CN114332067A - White matter high signal detection system and method - Google Patents
White matter high signal detection system and method Download PDFInfo
- Publication number
- CN114332067A CN114332067A CN202210005119.2A CN202210005119A CN114332067A CN 114332067 A CN114332067 A CN 114332067A CN 202210005119 A CN202210005119 A CN 202210005119A CN 114332067 A CN114332067 A CN 114332067A
- Authority
- CN
- China
- Prior art keywords
- image
- image data
- area
- signal
- white matter
- 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
- 210000004885 white matter Anatomy 0.000 title claims abstract description 192
- 238000000034 method Methods 0.000 title claims abstract description 74
- 238000001514 detection method Methods 0.000 title claims abstract description 33
- 238000013507 mapping Methods 0.000 claims abstract description 17
- 238000007781 pre-processing Methods 0.000 claims abstract description 16
- 238000012216 screening Methods 0.000 claims abstract description 13
- 238000004364 calculation method Methods 0.000 claims abstract description 10
- 238000003745 diagnosis Methods 0.000 abstract description 11
- 230000000875 corresponding effect Effects 0.000 description 31
- 206010072731 White matter lesion Diseases 0.000 description 9
- 238000010586 diagram Methods 0.000 description 9
- 208000034800 Leukoencephalopathies Diseases 0.000 description 8
- 238000011156 evaluation Methods 0.000 description 6
- 230000011218 segmentation Effects 0.000 description 6
- 230000000295 complement effect Effects 0.000 description 5
- 238000012937 correction Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000003902 lesion Effects 0.000 description 3
- 210000003625 skull Anatomy 0.000 description 3
- 208000029812 Cerebral Small Vessel disease Diseases 0.000 description 2
- 230000008499 blood brain barrier function Effects 0.000 description 2
- 210000001218 blood-brain barrier Anatomy 0.000 description 2
- 210000004556 brain Anatomy 0.000 description 2
- 239000003086 colorant Substances 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241001270131 Agaricus moelleri Species 0.000 description 1
- 208000029147 Collagen-vascular disease Diseases 0.000 description 1
- 208000037451 Leukoaraiosis Diseases 0.000 description 1
- 230000010455 autoregulation Effects 0.000 description 1
- 210000005013 brain tissue Anatomy 0.000 description 1
- 230000003727 cerebral blood flow Effects 0.000 description 1
- 208000026106 cerebrovascular disease Diseases 0.000 description 1
- 208000010877 cognitive disease Diseases 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- JXSJBGJIGXNWCI-UHFFFAOYSA-N diethyl 2-[(dimethoxyphosphorothioyl)thio]succinate Chemical compound CCOC(=O)CC(SP(=S)(OC)OC)C(=O)OCC JXSJBGJIGXNWCI-UHFFFAOYSA-N 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000001771 impaired effect Effects 0.000 description 1
- 238000002075 inversion recovery Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008506 pathogenesis Effects 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 210000001519 tissue Anatomy 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention relates to a white matter high signal detection system and a method, which comprises a database; an image preprocessing module; a calculation module; a processor; carrying out matching judgment for three times through a first method, a second method and a third method, comparing first image data to be detected with a plurality of reference image data, gradually screening the reference image data which is most matched with the first image data to be detected, outputting corresponding reference scoring data by combining known information such as the reference image data and a mapping relation table, and obtaining the condition of high signal load of white matter; according to the mode, the reference image data matched with the first image data to be detected can be accurately and quickly found out, and the white matter high signal score and the load condition in the first image data to be detected are output, so that a doctor can conveniently perform follow-up diagnosis and treatment, the judgment efficiency and accuracy are improved, and misjudgment caused by human factors is avoided.
Description
Technical Field
The invention relates to the technical field of medical images, in particular to a white matter high signal detection system and method capable of assisting diagnosis of doctors.
Background
White Matter Hyperintensity (WMH) is also known as leukoencephalopathy (Leukoaraiosis, LA). The imaging term was first proposed by the neurologist Hachinski. The disease is usually manifested by high signal of T2WI or T2 liquid attenuation inversion recovery sequence (FLAIR), equal signal or low signal of T1WI, etc.
With the continuous and intensive research, WMH is now widely considered as an expression of Cerebral Small Vessel Disease (CSVD), the pathogenesis of WMH is mainly related to impaired cerebral blood flow autoregulation, collagen vascular disease, Blood Brain Barrier (BBB) destruction and genetic factors, and the incidence of WMH is positively correlated with age, and the incidence of WMH in people over 60 years old exceeds 85%. There are also a number of studies that indicate that WMH is closely associated with the development of cognitive dysfunction and other cerebrovascular diseases. With the rapid development of imaging technology, the research of WMH has made some breakthrough progress in recent years.
The prior art only realizes image segmentation of white high signals, but does not further evaluate the white high signals on the basis, so that a doctor cannot be helped to perform preliminary diagnosis according to an evaluation result. Doctors usually need to recognize images according to the segmented images, namely, the white matter high signals are judged manually, time and labor are wasted, misjudgment often occurs, and the accuracy is easily influenced by human factors. Therefore, there is a need for a system and method for detecting high white matter signal that can assist the diagnosis of doctors.
Disclosure of Invention
The invention aims to provide a white matter high signal detection system and method capable of assisting diagnosis of doctors.
The invention relates to a white matter high signal detection system, which comprises
A database in which at least one reference image data and reference area data of different white matter high signals in the reference image data are prestored;
the image preprocessing module is used for preprocessing the MRI image data and outputting first image data;
a calculation module for calculating first area data of white matter high signals in the first image data;
and the processor judges whether the first image data is preliminarily matched with at least one reference image data through a first method, judges whether the positions of the white matter high signals in the first image data are matched with the positions of the different white matter high signals in the reference image data through a second method if the first image data is preliminarily matched, judges whether the first area data in the first image data are matched with the reference area data of the different white matter high signals in the reference image data through a third method if the first image data are matched with the reference area data of the different white matter high signals in the reference image data, outputs corresponding reference scoring data if the first image data are matched with the reference area data, and obtains the white matter high signal load condition.
