CN108197539A - A kind of Diagnosis of Crania By Means identification method - Google Patents
A kind of Diagnosis of Crania By Means identification method Download PDFInfo
- Publication number
- CN108197539A CN108197539A CN201711397107.4A CN201711397107A CN108197539A CN 108197539 A CN108197539 A CN 108197539A CN 201711397107 A CN201711397107 A CN 201711397107A CN 108197539 A CN108197539 A CN 108197539A
- Authority
- CN
- China
- Prior art keywords
- skull
- sample
- identified
- skull sample
- gender
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30016—Brain
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention discloses a kind of Diagnosis of Crania By Means identification methods.Method includes：The skull data of training sample are transformed under unified Frankfort coordinate system, and carry out dimension normalization；Each skull sample of normalized training sample set around Z axis is rotated, obtains multiple images of skull sample under different angle；Skull sample global characteristics are extracted using improved convolutional neural networks method, every image for calculating each skull sample is belonging respectively to the probability of male and female；Optimized parameter, design Diagnosis of Crania By Means identification function are obtained using least square method；The gender of unknown skull is differentiated using the function of structure.The present invention is easy to operate, and overcoming existing Diagnosis of Crania By Means identification method needs to have the shortcomings that the expert of professional knowledge participates in and cumbersome manual measurement；The problems such as overcome existing method is influenced greatly by skull change in size, and measurement accuracy is not high.
Description
Technical field
The invention belongs to area of pattern recognition, it is related to a kind of convolutional neural networks and least square method being utilized to carry out skull
The method do not classified.It is mainly used for the fields such as criminal investigation, archaeology, forensic anthropology.
Background technology
In living things feature recognition field, sex identification is the important research content in the field.Gender is determined in legal medical expert
There is certain application in the fields such as anthropology, Facial restoration and the unknown skull of identification, weight is provided for later correlative study
It will foundation.
When carrying out gender identification according to skeletal remains, basin bone and skull are the important research objects of Sex estimation.Basin bone is
Morphological differences region the most significant, but basin bone is easily influenced by outside environmental elements, and basin bone is caused to be unfavorable for protecting for a long time
It deposits and frangible, i.e., integrality cannot be guaranteed, and then cause basin bone that cannot play key effect in sex identification.Skull is
It is made of sclerous tissues, is not easy to be destroyed, can more fully preserved after death.So skull is as sex identification
Only choosing.Carrying out sex identification according to skull becomes the research hotspot of the related fields such as forensic anthropology, archaeology, criminal investigation.
Current existing sex appraisal method, mainly there is two kinds of morphological method and mensuration.Morphological method is mainly
By anthropologist's artificial observation skull general profile, the size of skull, angle, the thickness journey of true skull of crucial skull region
The difference progress sex identification of each feature such as degree and the skull degree of wear, such as the skull of male are larger, thick and heavy and sturdy,
Geisoma prosperity, margo supraorbitalis garden is blunt, and forehead relatively tilts, and cheekbone is high and sturdy, and zygomatic arch is thicker, grand outside mastoid process and pillow all to compare hair
It reaches；The skull of women is smaller, and smooth and careful, geisoma is undeveloped, and margo supraorbitalis is sharp thin, and forehead is more steep, and cheekbone it is low and
Very thin, zygomatic arch is thinner and more delicate, and mastoid process and the outer knuckle of pillow are undeveloped etc..Mensuration is typically in skull physical model, the X-ray of skull
Some measurement points are demarcated on photo and three-dimensional digital model, is studied according to anthropologist and defines some measurement indexes, according to these
Measurement index constructs discriminant function by statistical method.To sum up, morphological method is easily by expert's subjective factor and professional knowledge shadow
The inconsistency for causing judgement result is rung, subjective dependence is strong and without sufficient theory support；Measuring method is cumbersome, to measuring
The measurement accuracy requirement of item is high, is influenced greatly by skull size, it is also possible to can not there are the apparent feature of some gender differences
It measures.Some researches show that, measurement errors between most of feature different observers up to more than 10%.In addition, with the age
Increase, Skull research is without significant change, but size can change, and increases measurement difficulty.With computer technology and
The appearance of CT technologies, is possibly realized by computer aided measurement.But due to skull complexity, realize that accurate measure still is difficult.
In addition, the accuracy that current method identifies Diagnosis of Crania By Means is not high, substantially no more than 90%.
