CN1333000A - Chroma analysis method for cheloid proliferation of human body - Google Patents
Chroma analysis method for cheloid proliferation of human body Download PDFInfo
- Publication number
- CN1333000A CN1333000A CN01120275A CN01120275A CN1333000A CN 1333000 A CN1333000 A CN 1333000A CN 01120275 A CN01120275 A CN 01120275A CN 01120275 A CN01120275 A CN 01120275A CN 1333000 A CN1333000 A CN 1333000A
- Authority
- CN
- China
- Prior art keywords
- neural network
- artificial neural
- human body
- colorimetric analysis
- rgb
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Color Image Communication Systems (AREA)
- Spectrometry And Color Measurement (AREA)
Abstract
The chromaticity analysis method of hyperplastic scar of human body relates to a chromaticity detection method of image, and includes the following steps: using photoelectric test technology to obtain human scar tissue and colour image of normal skin, after A/D conversion, utilizing computer system and using module with artificial neural network function in the computer program to make chromaticity correction and chromaticity analysis, finally extracting characteristic chromaticity parameter so as to can objectively and accurately define scar healing time. Said invention is applicable to traumatic scars resulted from various scalds and burns, utilizes chromaticity analysis to provide scientific basis for evaluating therapeutic effect of scar and making therapeutic scheme.
Description
Technical field
The present invention relates to a kind of chroma detection method of image, relate in particular to a kind of chroma detection method of human body skin image.
Background technology
Cicatrix is the various wounds of human body, even if the mode of appearance of normal skin tissue that slight damage causes and histopathologic change are the inevitable physiological reactions that produces in wound or wound surface normal healing process after the wound.We can say do not have wound healing without cicatrization.But cicatrix is improper unhealthy tissue in essence.It does not possess the organizational structure and the physiological function of normal skin, does not more have the vigor of normal structure.And because its unusual structure and hard tough quality, often the pressuring nerve tip produces unbearably sufferings and sense of discomfort; The hypertrophy of scar tissue and the carrying out property contraction that thereupon produces then often form the joint part deformity and limbs are stretched easypro obstacle.So, cicatrix is brought huge misery having influenced patient's normal study, work and life in varying degrees.
The factor that influences scar hyperplasia is a lot, still imperfectly understands at present, is broadly divided into intrinsic factor and extrinsic factor two classes.Wherein Therapeutic Method is the important extrinsic factor that can be used for regulating and controlling the scar hyperplasia process so far.And judge that accurately the scar hyperplasia degree is the therapeutic effect of assessment scar hyperplasia and the key of working out therapeutic scheme targetedly.The present clinical objective reliable detection method of still not having is only judged with doctor's subjective experience, is badly in need of researching and developing out effective scientific analysis means.Up to now, attempted multiple cicatrix method of testing.For example, utilize the thickness of ultrasonic measurement cicatrix; Use the growth of laser detection cicatrix place blood capillary, so that scar hyperplasia information to be provided; Also can adopt infrared thermal imaging technique to monitor the subcutaneous blood capillary activity of cicatrix.But these methods or because of its apparatus expensive, condition harshness, or because of time-consuming oversize, test area is too little and can't be useful for clinical medicine.
Summary of the invention
The present invention is for overcoming the deficiencies in the prior art, provides a kind of photoelectricity test and computer colorimetric analysis technology used to realize colorimetric analysis method to the human body scar hyperplasia.Vascularity and this fact of blood fortune situation that it can reflect scar tissue well from the color or the ruddy degree of human body cicatrix, utilization photoelectricity test technology obtains the burns scar coloured image, utilizes computer system to carry out chromaticity correction and extract the feature colorimetric parameter that can reflect the scar hyperplasia degree by the artificial neural network module in the computer program again.These parameters can be distinguished the different healing times of wound surface well, become the key character index of judging the scar hyperplasia degree, observe scar hyperplasia from objective angle and change.