The invention relates to a white matter high signal detection system, wherein the first method comprises the following steps: the processor scales at least one piece of reference image data to a preset resolution and generates a reference scaled image, scales a first piece of image data to the preset resolution and generates a contrast scaled image, and the processor calculates the reference scaled image and the average gray value of all pixels of the contrast scaled image, and compares the first gray value and the average gray value of each pixel in the at least one piece of reference scaled image and the second gray value and the average gray value of each pixel in the contrast scaled image;
the processor calculates the absolute value of the difference value between a plurality of first pixel values a and a plurality of second pixel values b according to a first pixel value a of which the first gray value in at least one reference scaling image is greater than or equal to the gray average value and a second pixel value b of which the second gray value in the contrast scaling image is greater than or equal to the gray average value, and judges that at least one reference scaling image corresponding to the minimum value of the absolute values of the difference values between the contrast scaling image and the first pixel values a and the second pixel values b is matched, and first image data corresponding to the contrast scaling image is preliminarily matched with at least one corresponding reference image data.
The invention relates to a white matter high signal detection system, wherein the second method comprises the following steps: the processor carries out outline delineation of white matter high signals on first image data and at least one reference image data preliminarily matched with the first image data to respectively obtain a first signal image only containing the white matter high signal outline in the first image data and at least one reference signal image of the white matter high signal outline in the reference image data preliminarily matched with the first image data, the processor scales the first signal image and the at least one reference signal image to be matched with the size of an m multiplied by n grid, and the first signal image and the at least one reference signal image are respectively placed in the m multiplied by n grid according to corresponding positions;
the processor judges whether the distance between the geometric center g of the first signal image and the geometric center o of at least one reference signal image is smaller than or equal to a first variable, if not, the reference zoom image corresponding to the minimum value of the absolute value of the difference value of the first pixel value a and the second pixel value b is removed, and the first method is repeated;
and if so, determining that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the at least one reference image data.
The invention relates to a white matter high signal detection system, wherein the third method comprises the following steps: the processor calculates an area S of the first signal image1Area S of at least one reference signal image2；
The processor calculates an area S of the first signal image1Area S of reference signal image2The intersection area J is S1∩S2；
The processor judges that when the intersection area J is equal to the area S of only one reference signal image2When the absolute value of the difference is smaller than a first preset threshold value, the first signal image is judged to be matched with only one reference signal image, and first area data of white matter high signals in the first image data and only one reference image are judged to be matchedMatching reference area data of white matter high signals in the image data;
the processor judges the intersection area J and the area S of at least one reference signal image2When the absolute value of the difference is less than a first preset threshold, repeating the first step k times until the intersection area J and the area S of only one reference signal image are obtained2The absolute value of the difference is smaller than a first preset threshold;
the first step is: adjusting the m multiplied by n grids to (m + k) x (n + k) grids with the same total area, judging whether the distances between the geometric center of the first signal image and the geometric center of the at least one reference signal image are smaller than or equal to a second variable, if so, judging that the position of a white matter high signal in the first image data is matched with the position of a white matter high signal in the at least one reference signal image, if not, screening out the reference signal image with the distance between the geometric center of the first signal image and the geometric center of the first signal image larger than the second variable, and judging that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the rest reference image data; the processor determines whether the intersection area J and the area S of only one reference signal image2And if not, repeating the first step, and adding 1 to the value of k.
The invention relates to a white matter high signal detection system, wherein the third method comprises the following steps: the processor calculates an area S of the first signal image1Area S of at least one reference signal image2；
The processor calculates an area S of the first signal image1Area S of at least one reference signal image2The complete set of (a) is Q;
area S of the first signal image1Area S with respect to at least one reference signal image2Complementary collection area B of1＝CQS1；
Area S of reference signal image2Area S with respect to first signal image1Complementary collection area B of2＝CQS2；
The processor determines the area S of the reference signal image when only one reference signal image is present2Area S with respect to first signal image1Complementary collection area B of2When the first signal image is smaller than a second preset threshold value, judging that the first signal image is matched with only one reference signal image, and the first area data of the white matter high signal in the first image data is matched with the reference area data of the white matter high signal in only one reference image data;
the processor determines the area S of at least one reference signal image2Area S with respect to first signal image1Complementary collection area B of2When the area is smaller than a second preset threshold value, the first step is repeated for k times until the area S of only one reference signal image is larger than the second preset threshold value2Area S with respect to first signal image1Complementary collection area B of2Less than a second preset threshold:
the first step is: adjusting the m multiplied by n grids to (m + k) x (n + k) grids with the same total area, judging whether the distances between the geometric center of the first signal image and the geometric center of the at least one reference signal image are smaller than or equal to a second variable, if so, judging that the position of a white matter high signal in the first image data is matched with the position of a white matter high signal in the at least one reference signal image, if not, screening out the reference signal image with the distance between the geometric center of the first signal image and the geometric center of the first signal image larger than the second variable, and judging that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the rest reference image data; the processor determines whether there is only one reference signal image area S2Area S with respect to first signal image1Complementary collection area B of2If the value is smaller than the second preset threshold value, the first step is ended, otherwise, the first step is repeated, and the value of k is added by 1.
The invention relates to a white matter high signal detection system, wherein the step of outputting corresponding reference scoring data and obtaining the white matter high signal load condition comprises the following steps:
a white matter high signal mapping relation table is prestored in the database;
and the processor outputs corresponding white matter high signal scores and load conditions according to the white matter high signal mapping relation table and by combining the area S2 of the reference signal image.
The invention relates to a white matter high signal detection system, wherein a processor is electrically connected with an image preprocessing module and a calculation module.
The invention relates to a white matter high signal detection system, wherein the range of a first gray value and a second gray value is 0-63.
The invention relates to a white matter high signal detection system, wherein an improved Scheltens scale is adopted in a white matter high signal mapping relation table.
The invention relates to a white matter high signal detection method, which comprises the following steps
Storing at least one reference image data, reference area data of different white matter high signals in the reference image data;
preprocessing MRI image data and outputting first image data;
calculating first area data of white matter high signals in the first image data;
and judging whether the first image data is preliminarily matched with at least one reference image data, if so, judging whether the positions of the white matter high signals in the first image data are matched with the positions of the different white matter high signals in the reference image data, if so, judging whether the first area data in the first image data are matched with the reference area data of the different white matter high signals in the reference image data, if so, outputting corresponding reference scoring data, and obtaining the white matter high signal load condition.