Invention content
The object of the present invention is to provide a kind of automated method of Diagnosis of Crania By Means identification, with dependence in the prior art
Expert's subjective experience degree height, cumbersome manual measurement, measurement accuracy is not high and influenced the problem of big by skull size.
Diagnosis of Crania By Means identification function preparation method provided by the invention includes：
All skull samples in training skull sample set are transformed under the coordinate system A of Frankfort, and carry out by step 1
Dimension normalization makes its posture and size specification；Skull sample number N in training skull sample set is no less than 50；
Step 2 rotates the skull sample n after normalization around Z axis, and n=1,2 .3..., N obtain one per rotation alpha
The image of the skull sample is opened, 1≤α≤90 ° obtain multiple images of the skull sample under different rotary angle；
Step 3, using the global characteristics of improved convolutional neural networks method extraction skull sample n, n=1,2,
.3..., N, the improved convolutional neural networks method be followed successively by input, convolution, convolution, down-sampling, convolution, convolution, under adopt
Sample, full connection and output；The probability of the affiliated gender of every image of skull sample n is calculated using sigmoid functions, obtains skull
The probability vector p of sample nn, n=1,2 .3..., N；
Step 4 solves optimized parameter w using least square method to the probability vector of all skull samples*,
qn=0 or 1, qnAccording to
The affiliated gender difference values of skull sample n are different；
Step 5, design Diagnosis of Crania By Means identification function：D=pxw*- l, D are dependent variable, pxFor independent variable, l=0 or 1.
The present invention also provides a kind of Diagnosis of Crania By Means identification methods.The method provided includes：
Step 1, skull to be identified is transformed under the coordinate system A of Frankfort, and carries out dimension normalization, make its posture and
Size specification；
Step 2, it will be rotated through step 1 treated skull sample to be identified around Z axis, the cranium obtained per rotation alpha
The image of bone sample, 1≤α≤90 °, obtain different rotary angle under skull sample to be identified multiple images；
Step 3, the global characteristics of skull sample to be identified, the improvement are extracted using improved convolutional neural networks method
Convolutional neural networks method be followed successively by input, convolution, convolution, down-sampling, convolution, convolution, down-sampling, full connection and output；
The probability of the affiliated gender of every image of skull sample to be identified is calculated using sigmoid functions, obtains skull sample to be identified
Probability vector px；
Step 4, sex identification is carried out using the formula (7) that the present invention is built：
D=pxw*-l (7)
Assuming that l=0 represents male, l=1 represents women, if | D | value be less than the gender of skull sample to be identified if 0.5
For male, otherwise, the gender of skull sample to be identified is women；
Assuming that l=0 represents women, l=1 represents male, the value if D is more than the gender of skull sample to be identified if 0.5 and is
Man, otherwise the gender of skull sample to be identified is female.
The beneficial effects of the invention are as follows：
(1) the method applied in the present invention, overcome need in the prior art the expert with professional knowledge participate in and it is numerous
The shortcomings that trivial manual measurement；
(2) method that the present invention is taken, overcome the prior art is influenced greatly by skull change in size, and measurement accuracy is not high
The problems such as；
(3) method taken of the present invention has high degree of automation, and easy to operate, accuracy is higher, up to 94.4% with
On.
Description of the drawings
Fig. 1 is Frankfort coordinate system；
Fig. 2 is the three-dimensional cranium model after being normalized in embodiment；
Fig. 3 is the qualification result schematic diagram in embodiment 3.