In order to solve the problems of the technologies described above, technical scheme of the present invention is: this system comprises illuminator, CMYK pattern color printed strip, colored pickup system, A/D converting system and the chromaticity correction of artificial neural network technology is housed and the computer system of colorimetric analysis program; The colorimetric analysis method of this human body scar hyperplasia comprises the following steps: that (1) absorb the coloured image of CMYK color space colour code, human body cicatrix and normal skin respectively with color camera system, after the A/D conversion, convert them to corresponding RGB color space by the computer image treatment program, with the RGB User Colors data storage of the RGB actual color data of colour code image and human body cicatrix or normal skin image in memorizer; (2) with the theoretical color data input computer of the colored printing bar of CMYK color space, obtain the theoretical color data of the pairing RGB of this standard colour code by conversion formula, and it is stored in the memorizer; (3) utilize computer program to set up artificial neural network as output and input respectively the RGB actual color data of theoretical color data of above-mentioned RGB and colour code image, this neutral net is carried out learning training; (4) the RGB User Colors data of human body cicatrix and normal skin chromatic image are imported the above-mentioned artificial neural network that trains, then this artificial neural network output is human body cicatrix and the pairing RGB true colors data of normal skin image, promptly finish the chromaticity correction of scar hyperplasia, then with this data storage in memorizer; (5) utilize the computer image treatment program that above-mentioned RGB true colors data are converted to HSB color space data, and get that H and S value set up another artificial neural network as input in the HSB color space data, its output is the classification results of scar hyperplasia degree, determines the cicatrix wound healing time in view of the above; (6) cicatrix case by sufficient amount and normal skin sample are learnt this artificial neural network and are trained, and assess the scar hyperplasia degree with obtaining the successful artificial neural network of training, finish the colorimetric analysis of scar hyperplasia; (7) final result of colorimetric analysis is exported and deposit.
Described artificial neural network is respectively chromaticity correction artificial neural network and colorimetric analysis artificial neural network, and network structure all is 3 layers of BP network that contain input layer, hidden layer and output layer; The input layer of described chromaticity correction artificial neural network and output layer all have 3 unit, and its hidden layer has 7～9 implicit unit; The input layer of described colorimetric analysis artificial neural network has 4 unit, and hidden layer has 8～10 implicit unit; Output layer has 1 unit; Adopt S type transfer function between described input layer and the hidden layer, hidden layer and output layer then adopt linear transfer function.The hidden layer unit number of described chromaticity correction artificial neural network is the best with 8; The hidden layer unit number of described colorimetric analysis artificial neural network is the best with 9 then.
Compared with prior art, the invention has the beneficial effects as follows: simple and easy to do, can observe scar hyperplasia objectively and change, obtain and the corresponding to classification results of clinical practice, have good concordance and stability, the statistical analysis difference (P＜0.01) of utmost point significance is arranged.Thereby overcome the deficiency of only judging the scar hyperplasia degree clinically, and provide scientific basis for assessing the scar treatment effect and working out clinical protocol with subjective experience.
Utilization the inventive method continuous follow-up assessment tens of cicatrix patients, wherein the men and women half and half, the age, the cicatrix wound healing time was respectively 3 months ± 1 week, 6 months ± 1 week, 9 months ± 1 week and 12 months ± 1 week from many years old middle age of child to 40.The wound surface position relates to each position such as neck, breast, the back of the body, waist, upper arm, underarm, elbow, the back of the hand, thigh, shank, knee joint, foot.Cause the reason of wound cicatrix to be respectively hot water scald, oil scald, Colophonium scald, alkene material burn and fire burn etc.
Description of drawings
Fig. 1 is a workflow sketch map of the present invention;
Fig. 2 is a chromaticity correction artificial neural network structure sketch map of the present invention;
Fig. 3 is a colorimetric analysis artificial neural network structure sketch map of the present invention.
The specific embodiment
Below in conjunction with the drawings and specific embodiments the present invention is described in further detail:
In Fig. 1, in the computer storage of at first computer program that the present invention relates to being packed into, start this program after, profit With the image of this program reception after the A/D conversion; Described image is to use the colored CCD pickup system in certain standard The chromatic image of picked-up colour code and patient's scar and normal skin under the lighting condition; Described colour code belongs to commercial printing circle CMYK color space (using pure blue or green Cyan, pinkish red Magenta and yellow Yellow as primary colours) is through pickup system With the A/D conversion, the RGB color space that converts computer circle to (uses pure red Red, pure green Green and pure blue B lue As primary colours). The former is substractive color space, and the latter is for adding the colour space, and both are complementary colours. In addition, usually also use tone (Hue), saturation degree (Saturation) and brightness (Brightness) describes the three basic feature of color, form The HSB color space. These three kinds of color spaces communicate in theory, can mutually change through simple calculations, also can Utilize Photoshop software to realize conversion. The present invention takes in meter with the colored printing bar of 100 CMYK color spaces earlier In the calculation machine, obtain its corresponding RGB color space chromatic image. Can be obtained and 100 by Photoshop software The rgb value that CMYK is corresponding, the i.e. theoretical color data of RGB. Simultaneously, these 100 standard vittas are taken in computer After namely obtain its RGB actual color data corresponding in imaging system of the present invention. With above-mentioned 100 groups of RGB theories Color data and RGB actual color data deposit in the computer storage. Through computer program control and processing, utilize The corresponding R of standard colour code, the theoretical color data of G, B and actual image color data are set up as output and input respectively Artificial neural network is in order to carry out chromaticity correction. Chromatic image with patient's scar and normal skin turns to through color space again Change with chromaticity correction after as the input of another artificial neural network, with the branch of wound healing time as the scar proliferation degree The output of category feature parameter and neutral net. Artificial neural network behind abundant learning training can be in order to the colour to scar Image carries out colorimetric analysis, and obtains the classification results that the scar proliferation degree is pressed wound healing time, deposits this result in meter In the calculation machine memory, and in computer output equipment, export.
