CN113052795A - X-ray chest radiography image quality determination method and device - Google Patents
X-ray chest radiography image quality determination method and device Download PDFInfo
- Publication number
- CN113052795A CN113052795A CN202110198562.1A CN202110198562A CN113052795A CN 113052795 A CN113052795 A CN 113052795A CN 202110198562 A CN202110198562 A CN 202110198562A CN 113052795 A CN113052795 A CN 113052795A
- Authority
- CN
- China
- Prior art keywords
- result
- segmentation
- image
- model
- ray chest
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000002601 radiography Methods 0.000 title claims abstract description 50
- 238000000034 method Methods 0.000 title claims abstract description 29
- 230000011218 segmentation Effects 0.000 claims abstract description 192
- 210000004072 lung Anatomy 0.000 claims abstract description 120
- 238000000926 separation method Methods 0.000 claims abstract description 75
- 210000001991 scapula Anatomy 0.000 claims abstract description 72
- 238000001514 detection method Methods 0.000 claims abstract description 69
- 238000013136 deep learning model Methods 0.000 claims abstract description 33
- 238000012545 processing Methods 0.000 claims description 39
- 238000010606 normalization Methods 0.000 claims description 31
- 238000012163 sequencing technique Methods 0.000 claims description 7
- 238000011156 evaluation Methods 0.000 abstract description 7
- 210000000038 chest Anatomy 0.000 description 115
- 230000002159 abnormal effect Effects 0.000 description 9
- 238000012360 testing method Methods 0.000 description 9
- 238000013135 deep learning Methods 0.000 description 8
- 238000002372 labelling Methods 0.000 description 7
- 238000013139 quantization Methods 0.000 description 7
- 238000003745 diagnosis Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 210000003205 muscle Anatomy 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 210000000709 aorta Anatomy 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 2
- 210000004556 brain Anatomy 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000005034 decoration Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000011976 chest X-ray Methods 0.000 description 1
- 210000003109 clavicle Anatomy 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000000338 in vitro Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 230000008450 motivation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 210000001898 sternoclavicular joint Anatomy 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 210000000115 thoracic cavity Anatomy 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- 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
- G06T7/13—Edge detection
-
- 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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/30168—Image quality inspection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention discloses a method and a device for determining the quality of an X-ray chest radiography image, wherein the method comprises the following steps: acquiring an X-ray chest radiography image; inputting the images into a deep learning model for image separation and identification to obtain at least one of a lung lobe segmentation result, a spine segmentation result, a scapula segmentation result and a foreign matter detection result; the deep learning model includes at least one of the following: a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model and a foreign matter detection model; judging whether the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result meets corresponding preset conditions to obtain a judgment result; the image quality of the X-ray chest radiograph is determined based on the judgment result. The invention carries out full-automatic evaluation on the image quality of the X-ray chest radiography, and the image quality determination speed is high; and the image quality of the chest radiography can be controlled by a technician, so that the accuracy of radiography reading is indirectly improved.
Description
Technical Field
The invention relates to the field of X-ray chest radiography images, in particular to a method and a device for determining the quality of an X-ray chest radiography image.
Background
The X-ray chest radiography is the most widely applied medical image examination means at present, and the use scenes comprise emergency treatment, intensive care, general outpatient service, physical examination and the like. The quality of one chest film directly influences the diagnosis result, and if the quality of the chest film is not qualified, missed diagnosis and misdiagnosis are easy to occur.
The X-ray chest radiography normal position standard film comprises the following requirements:
1. the 4 th vertebral body below the thoracic vertebra is clearly visible without bilateral shadows;
2. the texture from the lung portal to the outside of the lung field is clearly displayed, and the scapula is projected outside the lung field;
3. the organ images can be continuously tracked from the neck to the organ bifurcation;
4. the edges of diaphragm muscles on two sides are sharp, and the lung tip part is clearly displayed;
5. the clavicle is overlapped with the 4 th rib, and the sternoclavicular joint is symmetrical left and right;
6. the heart, the aorta and the aorta are clearly displayed;
7. it does not allow any metal in vitro and other foreign bodies not penetrated by X-rays to appear.
Generally, hospitals have great requirements on radiography, and currently, the evaluation of image quality is mainly completed by a radiology technician, so that the manual evaluation efficiency is low, and some unqualified radiographs are easy to enter clinical radiography diagnosis, thereby affecting the safety and accuracy of the diagnosis.
Therefore, it is necessary to provide an intelligent X-ray chest radiography image quality determination method and apparatus.
Disclosure of Invention
The invention aims to provide an X-ray chest film image quality determining method aiming at the defects of the prior art.
The invention is realized by the following technical scheme:
in one aspect, the present invention provides a method for determining quality of an X-ray chest radiography image, the method comprising:
acquiring an X-ray chest radiography image;
inputting the image into a deep learning model to obtain at least one of a lung lobe segmentation result, a spine segmentation result, a scapula segmentation result and a foreign matter detection result; the deep learning model includes at least one of: a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model and a foreign matter detection model;
judging whether the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result meets corresponding preset conditions to obtain a judgment result; wherein, the preset conditions corresponding to the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result and the foreign matter detection result respectively include: the left and right edge threshold values of the lung lobe segmentation result are smaller than the preset left and right edge threshold values and/or the upper and lower edge threshold values are smaller than the preset upper and lower edge threshold values; the absolute value of the sum of the relative distances of each pixel in the spine segmentation result and the center of the image in the horizontal direction is larger than a preset offset; the ratio of the area of the scapula segmentation result to the area of the original image is larger than the preset area ratio; the confidence score of the bounding box in the foreign matter detection result is greater than a preset numerical value;
if so, the judgment result is an unqualified result;
if not, the judgment result is a qualified result;
and determining the image quality of the X-ray chest radiography based on the judgment result.
Further, the acquiring the X-ray chest radiography image comprises:
sorting the X-ray chest radiography image pixels according to the gray values;
taking the gray value of the pixels which are positioned at the first set value after sequencing as the minimum value, taking the gray value of the pixels which are positioned at the second set value after sequencing as the maximum value, and carrying out truncation normalization processing on the image according to the gray value;
and performing standard normalization processing on the normalized image.