Specific embodiment
The skull sample of the present invention is three-dimensional mesh data, facial physiology sites of Frankfort coordinate system A according to skull
It establishes, can be by point, left eye socket of the eye down contour point on the bilateral side door of left and right for example, with reference to Fig. 1, four Cranial features points of glabella point are true
Fixed, L is used respectivelyp、Rp、Mp、VpIt represents.Frankfurt plane is by Lp, Rp, Mp3 points determine.Wherein：Coordinate origin：With
For normal vector and cross point VpPlane and straight line LpRpIntersection point be denoted as the origin O' of Frankfurt plane；X-axis：Remember Lp、RpAnd
Mp3 points of planes formed are the XO'Y planes in coordinate system, point L on left side doorpPoint R on to auris dextra doorpDirection be X-axis side
To；Z axis：Origin O' will be crossed and be denoted as Z axis with XO'Y plane vertically upward directions；Y-axis：It will be both perpendicular to X-axis, Z axis
Straight line is denoted as the Y-axis of coordinate system, and the positive direction of Y-axis is determined by right-hand rule, after the dimension normalization is unified coordinate system,
Dimension normalization is carried out again.Set the L of all skull modelsp-RpThe distance between for unit 1, then to each vertex of skull
(x, y, z) carry out change of scale for (x/ | Lp-Rp|,y/|Lp-Rp|,z/|Lp-Rp|)。
The improved convolutional neural networks of the present invention are that the LeNet-5 models of convolutional neural networks are improved：
1) the input tomographic image size of standard LeNET5 models is 32 × 32, but in order to preserve the deep semantic of image and content information,
The input tomographic image expanded in size of the present invention, for example, input tomographic image size becomes 256 × 256；2) according to skull data set spy
Sign setting convolution kernel size, for example, convolution kernel is dimensioned to 17 × 17；3) it is added to one after the convolutional layer of master pattern
Layer convolutional layer avoids losing more depth informations of image.The improved convolutional neural networks detailed process of the present invention include input,
Convolution, convolution, down-sampling, convolution, convolution, down-sampling, full connection, output.
The present invention calculates the probability value that every image belongs to male and female using sigmoid functions, and calculation formula is as follows：
f(x)c=p (y=c | x)=sigmoid (W*X+b) (1)
Wherein, f (x)cIt is the probability function that sample is divided into male or female, X is global characteristics, and W is weight, and b is inclined
Shifting amount.C takes 0 or 1 according to gender.
The present invention makes the residual sum of squares (RSS) of quadratic loss function minimum using the regression model of least square method, to obtain most
Excellent parameter.Quadratic loss function is as follows：
Wherein, pjw-qjpnω-qnRepresent residual error, pnRepresent the probability of n-th of sample, w represents weight, qnRepresent training sample
The true classification of n-th of sample of this concentration.qnRepresent that training sample concentrates the true classification of n-th of sample, due to sex identification
Problem belongs to two classification problems, so representing two classifications with 0 and 1.
In order to obtain optimized parameter, minimize the summation Q of residuals squares, formula (2) can be converted into：
Wherein,
Formula (3) is further decomposed, above-mentioned formula can be converted into：
It is easily obtained according to formula (4)It is a constant, is influenced less on solving optimized parameter.Therefore, above-mentioned formula
It is converted to solve the parameter for minimizing S；
Derivation is carried out to parameter w based on Differential Geometry knowledge, it is known that the office of function can be obtained when the derivative of parameter w is zero
Portion's optimal solution, i.e. optimized parameter calculate as follows：
It can solve to obtain optimized parameter w by formula (6)*, design decision function：
D=pxw*-l (7)
Wherein, pxIt is the probability vector of multiple image constructions of skull to be identified, w*It is optimized parameter, l assumes that label,
Value is 0 or 1.
Heretofore described skull data are three-dimensional mesh data, are three-dimensional grid below for the skull data
Data are described.Used three-dimensional mesh data is by the way that head CT scan, the head for obtaining meeting dicom standard is broken
Face image data.
Embodiment 1：
The skull data of the embodiment are three-dimensional mesh data, are three-dimensional mesh data below for the skull data
It is described.Used three-dimensional mesh data is by head CT scan, obtaining the head cross-section diagram for meeting dicom standard
As data, denoising de-redundancy is carried out to CT data, reconstructs to obtain the three-dimensional grid number of skull using Marching Cubes methods
According to each three-dimensional mesh data includes about 100000 vertex and 200000 tri patch.Assuming that have 56 training samples,
Wherein 28 males, 28 women are below specific implementation steps：
Step 1：It, will be every under the three-dimensional mesh data unification to Frankfort coordinate system for the skull that training sample is concentrated
Distance metric unit of the distance of a left and right earhole central point of sample as the sample, i.e., to the coordinate of all the points on sample into
Row dimension normalization makes its posture and size specification；
Step 2：To each skull data that normalized training sample is concentrated, X-axis and Y-axis are kept, is often revolved around Z axis
Turn 18 ° of acquisitions, one image, obtain 20 images of skull sample under different angle；
Step 3：Using the global characteristics of improved convolutional neural networks method extraction skull sample n, n=1,2,
3 ..., 90, it establishes a complete network and is input, convolution, convolution, down-sampling, convolution, convolution, down-sampling, connects entirely, is defeated
Go out.Detailed process is：First, every layer of weight and offset parameter is initialized；Secondly, input layer gray level image size for 256 ×
256, it is calculated to simplify, pixel value is normalized between 0 and 1.First convolutional layer C1 be by six Feature Mappings, 17 ×
17 kernel and biasing are formed, and pass through the parameter of convolution kernel and activation primitive statistical nature mapping layer.In order to simplify Feature Mapping
Layer operation simultaneously saves time cost, and down-sampling layer is close to after two convolutional layers.The image size of second convolutional layer C2 is 224
The image size of × 224, down-sampling layer S3 are 112 × 112.Equally, it is 96 × 96 that third convolutional layer C4, which contains 6 sizes,
Image, the 4th convolution C5 have a subsample mapping graph that 6 sizes are 80 × 80, and down-sampling layer S6 image sizes become 40 ×
40.C1 to S6 layers are CNN characteristic extraction procedures, and full articulamentum C7 is the character representation layer of image, and size is 40 × 40 × 6.
