:
the method comprises the following steps: based on a medical image recognition system of a part to be detected of a patient, automatically acquiring and recognizing a pathological change region image of the patient by aligning an image acquisition module to the part to be detected of the patient;
step two: the image detection module carries out resolution detection on the lesion area of the patient collected and identified in the step one and rejects lesion images with unqualified resolution detection;
step three: the normalization processing module is used for carrying out normalization processing on the remaining lesion images in the step two to obtain an amplification factor, and the amplification factor is determined for the sample image according to the numerical value of the amplification factor;
step four: and the treatment evaluation module performs lesion detection and lesion analysis on the sample image amplified in the third step.
A medical image recognition system based on artificial intelligence comprises an image acquisition module, an image detection module, a normalization processing module and a treatment evaluation module; the image acquisition module, the image detection module, the normalization processing module and the treatment evaluation module are sequentially connected;
the image acquisition module is used for automatically acquiring and identifying a lesion area image required to be checked by a patient, and the lesion area image required to be checked by the patient is acquired by a CT scanner;
the image detection module is used for carrying out resolution detection on the lesion images acquired by the image acquisition module and rejecting the lesion images with unqualified resolution detection according to the resolution detection result;
the normalization processing module is used for carrying out normalization processing on the received qualified sample image;
and the treatment evaluation module is used for carrying out targeted lesion detection and lesion analysis on the lesion phenomenon of the lesion area amplified by the normalization processing module.
Further, the specific process of performing resolution detection on the lesion image includes: marking the received lesion image as a sample image i, wherein i is 1, 2, …, n, n is a positive integer, acquiring the spatial resolution and the density resolution of the sample image i, respectively marking the spatial resolution and the density resolution as KJi and MDi, and obtaining a display coefficient XSi of the sample image by performing numerical calculation on the spatial resolution KJi and the density resolution MDi;
comparing the display coefficient XSi with a display threshold XSmin: if the display coefficient XSi is smaller than or equal to the display threshold XSmin, judging that the corresponding sample image is unqualified; if the display coefficient XS is larger than a display threshold XSmin, judging that the corresponding sample image is qualified; and sending the lesion image qualified by the sample image to a normalization processing module.
Further, the specific process of the normalization process includes: marking sample data with the maximum display coefficient XSi value in a qualified sample image as a standard image, acquiring setting parameters of a CT scanner when the sample image is acquired, wherein the setting parameters of the CT scanner comprise Ai and Bi, the Ai is the distance from a focus to a scanning field center, the Bi is the distance from a bulb tube focus to a detector, the setting parameters of the CT scanner when the standard image is acquired are marked as standard parameters, and the standard parameters comprise Ab and Bb;
establishing a rectangular coordinate system by taking the distance from a focus to a scanning field center as an X axis and the distance from a bulb focus to a detector as a Y axis, marking a sample image according to set parameters in the rectangular coordinate system to obtain a sampling point i, wherein the coordinate of the sampling point i is (Ai, Bi), marking the sampling point corresponding to the standard image as a standard point, connecting the sampling point with the standard point to obtain a sampling line segment i, marking the length value of the sampling line segment i and the origin of the rectangular coordinate system as a reference value JZi of the sample image i, and obtaining an amplification coefficient FDi of the sample image by carrying out numerical calculation on the reference value JZi; and matching the sample image i with an amplification factor FDi, amplifying the sample image by an image processing technology, wherein the amplification factor is FDi, and sending the amplified sample image to an image detection module.
Further, the specific process of lesion detection and lesion analysis includes: marking the area value of the lesion area as MJ, carrying out image processing on a sample image through an image processing technology to obtain an average gray value HD of the lesion area of the sample image, and obtaining a lesion coefficient BX of the sample image through a formula BX of gamma 1 xMJ + gamma 2 xHD, wherein gamma 1 and gamma 2 are both proportional coefficients, and gamma 1 is more than gamma 2 and more than 1; acquiring a lesion coefficient of a patient in previous lesion detection for m times, performing variance calculation on the lesion coefficient of the patient in the previous lesion detection for m times to obtain a fluctuation coefficient BD, acquiring a fluctuation threshold BDmax and a lesion threshold BXmax, comparing the lesion coefficient BX and the fluctuation coefficient BD with the lesion threshold BXmax and the fluctuation threshold BDmax respectively, and judging whether the treatment effect of the patient is qualified or not according to the comparison result of the lesion coefficient BX and the fluctuation coefficient BD with the lesion threshold BXmax and the fluctuation threshold BDmax.