Main points of the present invention are to use artificial neural network technology and realize chromaticity correction and colorimetric analysis in computer program These two sport technique segments.
Fig. 2 illustrates and uses artificial neural network technology to carry out the colourity of scar proliferation in the computer program involved in the present invention Proofread and correct. After utilizing computer program in computer storage, to extract relevant data to call in internal memory, through in the computer program Have the artificial neural network technology module controls of chromaticity correction function and process these data, finally finish the chromaticity correction process, And the result of this process deposited in the computer storage for future use. The specific embodiment of this process is: use reference colour Mark corresponding actual RGB color data as the input of artificial neural network, the corresponding theoretical RGB number of colours of colour code According to the output as artificial neural network, can carry out learning training to this neutral net. Patient's scar or just with picked-up The artificial neural network that the rgb value input of normal skin image trains, then the output by artificial neural network can obtain this figure Resemble corresponding true rgb value. This network uses 3 layers of BP network that contain input layer 1, hidden layer 3 and output layer 3 to carry out Chromaticity correction. Input layer 1 and output layer 3 respectively contain 3 unit, and hidden layer 2 contains 8 unit. Input layer 1 and implicit Adopt S type transfer function between the layer 2,3 of hidden layer 2 and output layers adopt linear transfer function. For with input pattern Be mapped to the output mode of expectation, need to use earlier known mapping corresponding relation to train this network. For this reason, use reference colour Mark corresponding actual RGB color data as the input of artificial neural network, the corresponding theoretical RGB number of colours of colour code According to the output as artificial neural network, this neutral net is carried out learning training.
In the learning training process, neutral net realizes complexity between actual colour code and the theoretical colour code is reflected by the BP algorithm Penetrate the study of corresponding relation, and extract the feature of these mapped modes by means of the hidden layer unit. One group of actual RGB of every input Color data, neutral net will theoretical RGB color data be adjusted input block with implicit single according to a group of required output Connection weight coefficient between unit and output unit and the implicit unit, and optimization sums up and can fit from a large amount of learning samples Close all sample inputs---the best weight coefficient of output mode, it is deposited in the implicit unit. In order to check the study effect Really, also needing to use another to organize known input---the sample of output mode is tested the neutral net that trains. By surveying Examination can be understood neutral net to the grasp degree of complex mapping relation between actual colour code and the theoretical colour code, its universality and Shandong Rod.
Because the output mode that input pattern is mapped to expectation only need to be with known pattern drill network, and need not any Mathematical knowledge is described the mapping relations between input, the output. Therefore, the BP network can be learnt a large amount of mode map relations, Comprise many nonlinear relations. Say that in principle if input layer 1 unit number is M, output layer 3 unit number are N, Then the BP network can realize tieing up from M any mapping of the Euclidean space of N dimension. Patient's scar or normal skin figure with picked-up The artificial neural network that the input of the rgb value of elephant trains then can be obtained the true RGB of this image correspondence by the output of this network Value. So can finish chromaticity correction. The rgb value of patient's scar and normal skin image after artificial neural network is proofreaied and correct again Be converted to HSB color space value. In said process, constantly relevant data is deposited in the computer storage, be subsequently Can utilize relevant data to carry out colorimetric analysis.