Every image for calculating each skull sample is belonging respectively to the probability of male and female, f (x)c=p (y=c | x)=
Sigmoid (W*X+b), c=1 represent male, and c=0 represents women；
Step 4：To the probability value of every image of each sample that step 3 acquires, by multiple figures of each sample
As being regarded as a research object, probability vector is formed.Make the residual error of quadratic loss function using the regression model of least square method
Quadratic sum is minimum, to obtain optimized parameter.The embodiment represents male's skull with 1, and 0 represents women skull；Using optimized parameter
w*It is as follows to build decision function：
D=pxw*-l (7)；
In the embodiment：
WhereinRepresent the probability vector of the 1st sample,Represent the 2nd
The probability vector of sample,Represent the probability vector of the 3rd sample ...Represent the probability vector of the 56th sample, they
All it is by the formula in step 3：f(x)c=p (y=c | x)=sigmoid (W*X+b) is calculated, should because sample is more
The probability vector of first sample is only provided in example, Vector determines to be determined by skull training sample, in the embodiment, It finally obtains：w*=[-
0.0000；-0.0678；-0.3188；-0.2325；0.8005；0.7806；1.3203；-0.1481；0.4322；-0.2780；
0.2172；0.5314；0.6997；0.3086；0.3575；-0.1631；-0.1430；-0.1339；-0.2908；-0.0000].
Embodiment 2：
The embodiment is to differentiate Diagnosis of Crania By Means shown in Fig. 2 with the function that embodiment 1 is built：
Step 1：The three-dimensional mesh data of skull to be identified (Fig. 2) is transformed under unified Frankfort coordinate system and gone forward side by side
Row dimension normalization makes its posture and size specification；
Step 2：Obtain unknown skull image：Skull will be tested, keeps X-axis and Y-axis, 18 ° are often rotated around Z axis and obtains one
Image is opened, obtains 20 images of skull sample under different angle；
Step 3：The global characteristics of skull sample to be identified are extracted using improved convolutional neural networks method, calculates and waits to reflect
The Probability p of the affiliated gender of every image of other skull samplex=(0.6465,0.6108,0.5958,0.5865,0.5759,
0.5513,0.5481,0.5207,0.5119,0.5029,0.5214,0.5106,0.4748,0.4242,0.4269,
0.5508,0.5377,0.5116,0.5089,0.5737)；
Step 4：Further according to the formula (7) in specific embodiment：D=pxw*- l, it is assumed that label l takes 1, is calculated D's
Be worth is 0.7833, D=0.7833>0.5, so skull to be identified is male, it is consistent with anthropologist's qualification result.
Embodiment 3：
36 are used as test set known to embodiment selection gender, wherein 18 women, 18 males, using embodiment
1 function carries out sex identification, if as shown in figure 3, identify that women average accuracy is 94.7%, the average accuracy of male
94%, average accuracy 94.4%.
In short, the embodiment of the present invention announcement is its preferable embodiment, but it is not limited to this.The technology of this field
Personnel can understand the core concept of the present invention, without departing from technical scheme of the present invention easily according to above-described embodiment
Basis deformation or replacement, all within protection scope of the present invention.