Compared with the prior art, the white matter high signal detection system and the white matter high signal detection method are different in that the white matter high signal detection system and the white matter high signal detection method can accurately and quickly find out the reference image data matched with the first image data to be detected and output the white matter high signal score and the load condition in the first image data to be detected, so that a doctor can conveniently conduct follow-up diagnosis and treatment, the evaluation efficiency and accuracy are improved, and misjudgment caused by human factors is avoided.
The white matter high signal detection system and method of the present invention will be further explained with reference to the accompanying drawings.
Drawings
FIG. 1 is a flow chart of a white matter high signal detection system and method;
FIG. 2 is a diagram of a skull removal and segmentation process;
FIG. 3 is a schematic outline drawing;
FIG. 4 is a schematic view of a placement grid;
FIG. 5 is a schematic diagram of FIG. 4 after zooming in the first position;
FIG. 6 is a schematic diagram of FIG. 4 after zooming in and out of the second position;
FIG. 7 is a schematic diagram of the third position of FIG. 4 after zooming;
FIG. 8 is a schematic illustration of the fourth position of FIG. 4 after scaling;
FIG. 9 is a schematic illustration of the fifth position of FIG. 4 after scaling;
FIG. 10 is a schematic view of the superposition of FIGS. 5 and 6;
FIG. 11 is a schematic view of FIG. 5, FIG. 7 superimposed;
FIG. 12 is a mesh refinement diagram of FIG. 10;
FIG. 13 is a mesh refinement diagram of FIG. 11;
FIG. 14 is a mesh refinement diagram of FIG. 12;
fig. 15 is a mesh refinement diagram of fig. 13.
Detailed Description
As shown in FIGS. 1-15, referring to FIGS. 1 and 2, the white matter high signal detection system and method of the present invention comprises
A database in which at least one reference image data and reference area data of different white matter high signals in the reference image data are prestored;
the image preprocessing module is used for preprocessing the MRI image data and outputting first image data;
a calculation module for calculating first area data of white matter high signals in the first image data;
the processor judges whether the first image data is preliminarily matched with at least one reference image data through a first method, if not, the first method is repeated, if so, the second method is used for judging whether the positions of the white matter high signals in the first image data are matched with the positions of the different white matter high signals in the reference image data, if not, the first method and the second method are repeated, if so, the third method is used for judging whether the first area data in the first image data are matched with the reference area data of the different white matter high signals in the reference image data, if not, the first method, the second method and the third method are roughly repeated, and if so, corresponding reference scoring data are output, and the white matter high signal load condition is obtained.
The method comprises the steps of carrying out matching judgment for three times through a first method, a second method and a third method, comparing first image data to be detected with a plurality of reference image data, gradually screening the reference image data which is most matched with the first image data to be detected, outputting corresponding reference scoring data by combining known information such as the reference image data and a mapping relation table, and obtaining the white matter high signal load condition.
According to the method, the reference image data matched with the first image data to be detected can be accurately and quickly found out, and the white matter high signal score and the load condition in the first image data to be detected are output, so that the follow-up diagnosis and treatment of a doctor are facilitated, the evaluation efficiency and accuracy are improved, and the misjudgment caused by human factors is avoided.
Wherein the reference image data is a plurality of pieces of MRI image data which are stored in a database and are set according to different areas of white matter high signals.
Wherein the reference image data and the obtained MRI image are: the shooting is carried out at the same distance on the same equipment.
According to the invention, through the arrangement, the MRI image and the reference image data which are shot by the same person are matched most, or the user can match the reference image data of other persons with similar brain structures without being influenced by the background area and the brain size.
The MRI image data is to-be-detected MRI image data, and includes but is not limited to an MRI T1FLAIR sequence head scanning image and an MRI T2FLAIR sequence head scanning image.
Wherein the step of preprocessing the MRI image data by the image preprocessing module comprises: performing head motion correction processing on MRI image data by using SPM, wherein a head motion correction function in the SPM can determine parameters of time sequence images and reference registration thereof, and aligning each frame image in an experimental sequence with a reference image of the sequence on the basis to correct head motion;
the method uses N4BiasFieldCorrect bias field correction and a Gaussian filter-based method to carry out bias field correction processing on MRI image data so as to reduce misdiagnosis and improve the accuracy of diagnosis;
carrying out normalization processing on MRI image data by adopting Python to scale the data by the same order of magnitude, and improving the training speed and accuracy of the segmentation model;
removing skull, eye, skin edge tissues and the like in the MRI image data by using a skull removal algorithm to remove non-target images and obtain target images so as to further remove noise interference caused by the non-brain tissues and ensure the segmentation accuracy, which is shown in the middle diagram of FIG. 2;
and inputting the MRI image into a white matter high signal segmentation model trained in advance to output a white matter high signal segmentation result in the MRI, wherein the segmentation result can be seen in a right graph of fig. 2. Which is prior art and will not be described herein.
Wherein, the step of calculating the first area data of white matter high signal in the first image data by the calculating module is: placing a first signal image of a white matter high signal in first image data in a grid with the size equivalent to that of the first signal image, and calculating the number of the grids occupied by the first signal image so as to estimate first area data; or constructing a rectangular coordinate system for the first signal image of the white matter high signal in the first image data to obtain the coordinates of the outer contour of the first signal image, and obtaining first area data according to the Euclidean plane area calculation method; of course, the first area data may also be derived by other methods.
Wherein the "whether the first image data and at least one reference image data are preliminarily matched" is: whether the overall image structure and shape of the first image data and the at least one reference image data are similar.
Wherein the "whether the position of the white matter high signal in the first image data matches the position of the white matter high signal in the reference image data" is: whether the position of the white matter high signal in the first image data is in close proximity to the position of the different white matter high signal in the reference image data.
Wherein, whether the first area data in the first image data is matched with the reference area data of the different white matter high signal in the reference image data is as follows: whether the area of the white matter high signal in the first image data is at a very close area size to the area of the different white matter high signal in the reference image data.
The step of outputting the corresponding reference score data and obtaining the white matter high signal load condition may obtain the score and the load condition of the corresponding white matter high signal by using a white matter high signal mapping relation table, where the white matter high signal mapping relation table may use Fazekas scale (0-6 points), modified Scheltens scale (0-30 points), Ylikoski scale (0-48 points), Manolio scale (0-9 points), and the like, which are all evaluation scales based on MRI image examination, and are prior art and are not described herein.