Further, the determining the image quality of the X-ray chest radiograph based on the judgment result comprises:
when the lung lobe segmentation model, the spine segmentation model, the scapula segmentation model and the foreign matter detection model all obtain qualified results, judging the X-ray chest film to be a qualified film; or
When the lung lobe segmentation model obtains a qualified result and at least one unqualified result is obtained in the spine segmentation model and the scapula segmentation model, judging the X-ray chest film as a secondary film; or
When the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that foreign matters exist in the X-ray chest radiograph, judging whether the foreign matters are in a lung field range according to the foreign matter detection model and the lung lobe segmentation model, and if so, judging that the X-ray chest radiograph is an unqualified radiograph; if not, judging the X-ray chest film as a secondary film; or when the lung lobe segmentation model obtains an unqualified result, judging the X-ray chest film as an unqualified film.
In another aspect, the present invention further provides an apparatus for determining the quality of an X-ray chest image, the apparatus comprising:
the acquisition module is used for acquiring an X-ray chest radiography image;
a separation result acquisition unit, configured to input the image into a deep learning model to obtain at least one of a lung lobe segmentation result, a spine segmentation result, a scapula segmentation result, and a foreign object detection result; the deep learning model includes at least one of: a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model and a foreign matter detection model;
the separation result judging unit is used for judging whether the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result meets corresponding preset conditions to obtain a judgment result; wherein, the preset conditions corresponding to the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result and the foreign matter detection result respectively include: the left and right edge threshold values of the lung lobe segmentation result are smaller than the preset left and right edge threshold values and/or the upper and lower edge threshold values are smaller than the preset upper and lower edge threshold values; the absolute value of the sum of the relative distances of each pixel in the spine segmentation result and the center of the image in the horizontal direction is larger than a preset offset; the ratio of the area of the scapula segmentation result to the area of the original image is larger than the preset area ratio; the confidence score of the bounding box in the foreign matter detection result is greater than a preset numerical value;
the unqualified result judging unit is used for determining that the judging result is an unqualified result when the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result meets corresponding preset conditions;
the qualified result judging unit is used for determining that the judging result is a qualified result when the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result does not meet corresponding preset conditions;
and the image quality determining module is used for determining the image quality of the X-ray chest radiography based on the judgment result.
Further, the obtaining module comprises:
the sorting unit is used for sorting the X-ray chest film image pixels according to the gray values;
the normalization processing unit is used for taking the gray value of the pixels which are positioned at the first set value after being sorted as a first numerical value, taking the gray value of the pixels which are positioned at the second set value after being sorted as a second numerical value, and performing truncated normalization processing on the image according to the first numerical value and the second numerical value;
and the standard normalization processing unit is used for performing standard normalization processing on the normalized image.
Further, the image quality determination module includes:
the qualified plate judging unit is used for judging the X-ray chest plate to be a qualified plate when the lung lobe segmentation model, the spine segmentation model, the scapula segmentation model and the foreign matter detection model all obtain qualified results;
the second-level film judging unit is used for judging the X-ray chest film as a second-level film when the lung lobe segmentation model obtains a qualified result and at least one unqualified result is obtained in the spine segmentation model and the scapula segmentation model; the X-ray chest film processing module is also used for judging whether the foreign matters exist in the lung field range according to the foreign matter detection model and the lung lobe segmentation model when the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that the foreign matters exist in the X-ray chest film, and if not, judging that the X-ray chest film is a secondary film;
the unqualified slice judging unit is used for judging whether the foreign matter is in the lung field range according to the foreign matter detection model and the lung lobe segmentation model when the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that the foreign matter exists in the X-ray chest slice, and if so, judging that the X-ray chest slice is an unqualified slice; and when the lung lobe segmentation model obtains an unqualified result, judging the X-ray chest radiograph as an unqualified radiograph.
In a third aspect, the present invention also provides a computer device comprising a processor and a memory, wherein at least one instruction, at least one program, a set of codes, or a set of instructions is stored in the memory, and the at least one instruction, the at least one program, the set of codes, or the set of instructions is loaded and executed by the processor to implement the X-ray chest image quality determination method as described above.
The technical scheme provided by the embodiment of the invention has the following beneficial effects:
(1) according to the invention, the evaluation of the chest radiography is divided into several sub-problems of whether foreign matters appear in the lung field or not, whether scapulae appear in the lung field or not, whether the center of the image is aligned or not, whether diaphragm muscles on two sides exceed the image range or not and the like, the image quality of the chest radiography is rapidly graded based on the deep learning technology, and not only can the technician be reminded whether the image quality is qualified or not, but also the technician can be reminded of which unqualified chest radiography belongs to.
(2) The invention can automatically evaluate the quality of X-ray chest film image, and the image quality can be determined quickly.
(3) The rapid evaluation method for X-ray chest radiography image quality grading can help technicians control the image quality of chest radiography and indirectly improve the accuracy of radiography reading.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings needed for the embodiment or the prior art description will be briefly introduced below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive labor.
