CN101021946A - Image processing method for screening esophageal micropathology - Google Patents
Image processing method for screening esophageal micropathology Download PDFInfo
- Publication number
- CN101021946A CN101021946A CNA2006100227294A CN200610022729A CN101021946A CN 101021946 A CN101021946 A CN 101021946A CN A2006100227294 A CNA2006100227294 A CN A2006100227294A CN 200610022729 A CN200610022729 A CN 200610022729A CN 101021946 A CN101021946 A CN 101021946A
- Authority
- CN
- China
- Prior art keywords
- value
- pixel
- gray
- vibration frequency
- average
- 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
Abstract
This invention relates to a medical image process technology applying an image process method of filtering tiny pathological changes of oesophaguses including the following steps: a, reading in an oesophagus picture to be analyzed, b, determining the image analysis sphere, c, picking up the R, G and B values of each pixel point in the sphere, d, computing the hue value, saturation value and brightness value of each point, computing the mean hue value and mean saturation value of the sphere to select the pixel points higher than the mean hue and saturation values obviously as the pathological changes point, f, displaying the picture and displaying the changes point especially.
Description
Technical field
Patent of the present invention relates to the medical imaging treatment technology.
Background technology
Common optics or electronic gastroscopy have rapidly and efficiently, look at straight accurately location, can carry out preliminary examination to esophageal micropathology (comprising the early stage cancer of the esophagus), are to diagnose feasible, the most effective inspection method of esophageal micropathology at present clinically.But, cause in the checking process particularly for some malignant tumour pathologies,, in identification, having big difficulty as the cancer of the esophagus to there being certain difficulty on the quick identification of minute lesion because people's naked eyes are differentiated the limitation of ability to color, degree of roughness etc.
These pathologies if can early detection reach early stage treatment, and the patient is effected a radical cure and long-term surviving fully.Simultaneously, because China is populous and the relatively limited property of medical resource, the patient's quantity that will carry out endoscopy and generaI investigation in certain working time is very big, can not be by reducing workload, sacrificing the diagnosis that work efficiency improves esophageal micropathology, this just must cause the rate of missed diagnosis of esophageal micropathology (comprising the early stage cancer of the esophagus) to increase.
Therefore, seek and a kind ofly can just seem very important the method that minute lesion is carried out rapid screening.It can play the effect of auxiliary diagnosis to inspecting doctor's visual inspection, helps improving the diagnosis of minute lesion.
Image texture characteristic is analyzed, and color and luster mutation analysis etc. all is ripe image processing techniques.Utilize the Computer Analysis picture, have fast operation, can discern the feature of some pathology to a certain extent.
Realize many researchs had been carried out in automatic diagnosis that may pathology by computer image processing technology, but focus mostly in pathological section, X-ray plain film or CT, MRI sheet interpretation, analyze less to the interpretation of oesophagus image under the scope.Present domestic with computer image processing technology, comprise that the research that the greyscale transformation technology is carried out auxiliary diagnosis to the oesophagus pathology mainly concentrates on two aspects: 1, to the graphical analysis of oesophagus pathology pathological biopsy sample diagnosis; 2, X-ray sheet result is carried out the image analysing computer diagnosis.
Abroad have by color and cut apart and the crooked report that endoscopic images is carried out analyzing and processing such as calculate, and the report of quantitatively esophagitis being diagnosed by color.Still the oesophagus pathology is not carried out the report of express-analysis examination.
Summary of the invention
Technical matters to be solved by this invention is to provide a kind of Computer Image Processing method, the preliminary automatic identification that realizes the part esophageal micropathology.