In the method of substantially repeating the first method, the second method, and the third method, reference may be made to the first step.
As a further explanation of the present invention, referring to fig. 1 and 2, the first method is: the processor scales at least one piece of reference image data to a preset resolution and generates a reference scaled image, scales a first piece of image data to the preset resolution and generates a contrast scaled image, and the processor calculates the reference scaled image and the average gray value of all pixels of the contrast scaled image, and compares the first gray value and the average gray value of each pixel in the at least one piece of reference scaled image and the second gray value and the average gray value of each pixel in the contrast scaled image;
the processor calculates the absolute value of the difference value between a plurality of first pixel values a and a plurality of second pixel values b according to a first pixel value a of which the first gray value in at least one reference scaling image is greater than or equal to the gray average value and a second pixel value b of which the second gray value in the contrast scaling image is greater than or equal to the gray average value, and judges that at least one reference scaling image corresponding to the minimum value of the absolute values of the difference values between the contrast scaling image and the first pixel values a and the second pixel values b is matched, and first image data corresponding to the contrast scaling image is preliminarily matched with at least one corresponding reference image data.
According to the first method, the reference image data and the first image data are firstly zoomed to the same resolution, and then the size relation among the zoomed reference image data, the gray value of the first image data and the average gray value is calculated and compared, so that after the primary screening of the first method, the reference image data which does not meet the related gray value relation with the first image data is screened out, the reference image data which meets the related gray value relation with the first image data is left, and the next screening and matching are carried out.
Wherein, the definition of the first gray value and the second gray value can be: because the colors and the brightness of each point of the scenery are different, each point on the shot black-and-white picture or the black-and-white image reproduced by the television receiver presents different gray colors.
The first gray value and the second gray value are as follows: the logarithmic relationship between white and black is divided into several levels, called "gray scale". The first and second gray scale values generally range from 0 to 255, white is 255 and black is 0.
Of course, the range of the first and second gray scale values is preferably from 0 to 63, with white being 63 and black being 0. That is, the preferred range used in the present invention is a combination of four levels of the above range.
At least one of the reference image data and the first image data has a gray scale corresponding to a resolution thereof, and the resolution thereof is made to be the same as the gray scale for the convenience of calculation.
The preset resolution is 6 × 6 to 720 × 720, preferably 256 × 256. For example, the resolution of at least one of the reference image data is 720 × 720, the resolution of the first image data is 800 × 800, the at least one of the reference image data is converted into a reference scaled image with a resolution of 256 × 256 by an image processing tool, and the first image data is converted into a contrast scaled image with a resolution of 256 × 256, where the reference scaled image and the contrast scaled image each have 65536 pixel points and 64 gray scales. The processor calculates the gray average value of 131072 pixel points in the reference zoom image and the contrast zoom image, and the calculation method comprises the following steps: and adding the first gray values of 65536 pixel points of the reference scaled image and the second gray values of 65536 pixel points in the contrast scaled image, and dividing the sum by 131072, wherein the average value of the calculated gray levels is 30. The processor calculates 23655, 32120, 35330, 52123 and 61300 first pixel values a of which the first gray scale values are greater than or equal to the gray scale average value 30 in at least one reference scaling image respectively, 35120 second pixel values b of which the second gray scale values are greater than or equal to the gray scale average value 30 in the contrast scaling image, 11465, 3000, 210, 17003 and 26180 absolute values of differences with the first pixel values a respectively, and 35330 and 35120 which are minimum values are 210, and judges that the contrast scaling image is matched with the reference scaling image of which the second pixel value b is 35330, namely the first image data corresponding to the contrast scaling image is matched with the reference image data corresponding to the reference scaling image.
For further explanation of the present invention, referring to fig. 1 to 9, the second method is: the processor carries out outline delineation of white matter high signals on first image data and at least one reference image data preliminarily matched with the first image data to respectively obtain a first signal image only containing the white matter high signal outline in the first image data and at least one reference signal image of the white matter high signal outline in the reference image data preliminarily matched with the first image data, the processor scales the first signal image and the at least one reference signal image to be matched with the size of an m multiplied by n grid, and the first signal image and the at least one reference signal image are respectively placed in the m multiplied by n grid according to corresponding positions;
the processor judges whether the distance between the geometric center g of the first signal image and the geometric center o of at least one reference signal image is smaller than or equal to a first variable, if not, the reference zoom image corresponding to the minimum value of the absolute value of the difference value of the first pixel value a and the second pixel value b is removed, and the first method is repeated;
and if so, determining that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the at least one reference image data.
According to the second method, the first image data and the reference image data left after screening by the first method are subjected to white matter high signal contour delineation to generate a first signal image and a plurality of reference signal images, the first signal image and the plurality of reference signal images are placed in a grid matched with the size of the first signal image, whether the geometric center of the first signal image is close to the geometric center of the reference signal image is judged, the reference signal image with the geometric center far away from the geometric center of the first signal image is screened, the reference signal image with the geometric center close to the geometric center of the first signal image is left, and the position of the reference signal image is judged to be matched, so that further subsequent judgment is carried out.
The processor can automatically perform contour delineation on the white matter high signal, and acquire a geometric center of the first signal image and a geometric center of the at least one reference signal image, which is the prior art and is not described herein again.
Wherein, the side length of each grid in the m multiplied by n grids is x, the unit is mm, and the total area of the whole grid is mnx2In units of mm2The first variable is the maximum distance between two points in each of the m x n grids, i.e. the distance between two points in each grid
Wherein m and n are integers of 3 or more, and can be based on white matter high signalThe contour shape, the required degree of refinement, adjusts the size of m, n, with an initial value of m x n of preferably 4 x 4, the total area mnx of the m x n grid2Is preferably 16x2。
The process of scaling the first signal image and the at least one reference signal image to match the size of the mxn grid by the processor and placing the scaled first signal image and the at least one reference signal image in the mxn grid according to the corresponding positions is shown in fig. 3 to 9, the upper left corner of fig. 3 is the reference signal image of "the first signal image only containing the white matter high signal contour in the first image data" and the contour of the white matter high signal in the 4 reference image data preliminarily matched with the first image data in fig. 3 except the upper left corner ", fig. 4 is the reference signal image in which the images are placed in the grid of 4 × 4 according to the relative positions, fig. 5 is the first amplified signal image, and fig. 6 to 9 are the 4 amplified reference signal images.