FIG. 1 is a flow chart of an X-ray chest image quality determination method of the present invention;
FIG. 2 is a flow chart of the present invention for normalizing an X-ray chest radiograph;
FIG. 3 is a flowchart of inputting the normalized image into a lung lobe segmentation model for image separation and recognition according to the present invention;
FIG. 4 is a flowchart of inputting the normalized image into the spine segmentation model for image separation and recognition according to the present invention;
FIG. 5 is a flowchart of the present invention for inputting normalized images into a scapula segmentation model for image separation and recognition;
FIG. 6 is a flowchart of inputting the normalized image into a foreign object detection model for image separation and recognition according to the present invention;
FIG. 7 is a schematic diagram showing a configuration of an apparatus for determining the quality of an X-ray chest image according to the present invention;
FIG. 8 is a schematic diagram of a normalization module according to the present invention;
FIG. 9 is a schematic view of a first structure of a separation result obtaining module according to the present invention;
FIG. 10 is a schematic view of a second structure of the separation result obtaining module according to the present invention;
FIG. 11 is a schematic view of a third structure of the separation result obtaining module according to the present invention;
FIG. 12 is a diagram showing a fourth structure of the separation result obtaining module according to the present invention;
FIG. 13 is a schematic diagram of an embodiment of an image quality determination module according to the present invention;
FIG. 14 is an X-ray chest image (a) and an annotated lung lobe image (b) of the present invention;
FIG. 15 is an X-ray chest image (a) and an annotated spine image (b) of the present invention;
FIG. 16 is an X-ray chest radiograph image (a) and an annotated scapula image (b) in the present invention;
FIG. 17 is an X-ray chest radiograph image (a) and a foreign body image (b) after labeling in the present invention;
FIG. 18 shows an X-ray chest radiograph (a) and an output image (b) of the lung lobe segmentation model before the input of the model in the present invention;
FIG. 19 is an X-ray chest radiograph image (a) and an output image (b) of the spine segmentation model before the model is input in the present invention;
FIG. 20 shows an X-ray chest radiograph (a) and an output image (b) of the scapula segmentation model before the input of the model in the present invention;
fig. 21 shows an X-ray chest radiograph image (a) before a foreign object detection model is input and an output image (b) of the model in the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the present application, and it is obvious 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.
In the several embodiments provided in the present invention, the described device embodiments are only illustrative, for example, the division of the modules is only one logical function division, and there may be other division manners in actual implementation, for example, a plurality of modules or components may be combined or integrated into another device, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of modules or units through some interfaces, and may be in an electrical or other form.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical modules, may be located in one place, or may be distributed on a plurality of network modules. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
Example 1
As shown in fig. 1, the present embodiment provides a flow chart of a method for determining quality of an X-ray chest image, and the present specification provides the operation steps of the method as described in the embodiment or the flow chart, but may include more or less operation steps based on conventional or non-creative labor. The order of steps recited in the embodiments is merely one manner of performing the steps in a multitude of orders and does not represent the only order of execution. Specifically, as shown in fig. 1, the method includes:
s101, carrying out normalization processing on an X-ray chest radiography image;
s102, inputting the normalized image into a depth learning model for image separation and identification to obtain at least one separation result; the deep learning model includes at least one of: a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model and a foreign matter detection model;
the deep learning model is trained and completed based on a uniform deep learning platform, and the quantity and quality of labeled data influence the accuracy of the model. The data labeling method in the embodiment can be manual labeling or labeling by an intelligent machine.
As shown in fig. 14, a is an X-ray chest radiograph image, and b is an annotated lung lobe image;
FIG. 15 shows a chest X-ray image and a labeled spine image;
FIG. 16 shows an X-ray chest image (a) and an annotated scapula image (b);
as shown in fig. 17, a is an X-ray chest radiograph image, and b is a marked foreign body image.
Deep learning network structures include, but are not limited to, split networks (V-Net, LinkNet, FC-DenseNet, etc.) and detection networks (Faster-RCNN, RetinaNet, YoLO V3, etc.). The lung lobe segmentation model adopts FC-DenseNet, the spine segmentation model and the scapula segmentation model adopt LinkNet and FC-DenseNet, and the foreign matter detection model adopts RetinaNet.
Deep learning is a method based on characterization learning of data in machine learning. An observation (e.g., an image) may be represented using a number of ways, such as a vector of intensity values for each pixel, or more abstractly as a series of edges, a specially shaped region, etc. And tasks are easier to learn from the examples using some specific representation methods. The benefit of deep learning is to replace the manual feature acquisition with unsupervised or semi-supervised feature learning and hierarchical feature extraction efficient algorithms.
Deep learning is a new field in machine learning research, and the motivation is to establish and simulate a neural network for human brain to analyze and learn, which simulates the mechanism of human brain to interpret data.
S103, determining the image quality of the X-ray chest radiograph based on the separation result.
As shown in fig. 2, the normalizing the X-ray chest radiography image includes:
s1011, sequencing the pixels of the X-ray chest radiography image in an ascending order according to the gray values;
s1012, taking the gray value of the pixel which is positioned at the first set value after sequencing as a first numerical value, namely a minimum value, taking the gray value of the pixel which is positioned at the second set value after sequencing as a second numerical value, namely a maximum value, and performing maximum and minimum normalization processing with truncation on the image according to the gray value; wherein the first set value is greater than or equal to 0% and less than or equal to 10%, the second set value is greater than or equal to 90% and less than or equal to 100%, in a preferred embodiment, the gray value of the pixel located at the 5 th% after sorting is taken as the minimum value, and the gray value of the pixel located at the 95 th% after sorting is taken as the maximum value;
it should be noted that, if the pixels of the X-ray chest radiography image are sorted in descending order according to the gray values, the second set value is greater than or equal to 0% and less than or equal to 10%, and the first set value is greater than or equal to 90% and less than or equal to 100%.
S1013, standard normalization processing is carried out on the image after normalization processing, namely the image data after normalization processing is subtracted from the average value on the training data set, and then the image data is divided by the standard deviation of the training data;
the specific calculation formula is as follows:
x*＝(x-μ)/σ；
where x is the maximum and minimum normalized image data, where μ is the mean of the training data and σ is the standard deviation of the training data.
Example 2
This example is based on example 1. As shown in fig. 3, when the deep learning model is a lung lobe segmentation model, the inputting the normalized image into the deep learning model for image separation and recognition to obtain at least one separation result includes:
s201, inputting the normalized image into a lung lobe segmentation model for image separation and identification to obtain a lung lobe separation result; as shown in fig. 18, where a is an image before input into the lung lobe segmentation model; b is an output image after the lung lobe segmentation model is input;
s202, judging whether the lung lobe separation result is smaller than a preset lung lobe edge threshold value or not;
s203, if yes, judging that the position of a lung lobe in the X-ray chest film is abnormal, and obtaining a first unqualified result; if not, judging that the position of the lung lobe in the X-ray chest film is normal, and obtaining a first qualified result.