The present invention solves the problems of the technologies described above the technical scheme that is adopted to be, a kind of image processing method that screens esophageal micropathology may further comprise the steps:
A, read in oesophagus picture to be analyzed;
B, determine the graphical analysis scope;
C, R value, G value, the B value of taking out each pixel in the analyst coverage;
Tone value, intensity value, the brightness value of each pixel in d, the computational analysis scope;
The average color tone pitch of computational analysis scope, average staturation value; Select apparently higher than the pixel of average color tone pitch or average staturation value, be the pathology point;
Each gray values of pixel points vibration frequency in e, the computational analysis scope;
The average gray value vibration frequency of computational analysis scope; Select the pixel of gray-scale value vibration frequency, be the pathology point apparently higher than the average gray value vibration frequency;
F, Show Picture, and highlight the pathology point.
The invention has the beneficial effects as follows that by computer image processing technology, energy express-analysis pathology and normal esophageal tissue identify out with pathology and edge thereof, and be not limited to the concrete oesophagus pathology of certain class.In the esophagus detecting of gastroscope, can play real-time supplementary function to the naked eyes diagnosis of present oesophagus pathology.To the various pathologies of oesophagus, comprise that normal esophageal mucous membrane, esophagitis, the early stage cancer of the esophagus, advanced esophageal carcinoma, oesophagus mucous membrane atypical hyperplasia etc. all have the value of auxiliary diagnosis.
Description of drawings
Fig. 1 is the analyzed area synoptic diagram of algorithm 1 among the embodiment;
Fig. 2 is the synoptic diagram of pixel in the analyst coverage of algorithm 2 among the embodiment.
Embodiment
Because the BMP form can directly carry out graphical analysis, does not need format conversion, so adopt the image of 256 true color BMP forms in the present invention, analyzes.
The normal structure of biosome organ all has the height consistance on structure and function, the color, compactness, smoothness etc. of exhibit tissue have similar physical characteristics in appearance at naked eyes.Normal mucosa is organized in the image, and R, G, B ratio rate of change is less or constant substantially, also should be less or constant substantially and convert the rate of change of the tone value (H) that forms, intensity value (S), brightness value (I) by R, G, B.
Minute lesion then shows as diseased region mucous membrane color and luster, roughness changes:
1. diseased region mucous membrane color and luster changes, and the tone value, the intensity value that mainly show as diseased region depart from average color tone pitch, the average staturation value in zone.
2. the diseased region roughness changes, and mainly shows as neighbor pixel gray-scale value vibration frequency increase in the diseased region.
Pixel in detecting picture satisfies any situation in above-mentioned two points, then is considered as the pathology point.Change or roughening because pathology shows as color, that is: the coarse color that not necessarily has of pathology changes, and is same, and pathology the color change occurs and also not necessarily produces roughening.Color (the red intensification) appears in the inflammatory disorders district as the oesophagus inflammation, but not coarse.The color change of oesophagus atypical hyperplasia diseased region is little, and mainly is that roughening changes.
Because oesophagus is the hollow tube chamber, tube wall is around circular being positioned at, and central authorities do not have mucous membrane to exist, and the characteristics of image that calculates central space is nonsensical.Therefore the scope of analyzing should be defined in the annular region of image border a+b+c circle ring area as shown in Figure 1.
After having determined analyst coverage, adopt algorithm in following two, the point of the pathology in the analyst coverage screened:
Algorithm 1: be used to screen the pathology point that the mucous membrane color and luster changes
(1) tone value, the intensity value of each pixel in the computational analysis scope;
The average color tone pitch of (2) computational analysis scope, average staturation value; Select apparently higher than the pixel of average color tone pitch or average staturation value, be the pathology point.
The image that the process scope collects, the mucous membrane color and luster is by R, G, the decision of B value of each pixel, the difference rather than the mean value that promptly will compare three values respectively, promptly the brightness value I in the algorithm 1 (average gray I=(R+G+B)/3) index is insensitive, need not so give up.
Because R, G, B fluctuation can be very big, but its average gray value may fluctuate not quite, so insensitive.Show that at R, G, B color difference is very big as point (100,100,100) and point (50,200,50), but average gray value is the same, is 100.