Wherein the geometric center of the first signal image is indicated by the letter g, see fig. 5. The geometric centers of at least one of the reference signal images are all indicated by the letter o, see fig. 6-9.
For example, in an example of the first method, the second pixel values b of the contrast scaled image with the second gray scale value greater than or equal to the gray scale average value 30 are 35120, the absolute values of the differences between the second gray scale value and the plurality of first pixel values a are 11465, 3000, 210, 17003 and 26180, respectively, and the minimum value is 35330-.
As a further explanation of the present invention, referring to fig. 1, 10, 11, 12, 13, 14, 15, the third method is: the processor calculates an area S of the first signal image1Area S of at least one reference signal image2；
The processorCalculating the area S of the first signal image1Area S of reference signal image2The intersection area J is S1∩S2；
The processor judges that when the intersection area J is equal to the area S of only one reference signal image2When the absolute value of the difference is smaller than a first preset threshold value, judging that the first signal image is matched with only one reference signal image, and matching first area data of a white matter high signal in the first image data with reference area data of a white matter high signal in only one reference image data;
the processor judges the intersection area J and the area S of at least one reference signal image2When the absolute value of the difference is less than a first preset threshold, repeating the first step k times until the intersection area J and the area S of only one reference signal image are obtained2The absolute value of the difference is smaller than a first preset threshold;
the first step is: adjusting the m multiplied by n grids to (m + k) x (n + k) grids with the same total area, judging whether the distances between the geometric center of the first signal image and the geometric center of the at least one reference signal image are smaller than or equal to a second variable, if so, judging that the position of a white matter high signal in the first image data is matched with the position of a white matter high signal in the at least one reference signal image, if not, screening out the reference signal image with the distance between the geometric center of the first signal image and the geometric center of the first signal image larger than the second variable, and judging that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the rest reference image data; the processor determines whether the intersection area J is equal to the area S of only one reference signal image that matches a white matter high signal position in the first image data2And if not, repeating the first step, and adding 1 to the value of k.
According to the third method, the area of the first signal image is calculated, the area of the reference image data left after screening by the first method and the second method is calculated, intersection is taken, and when the area of the first signal image and the area of one reference image data only meet the correlation intersection relationship, the fact that only one reference image data is left is judged, and the fact that the area of the first signal image is matched with the area of the reference image data is judged. If the judgment is not satisfied, the area of at least one reference image data satisfies the correlation intersection relationship, the grid is refined and then returned to the second method, and then the plurality of reference image data are screened again by judging whether the geometric center is close enough or not until the area of only one reference image data satisfies the correlation intersection relationship, so that the final reference image data which is most matched with the first signal image is screened out.
Wherein, the side length of each grid in the m multiplied by n grids is x, and the total area of the whole grid is mnx2After the m × n grid is divided into (m + k) × (n + k) grids, the side length of each grid is
The second variable is the maximum distance between two points in each of the grids of (m + k) × (n + k), i.e. the second variable is
Wherein k is: the number of grids is changed from m × n to (m + k) × (n + k) at the k-th cycle of the first step, for example, the number of grids is changed from m × n to (m +1) × (n +1) at the first cycle of the first step; when the first step is circulated for the second time, the grid number is changed from m multiplied by n to (m +2) multiplied by (n + 2); when the first step is circulated for the third time, the grid number is changed from m multiplied by n to (m +3) multiplied by (n + 3);
......
and so on.
Wherein an area S of the first signal image1May be the first area data.
Wherein an area S of the first signal image1Area S of reference signal image2The unit of the intersection area J is mm2。
Wherein the first preset threshold value is [0 ],0.5]preferably 0.25, in mm2。
Wherein the "area of at least one reference signal image S2"the area of each reference signal image is S2。
And if no reference image data matched with the white matter high signal position in the first image data exists, screening out a reference signal image with the distance between the geometric center and the geometric center of the first signal image larger than a second variable in the database, and repeating the first method and the second method. And if the number of times of repeating the k is larger than the number of the reference image data, outputting a signal with a query result of 0 to the user.
For example, FIG. 5 is a first signal image that has been placed in a 4 x 4 grid containing only contours of white matter high signals in the first image data, FIGS. 6 to 9 are 4 reference signal images which are placed in a 4 × 4 grid and only contain the contour of the white matter high signal in the reference image data matched with the first image data, the grid coordinates of the geometric center of the first signal image are (2, 3), the grid coordinates of the geometric centers of the 4 reference signal images are (2, 3), (3, 2), (1, 1), the geometric center of the first signal image is located a small distance from the geometric center of the reference image data in fig. 6 and 7, and is smaller than the first variable, determining that the position of the white matter high signal in the first signal image matches the position of the white matter high signal in the reference image data of fig. 6 and 7;
referring to fig. 10, the processor calculates an area S of the first signal image of fig. 51Area S of reference signal image in FIG. 62And calculating the intersection area J in FIG. 10 as S1∩S2；
Referring to fig. 11, the processor calculates an area S of the first signal image of fig. 51Area S of reference signal image in FIG. 72And calculating the intersection area J in FIG. 11 as S1∩S2；
The processor determines the intersection area J and the area S of the reference signal image in FIG. 6 or FIG. 72The absolute values of the differences are all smaller than a first preset threshold valueThe first step is carried out once, and k is 1;
the first signal image and the reference signal image in fig. 6 are respectively placed in a plurality of (m +1) × (n +1) grids, i.e., 5 × 5 grids, according to the corresponding positions, see fig. 12;
the first signal image and the reference signal image in fig. 7 are respectively placed in a plurality of (m +1) × (n +1) grids, i.e., 5 × 5 grids, according to the corresponding positions, see fig. 13;
the processor determines that the distances between the geometric center of the first signal image in fig. 5 and the geometric centers of the reference signal images in fig. 6 and 7 are both less than or equal to a second variable, i.e., less than
Determining that a white matter high signal in the first image data matches a location of a white matter high signal in the reference image data; the processor determines the area of the intersection J and the area S of the reference signal image in FIG. 6 or FIG. 72If the absolute values of the differences are all smaller than a first preset threshold, the first step is repeated for the second time, k is 22The absolute value of the difference is less than a first preset threshold, for example, referring to fig. 14 and 15, when the first step is repeated 12 th time, i.e., when the grid is divided into (m +12) × (n +12), i.e., 16 × 16 grids, the distance between the geometric center of the first signal image in fig. 5 and the geometric center of the reference signal image in fig. 6 is less than a first variable
The distance between the geometric center of the first signal image in fig. 5 and the geometric center of the reference signal image in fig. 7 is larger than the first variable
The reference signal image in fig. 7 is screened out and the subsequent steps are performed. It is only assumed here that comparing the reference signal images in fig. 6 and 7, in practice, there may be a plurality of reference information images compared with the first signal image.