In a specific embodiment, for the lung lobe segmentation model, batch test 493 cases of image data, 443 cases of normal image data and 50 cases of rib angle missing are performed, and according to the lung lobe segmentation result, statistics of pixel values of 443 cases of normal chest images from the upper, lower, left and right edges of the image are performed, so that a preset left and right edge threshold value is a1 (in this embodiment, a1 is 0.012-0.035), and a preset upper and lower edge threshold value is a2 (in this embodiment, a2 is 0.031-0.062). If the left and right edge threshold values of the test image are smaller than a preset left and right edge threshold value A1 and/or the upper and lower edge threshold values of the test image are smaller than a preset upper and lower edge threshold value A2, judging that the position of the lung lobe in the X-ray chest film is abnormal; otherwise, judging that the position of the lung lobe in the X-ray chest film is normal. It should be noted that the preset left-right edge threshold a1 and the preset upper-lower edge threshold a2 can be set according to specific test data.
As shown in fig. 4, when the deep learning model is a spine segmentation model, the inputting the normalized image into the deep learning model for image separation and recognition to obtain at least one separation result includes:
s301, inputting the image subjected to standard normalization processing into a spine segmentation model for image separation and identification to obtain a spine separation result; as shown in fig. 19, where a is an image before inputting the spine segmentation model; b is an output image after the spine segmentation model is input;
s302, judging whether the absolute value of the spine separation result is larger than a preset offset or not;
s303, if yes, judging that the position of a spine in the X-ray chest radiograph is abnormal, and obtaining a second unqualified result; if not, judging that the position of the spine in the X-ray chest radiography is normal, and obtaining a second qualified result.
In one specific embodiment, 490 cases of data, 400 cases of normal data and 90 cases of image center offset are tested in batch for the spine segmentation model, and the relative distance of each pixel in the segmentation result from the image center is calculated and summed in the horizontal direction. This can provide 490 signed values, and by taking ± B (in this embodiment, B is 0.02 to 0.07) as a threshold, an accuracy of 95% or more can be ensured. If the result of testing the image is less than-B, the image is shifted to the left from the center of the image, which indicates that the image is abnormal; if the quantization result of the test image is greater than B, then the image is shifted to the right from the center of the image, again indicating an image anomaly. Otherwise, judging that the position of the spine in the X-ray chest radiography is normal. It should be noted that the preset offset threshold B may be set according to specific test data.
As shown in fig. 5, when the deep learning model is a scapula segmentation model, the inputting the normalized image into the deep learning model for image separation and recognition to obtain at least one separation result includes:
s401, inputting the normalized image into a scapula segmentation model for image separation and identification to obtain a scapula separation result; as shown in fig. 20, where a is an image before the scapula segmentation model is input; b is an output image after the scapula segmentation model is input;
s402, judging whether the scapula separation result is larger than a preset area ratio or not;
s403, if yes, judging that the position of the scapula in the X-ray chest radiograph is abnormal, and obtaining a third unqualified result; if not, judging that the position of the scapula in the X-ray chest radiography is normal, and obtaining a third qualified result.
In a specific embodiment, for the scapula segmentation model, 500 cases were tested in batch, 100 cases were scapulae appeared in the lung field range, and 400 cases were normal chest slices. The ratio of the area of the segmentation result to the original image is calculated from the scapula segmentation model, and when the preset ratio is C (C is 0.009-0.028 in the present embodiment), the accuracy can be 90% or more. If the quantization result of the test image is larger than the preset ratio, judging that the position of the scapula in the X-ray chest film is abnormal; otherwise, judging that the position of the scapula in the X-ray chest film is normal. It should be noted that the preset ratio C can be set according to specific test data.
As shown in fig. 6, when the deep learning model is a foreign object detection model, the inputting the normalized image into the deep learning model for image separation and recognition to obtain at least one separation result includes:
s501, inputting the images subjected to normalization processing into a foreign matter detection model for image separation and identification to obtain foreign matter separation results, namely a plurality of boundary frames; as shown in fig. 21, where a is an image before the foreign object detection model is input; b is an output image after the foreign matter detection model is input;
s502, judging whether the confidence score of the bounding box is larger than a preset numerical value or not; in a specific embodiment, the predetermined value of the confidence score is D (in this embodiment D is 0.2-0.8); it should be noted that the preset value D can be set according to actual situations.
S503, if yes, judging that foreign matters exist in the X-ray chest radiograph; further, the bounding boxes with the confidence score larger than the preset value can be processed by adopting a non-maximum suppression algorithm to obtain all candidate bounding boxes; if not, judging that no foreign matter exists in the X-ray chest radiograph, and obtaining a fourth qualified result.
In a specific embodiment, when the deep learning model includes a lung lobe segmentation model and a foreign object detection model;
the determining the image quality of the X-ray chest radiograph based on the separation result comprises:
when the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that foreign matters exist in the X-ray chest radiograph, judging whether the foreign matters are in a lung field range according to the foreign matter detection model and the lung lobe segmentation model, and if so, judging that the X-ray chest radiograph is an unqualified radiograph; if not, judging the X-ray chest film to be a secondary film.
In a specific embodiment, when the deep learning model includes a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model, and a foreign object detection model;
correspondingly, the determining the image quality of the X-ray chest radiography based on the separation result comprises the following steps:
when the lung lobe segmentation model, the spine segmentation model, the scapula segmentation model and the foreign matter detection model all obtain qualified results, judging the X-ray chest film to be a qualified film;
when the lung lobe segmentation model obtains a qualified result and at least one unqualified result is obtained in the spine segmentation model and the scapula segmentation model, judging the X-ray chest film as a secondary film;
when the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that foreign matters exist in the X-ray chest radiograph, judging whether the foreign matters are in a lung field range according to the foreign matter detection model and the lung lobe segmentation model, and if so, judging that the X-ray chest radiograph is an unqualified radiograph; if not, judging the X-ray chest film as a secondary film;
and when the lung lobe segmentation model obtains an unqualified result, judging the X-ray chest film as an unqualified film no matter whether the spine segmentation model, the scapula segmentation model and the foreign matter detection model obtain qualified or unqualified results.