Further, the process of comparing the lesion coefficient BX and the fluctuation coefficient BD with the lesion threshold BXmax and the fluctuation threshold BDmax includes:
if the lesion coefficient BX is smaller than the lesion threshold BXmax and the fluctuation coefficient BD is smaller than the fluctuation threshold BDmax, judging that the treatment effect of the patient is qualified, and sending a qualified signal to a mobile phone terminal of a doctor by a treatment evaluation module;
if the lesion coefficient BX is smaller than the lesion threshold BXmax and the fluctuation coefficient BD is larger than or equal to the fluctuation threshold BDmax, judging that the treatment effect of the patient is qualified, and sending an observation signal to a mobile phone terminal of a doctor by a treatment evaluation module;
and if the lesion coefficient BX is greater than or equal to the lesion threshold BXmax, judging that the treatment effect of the patient is unqualified, and sending an unqualified signal to a mobile phone terminal of a doctor by the treatment evaluation module.
The invention has the following beneficial effects:
1. the image detection module is used for detecting the resolution of the acquired medical image to obtain a display coefficient, and the scanning effect of the CT scanner is judged according to the numerical value of the display coefficient, so that the images with unqualified resolution are removed from a batch of sample images, the identification procedure of the medical images with unqualified resolution is saved, and the images with unqualified resolution are rescanned;
2. the standard image with the best display effect can be obtained through numerical screening of the display coefficients by the normalization processing module, the CT scanner setting parameters of the standard image are compared with the CT scanner setting parameters of the sample image, so that the sample image is converted into the standard image through the difference of the setting parameters, all the converted sample images can obtain similar display coefficients, the sample images with similar display coefficients are subjected to lesion area identification, and the accuracy of the identification result of the sample image is improved;
3. can carry out the analysis through treatment evaluation module to the pathological change region after the enlargeing and obtain the pathological change coefficient, combine the pathological change coefficient of patient's current phase when carrying out the pathological change and detect and carry out the analysis and obtain the fluctuation coefficient, judge the treatment effect of patient from the numerical value according to pathological change coefficient and fluctuation coefficient, can design different counter measures according to the treatment effect of difference, the pertinence makes treatment scheme to the patient.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic block diagram of a system according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a method according to a second embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example one
As shown in fig. 1, a medical image recognition system based on artificial intelligence comprises an image acquisition module, an image detection module, a normalization processing module and a treatment evaluation module; the image acquisition module, the image detection module, the normalization processing module and the treatment evaluation module are sequentially connected.
The image acquisition module is used for automatically acquiring and identifying a lesion area image required to be checked by a patient, the lesion area image required to be checked by the patient is acquired by a CT scanner, the CT scanner is a fully functional disease detection instrument and is short for the technical abbreviation of computer tomography, and the working process of the CT scanner comprises the following steps: according to the different absorption and transmittance of different tissues of human body to X-ray, the instrument with very high sensitivity is used to measure human body, then the data obtained by measurement is inputted into the electronic computer, and after the electronic computer processes the data, the cross section or stereo image of the examined part of human body can be taken, and the tiny lesion of any part in human body can be found.
The image detection module is used for carrying out resolution detection on the lesion images acquired by the image acquisition module and rejecting the lesion images with unqualified resolution detection according to resolution detection results; the specific process of performing resolution detection on the lesion image comprises the following steps: marking the received lesion image as a sample image i, wherein i is 1, 2, …, n, n is a positive integer, acquiring the spatial resolution and the density resolution of the sample image i and marking the spatial resolution and the density resolution as KJi and MDi respectively, wherein the spatial resolution refers to the size or the size of a minimum unit which can be distinguished in detail on a remote sensing image and is an index for representing the details of an image-resolved ground target. Generally expressed by pixel size, image resolution or field angle; the density resolution represents the minimum density difference which can be displayed in the image, and has the capability of distinguishing different tissue densities, and the density resolution of the CT to the tissue is higher than that of the X-ray photography. The density resolution of the CT is 0.5%, that is, the density difference between two tissues is equal to or greater than 0.5%, a display coefficient XSi of the sample image is obtained by a formula XSi ═ α 1 × KJi + α 2 × MDi, the display coefficient is a value reflecting the resolution of the sample image, the higher the value of the display coefficient is, the higher the resolution of the corresponding sample image is, wherein α 1 and α 2 are both proportional coefficients, and α 1 > α 2 > 1; comparing the display coefficient XSi with a display threshold XSmin: if the display coefficient XSi is smaller than or equal to the display threshold XSmin, judging that the corresponding sample image is unqualified; if the display coefficient XS is larger than a display threshold XSmin, judging that the corresponding sample image is qualified; sending the lesion image qualified by the sample image to a normalization processing module; the resolution detection is carried out on the acquired medical images to obtain the display coefficient, the scanning effect of the CT scanner is judged through the numerical value of the display coefficient, so that the images with unqualified resolution are removed from a batch of sample images, the identification procedure of the medical images with unqualified resolution is saved, and the images with unqualified resolution are rescanned.