Claims (2)
1. a kind of Diagnosis of Crania By Means identifies function preparation method, which is characterized in that method includes：
All skull samples in training skull sample set are transformed under the coordinate system A of Frankfort, and carry out scale by step 1
Normalization；Skull sample number N in training skull sample set is no less than 50；
Step 2 rotates the skull sample n after normalization around Z axis, n=1,2 .3..., N, and one is obtained per rotation alpha should
The image of skull sample, 1≤α≤90 ° obtain multiple images of the skull sample under different rotary angle；
Step 3, using the global characteristics of improved convolutional neural networks method extraction skull sample n, n=1,2 .3..., N,
The improved convolutional neural networks method is followed successively by input, the first convolution, the second convolution, down-sampling, third convolution, Volume Four
Product, down-sampling, full connection and output；The probability of the affiliated gender of every image of skull sample n is calculated using sigmoid functions,
The multiple probability of gained form the probability vector p of skull sample nn, n=1,2 .3..., N；
Step 4 solves optimized parameter w using least square method to the probability vector of all skull samples*,
qn=0 or 1, qnValue
According to skull sample n affiliated gender differences, value is different；
Step 5, design Diagnosis of Crania By Means identification function：D=pxw*- l, D are dependent variable, pxFor independent variable, l=0 or 1.
2. a kind of Diagnosis of Crania By Means identification method, which is characterized in that method includes：
Step 1, skull to be identified is transformed under the coordinate system A of Frankfort, and carries out dimension normalization；
Step 2, it will be rotated through step 1 treated skull sample to be identified around Z axis, the skull sample obtained per rotation alpha
This image, 1≤α≤90 ° obtain multiple images of skull sample to be identified under different rotary angle；
Step 3, the global characteristics of skull sample to be identified, the improved volume are extracted using improved convolutional neural networks method
Product neural network method be followed successively by input, the first convolution, the second convolution, down-sampling, third convolution, Volume Four product, down-sampling, entirely
Connection and output；The probability of the affiliated gender of every image of skull sample to be identified is calculated using sigmoid functions, obtains waiting to reflect
The probability vector p of other skull samplex；
Step 4, sex identification is carried out to skull sample to be identified using the formula (7) of claim 1 the method structure：
D=pxw*-l (7)
Assuming that l=0 represents male, l=1 represents women, if | D | value to be less than the gender of skull sample to be identified if 0.5 be man
Property, otherwise, the gender of skull sample to be identified is women；
Assuming that l=0 represents women, l=1 represents male, if the value of D, to be more than the gender of skull sample to be identified if 0.5 be man, no
Then the gender of skull sample to be identified is female.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711397107.4A CN108197539A (en) | 2017-12-21 | 2017-12-21 | A kind of Diagnosis of Crania By Means identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711397107.4A CN108197539A (en) | 2017-12-21 | 2017-12-21 | A kind of Diagnosis of Crania By Means identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108197539A true CN108197539A (en) | 2018-06-22 |
Family
ID=62583316
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711397107.4A Pending CN108197539A (en) | 2017-12-21 | 2017-12-21 | A kind of Diagnosis of Crania By Means identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108197539A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109118501A (en) * | 2018-08-03 | 2019-01-01 | 上海电气集团股份有限公司 | Image processing method and system |
| CN111462055A (en) * | 2020-03-19 | 2020-07-28 | 沈阳先进医疗设备技术孵化中心有限公司 | Skull detection method and device |
| CN112907537A (en) * | 2021-02-20 | 2021-06-04 | 司法鉴定科学研究院 | Skeleton sex identification method based on deep learning and on-site virtual simulation technology |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1801213A (en) * | 2005-09-30 | 2006-07-12 | 铁岭市公安局213研究所 | Method and apparatus for three-dimensional cranium body source identification |
| CN102521875A (en) * | 2011-11-25 | 2012-06-27 | 北京师范大学 | Partial least squares recursive craniofacial reconstruction method based on tensor space |
| CN102831443A (en) * | 2012-07-27 | 2012-12-19 | 北京师范大学 | Skull sex determining method based on spatial analysis |