In a specific embodiment, when the deep learning model includes a plurality of models, the method further includes:
carrying out undifferentiated quantization processing on a plurality of separation results;
and determining the image quality of the X-ray chest film based on the separation result after the undifferentiated quantization processing.
The present invention also provides a computer device comprising a processor and a memory, said memory having stored therein at least one instruction, at least one program, set of codes or set of instructions, said at least one instruction, said at least one program, said set of codes or set of instructions being loaded and executed by said processor to implement the X-ray chest image quality determination method as described above.
Example 3
As shown in fig. 7, the present embodiment discloses an X-ray chest radiography image quality determining apparatus, which includes:
a normalization processing module 701, configured to perform normalization processing on the X-ray chest radiography image;
a separation result obtaining module 702, configured to input the normalized image into the deep learning model for image separation and identification, so as to obtain at least one separation result; the deep learning model includes at least one of: a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model and a foreign matter detection model;
an image quality determination module 703 for determining the image quality of the X-ray chest radiograph based on the separation result.
In a specific embodiment, as shown in fig. 8, the normalization processing module 701 includes:
the sorting unit 7011 is configured to sort the pixels of the X-ray chest radiograph image in an ascending order according to the gray values;
a normalization processing unit 7012, configured to take the gray value of the sorted pixels located at the first setting value as a first numerical value, that is, a minimum value, take the gray value of the sorted pixels located at the second setting value as a second numerical value, that is, a maximum value, and perform maximum and minimum normalization processing with truncation on the image according to the gray value; wherein the first set value is greater than or equal to 0% and less than or equal to 10%, the second set value is greater than or equal to 90% and less than or equal to 100%, in a preferred embodiment, the gray value of the pixel located at the 5 th% after sorting is taken as the minimum value, and the gray value of the pixel located at the 95 th% after sorting is taken as the maximum value;
it should be noted that, if the pixels of the X-ray chest radiography image are sorted in descending order according to the gray values, the second set value is greater than or equal to 0% and less than or equal to 10%, and the first set value is greater than or equal to 90% and less than or equal to 100%.
And a standard normalization processing unit 7013, configured to perform standard normalization processing on the normalized image.
The deep learning model is trained and completed based on a uniform deep learning platform, and the quantity and quality of labeled data influence the accuracy of the model. The data labeling method in the embodiment can be manual labeling or labeling by an intelligent machine.
Example 4
On the basis of embodiment 3, for different deep learning models, the separation result obtaining module in the apparatus of the present invention is as follows:
as shown in fig. 9, when the deep learning model is a lung lobe segmentation model, the separation result obtaining module 702 includes:
a lung lobe separation result obtaining unit 901, configured to input the image after the standard normalization processing into a lung lobe segmentation model for image separation and identification, so as to obtain a lung lobe separation result;
a lung lobe separation result determining unit 902, configured to determine whether a lung lobe separation result is smaller than a preset lung lobe edge threshold;
a first unqualified result determination unit 903, configured to determine that a lung lobe position in the X-ray chest radiograph is abnormal when a lung lobe separation result is smaller than a preset lung lobe edge threshold, so as to obtain a first unqualified result;
a first qualified result determining unit 904, configured to determine that a lung lobe position in the X-ray chest film is normal when a lung lobe separation result is greater than or equal to a preset lung lobe edge threshold, so as to obtain a first qualified result;
as shown in fig. 10, when the deep learning model is a spine segmentation model, the separation result obtaining module 702 includes:
a spine separation result obtaining unit 1001 configured to input the image after the standard normalization processing into a spine segmentation model to perform image separation identification, so as to obtain a spine separation result;
a spine separation result determination unit 1002, configured to determine whether an absolute value of a spine separation result is greater than a preset offset;
a second unqualified result judgment unit 1003, configured to judge that a position of a spine in the X-ray chest radiograph is abnormal when an absolute value of a spine separation result is greater than a preset offset, so as to obtain a second unqualified result;
a second qualified result judging unit 1004, configured to judge that a position of a spine in the X-ray chest radiograph is normal when an absolute value of a spine separation result is less than or equal to a preset offset, so as to obtain a second qualified result;
as shown in fig. 11, when the deep learning model is a scapula segmentation model, the separation result obtaining module 702 includes:
a scapula separation result acquisition unit 1101, configured to input the image after the standard normalization processing into a scapula segmentation model for image separation and identification, so as to obtain a scapula separation result;
a scapula separation result determination unit 1102 for determining whether the scapula separation result is greater than a preset area ratio;
a third nonconforming result judging unit 1103, configured to, when the scapula separation result is greater than the preset area ratio, judge that the scapula position in the X-ray chest radiograph is abnormal, to obtain a third nonconforming result;
a third qualified result determining unit 1104, configured to determine that the position of the scapula in the X-ray chest radiograph is normal when the scapula separation result is smaller than or equal to the preset area ratio, so as to obtain a third qualified result;
as shown in fig. 12, when the deep learning model is a foreign object detection model, the separation result obtaining module 702 includes:
a foreign matter separation result obtaining unit 1201, configured to input the image after the standard normalization processing into a foreign matter detection model for image separation and identification, so as to obtain a foreign matter separation result, that is, a plurality of bounding boxes;
a foreign matter separation result determination unit 1202 configured to determine whether the confidence score of the bounding box is greater than a preset value;
a foreign matter determination unit 1203, configured to determine that a foreign matter exists in the X-ray chest radiograph when the confidence score of the bounding box is greater than a preset value;
a fourth qualified result determining unit 1204, configured to determine that no foreign object exists in the X-ray chest radiograph when the confidence score of the bounding box is smaller than or equal to a preset value, so as to obtain a fourth qualified result.