The normalization processing module is used for carrying out normalization processing on the received qualified sample image, and the specific process of the normalization processing comprises the following steps: marking the sample data with the maximum numerical value of a display coefficient XSi in a qualified sample image as a standard image, acquiring the setting parameters of a CT scanner when the sample image is acquired, wherein the setting parameters of the CT scanner comprise Ai and Bi, wherein Ai is the distance from a focus to the center of a scanning field, Bi is the distance from a focus of a bulb to a detector, the setting parameters of the CT scanner when the standard image is acquired are marked as the standard parameters, the standard parameters comprise Ab and Bb, a rectangular coordinate system is established by taking the distance from the focus to the center of the scanning field as an X axis and the distance from the focus of the bulb to the detector as a Y axis, marking the sample image according to the setting parameters in the rectangular coordinate system to obtain a sampling point i, the coordinates of the sampling point i are (Ai, Bi), marking the sampling point corresponding to the standard image as the standard point, connecting the sampling point to the standard point to obtain a sampling line segment i, and marking the length value of the origin of the sampling line segment i and the rectangular coordinate system as the reference value JZi of the sample image i, by the formula
Obtaining an amplification factor FDi of the sample image, wherein beta is a proportionality factor, beta is more than 0, e is a natural constant, and the value of e is 2.7182; matching the sample image i with an amplification factor FDi, amplifying the sample image by an image processing technology, wherein the amplification factor is FDi, sending the amplified sample image to an image detection module,the standard image with the best display effect is obtained through numerical screening of the display coefficients, the CT scanner setting parameters of the standard image are compared with the CT scanner setting parameters of the sample image, so that the sample image is converted into the standard image through the difference of the setting parameters, all the converted sample images can obtain similar display coefficients, the sample images with the similar display coefficients are subjected to lesion area identification, and the accuracy of the identification result of the sample image is improved.
The treatment evaluation module is used for carrying out targeted lesion detection and lesion analysis on the lesion phenomenon of the lesion area amplified by the normalization processing module; the specific process of lesion detection and lesion analysis comprises: marking the area value of the lesion area as MJ, carrying out image processing on a sample image through an image processing technology to obtain an average gray value HD of the lesion area of the sample image, and obtaining a lesion coefficient BX of the sample image through a formula BX (gamma 1 × MJ + gamma 2 × HD), wherein gamma 1 and gamma 2 are both proportionality coefficients, and gamma 1 is more than gamma 2 and is more than 1; acquiring a lesion coefficient of a patient during lesion detection m times, performing variance calculation on the lesion coefficient of the patient during lesion detection m times to obtain a fluctuation coefficient BD, acquiring a fluctuation threshold BDmax and a lesion threshold BXmax, and comparing the lesion coefficient BX and the fluctuation coefficient BD with the lesion threshold BXmax and the fluctuation threshold BDmax respectively: if the lesion coefficient BX is smaller than the lesion threshold BXmax and the fluctuation coefficient BD is smaller than the fluctuation threshold BDmax, judging that the treatment effect of the patient is qualified, and sending a qualified signal to a mobile phone terminal of a doctor by a treatment evaluation module; if the lesion coefficient BX is smaller than the lesion threshold BXmax and the fluctuation coefficient BD is greater than or equal to the fluctuation threshold BDmax, judging that the treatment effect of the patient is qualified, and sending an observation signal to a mobile phone terminal of a doctor by a treatment evaluation module; if the lesion coefficient BX is larger than or equal to the lesion threshold BXmax, the treatment effect of the patient is judged to be unqualified, the treatment evaluation module sends an unqualified signal to a mobile phone terminal of a doctor, the amplified lesion area is analyzed to obtain a lesion coefficient, the lesion coefficient is analyzed in combination with the current period of the patient to detect the lesion to obtain a fluctuation coefficient, the treatment effect of the patient is judged according to the values of the lesion coefficient and the fluctuation coefficient, different countermeasures can be designed according to different treatment effects, and a treatment scheme is established for the patient in a targeted manner.