Fig. 3 illustrates end user's artificial neural networks of the present invention and carries out the colorimetric analysis of scar proliferation. In Computer Memory Unit Extract through the relevant data behind the chromaticity correction, after data are called in internal memory, have the colorimetric analysis merit through in the computer program The artificial neural network technology module controls of energy and these data of processing are finally finished the colorimetric analysis process, thereby are obtained scar Hyperplasia degree is pressed the classification results of wound healing time. The specific embodiment of this process is: this network also is 3 layers of BP Network. Wherein input layer 4 contains 4 unit, and output layer 6 contains 1 unit, and hidden layer 5 contains 9 unit. All the other knots The structure parameter is identical with the BP network with chromaticity correction. Because the scar chromatic value is relevant with the normal skin chromatic value, so not only select The tone H of scar image1With saturation degree S1, also select simultaneously the tone H of normal skin2With saturation degree S2As being used for look The input of the artificial neural network that degree is analyzed; With the characteristic of division parameter of wound healing time as the scar proliferation degree, draw Be divided into 4 classes, corresponding 3 months, 6 months, 9 months, 12 months and with this output as artificial neural network respectively. Scar case by selecting sufficient amount and normal skin sample are learnt this artificial neural network and are trained. Obtain Train successful artificial neural network namely to can be used for the colorimetric analysis of scar proliferation, obtain the scar proliferation degree by wound healing The classification results of time.
Claims (5)
1. colorimetric analysis method of utilizing computer system to realize the human body scar hyperplasia, it is characterized in that: this system comprises illuminator, CMYK pattern color printed strip, colored pickup system, A/D converting system and the chromaticity correction of artificial neural network technology is housed and the computer system of colorimetric analysis program; The colorimetric analysis method of this human body scar hyperplasia comprises the following steps:
(1) absorbs the coloured image of CMYK color space colour code, human body cicatrix and normal skin respectively with color camera system, after the A/D conversion, convert them to corresponding RGB color space by the computer image treatment program, with the RGB User Colors data storage of the RGB actual color data of colour code image and human body cicatrix or normal skin image in memorizer;
(2) with the theoretical color data input computer of the colored printing bar of CMYK color space, obtain the theoretical color data of the pairing RGB of this standard colour code by conversion formula, and it is stored in the memorizer;
(3) with the RGB actual color data of theoretical color data of above-mentioned RGB and colour code image respectively as output with import, utilize computer program to set up artificial neural network, this neutral net is carried out learning training;
(4) the RGB User Colors data of human body cicatrix and normal skin chromatic image are imported the above-mentioned artificial neural network that trains, then this artificial neural network output is human body cicatrix and the pairing RGB true colors data of normal skin image, promptly finish the chromaticity correction of scar hyperplasia and normal skin, then with this data storage in memorizer;
(5) utilize the computer image treatment program that above-mentioned RGB true colors data are converted to HSB color space data, and get that H and S value set up another artificial neural network as input in the HSB color space data, its output is the classification results of scar hyperplasia degree, determines the cicatrix wound healing time in view of the above;
(6) cicatrix case by sufficient amount and normal skin sample are learnt this artificial neural network and are trained, and assess the scar hyperplasia degree with obtaining the successful artificial neural network of training, finish the colorimetric analysis of scar hyperplasia;
(7) final result of colorimetric analysis is exported and deposit.
2. the colorimetric analysis method of human body scar hyperplasia according to claim 1, it is characterized in that described artificial neural network is respectively chromaticity correction artificial neural network and colorimetric analysis artificial neural network, network structure all is 3 layers of BP network that contain input layer, hidden layer and output layer; The input layer of described chromaticity correction artificial neural network and output layer all have 3 unit, and its hidden layer has 7～9 implicit unit; The input layer of described colorimetric analysis artificial neural network has 4 unit, and hidden layer has 8～10 implicit unit; Output layer has 1 unit.
3. the colorimetric analysis method of human body scar hyperplasia according to claim 2 is characterized in that adopting S type transfer function between described input layer and the hidden layer, and hidden layer and output layer then adopt linear transfer function.
4. the colorimetric analysis method of human body scar hyperplasia according to claim 2 is characterized in that the hidden layer of described chromaticity correction artificial neural network has 8 implicit unit.