In a specific embodiment, when the deep learning model includes a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model, and a foreign object detection model;
as shown in fig. 13, the image quality determination module 703 includes:
a qualified slice judging unit 7031, configured to judge that the X-ray chest slice is a qualified slice when all of the lung lobe segmentation model, the spine segmentation model, the scapula segmentation model, and the foreign object detection model obtain qualified results;
a secondary slice determining unit 7032, configured to determine that the X-ray chest slice is a secondary slice when the lung lobe segmentation model obtains a qualified result and at least one unqualified result is obtained in the spine segmentation model and the scapula segmentation model; the X-ray chest film processing module is also used for judging whether the foreign matters exist in the lung field range according to the foreign matter detection model and the lung lobe segmentation model when the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that the foreign matters exist in the X-ray chest film, and if not, judging that the X-ray chest film is a secondary film;
a non-qualified slice determining unit 7033, configured to, when the lung lobe segmentation model obtains a qualified result and the foreign object detection model determines that a foreign object exists in the X-ray chest slice, determine whether the foreign object is in a lung field range according to the foreign object detection model and the lung lobe segmentation model, and if so, determine that the X-ray chest slice is a non-qualified slice; and when the lung lobe segmentation model obtains an unqualified result, judging the X-ray chest radiograph as an unqualified radiograph.
In a specific embodiment, when the deep learning model includes a plurality of models, the apparatus further includes:
the undifferentiated quantization processing module is used for carrying out undifferentiated quantization processing on the plurality of separation results;
and the image quality determining module is used for determining the image quality of the X-ray chest radiography based on the separation result after the undifferentiated quantization processing.
According to the invention, the evaluation of the chest radiography is divided into several sub-problems of whether foreign matters appear in the lung field or not, whether scapulae appear in the lung field or not, whether the center of the image is aligned or not, whether diaphragm muscles on two sides exceed the image range or not and the like, the image quality of the chest radiography is rapidly graded based on the deep learning technology, and not only can the technician be reminded whether the image quality is qualified or not, but also the technician can be reminded of which unqualified chest radiography belongs to. The rapid evaluation method for grading the image quality of the X-ray chest radiography can help a technician control the image quality of the chest radiography, and indirectly improves the accuracy of radiography.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The above description is a preferred embodiment of the present invention, and the technical solutions of the present invention are further described in detail, and are not intended to limit the protection scope of the present invention, it should be noted that, for those skilled in the art, many modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations are also regarded as the protection scope of the present invention.
Claims (7)
1. An X-ray chest image quality determination method, characterized in that the method comprises:
acquiring an X-ray chest radiography image;
inputting the image into a deep learning model to obtain at least one of a lung lobe segmentation result, a spine segmentation result, a scapula segmentation result and a foreign matter detection result; the deep learning model includes at least one of: a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model and a foreign matter detection model;
judging whether the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result meets corresponding preset conditions to obtain a judgment result; wherein, the preset conditions corresponding to the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result and the foreign matter detection result respectively include: the left and right edge threshold values of the lung lobe segmentation result are smaller than the preset left and right edge threshold values and/or the upper and lower edge threshold values are smaller than the preset upper and lower edge threshold values; the absolute value of the sum of the relative distances of each pixel in the spine segmentation result and the center of the image in the horizontal direction is larger than a preset offset; the ratio of the area of the scapula segmentation result to the area of the original image is larger than the preset area ratio; the confidence score of the bounding box in the foreign matter detection result is greater than a preset numerical value;
if so, the judgment result is an unqualified result;
if not, the judgment result is a qualified result;
and determining the image quality of the X-ray chest radiography based on the judgment result.
2. The method of claim 1, wherein said acquiring an X-ray chest image comprises:
sorting the X-ray chest radiography image pixels according to the gray values;
taking the gray value of the pixels which are positioned at the first set value after sequencing as the minimum value, taking the gray value of the pixels which are positioned at the second set value after sequencing as the maximum value, and carrying out truncation normalization processing on the image according to the gray value;
and performing standard normalization processing on the normalized image.
3. The method of claim 1, wherein said determining the image quality of the X-ray chest radiograph based on the determination comprises:
when the lung lobe segmentation model, the spine segmentation model, the scapula segmentation model and the foreign matter detection model all obtain qualified results, judging the X-ray chest film to be a qualified film; or
When the lung lobe segmentation model obtains a qualified result and at least one unqualified result is obtained in the spine segmentation model and the scapula segmentation model, judging the X-ray chest film as a secondary film; or
When the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that foreign matters exist in the X-ray chest radiograph, judging whether the foreign matters are in a lung field range according to the foreign matter detection model and the lung lobe segmentation model, and if so, judging that the X-ray chest radiograph is an unqualified radiograph; if not, judging the X-ray chest film as a secondary film; or when the lung lobe segmentation model obtains an unqualified result, judging the X-ray chest film as an unqualified film.
4. An X-ray chest image quality determination apparatus, comprising:
the acquisition module is used for acquiring an X-ray chest radiography image;
a separation result acquisition unit, configured to input the image into a deep learning model to obtain at least one of a lung lobe segmentation result, a spine segmentation result, a scapula segmentation result, and a foreign object detection result; the deep learning model includes at least one of: a lung lobe segmentation model, a spine segmentation model, a scapula segmentation model and a foreign matter detection model;
the separation result judging unit is used for judging whether the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result meets corresponding preset conditions to obtain a judgment result; wherein, the preset conditions corresponding to the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result and the foreign matter detection result respectively include: the left and right edge threshold values of the lung lobe segmentation result are smaller than the preset left and right edge threshold values and/or the upper and lower edge threshold values are smaller than the preset upper and lower edge threshold values; the absolute value of the sum of the relative distances of each pixel in the spine segmentation result and the center of the image in the horizontal direction is larger than a preset offset; the ratio of the area of the scapula segmentation result to the area of the original image is larger than the preset area ratio; the confidence score of the bounding box in the foreign matter detection result is greater than a preset numerical value;
the unqualified result judging unit is used for determining that the judging result is an unqualified result when the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result meets corresponding preset conditions;
the qualified result judging unit is used for determining that the judging result is a qualified result when the lung lobe segmentation result, the spine segmentation result, the scapula segmentation result or the foreign matter detection result does not meet corresponding preset conditions;
and the image quality determining module is used for determining the image quality of the X-ray chest radiography based on the judgment result.