Example two
As shown in fig. 2, a medical image recognition method based on artificial intelligence includes the following steps:
the method comprises the following steps: based on a medical image recognition system of a part to be detected of a patient, automatically acquiring and recognizing a lesion area image of the patient by aligning an image acquisition module to the part to be detected of the patient;
step two: the image detection module carries out resolution detection on the lesion area of the patient collected and identified in the step one to obtain a display coefficient of a sample image, judges a resolution detection result according to a comparison result of the display coefficient and a display threshold value, and eliminates lesion images with unqualified resolution detection;
step three: the normalization processing module is used for carrying out normalization processing on the remaining lesion images in the step two to obtain an amplification factor, determining the amplification factor for the sample images according to the numerical value of the amplification factor, and converting the sample images into standard images according to the difference of the set parameters, so that all the converted sample images can obtain similar display coefficients;
step four: and the treatment evaluation module performs lesion detection and lesion analysis on the amplified sample image in the third step, judges the treatment effect of the patient according to the numerical values of the lesion coefficient and the fluctuation coefficient, can design different countermeasures according to different treatment effects, and develops a treatment scheme for the patient in a targeted manner.
The normalization processing module can obtain a standard image with the best display effect through numerical screening of the display coefficients, compare the CT scanner setting parameters of the standard image with the CT scanner setting parameters of the sample image, convert the sample image into the standard image through the difference of the setting parameters, enable all sample images obtained after conversion to obtain similar display coefficients, and identify the lesion area of the sample images with similar display coefficients, so that the accuracy of the identification result of the sample image is improved.
A medical image recognition system based on artificial intelligence, when working, based on the medical image recognition system of the part to be detected of the patient, automatically collects and recognizes the image of the lesion area of the patient by aligning the image collection module to the part to be detected of the patient; the image detection module carries out resolution detection on the lesion area of the patient collected and identified in the step one to obtain a display coefficient of a sample image, judges a resolution detection result according to a comparison result of the display coefficient and a display threshold value, and eliminates lesion images with unqualified resolution detection; and the normalization processing module is used for carrying out normalization processing on the remaining lesion images in the step two to obtain an amplification factor, determining the amplification factor for the sample image according to the numerical value of the amplification factor, and converting the sample image into a standard image according to the difference of the set parameters, so that all the converted sample images can obtain similar display factors.
The foregoing is merely illustrative and explanatory of the present invention and various modifications, additions or substitutions may be made to the specific embodiments described by those skilled in the art without departing from the scope of the invention as defined in the accompanying claims.
The formulas are obtained by acquiring a large amount of data and performing software simulation, and the coefficients in the formulas are set by the technicians in the field according to actual conditions; such as: the formula XSi ═ α 1 × KJi + α 2 × MDi; collecting multiple groups of sample data and setting corresponding display coefficients for each group of sample data by a person skilled in the art; substituting the set display coefficient and the acquired sample data into formulas, forming a linear equation set of two variables by any two formulas, screening the calculated coefficients and taking the mean value to obtain values of alpha 1 and alpha 2 which are respectively 2.95 and 1.87;
the size of the coefficient is a specific numerical value obtained by quantizing each parameter, so that the subsequent comparison is convenient, and the size of the coefficient depends on the number of sample data and the corresponding display coefficient is preliminarily set for each group of sample data by a person skilled in the art; it is sufficient that the image parameters are not in proportion to the quantized values, for example, the display coefficients are in proportion to the spatial resolution values.
In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.
Claims (6)
1. A medical image recognition method based on artificial intelligence is characterized by comprising the following steps:
the method comprises the following steps: based on a medical image recognition system of a part to be detected of a patient, automatically acquiring and recognizing a lesion area image of the patient by aligning an image acquisition module to the part to be detected of the patient;
step two: the image detection module carries out resolution detection on the lesion area of the patient collected and identified in the step one and rejects lesion images with unqualified resolution detection;
step three: the normalization processing module is used for carrying out normalization processing on the remaining lesion images in the step two to obtain an amplification factor, and the amplification factor is determined for the sample image according to the numerical value of the amplification factor;
step four: and the treatment evaluation module performs lesion detection and lesion analysis on the sample image amplified in the third step.