Idiographic flow is as follows:
Read in 256 true color (0～255) BMP format-pattern;
Determine first analyst coverage;
Three color component R, G of each pixel in first analyst coverage, B through following formula, are converted into tone value (H), intensity value (S), brightness value (I):
I＝(R+G+B)/3；
As R＞B≤G H=(G-B)/(3* (I-B)); S=1-B/I;
As G＞R≤B H=(B-R)/(3* (I-R)); S=1-R/I;
As B＞G≤R H=(R-G)/(3* (I-G)); S=1-G/I;
In order to get rid of the inhomogenous interference of light source irradiation, it is not that pathology causes that the tone value that promptly has, intensity value change, but due to the light source irradiation heterogeneity.As shown in Figure 1, analyst coverage is divided into several littler zones, promptly analyst coverage is divided into first analyst coverage, second analyst coverage; First analyst coverage is whole annulus, i.e. the a+b+c zone; Second analyst coverage is the subregion in the a+b+c zone.
First analyst coverage is divided into 3 concentric rings, cuts apart annulus by the angle of π/6 then, form a1-a12 subregion in a zone, b1-b12 subregion in the b zone, c1-c12 subregion in the c zone, totally 36 little fan rings.Each little fan ring is as second analyst coverage at pixel place.
Calculate average color tone pitch, the average staturation value of whole annulus; The average color tone pitch of each little fan ring, average staturation value;
First analyst coverage at the tone value of compared pixels point, intensity value and its place and the average color tone pitch of second analyst coverage, average staturation value;
The tone value of compared pixels point or intensity value must satisfy simultaneously apparently higher than the average color tone pitch or the average staturation value of this place first analyst coverage and second analyst coverage just being regarded as the pathology point.
Can draw by experiment, tone value or intensity value just can better must filter out the pathology point greater than its mean value more than 30%, and be wherein more accurate more than 50% greater than average.
And the pathology point screened, show with other normal pixel point difference.Pathology is put available scintillation fluor, and diseased region can identify with profile, and these two kinds of ways relatively expend system time, influence result's display speed.Preferably, give the special R of pathology point, G, B value, as:
R, G, the B value of giving the pathology point of tonal anomaly are (255,0,0), and such pathology point adopts red display; R, G, the B value of giving the pathology point of saturation anomaly are (0,255,0), and such pathology point adopts green the demonstration.
Evidence, this method to light, in, after the picture of severe esophagitis analyzes, can be presented at the pathology in the analyst coverage and the juncture area of normal structure preferably.Analysis result shows along inflammation mucous membrane edge a bit or sheet saturation anomaly zone and point or sheet tonal anomaly zone, for the diseased region of esophagitis takes place.Can distinguish the esophagitis of the different orders of severity generally.After the picture of the early, middle and late phase cancer of the esophagus analyzed, the juncture area of lesions showed zone and normal structure preferably, analysis result show to be had a sheet saturation anomaly zone and puts sheet tonal anomaly zone along pathology mucous membrane edge.For advanced esophageal carcinoma, can demonstrate and seem to be normal pathology edge, this helps determining in pathological biopsy under the gastroscope and the operation excision extension of pathology.In the regional biopsy of H, S display abnormality, whether the decidable cancerous tissue has potential infiltration growth tendency, and also the decidable junctional area is organized the situation of atypical hyperplasia.
Algorithm 2: be used to screen the pathology point that roughness changes
The variation of diseased region roughness is embodied by the increase of neighbor pixel gray-scale value vibration frequency.The gray-scale value vibration frequency is by 2 decisions: 1, the fluctuation of neighbor point gray-scale value increases; 2, the pixel distance of gray-scale value fluctuation increase is near.
By above-mentioned 2 judgements, can filter out the pathology point preferably, reject the pixel that causes the gray-scale value fluctuation merely because of error.
The gray-scale value fluctuation is that difference is big more by the difference decision of the gray-scale value of neighbor pixel, and it is big more to fluctuate.Gray values of pixel points determines that by its brightness value present embodiment adopts average gray value at this, calculates: I=(R+G+B)/3.
For normal structure, horizontal or vertical continuous 2 difference generally is no more than 2, even surpass 2, does not also have a plurality of points and surpasses 2 simultaneously.And diseased region, horizontal or vertically continuous 2 difference not only surpasses 2, and has a plurality of points to surpass 2 simultaneously.The standard of difference is made as 2, is to be drawn by experiment, and promptly this value gets final product greater than the difference of continuous 2 pixels in the normal tissue regions.
Idiographic flow is as follows:
Read in 256 true color (0～255) BMP format-pattern;
Determine analyst coverage;
Calculate the average gray of each pixel successively along horizontal (X-axis) and vertical (Y-axis), I=(R+G+B)/3;
Each gray values of pixel points vibration frequency is calculated by following steps:
(1) calculates the laterally difference of average gray between each neighbor pixel, select difference greater than 2 point;
(2) to the pixel in analyzed area assignment again, difference is 10 greater than 2 pixel initialize, and other pixel assignment is 0, and promptly the gray-scale value vibration frequency is 0; , as | I
XnYn-I
XnYn+1|＞2, then XnYn point initialize is 10;
(3) difference recomputates the gray-scale value vibration frequency greater than 2 pixel, for:
Initial value+∑ (be expert in the difference of assignment greater than the initial value/distance of 2 pixel);
(4) the horizontal mean value of all gray values of pixel points vibration frequencies in the computational analysis scope;
(5) the gray-scale value vibration frequency is the pathology point apparently higher than the pixel of horizontal mean value;
(6) difference of calculating vertical each neighbor pixel is selected difference greater than 2 point;
(7) to the pixel in analyzed area assignment again, difference is 10 greater than 2 pixel initialize, and other pixel assignment is 0, and promptly the gray-scale value vibration frequency is 0;
(8) difference recomputates the gray-scale value vibration frequency greater than 2 pixel, is: initial value+∑ (in the column difference of assignment greater than the initial value/distance of 2 pixel);
(9) the horizontal mean value of all gray values of pixel points vibration frequencies in the computational analysis scope;
(10) the gray-scale value vibration frequency is the pathology point apparently higher than the pixel of vertical mean value;
The gray-scale value vibration frequency is considered as the pathology point apparently higher than the pixel of horizontal or vertical arbitrary mean value.
Can draw by experiment, described apparently higher than, be the gray-scale value vibration frequency and just can better must filter out the pathology point more than 30% greater than mean value, wherein more accurate more than 50% greater than mean.
As Fig. 2 is example, and analyst coverage is horizontal X1-X15, vertical Y1-Y9, totally 135 pixels.Difference is a white greater than 2 point, and other point is a black, and the some gray-scale value vibration frequency assignment of black is 0.Assignment from left to right successively.
Laterally to be example, the gray-scale value vibration frequency of white each point is:
X2(Y2-Y8)：10
X3(Y2-Y8)：10+10/1＝20
X4(Y2-Y8)：10+10/1+10/2＝25
X5(Y2-Y8)：10+10/1+10/2+10/4＝27.5
X6(Y2-Y8)：10+10/1+10/2+10/4+10/5＝29.5
X7(Y2-Y8)：10+10/1+10/2+10/4+10/5+10/6＝31.1
X8(Y2-Y8)：10+10/1+10/2+10/4+10/5+10/6+10/7＝32.5
X14?Y9：10
All gray values of pixel points vibration frequencies are: (10+20+25+27.5+29.5+31.1+32.5) * 7=1229.2; The horizontal mean value of all gray values of pixel points vibration frequencies: 1229.2/135=9.1
Filter out 50% pixel greater than mean value, promptly the gray-scale value vibration frequency promptly filters out X3 (Y2-Y8), X4 (Y2-Y8), X5 (Y2-Y8), X6 (Y2-Y8), X7 (Y2-Y8), X7 (Y2-Y8) greater than 13.5 pixel, is the pathology point.
Effectively reject the pixel that causes the gray-scale value fluctuation merely because of error, as pixel X14 Y9.
Vertical algorithm is the same, highlights the disease point at last.
R, G, the B value of giving the pathology point of tonal anomaly are (255,0,255), and such pathology point adopts purple to show; R, G, the B value of giving the pathology point of saturation anomaly are (0,0,255), and such pathology point adopts blue the demonstration.
This algorithm combines the advantage of spatial autocorrelation algorithm and neighborhood characteristics statistic algorithm, has avoided loaded down with trivial details loop computation, helps realtime graphic and handles.Present embodiment has been considered the gray-scale value vibration frequency along X-axis, Y-axis, can certainly handle the gray scale fluctuation of other angle.
In the oesophagus mucous membrane, the severe atypical hyperplasia is a kind of minute lesion that easily transforms to canceration, and what have itself is exactly precancerous lesion.But this type of pathology is not only rare, and the naked eyes of pathology are seen and normal mucous membrane difference is little, so easily fail to pinpoint a disease in diagnosis.After the method for the invention is analyzed, show that violet region and blue region are the atypical hyperplasia mucous membrane in the analyst coverage, show as a sheet hyperplasia.Show that this method can know the coarse mucous membrane that shows atypical hyperplasia.
Endoscopic images is by the computing simultaneously of above-mentioned 2 kinds of algorithms, can differentiate and identify the various minute lesions of oesophagus quickly and efficiently, the different excessively color identification of complication accommodation comes out, these colors are different from the color of pathology itself, remarkable difference is also arranged mutually, therefore not only pathology can be demonstrated better, different pathology types can also be shown.
The above only is preferred embodiment of the present invention, does not therefore limit the protection domain and the usable range of this patent.
Claims (8)
1, a kind of image processing method that screens esophageal micropathology is characterized in that, may further comprise the steps:
A, read in oesophagus picture to be analyzed;
B, determine the graphical analysis scope;
C, R value, G value, the B value of taking out each pixel in the analyst coverage;
Tone value, intensity value, the brightness value of each pixel in d, the computational analysis scope;
The average color tone pitch of computational analysis scope, average staturation value;
Select the pixel of tone value, be the pathology point apparently higher than the average color tone pitch;
Select the pixel of intensity value, be the pathology point apparently higher than the average staturation value;
Each gray values of pixel points vibration frequency in e, the computational analysis scope;
The average gray value vibration frequency of computational analysis scope;
Select the pixel of gray-scale value vibration frequency, be the pathology point apparently higher than the average gray value vibration frequency;
F, Show Picture, and highlight the pathology point.
2, a kind of according to claim 1 image processing method that screens esophageal micropathology is characterized in that, analyst coverage described in the steps d comprises first analyst coverage, second analyst coverage, and described second analyst coverage is the subregion of first analyst coverage;
When the tone value of pixel average color tone pitch, be the pathology point apparently higher than its place first analyst coverage and second analyst coverage;
When the intensity value of pixel average staturation value, be the pathology point apparently higher than its place first analyst coverage and second analyst coverage.
3, a kind of as claimed in claim 1 or 2 image processing method that screens esophageal micropathology is characterized in that, each gray values of pixel points vibration frequency in the described computational analysis scope of step e specifically may further comprise the steps:
E1, calculate the difference of average gray between neighbor pixel, select the pixel of difference overgauge difference;
E2, to pixel assignment again, the pixel initialize of difference overgauge difference, other pixel assignment is 0, promptly the gray-scale value vibration frequency is 0;
The pixel of e3, difference overgauge difference calculates the gray-scale value vibration frequency, for:
Initial value+∑ (difference of assignment is greater than the initial value/distance of 2 pixel).
4, as a kind of image processing method that screens esophageal micropathology as described in the claim 3, it is characterized in that described standard difference is 2; Initial value is 10.
5, as a kind of image processing method that screens esophageal micropathology as described in the claim 3, it is characterized in that described step e is specially:
The horizontal gray-scale value vibration frequency of each pixel and vertical gray-scale value vibration frequency in the computational analysis scope;
The horizontal average gray value vibration frequency of computational analysis scope and vertical average gray value vibration frequency;
Select the pixel of horizontal gray-scale value vibration frequency, be the pathology point apparently higher than horizontal average gray value vibration frequency;
Select the pixel of vertical gray-scale value vibration frequency, be the pathology point apparently higher than vertical average gray value vibration frequency.
6, a kind of according to claim 1 image processing method that screens esophageal micropathology is characterized in that, in the steps d, described tone value is that tone value is greater than the pixel of average color tone pitch more than 30% apparently higher than the pixel of average color tone pitch;
Described intensity value is that intensity value is greater than the pixel of average staturation value more than 30% apparently higher than the average staturation value;
Among the step e, described gray-scale value vibration frequency apparently higher than the average gray value vibration frequency is, the gray-scale value vibration frequency is greater than the average gray value vibration frequency more than 30%.
7, a kind of according to claim 1 image processing method that screens esophageal micropathology is characterized in that, the described oesophagus picture of step a is 256 true color BMP format-patterns.
8, a kind of according to claim 1 image processing method that screens esophageal micropathology is characterized in that, step f, and the described pathology point that highlights is to give the color different with normal structure to the pathology point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100227294A CN100565582C (en) | 2006-12-31 | 2006-12-31 | A kind of image processing method that screens esophageal micropathology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100227294A CN100565582C (en) | 2006-12-31 | 2006-12-31 | A kind of image processing method that screens esophageal micropathology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101021946A true CN101021946A (en) | 2007-08-22 |
| CN100565582C CN100565582C (en) | 2009-12-02 |
Family
ID=38709702
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100227294A Expired - Fee Related CN100565582C (en) | 2006-12-31 | 2006-12-31 | A kind of image processing method that screens esophageal micropathology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100565582C (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103227684A (en) * | 2011-12-14 | 2013-07-31 | 波音公司 | Ambient signal identification |
| CN104246828A (en) * | 2012-02-23 | 2014-12-24 | 史密夫和内修有限公司 | Video endoscopic system |
| CN105705075A (en) * | 2013-10-28 | 2016-06-22 | 富士胶片株式会社 | Image processing device and operation method therefor |
| CN112089397A (en) * | 2020-07-05 | 2020-12-18 | 唐婧 | Inflammation analysis system using swallowable detection platform |
| CN113129356A (en) * | 2020-01-16 | 2021-07-16 | 安翰科技(武汉)股份有限公司 | Capsule endoscope system, image staining area recognition method thereof and computer-readable storage medium |
| CN115311268A (en) * | 2022-10-10 | 2022-11-08 | 武汉楚精灵医疗科技有限公司 | Esophagus endoscope image identification method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1331454A (en) * | 2000-06-24 | 2002-01-16 | 王建华 | Dynamic monitor method for digital medium image |
-
2006
- 2006-12-31 CN CNB2006100227294A patent/CN100565582C/en not_active Expired - Fee Related
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103227684A (en) * | 2011-12-14 | 2013-07-31 | 波音公司 | Ambient signal identification |
| CN103227684B (en) * | 2011-12-14 | 2017-05-17 | 波音公司 | Method for identifying test signal data central ambient signal data |
| CN104246828A (en) * | 2012-02-23 | 2014-12-24 | 史密夫和内修有限公司 | Video endoscopic system |
| US9715727B2 (en) | 2012-02-23 | 2017-07-25 | Smith & Nephew, Inc. | Video endoscopic system |
| CN104246828B (en) * | 2012-02-23 | 2018-11-23 | 史密夫和内修有限公司 | Video-endoscope system |
| US10783626B2 (en) | 2012-02-23 | 2020-09-22 | Smith & Nephew, Inc. | Video endoscopic system |
| CN105705075A (en) * | 2013-10-28 | 2016-06-22 | 富士胶片株式会社 | Image processing device and operation method therefor |
| CN113129356A (en) * | 2020-01-16 | 2021-07-16 | 安翰科技(武汉)股份有限公司 | Capsule endoscope system, image staining area recognition method thereof and computer-readable storage medium |
| CN112089397A (en) * | 2020-07-05 | 2020-12-18 | 唐婧 | Inflammation analysis system using swallowable detection platform |
| CN112089397B (en) * | 2020-07-05 | 2021-07-13 | 陈伟 | Inflammation analysis system using swallowable detection platform |
| CN115311268A (en) * | 2022-10-10 | 2022-11-08 | 武汉楚精灵医疗科技有限公司 | Esophagus endoscope image identification method and device |
| CN115311268B (en) * | 2022-10-10 | 2022-12-27 | 武汉楚精灵医疗科技有限公司 | Esophagus endoscope image identification method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100565582C (en) | 2009-12-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Roblyer et al. | Objective detection and delineation of oral neoplasia using autofluorescence imaging | |
| Inoue et al. | Magnification endoscopy in esophageal squamous cell carcinoma: a review of the intrapapillary capillary loop classification | |
| CN100565582C (en) | A kind of image processing method that screens esophageal micropathology | |
| EP1568307B1 (en) | Image processing device and image processing method | |
| US7907775B2 (en) | Image processing apparatus, image processing method and image processing program | |
| US8472682B2 (en) | Medical image processing | |
| CN102973231B (en) | Image processing apparatus and image processing method | |
| US8428320B2 (en) | Method of reconstituting cellular spectra useful for detecting cellular disorders | |
| JP2020519851A5 (en) | ||
| WO2013187148A1 (en) | Image processing device, microscope system, endoscope system, and image processing method | |
| JP2010113616A (en) | Apparatus, program and method for processing image | |
| WO2010042211A1 (en) | Methods for tissue classification in cervical imagery | |
| WO2006106509A2 (en) | Medical imaging method and system | |
| JP2004254742A (en) | Medical image processor and method of judging malignancy | |
| Yang et al. | Development of an integrated multimodal optical imaging system with real-time image analysis for the evaluation of oral premalignant lesions | |
| CN103218802B (en) | Image processing apparatus and image processing method | |
| TWI537762B (en) | Application of hyperfamily imaging to identification of cancerous lesions | |
| JP2010252895A (en) | Method of deriving threshold value for identifying melanonychia | |
| JP2005352571A (en) | Image processing device | |
| CN107529962B (en) | Image processing apparatus, image processing method, and recording medium | |
| Shi et al. | A CT-based radiomics nomogram model for differentiating primary malignant melanoma of the esophagus from esophageal squamous cell carcinoma | |
| JPH1119077A (en) | Method and device for detection of tumor shadow in radiation image | |
| Pallavi et al. | Automated analysis of cervix images to grade the severity of cancer | |
| KR20150059931A (en) | Diagnostic method of melanoma by the multi-wavelength imaging and image processing | |
| Meyer-Veit et al. | Hyperspectral Wavelength Analysis with U-Net for Larynx Cancer Detection. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| ASS | Succession or assignment of patent right |
Owner name: CHENGDU BOSLOY MEDICAL TECHNOLOGY CO., LTD. Free format text: FORMER OWNER: WEST CHINA HOSPITAL OF SICHUAN UNIVERSITY Effective date: 20111117 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 610041 CHENGDU, SICHUAN PROVINCE TO: 610000 CHENGDU, SICHUAN PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20111117 Address after: West high tech Zone Fucheng Road in Chengdu city of Sichuan province 610000 No. 399 No. 5 No. 1004-1008 Patentee after: Chengdu Bosch Medical Technology Co., Ltd. Address before: 610041 Wuhou District, Chengdu, China, No. Guoxue lane, No. 37 Patentee before: West China Hospital of Sichuan University |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20091202 Termination date: 20161231 |