5. The apparatus of claim 4, wherein the obtaining module comprises:
the sorting unit is used for sorting the X-ray chest film image pixels according to the gray values;
the normalization processing unit is used for taking the gray value of the pixels which are positioned at the first set value after being sorted as a first numerical value, taking the gray value of the pixels which are positioned at the second set value after being sorted as a second numerical value, and performing truncated normalization processing on the image according to the first numerical value and the second numerical value;
and the standard normalization processing unit is used for performing standard normalization processing on the normalized image.
6. The apparatus of claim 4, wherein the image quality determination module comprises:
the qualified plate judging unit is used for judging the X-ray chest plate to be a qualified plate when the lung lobe segmentation model, the spine segmentation model, the scapula segmentation model and the foreign matter detection model all obtain qualified results;
the second-level film judging unit is used for judging the X-ray chest film as a second-level film when the lung lobe segmentation model obtains a qualified result and at least one unqualified result is obtained in the spine segmentation model and the scapula segmentation model; the X-ray chest film processing module is also used for judging whether the foreign matters exist in the lung field range according to the foreign matter detection model and the lung lobe segmentation model when the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that the foreign matters exist in the X-ray chest film, and if not, judging that the X-ray chest film is a secondary film;
the unqualified slice judging unit is used for judging whether the foreign matter is in the lung field range according to the foreign matter detection model and the lung lobe segmentation model when the lung lobe segmentation model obtains a qualified result and the foreign matter detection model judges that the foreign matter exists in the X-ray chest slice, and if so, judging that the X-ray chest slice is an unqualified slice; and when the lung lobe segmentation model obtains an unqualified result, judging the X-ray chest radiograph as an unqualified radiograph.
7. A computer device, characterized in that it comprises a processor and a memory, in which at least one instruction, at least one program, a set of codes or a set of instructions is stored, which is loaded and executed by the processor to implement the X-ray chest image quality determination method according to any of the claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110198562.1A CN113052795B (en) | 2018-12-28 | 2018-12-28 | X-ray chest radiography image quality determination method and device |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811626384.2A CN109859168B (en) | 2018-12-28 | 2018-12-28 | X-ray chest radiography image quality determination method and device |
| CN202110198562.1A CN113052795B (en) | 2018-12-28 | 2018-12-28 | X-ray chest radiography image quality determination method and device |
Related Parent Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811626384.2A Division CN109859168B (en) | 2018-12-28 | 2018-12-28 | X-ray chest radiography image quality determination method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113052795A true CN113052795A (en) | 2021-06-29 |
| CN113052795B CN113052795B (en) | 2023-12-08 |
Family
ID=66893074
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811626384.2A Active CN109859168B (en) | 2018-12-28 | 2018-12-28 | X-ray chest radiography image quality determination method and device |
| CN202110198562.1A Active CN113052795B (en) | 2018-12-28 | 2018-12-28 | X-ray chest radiography image quality determination method and device |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811626384.2A Active CN109859168B (en) | 2018-12-28 | 2018-12-28 | X-ray chest radiography image quality determination method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN109859168B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113555089A (en) * | 2021-07-14 | 2021-10-26 | 江苏宏创信息科技有限公司 | Artificial intelligence medical image quality control method applied to clinical image |
| CN113689425A (en) * | 2021-09-15 | 2021-11-23 | 心医国际数字医疗系统(大连)有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
| CN116092642A (en) * | 2023-03-07 | 2023-05-09 | 福建智康云医疗科技有限公司 | Medical image quality control method and system |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11436720B2 (en) * | 2018-12-28 | 2022-09-06 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for generating image metric |
| CN110264449B (en) * | 2019-06-14 | 2022-05-20 | 辽宁万象联合医疗科技有限公司 | Quality control method of medical radiological image chest radiography based on artificial intelligence |
| CN110866898B (en) * | 2019-10-30 | 2022-10-14 | 上海联影智能医疗科技有限公司 | Mammary gland image processing method, equipment and storage medium |
| CN111476777B (en) * | 2020-04-07 | 2023-08-22 | 上海联影智能医疗科技有限公司 | Chest radiography image processing method, system, readable storage medium and apparatus |
| CN111709906A (en) * | 2020-04-13 | 2020-09-25 | 北京深睿博联科技有限责任公司 | Medical image quality evaluation method and device |
| CN111640093A (en) * | 2020-05-20 | 2020-09-08 | 上海联影智能医疗科技有限公司 | Quality control method for medical image and computer-readable storage medium |
| CN111753896A (en) * | 2020-06-16 | 2020-10-09 | 江西中科九峰智慧医疗科技有限公司 | Chest radiography exposure problem classification method and system based on deep learning |
| CN112263269A (en) * | 2020-09-22 | 2021-01-26 | 北京赛迈特锐医疗科技有限公司 | Intelligent detection system and method for urinary system X-ray plain stone |
| CN112168193A (en) * | 2020-10-14 | 2021-01-05 | 北京赛迈特锐医疗科技有限公司 | System and method for acquiring patella anatomical parameters based on patella axial position X-ray image |
| CN112308853A (en) * | 2020-10-20 | 2021-02-02 | 平安科技（深圳）有限公司 | Electronic equipment, medical image index generation method and device and storage medium |
| CN113409924A (en) * | 2021-07-01 | 2021-09-17 | 上海市第一人民医院 | Artificial intelligence-based lung examination image auxiliary marking method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016066444A1 (en) * | 2014-10-30 | 2016-05-06 | Koninklijke Philips N.V. | Device and method for determining image quality of a radiogram image |
| CN105957094A (en) * | 2016-06-15 | 2016-09-21 | 深圳维影医疗科技有限公司 | contour detection based rapid automatic segmentation method and system for lung field of chest radiograph |
| US20170039711A1 (en) * | 2015-08-07 | 2017-02-09 | Arizona Board Of Regents On Behalf Of Arizona State University | System and method for detecting central pulmonary embolism in ct pulmonary angiography images |
| CN107590809A (en) * | 2017-06-30 | 2018-01-16 | 上海联影医疗科技有限公司 | Lung dividing method and medical image system |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8660372B2 (en) * | 2010-05-10 | 2014-02-25 | Board Of Regents Of The University Of Texas System | Determining quality of an image or video using a distortion classifier |
| CN103745470A (en) * | 2014-01-08 | 2014-04-23 | 兰州交通大学 | Wavelet-based interactive segmentation method for polygonal outline evolution medical CT (computed tomography) image |
| US10049457B2 (en) * | 2016-08-29 | 2018-08-14 | CephX Technologies Ltd. | Automated cephalometric analysis using machine learning |
| CN108596198A (en) * | 2017-07-26 | 2018-09-28 | 江西中科九峰智慧医疗科技有限公司 | A kind of recognition methods of pneumothorax x-ray image and system based on deep learning |
| CN107808377B (en) * | 2017-10-31 | 2019-02-12 | 北京青燕祥云科技有限公司 | The positioning device of lesion in a kind of lobe of the lung |
| CN108364017B (en) * | 2018-01-24 | 2019-11-05 | 华讯方舟科技有限公司 | A kind of picture quality classification method, system and terminal device |
| CN108460774A (en) * | 2018-03-29 | 2018-08-28 | 河北北方学院 | A kind of lung's Automatic medical image segmentation method |
| CN108648182B (en) * | 2018-04-27 | 2022-02-11 | 南京信息工程大学 | Breast cancer nuclear magnetic resonance image tumor region segmentation method based on molecular subtype |
| CN108806776A (en) * | 2018-06-14 | 2018-11-13 | 暨南大学附属第医院（广州华侨医院） | A method of the Multimodal medical image based on deep learning |
-
2018
- 2018-12-28 CN CN201811626384.2A patent/CN109859168B/en active Active
- 2018-12-28 CN CN202110198562.1A patent/CN113052795B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016066444A1 (en) * | 2014-10-30 | 2016-05-06 | Koninklijke Philips N.V. | Device and method for determining image quality of a radiogram image |
| US20170039711A1 (en) * | 2015-08-07 | 2017-02-09 | Arizona Board Of Regents On Behalf Of Arizona State University | System and method for detecting central pulmonary embolism in ct pulmonary angiography images |
| CN105957094A (en) * | 2016-06-15 | 2016-09-21 | 深圳维影医疗科技有限公司 | contour detection based rapid automatic segmentation method and system for lung field of chest radiograph |
| CN107590809A (en) * | 2017-06-30 | 2018-01-16 | 上海联影医疗科技有限公司 | Lung dividing method and medical image system |
Non-Patent Citations (2)
| Title |
|---|
| 刘炎;邱天爽;郭冬梅;: "基于柔性形态学与聚类算法的X光胸片肺部分割", 中国生物医学工程学报, no. 05 * |
| 贾广河;: "X光胸片图像分割和特征识别技术研究", 健康之路, no. 03 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113555089A (en) * | 2021-07-14 | 2021-10-26 | 江苏宏创信息科技有限公司 | Artificial intelligence medical image quality control method applied to clinical image |
| CN113689425A (en) * | 2021-09-15 | 2021-11-23 | 心医国际数字医疗系统(大连)有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
| CN116092642A (en) * | 2023-03-07 | 2023-05-09 | 福建智康云医疗科技有限公司 | Medical image quality control method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109859168B (en) | 2021-02-19 |
| CN109859168A (en) | 2019-06-07 |
| CN113052795B (en) | 2023-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109859168B (en) | X-ray chest radiography image quality determination method and device | |
| US11922626B2 (en) | Systems and methods for automatic detection and quantification of pathology using dynamic feature classification | |
| US20210338179A1 (en) | Mammographic image processing method and apparatus, system and medium | |
| CN111325739B (en) | Method and device for detecting lung focus and training method of image detection model | |
| Hao et al. | Inaccurate labels in weakly-supervised deep learning: Automatic identification and correction and their impact on classification performance | |
| US10755413B1 (en) | Method and system for medical imaging evaluation | |
| KR20200095504A (en) | 3D medical image analysis method and system for identifying vertebral fractures | |
| US10438353B2 (en) | Evaluation of an X-ray image of a breast produced during a mammography | |
| CN111325745B (en) | Fracture region analysis method and device, electronic equipment and readable storage medium | |
| CN105640577A (en) | Method and system automatically detecting local lesion in radiographic image | |
| CN109919254B (en) | Breast density classification method, system, readable storage medium and computer device | |
| US20180053297A1 (en) | Methods and Apparatuses for Detection of Abnormalities in Low-Contrast Images | |
| CN112699868A (en) | Image identification method and device based on deep convolutional neural network | |
| CN111986182A (en) | Auxiliary diagnosis method, system, electronic device and storage medium | |
| CN111883256A (en) | Tuberculosis patient early warning system and method based on electronic medical record data | |
| EP1431901A2 (en) | Computer-assisted reconciliation of multiple image reads | |
| CN112241961A (en) | Chest X-ray film auxiliary diagnosis method and system based on deep convolutional neural network | |
| CN113763340A (en) | Automatic grading method based on multitask deep learning ankylosing spondylitis | |
| CN115187566A (en) | Intracranial aneurysm detection method and device based on MRA image | |
| CN114757873A (en) | Rib fracture detection method and device, terminal equipment and readable storage medium | |
| CN113592834A (en) | X-ray chest radiography image quality determination method and device | |
| CN116977338B (en) | Chromosome case-level abnormality prompting system based on visual semantic association | |
| CN116228660A (en) | Method and device for detecting abnormal parts of chest film | |
| Ummah et al. | Covid-19 and Tuberculosis Detection in X-Ray of Lung Images with Deep Convolutional Neural Network. | |
| Lian et al. | Machine learning in medicine—focus on radiology |
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 |