2. A medical image recognition system based on artificial intelligence is characterized by comprising an image acquisition module, an image detection module, a normalization processing module and a treatment evaluation module; the image acquisition module, the image detection module, the normalization processing module and the treatment evaluation module are sequentially connected;
the image acquisition module is used for automatically acquiring and identifying a lesion area image required to be checked by a patient, and the lesion area image required to be checked by the patient is acquired by a CT scanner;
the image detection module is used for carrying out resolution detection on the lesion images acquired by the image acquisition module and rejecting the lesion images with unqualified resolution detection according to resolution detection results;
the normalization processing module is used for carrying out normalization processing on the received qualified sample image;
and the treatment evaluation module is used for carrying out targeted lesion detection and lesion analysis on the lesion phenomenon of the lesion area amplified by the normalization processing module.
3. The artificial intelligence based medical image recognition system of claim 2, wherein the specific process of performing resolution detection on the lesion image comprises: marking the received lesion image as a sample image i, wherein i is 1, 2, …, n and n are positive integers, acquiring the spatial resolution and the density resolution of the sample image i, respectively marking the spatial resolution and the density resolution as KJi and MDi, and obtaining a display coefficient XSi of the sample image by performing numerical calculation on the spatial resolution KJi and the density resolution MDi;
comparing the display coefficient XSi with a display threshold XSmin: if the display coefficient XSi is smaller than or equal to the display threshold value XSmin, determining that the corresponding sample image is unqualified; if the display coefficient XS is larger than a display threshold XSmin, judging that the corresponding sample image is qualified; and sending the lesion image qualified by the sample image to a normalization processing module.
4. The artificial intelligence based medical image recognition system of claim 3, wherein the normalization process comprises: marking sample data with the maximum value of a display coefficient XSi in a qualified sample image as a standard image, acquiring setting parameters of a CT scanner when the sample image is acquired, wherein the setting parameters of the CT scanner comprise Ai and Bi, the Ai is the distance from a focus to a scanning field center, the Bi is the distance from a bulb tube focus to a detector, the setting parameters of the CT scanner when the standard image is acquired are marked as standard parameters, and the standard parameters comprise Ab and Bb;
establishing a rectangular coordinate system by taking the distance from a focus to a scanning field center as an X axis and the distance from a bulb focus to a detector as a Y axis, marking a sample image according to set parameters in the rectangular coordinate system to obtain a sampling point i, wherein the coordinate of the sampling point i is (Ai, Bi), marking the sampling point corresponding to the standard image as a standard point, connecting the sampling point with the standard point to obtain a sampling line segment i, marking the length value of the sampling line segment i and the origin of the rectangular coordinate system as a reference value JZi of the sample image i, and obtaining an amplification coefficient FDi of the sample image by carrying out numerical calculation on the reference value JZi; and matching the sample image i with an amplification factor FDi, amplifying the sample image by an image processing technology, wherein the amplification factor is FDi, and sending the amplified sample image to an image detection module.
5. The system of claim 4, wherein the lesion detection and lesion analysis comprises: marking the area value of the lesion area as MJ, carrying out image processing on a sample image through an image processing technology to obtain an average gray value HD of the lesion area of the sample image, and obtaining a lesion coefficient BX of the sample image through a formula BX of gamma 1 xMJ + gamma 2 xHD, wherein gamma 1 and gamma 2 are both proportional coefficients, and gamma 1 is more than gamma 2 and more than 1; obtaining a lesion coefficient of a patient during lesion detection m times, performing variance calculation on the lesion coefficient of the patient during lesion detection m times to obtain a fluctuation coefficient BD, obtaining a fluctuation threshold BDmax and a lesion threshold BXmax, comparing the lesion coefficient BX and the fluctuation coefficient BD with the lesion threshold BXmax and the fluctuation threshold BDmax respectively, and judging whether the treatment effect of the patient is qualified or not according to the comparison result of the lesion coefficient BX and the fluctuation coefficient BD with the lesion threshold BXmax and the fluctuation threshold BDmax.