As a variant of the invention, with reference to fig. 1, 10, 11, 12, 13, 14, 15, the third method is: the processor calculates an area S of the first signal image1Area S of at least one reference signal image2；
The processor calculates an area S of the first signal image1Area S of at least one reference signal image2The complete set of (a) is Q;
area S of the first signal image1Area S with respect to at least one reference signal image2Complementary collection area B of1＝CQS1；
Area S of reference signal image2Area S with respect to first signal image1Complementary collection area B of2＝CQS2；
The processor determines the area S of the reference signal image when only one reference signal image is present2Area S with respect to first signal image1Complementary collection area B of2When the first signal image is smaller than a second preset threshold value, judging that the first signal image is matched with only one reference signal image, and the first area data of the white matter high signal in the first image data is matched with the reference area data of the white matter high signal in only one reference image data;
the processor determines the area S of at least one reference signal image2Area S with respect to first signal image1Complementary collection area B of2When the area is smaller than a second preset threshold value, the first step is repeated for k times until the area S of only one reference signal image is larger than the second preset threshold value2Area S with respect to first signal image1Complementary collection area B of2Less than a second preset threshold:
the first step is: adjusting the m multiplied by n grids to (m + k) multiplied by (n + k) grids with the same total area, judging whether the distances between the geometric center of the first signal image and the geometric center of the at least one reference signal image are smaller than or equal to a second variable, if so, judging that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the at least one reference signal image, and if not, judging that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the at least one reference signal imageIf not, screening out a reference signal image of which the distance between the geometric center and the geometric center of the first signal image is greater than a second variable, and judging that the white matter high signal in the first image data is matched with the white matter high signal in the residual reference image data in position; the processor determines whether only one reference signal image area S matches a white matter high signal position in the first image data2Area S with respect to first signal image1Complementary collection area B of2If the value is smaller than the second preset threshold value, the first step is ended, otherwise, the first step is repeated, and the value of k is added by 1.
According to the third method, firstly, the area of the first signal image and the area of the reference image data left after screening by the first method and the second method are calculated, a complementary set is taken, when the area of the first signal image and the area of one reference image data only meet the related complementary set relationship, only one reference image data is left, and the first signal image is judged to be matched with the area of the reference image data. If the judgment is not satisfied, the area of the reference image data satisfies the relevant complement relationship, the grid is refined and then returned to the second method, and the reference image data is screened again by judging whether the geometric center is close enough or not until the area of the remaining reference image data satisfies the relevant complement relationship, so that the final reference image data which is most matched with the first signal image is screened out.
Wherein the complementary area B2In mm unit2。
Wherein the second preset threshold is [0, 0.5 ]]Preferably 0.25, in mm2。
As a further explanation of the present invention, referring to fig. 1, the step of "outputting corresponding reference score data and deriving white matter high signal load condition" comprises:
a white matter high signal mapping relation table is prestored in the database;
and the processor outputs corresponding white matter high signal scores and load conditions according to the white matter high signal mapping relation table and by combining the area S2 of the reference signal image.
According to the method, the white matter high signal score and the load condition in the first image data to be detected can be output, so that a doctor can conveniently perform subsequent diagnosis and treatment, the evaluation efficiency and accuracy are improved, and misjudgment caused by human factors is avoided.
Wherein the white matter high signal mapping relation table can evaluate the degree and range of white matter looseness, wherein the corresponding relation of the area of the white matter high signal and the white matter high signal score is contained, and in the third method, the area S of the first signal image is used1Area S of reference signal image2Matching is performed, and the area S of the reference signal image which is screened and matched2It is known that the corresponding white matter high signal score and load condition can be found through the white matter high signal mapping relation table.
The white matter high signal mapping relation table can adopt a Fazekas scale (0-6 points), a modified Scheltens scale (0-30 points), a Ylikoski scale (0-48 points), a Manolio scale (0-9 points) and the like, which are all evaluation scales based on MRI image examination, and are prior art and are not described herein.
For example, a modified Scheltens scale (0-30 points) as shown in Table 1 can be used, which scores paraventricular and deep lesions (white matter high signals) separately. Since the first signal image of the white matter high signal contour in the first image data to be detected is matched with the reference signal image of the white matter high signal contour in certain reference image data, and the shape, position and area of the reference signal image of the latter are known, the corresponding white matter high signal score and load condition can be found out by combining the improved Scheltens scale according to the known information.
It should be noted that, what is found out in the above manner is only the corresponding white matter high signal score and load condition, which is only used as a reference for a doctor to perform a subsequent diagnosis and treatment, and not a disease diagnosis and treatment result.
TABLE 1
As a further explanation of the invention, the processor is electrically connected with the image preprocessing module and the computing module.
Through the arrangement, the processor can be electrically connected with the image preprocessing module and the computing module, so that signal transmission is realized.
As a further explanation of the invention, the first gray scale value and the second gray scale value range from 0 to 63.
According to the invention, through the arrangement, the calculation is more convenient under the condition of the same effect, and the calculation efficiency is improved.
As a further explanation of the invention, the white matter high signal mapping relation table adopts a modified Scheltens scale.
The invention adopts the improved Scheltens scale as the white matter high signal mapping relation table, and realizes the following advantages: the reliability and the effectiveness are good, when the leukoencephalopathy is evaluated, the lesion size and the lesion part are included, the number of deep leukoencephalopathy variables is also evaluated, the deep leukoencephalopathy variables have good correlation with the area of the leukoencephalopathy, and the deep leukoencephalopathy is more sensitive to longitudinally observing the progress of the leukoencephalopathy and the relation between the progress and the clinical manifestation of the leukoencephalopathy.