| US9165360B1 (en) * | 2012-09-27 | 2015-10-20 | Zepmed, Llc | Methods, systems, and devices for automated analysis of medical scans |
| CN105760344A (en) * | 2016-01-29 | 2016-07-13 | 杭州电子科技大学 | Distributed principal component analysis neural network modeling method for chemical exothermic reaction |
-
2017
- 2017-12-21 CN CN201711397107.4A patent/CN108197539A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1801213A (en) * | 2005-09-30 | 2006-07-12 | 铁岭市公安局213研究所 | Method and apparatus for three-dimensional cranium body source identification |
| CN102521875A (en) * | 2011-11-25 | 2012-06-27 | 北京师范大学 | Partial least squares recursive craniofacial reconstruction method based on tensor space |
| CN102831443A (en) * | 2012-07-27 | 2012-12-19 | 北京师范大学 | Skull sex determining method based on spatial analysis |
| US9165360B1 (en) * | 2012-09-27 | 2015-10-20 | Zepmed, Llc | Methods, systems, and devices for automated analysis of medical scans |
| CN105760344A (en) * | 2016-01-29 | 2016-07-13 | 杭州电子科技大学 | Distributed principal component analysis neural network modeling method for chemical exothermic reaction |
Non-Patent Citations (2)
| Title |
|---|
| RONG RONG REN 等: "Automatic Sex Identification Based on Convolution Neural Network and Least Square Method", 《2016 INTERNATIONAL CONFERENCE ON ELECTRONIC INFORMATION TECHNOLOGY AND INTELLECTUALIZATION (ICEITI 2016)》 * |
| 赵倩娜: "计算机辅助三维颅骨性别鉴定方法研究", 《中国优秀硕士学位论文全文数据库 医药卫生科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109118501A (en) * | 2018-08-03 | 2019-01-01 | 上海电气集团股份有限公司 | Image processing method and system |
| CN111462055A (en) * | 2020-03-19 | 2020-07-28 | 沈阳先进医疗设备技术孵化中心有限公司 | Skull detection method and device |
| CN111462055B (en) * | 2020-03-19 | 2024-03-08 | 东软医疗系统股份有限公司 | Skull detection method and device |
| CN112907537A (en) * | 2021-02-20 | 2021-06-04 | 司法鉴定科学研究院 | Skeleton sex identification method based on deep learning and on-site virtual simulation technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pishchulin et al. | Building statistical shape spaces for 3d human modeling | |
| CN107369160B (en) | Choroid neogenesis blood vessel segmentation algorithm in OCT image | |
| Claes et al. | Craniofacial reconstruction using a combined statistical model of face shape and soft tissue depths: methodology and validation | |
| Claes et al. | Computerized craniofacial reconstruction: conceptual framework and review | |
| Quellec et al. | Three-dimensional analysis of retinal layer texture: identification of fluid-filled regions in SD-OCT of the macula | |
| Scandrett et al. | A person-specific, rigorous aging model of the human face | |
| CN108932536A (en) | Human face posture method for reconstructing based on deep neural network | |
| Pluim et al. | The truth is hard to make: Validation of medical image registration | |
| Claes et al. | Bayesian estimation of optimal craniofacial reconstructions | |
| Duan et al. | Craniofacial reconstruction based on multi-linear subspace analysis | |
| Zhang et al. | A supervised texton based approach for automatic segmentation and measurement of the fetal head and femur in 2D ultrasound images | |
| CN108197539A (en) | A kind of Diagnosis of Crania By Means identification method | |
| CN111062340A (en) | Abnormal gait behavior identification method based on virtual posture sample synthesis | |
| CN106327479A (en) | Apparatus and method for identifying blood vessels in angiography-assisted congenital heart disease operation | |
| CN110880366A (en) | Medical image processing system | |
| CN115830016B (en) | Medical image registration model training method and equipment | |
| Wu et al. | Reconstructing 3D lung shape from a single 2D image during the deaeration deformation process using model-based data augmentation | |
| CN113096137A (en) | Adaptive segmentation method and system for OCT (optical coherence tomography) retinal image field | |
| Pereira-Pedro et al. | Landmarking endocasts | |
| Piras et al. | Current options for visualization of local deformation in modern shape analysis applied to paleobiological case studies | |
| Nguyen et al. | A statistical shape modeling approach for predicting subject-specific human skull from head surface | |
| CN115115676A (en) | Image registration method, device, equipment and storage medium | |
| Berar et al. | 3D statistical facial reconstruction | |
| CN107680070A (en) | A kind of layering weight image interfusion method based on original image content | |
| Pei et al. | The craniofacial reconstruction from the local structural diversity of skulls |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180622 |