6. The artificial intelligence based medical image identification system according to claim 5, wherein the comparison of the lesion coefficient BX and the fluctuation coefficient BD with the lesion threshold BXmax and the fluctuation threshold BDmax comprises:
if the lesion coefficient BX is smaller than the lesion threshold BXmax and the fluctuation coefficient BD is smaller than the fluctuation threshold BDmax, judging that the treatment effect of the patient is qualified, and sending a qualified signal to a mobile phone terminal of a doctor by a treatment evaluation module;
if the lesion coefficient BX is smaller than the lesion threshold BXmax and the fluctuation coefficient BD is larger than or equal to the fluctuation threshold BDmax, judging that the treatment effect of the patient is qualified, and sending an observation signal to a mobile phone terminal of a doctor by a treatment evaluation module;
and if the lesion coefficient BX is greater than or equal to the lesion threshold BXmax, judging that the treatment effect of the patient is unqualified, and sending an unqualified signal to a mobile phone terminal of the doctor by the treatment evaluation module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210396421.5A CN114782352A (en) | 2022-04-15 | 2022-04-15 | Medical image recognition system and method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210396421.5A CN114782352A (en) | 2022-04-15 | 2022-04-15 | Medical image recognition system and method based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114782352A true CN114782352A (en) | 2022-07-22 |
Family
ID=82429157
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210396421.5A Pending CN114782352A (en) | 2022-04-15 | 2022-04-15 | Medical image recognition system and method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114782352A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116206216A (en) * | 2023-05-06 | 2023-06-02 | 山东省国土空间数据和遥感技术研究院(山东省海域动态监视监测中心) | Vector geographic information acquisition method and system based on remote sensing image |
-
2022
- 2022-04-15 CN CN202210396421.5A patent/CN114782352A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116206216A (en) * | 2023-05-06 | 2023-06-02 | 山东省国土空间数据和遥感技术研究院(山东省海域动态监视监测中心) | Vector geographic information acquisition method and system based on remote sensing image |
| CN116206216B (en) * | 2023-05-06 | 2023-08-01 | 山东省国土空间数据和遥感技术研究院(山东省海域动态监视监测中心) | Vector geographic information acquisition method and system based on remote sensing image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| DK2973397T3 (en) | Tissue-object-based machine learning system for automated assessment of digital whole-slide glass | |
| Cruz et al. | Determination of blood components (WBCs, RBCs, and Platelets) count in microscopic images using image processing and analysis | |
| EP0628186B1 (en) | Method for identifying normal biomedical specimens | |
| US9177379B1 (en) | Method and system for identifying anomalies in medical images | |
| CN103249358A (en) | Medical image processing device | |
| CN115760858B (en) | Deep learning-based kidney pathological section cell identification method and system | |
| GB2093586A (en) | Apparatus for classification of cells | |
| CN114782352A (en) | Medical image recognition system and method based on artificial intelligence | |
| CN116993736B (en) | Intelligent identification system for osteoporosis area | |
| CN113706518A (en) | Femoral head necrosis stage analysis method and system based on artificial intelligence | |
| CN113936005A (en) | DNA index calculation method and device, computer equipment and storage medium | |
| CN117152152A (en) | Production management system and method for detection kit | |
| CN111398138A (en) | Optical detection system and method of dry type blood cell analysis device | |
| CN115019972A (en) | Medical image data sharing system based on big data analysis | |
| CN113130050B (en) | Medical information display method and display system | |
| CN113469942B (en) | CT image lesion detection method | |
| CN111402242B (en) | Medical image refinement processing system and method | |
| CN114494299A (en) | Temperature detection device | |
| CN113509191A (en) | Method, device and equipment for analyzing mammary gland molybdenum target X-ray image | |
| CN116721240B (en) | AI system and analysis method for skin scar measurement and subcutaneous hard mass perception test | |
| CN106204609B (en) | The processing of Laser scanning confocal microscope lung image and analysis system and method | |
| CN110349662A (en) | The outliers across image collection that result is accidentally surveyed for filtering pulmonary masses find method and system | |
| CN113313685B (en) | Renal tubular atrophy region identification method and system based on deep learning | |
| CN113723441B (en) | Intelligent analysis system and method for lip gland pathology | |
| CN117351221A (en) | Method for extracting key points of paper simulation verification electrocardiogram |
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 |