The invention relates to a white matter high signal detection method, which is shown in figure 1 and stores at least one reference image data and reference area data of different white matter high signals in the reference image data;
preprocessing MRI image data and outputting first image data;
calculating first area data of white matter high signals in the first image data;
and judging whether the first image data is preliminarily matched with at least one reference image data, if so, judging whether the positions of the white matter high signals in the first image data are matched with the positions of the different white matter high signals in the reference image data, if so, judging whether the first area data in the first image data are matched with the reference area data of the different white matter high signals in the reference image data, if so, outputting corresponding reference scoring data, and obtaining the white matter high signal load condition.
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solution of the present invention by those skilled in the art should fall within the protection scope defined by the claims of the present invention without departing from the spirit of the present invention.
Claims (10)
1. A white matter high signal detection system, characterized by: comprises that
A database in which at least one reference image data and reference area data of different white matter high signals in the reference image data are prestored;
the image preprocessing module is used for preprocessing the MRI image data and outputting first image data;
a calculation module for calculating first area data of white matter high signals in the first image data;
and the processor judges whether the first image data is preliminarily matched with at least one reference image data through a first method, judges whether the positions of the white matter high signals in the first image data are matched with the positions of the different white matter high signals in the reference image data through a second method if the first image data is preliminarily matched, judges whether the first area data in the first image data are matched with the reference area data of the different white matter high signals in the reference image data through a third method if the first image data are matched with the reference area data of the different white matter high signals in the reference image data, outputs corresponding reference scoring data if the first image data are matched with the reference area data, and obtains the white matter high signal load condition.
2. The white matter high signal detection system of claim 1, wherein:
the first method comprises the following steps: the processor scales at least one piece of reference image data to a preset resolution and generates a reference scaled image, scales a first piece of image data to the preset resolution and generates a contrast scaled image, and the processor calculates the reference scaled image and the average gray value of all pixels of the contrast scaled image, and compares the first gray value and the average gray value of each pixel in the at least one piece of reference scaled image and the second gray value and the average gray value of each pixel in the contrast scaled image;
the processor calculates the absolute value of the difference value between a plurality of first pixel values a and a plurality of second pixel values b according to a first pixel value a of which the first gray value in at least one reference scaling image is greater than or equal to the gray average value and a second pixel value b of which the second gray value in the contrast scaling image is greater than or equal to the gray average value, and judges that at least one reference scaling image corresponding to the minimum value of the absolute values of the difference values between the contrast scaling image and the first pixel values a and the second pixel values b is matched, and first image data corresponding to the contrast scaling image is preliminarily matched with at least one corresponding reference image data.
3. A white matter high signal detection system according to claim 2, characterized in that:
the second method comprises the following steps: the processor carries out outline delineation of white matter high signals on first image data and at least one reference image data preliminarily matched with the first image data to respectively obtain a first signal image only containing the white matter high signal outline in the first image data and at least one reference signal image of the white matter high signal outline in the reference image data preliminarily matched with the first image data, the processor scales the first signal image and the at least one reference signal image to be matched with the size of an m multiplied by n grid, and the first signal image and the at least one reference signal image are respectively placed in the m multiplied by n grid according to corresponding positions;
the processor judges whether the distance between the geometric center g of the first signal image and the geometric center o of at least one reference signal image is smaller than or equal to a first variable, if not, the reference zoom image corresponding to the minimum value of the absolute value of the difference value of the first pixel value a and the second pixel value b is removed, and the first method is repeated;
and if so, determining that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the at least one reference image data.
4. A white matter high signal detection system according to claim 3, characterized in that:
the third method comprises the following steps: the processor calculates an area S of the first signal image1Area S of at least one reference signal image2；
The processor calculates an area S of the first signal image1Area S of reference signal image2The intersection area J is S1∩S2；
The processor judges that when the intersection area J is equal to the area S of only one reference signal image2When the absolute value of the difference is smaller than a first preset threshold value, judging that the first signal image is matched with only one reference signal image, and matching first area data of a white matter high signal in the first image data with reference area data of a white matter high signal in only one reference image data;
the processor judges the intersection area J and the area S of at least one reference signal image2When the absolute value of the difference is less than a first preset threshold, repeating the first step k times until the intersection area J and the area S of only one reference signal image are obtained2The absolute value of the difference is smaller than a first preset threshold;
the first step is: adjusting the m multiplied by n grids to (m + k) x (n + k) grids with the same total area, judging whether the distances between the geometric center of the first signal image and the geometric center of the at least one reference signal image are smaller than or equal to a second variable, if so, judging that the position of a white matter high signal in the first image data is matched with the position of a white matter high signal in the at least one reference signal image, if not, screening out the reference signal image with the distance between the geometric center of the first signal image and the geometric center of the first signal image larger than the second variable, and judging that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the rest reference image data; the processor determines whether the intersection area J intersects white in the first image data with only oneArea S of reference signal image with high-quality signal position matching2And if not, repeating the first step, and adding 1 to the value of k.
5. A white matter high signal detection system according to claim 3, characterized in that:
the third method comprises the following steps: the processor calculates an area S of the first signal image1Area S of at least one reference signal image2；
The processor calculates an area S of the first signal image1Area S of at least one reference signal image2The complete set of (a) is Q;
area S of the first signal image1Area S with respect to at least one reference signal image2Complementary collection area B of1＝CQS1；
Area S of reference signal image2Area S with respect to first signal image1Complementary collection area B of2＝CQS2；
The processor determines the area S of the reference signal image when only one reference signal image is present2Area S with respect to first signal image1Complementary collection area B of2When the first signal image is smaller than a second preset threshold value, judging that the first signal image is matched with only one reference signal image, and the first area data of the white matter high signal in the first image data is matched with the reference area data of the white matter high signal in only one reference image data;
the processor determines the area S of at least one reference signal image2Area S with respect to first signal image1Complementary collection area B of2When the area is smaller than a second preset threshold value, the first step is repeated for k times until the area S of only one reference signal image is larger than the second preset threshold value2Area S with respect to first signal image1Complementary collection area B of2Less than a second preset threshold:
the first step is: adjusting the m x n grid to be general to itThe processor judges whether the distances between the geometric center of the first signal image and the geometric center of at least one reference signal image are smaller than or equal to a second variable, if so, the processor judges that the position of a white matter high signal in the first image data is matched with the position of a white matter high signal in at least one reference signal image, if not, the processor screens out the reference signal image of which the distance between the geometric center of the first signal image and the geometric center of the first signal image is larger than the second variable, and judges that the position of the white matter high signal in the first image data is matched with the position of the white matter high signal in the rest of reference image data; the processor determines whether only one reference signal image area S matches a white matter high signal position in the first image data2Area S with respect to first signal image1Complementary collection area B of2If the value is smaller than the second preset threshold value, the first step is ended, otherwise, the first step is repeated, and the value of k is added by 1.
6. A white matter high signal detection system according to claim 4 or 5, characterized in that:
the step of outputting corresponding reference scoring data and obtaining the white matter high signal load condition comprises the following steps:
a white matter high signal mapping relation table is prestored in the database;
and the processor outputs corresponding white matter high signal scores and load conditions according to the white matter high signal mapping relation table and by combining the area S2 of the reference signal image.
7. The white matter high signal detection system of claim 6, wherein:
the processor is electrically connected with the image preprocessing module and the computing module.
8. The white matter high signal detection system of claim 7, wherein:
the range of the first gray value and the second gray value is 0-63.
9. The white matter high signal detection system of claim 8, wherein:
the white matter high signal mapping relation table adopts an improved Scheltens scale.
10. The method for analyzing a white matter high signal detection system according to claims 1 to 9, comprising the steps of:
storing at least one reference image data, reference area data of different white matter high signals in the reference image data;
preprocessing MRI image data and outputting first image data;
calculating first area data of white matter high signals in the first image data;
and judging whether the first image data is preliminarily matched with at least one reference image data, if so, judging whether the positions of the white matter high signals in the first image data are matched with the positions of the different white matter high signals in the reference image data, if so, judging whether the first area data in the first image data are matched with the reference area data of the different white matter high signals in the reference image data, if so, outputting corresponding reference scoring data, and obtaining the white matter high signal load condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210005119.2A CN114332067A (en) | 2022-01-04 | 2022-01-04 | White matter high signal detection system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210005119.2A CN114332067A (en) | 2022-01-04 | 2022-01-04 | White matter high signal detection system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114332067A true CN114332067A (en) | 2022-04-12 |
Family
ID=81024074
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210005119.2A Pending CN114332067A (en) | 2022-01-04 | 2022-01-04 | White matter high signal detection system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114332067A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109859215A (en) * | 2019-01-30 | 2019-06-07 | 北京慧脑云计算有限公司 | A kind of automatic segmenting system of white matter high signal intensity based on Unet model and its method |
| CN109886944A (en) * | 2019-02-02 | 2019-06-14 | 浙江大学 | A kind of white matter high signal intensity detection and localization method based on multichannel chromatogram |
| US20200302602A1 (en) * | 2019-03-22 | 2020-09-24 | Nehal PARIKH | Objective analysis of medical image data for white matter abnormality |
| KR102300228B1 (en) * | 2020-12-30 | 2021-09-10 | 뉴로핏 주식회사 | A method for providing disease information and device performing the same |
| CN113724243A (en) * | 2021-09-14 | 2021-11-30 | 首都医科大学附属北京天坛医院 | Image processing method, image processing device, electronic equipment and storage medium |
-
2022
- 2022-01-04 CN CN202210005119.2A patent/CN114332067A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109859215A (en) * | 2019-01-30 | 2019-06-07 | 北京慧脑云计算有限公司 | A kind of automatic segmenting system of white matter high signal intensity based on Unet model and its method |
| CN109886944A (en) * | 2019-02-02 | 2019-06-14 | 浙江大学 | A kind of white matter high signal intensity detection and localization method based on multichannel chromatogram |
| US20200302602A1 (en) * | 2019-03-22 | 2020-09-24 | Nehal PARIKH | Objective analysis of medical image data for white matter abnormality |
| KR102300228B1 (en) * | 2020-12-30 | 2021-09-10 | 뉴로핏 주식회사 | A method for providing disease information and device performing the same |
| CN113724243A (en) * | 2021-09-14 | 2021-11-30 | 首都医科大学附属北京天坛医院 | Image processing method, image processing device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110276356B (en) | Fundus image microaneurysm identification method based on R-CNN | |
| CN105513077B (en) | A kind of system for diabetic retinopathy screening | |
| US9076238B2 (en) | Intelligent weighted blending for ultrasound image stitching | |
| JP2004032684A (en) | Automated method and apparatus for detecting mass or substantial tissue deformation in medical image using computer | |
| CN110288698B (en) | Meniscus three-dimensional reconstruction system based on MRI | |
| KR102206621B1 (en) | Programs and applications for sarcopenia analysis using deep learning algorithms | |
| Li et al. | Vessel recognition of retinal fundus images based on fully convolutional network | |
| Hatanaka et al. | Improvement of automatic hemorrhage detection methods using brightness correction on fundus images | |
| Hatanaka et al. | Improvement of automated detection method of hemorrhages in fundus images | |
| Hatanaka et al. | CAD scheme to detect hemorrhages and exudates in ocular fundus images | |
| JP3195136B2 (en) | Image processing device | |
| WO2011050454A1 (en) | Bone imagery segmentation method and apparatus | |
| Ou et al. | Automatic radiographic quantification of joint space narrowing progression in rheumatoid arthritis using POC | |
| Sagar et al. | A novel integrated approach using dynamic thresholding and edge detection (idted) for automatic detection of exudates in digital fundus retinal images | |
| Wisaeng et al. | Automatic detection of exudates in retinal images based on threshold moving average models | |
| Kawathekar et al. | Severity analysis of Osteoarthritis of knee joint from X-ray images: A Literature review | |
| Kusuma et al. | Retracted: Heart Abnormalities Detection Through Iris Based on Mobile | |
| EP3977400A1 (en) | A computer based method for classifying a mass of an organ as a cyst | |
| CN114332067A (en) | White matter high signal detection system and method | |
| Septiarini et al. | Peripapillary atrophy detection in fundus images based on sectors with scan lines approach | |
| KR20210050790A (en) | Apparatus and methods for classifying neurodegenerative diseases image of amyloid-positive based on deep-learning | |
| CN112991289B (en) | Processing method and device for standard section of image | |
| CN111292285A (en) | Automatic screening method for diabetes mellitus based on naive Bayes and support vector machine | |
| CN115294329A (en) | Focus image identification method based on deep learning model | |
| Cao et al. | Liver fibrosis identification based on ultrasound images |
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 |