5. the colorimetric analysis method of human body scar hyperplasia according to claim 2 is characterized in that the hidden layer of described colorimetric analysis artificial neural network has 9 implicit unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN01120275A CN1333000A (en) | 2001-07-13 | 2001-07-13 | Chroma analysis method for cheloid proliferation of human body |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN01120275A CN1333000A (en) | 2001-07-13 | 2001-07-13 | Chroma analysis method for cheloid proliferation of human body |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1333000A true CN1333000A (en) | 2002-01-30 |
Family
ID=4664034
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN01120275A Pending CN1333000A (en) | 2001-07-13 | 2001-07-13 | Chroma analysis method for cheloid proliferation of human body |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1333000A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101894212A (en) * | 2010-07-05 | 2010-11-24 | 上海交通大学医学院 | Skin wound comprehensive detection method |
| CN101564303B (en) * | 2008-04-21 | 2010-12-01 | 中国医学科学院整形外科医院 | Multifunctional scar ultrasonic inspection system |
| US8000777B2 (en) | 2006-09-19 | 2011-08-16 | Kci Licensing, Inc. | System and method for tracking healing progress of tissue |
| CN107411713A (en) * | 2017-07-27 | 2017-12-01 | 四川省肿瘤医院 | Method based on neck Skin graft color quantizing under RGB patterns |
| CN109859117A (en) * | 2018-12-30 | 2019-06-07 | 南京航空航天大学 | A kind of image color correction method directly correcting rgb value using neural network |
| CN110889847A (en) * | 2019-12-10 | 2020-03-17 | 苏州大学 | Nuclear radiation damage assessment system and method based on infrared imaging |
-
2001
- 2001-07-13 CN CN01120275A patent/CN1333000A/en active Pending
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8000777B2 (en) | 2006-09-19 | 2011-08-16 | Kci Licensing, Inc. | System and method for tracking healing progress of tissue |
| CN101516262B (en) * | 2006-09-19 | 2012-01-04 | 凯希特许有限公司 | System for tracking healing progress of tissue |
| US8588893B2 (en) | 2006-09-19 | 2013-11-19 | Kci Licensing, Inc. | System and method for tracking healing progress of tissue |
| CN101564303B (en) * | 2008-04-21 | 2010-12-01 | 中国医学科学院整形外科医院 | Multifunctional scar ultrasonic inspection system |
| CN101894212A (en) * | 2010-07-05 | 2010-11-24 | 上海交通大学医学院 | Skin wound comprehensive detection method |
| CN101894212B (en) * | 2010-07-05 | 2012-02-15 | 上海交通大学医学院 | Skin wound comprehensive detection method |
| CN107411713A (en) * | 2017-07-27 | 2017-12-01 | 四川省肿瘤医院 | Method based on neck Skin graft color quantizing under RGB patterns |
| CN109859117A (en) * | 2018-12-30 | 2019-06-07 | 南京航空航天大学 | A kind of image color correction method directly correcting rgb value using neural network |
| CN110889847A (en) * | 2019-12-10 | 2020-03-17 | 苏州大学 | Nuclear radiation damage assessment system and method based on infrared imaging |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102331301B (en) | Dental shade mapping | |
| JP4487535B2 (en) | Health measurement system and program | |
| CN103106669B (en) | Chinese medicine tongue picture is as environmental suitability color reproduction method | |
| CN1333000A (en) | Chroma analysis method for cheloid proliferation of human body | |
| CN113610844A (en) | Intelligent skin care method, device, equipment and storage medium | |
| CN112259240B (en) | Construction method of mathematical model of cold-heat deficiency-excess of tongue diagnosis | |
| CN113707305A (en) | Health scheme recommendation method, device, equipment and medium based on deep learning | |
| Abhaya et al. | Variation of the lateral cord of brachial plexus piercing coracobrachialis muscle | |
| CN111640496B (en) | Traditional Chinese medicine remote consultation system based on artificial intelligence | |
| CN106053024A (en) | LED light source preference prediction method for monochromatic objects | |
| CN105748027B (en) | Imaging method and system applied to endoscope | |
| CN113855067A (en) | Visual image and medical image fusion recognition and autonomous positioning scanning method | |
| CN109065151A (en) | Intelligence treats Non-surgical wound processing system | |
| Wray et al. | A model of color vision based on cortical reentry | |
| CN102419861A (en) | Color image correcting method based on topology subdivision of uniform color space | |
| CA3182631A1 (en) | Image display system and image display method | |
| CN109199334A (en) | A kind of discriminating conduct of tongue picture constitution and equipment based on deep neural network | |
| CN115170514B (en) | Medical and medical diagnosis and treatment effect intelligent evaluation system based on multidimensional characteristic data analysis | |
| Lu et al. | Dccn: a deep-color correction network for traditional Chinese medicine tongue images | |
| CN109934460A (en) | Ship's space Evaluating Luminous Environment and system based on Method of extenics | |
| CN106618551B (en) | A kind of intelligent terminal for being used for the connection of three lead electrocardioelectrodes and differentiating | |
| CN212326353U (en) | Novel digital tongue depressor | |
| CN108904978B (en) | Beauty device based on simulation biological micro-current | |
| Leow et al. | Lifelikeness in multilayered digital prostheses | |
| CN112365559A (en) | Infrared image coloring method for generating countermeasure network based on structural similarity |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |