CN110288589A - Hematoma Enlargement in Spontaneous prediction technique and device - Google Patents
Hematoma Enlargement in Spontaneous prediction technique and device Download PDFInfo
- Publication number
- CN110288589A CN110288589A CN201910583573.4A CN201910583573A CN110288589A CN 110288589 A CN110288589 A CN 110288589A CN 201910583573 A CN201910583573 A CN 201910583573A CN 110288589 A CN110288589 A CN 110288589A
- Authority
- CN
- China
- Prior art keywords
- hemotoncus
- preset
- image
- sample
- hematoma
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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
-
- 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
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
- G06T2207/30104—Vascular flow; Blood flow; Perfusion
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Quality & Reliability (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The present embodiment is related to Medical Imaging Technology field, provides a kind of hematoma Enlargement in Spontaneous prediction technique and device, which comprises obtains clinical examination data and the CT image comprising hemotoncus;CT image is inputted preset automatic parted pattern to be split, obtains segmentation result, and calculate volume of hematoma according to segmentation result;Segmentation result, volume of hematoma and clinical examination data are formed into test sample;The matching degree in preset multiple support samples between each support sample and test sample is calculated, and according to multiple matching degrees, predicts whether hemotoncus can expand.Compared with prior art, hematoma Enlargement in Spontaneous prediction technique provided in this embodiment and device can predict whether hemotoncus can expand in time, to carry out respective handling, to ensure the life security of patient.
Description
Technical field
The present invention relates to Medical Imaging Technology fields, in particular to a kind of hematoma Enlargement in Spontaneous prediction technique and device.
Background technique
Cerebral hemorrhage (Intracerebral Haemorrhage, abbreviation ICH) refers to the intracerebral hemorrhage caused by angiorrhoxis,
Medically signified cerebral hemorrhage is mainly spontaneous non-injury-ness brain hemorrhage, i.e. spontaneous cerebral hemorrhage, and spontaneous cerebral hemorrhage is logical
It is often as caused by the factors such as hypertension, hyperglycemia, hyperlipidemia and smoking.The disease incidence is unexpected, and the state of an illness is dangerous, treatment cost
With, recurrence rate, disability rate and the death rate it is all very high, patients with cerebral hemorrhage more than 40% can be dead in one month, the trouble of survival
In person 80% need by other people nursing and live, mitigate ICH disease burden can be effectively reduced China medical treatment take
Whole burden in business.
In the prior art, pass through follow-up CT scan (Computed Tomography, abbreviation CT) and base
Line CT is compared, to judge whether hemotoncus expands.But follow-up CT examination is opposite with baseline CT examination time interval
It is longer, when judging to occur hematoma Enlargement in Spontaneous by such mode, imply that conditions of patients is likely to deteriorate even dead, use
This method determines whether that hematoma Enlargement in Spontaneous is not prompt enough, and will lead to some patientss, sb.'s illness took a turn for the worse during this period or even dead.
Summary of the invention
The purpose of the present invention is to provide a kind of hematoma Enlargement in Spontaneous prediction technique and devices, are expanded with improving hemotoncus in the prior art
Not in time, leading to some patientss, sb.'s illness took a turn for the worse during this period or even dead situation for big judgement.
To achieve the goals above, technical solution used in the embodiment of the present invention is as follows:
In a first aspect, the embodiment of the present invention provides a kind of hematoma Enlargement in Spontaneous prediction technique, which comprises obtain clinical inspection
Look into data and the CT scan CT image comprising hemotoncus;The CT image is inputted into preset automatic parted pattern
It is split, obtains segmentation result, and calculate volume of hematoma according to the segmentation result;By the segmentation result, hemotoncus body
The long-pending and described clinical examination data form test sample;Calculate each support sample and institute in preset multiple support samples
The matching degree between test sample is stated, and according to multiple matching degrees, predicts whether the hemotoncus can expand.
Second aspect, the embodiment of the present invention provide a kind of hematoma Enlargement in Spontaneous prediction meanss, and described device includes: acquisition module,
CT scan CT image for obtaining clinical examination data and comprising hemotoncus;Processing module, being used for will be described
CT image inputs preset automatic parted pattern and is split, and obtains segmentation result, and calculate bleeding according to the segmentation result
Swollen volume;The segmentation result, volume of hematoma and the clinical examination data are formed into test sample;Calculate preset multiple
The matching degree in sample between each support sample and the test sample is supportted, and according to multiple matching degrees, described in prediction
Whether hemotoncus can expand.
Compared with the existing technology, a kind of hematoma Enlargement in Spontaneous prediction technique and device provided by the embodiment of the present invention, pass through by
CT image comprising hemotoncus inputs preset automatic parted pattern and is split, and obtains segmentation result, and according to segmentation result meter
Volume of hematoma is calculated, segmentation result, volume of hematoma and clinical examination data are then formed into test sample, then calculate preset multiple
The matching degree of each support sample and test sample in sample is supported, and according to multiple matching degrees, predicts whether hemotoncus can occur
Expand.Predict that whether hemotoncus can expand, can predict in advance is according to existing image data and clinical examination data
It is no that hematoma Enlargement in Spontaneous can occur, and perform corresponding processing, it avoids and determines hematoma Enlargement in Spontaneous not in time, cause some patientss in this phase
Between sb.'s illness took a turn for the worse even dead situation.
To enable the above objects, features and advantages of the present invention to be clearer and more comprehensible, preferred embodiment is cited below particularly, and cooperate
Appended attached drawing, is described in detail below.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 shows the block diagram of electronic equipment provided in an embodiment of the present invention.
Fig. 2 shows hematoma Enlargement in Spontaneous prediction techniques provided in an embodiment of the present invention.
Fig. 3 shows the structure chart of segmentation network in mixing cavity provided in an embodiment of the present invention.
Fig. 4 shows provided in an embodiment of the present invention 1,2,3,5 voidage schematic diagram.
Fig. 5 is that Fig. 2 shows the sub-step flow charts of step S2.
Fig. 6 shows segmentation result schematic diagram provided in an embodiment of the present invention.
Fig. 7 is that Fig. 2 shows the first sub-step flow charts of step S4.
Fig. 8 is that Fig. 7 shows the sub-step flow chart of step S41.
Fig. 9 shows the schematic diagram of the twin network of three-dimensional provided in an embodiment of the present invention.
Figure 10 is that Fig. 2 shows the second sub-step flow charts of step S4.
Figure 11 shows the block diagram of hematoma Enlargement in Spontaneous prediction meanss provided in an embodiment of the present invention.
Icon: 100- electronic equipment;101- processor;102- memory;103- bus;104- communication interface;200- blood
It is swollen to expand prediction meanss;201- obtains module;202- processing module.
Specific embodiment
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, is not intended to limit claimed invention to the detailed description of the embodiment of the present invention provided in the accompanying drawings below
Range, but it is merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, art technology is not having with personnel
Every other embodiment obtained under the premise of creative work is made, shall fall within the protection scope of the present invention.
It should also be noted that similar label and letter indicate similar terms in following attached drawing, therefore, once a certain Xiang Yi
It is defined in a attached drawing, does not then need that it is further defined and explained in subsequent attached drawing.Meanwhile of the invention
In description, term " first ", " second " etc. are only used for distinguishing description, are not understood to indicate or imply relative importance.
It is common variation in Cerebral Hemorrhage Disease that hematoma Enlargement in Spontaneous, which occurs, for early stage (after the onset of being often referred in 24 hours) after cerebral hemorrhage,
Hematoma Enlargement in Spontaneous according to volume of hematoma between follow-up CT and baseline CT whether be greater than 33% or 6 milliliters determine, if expanded beyond
33% or 6 milliliters, then be determined as hematoma Enlargement in Spontaneous, occur hematoma Enlargement in Spontaneous imply conditions of patients be likely to deteriorate it is even dead,
And follow-up CT examination and baseline CT examination time interval are relatively long, determine whether that hematoma Enlargement in Spontaneous is not prompt enough with this method, meeting
Leading to some patientss, sb.'s illness took a turn for the worse during this period or even dead.
The technical problem to be solved by the present invention is in view of the above-mentioned problems, provide a kind of hematoma Enlargement in Spontaneous prediction technique, core
Heart improvement is, after being diagnosed as patients with cerebral hemorrhage, according to existing image data and clinical examination data to hemotoncus whether
It can expand and be predicted, whether can occur hematoma Enlargement in Spontaneous, predict whether patient can occur hemotoncus expansion in time if can predict in advance
It greatly, is the key that further diagnosing and treating is carried out to ICH patient.
Hematoma Enlargement in Spontaneous prediction technique provided in an embodiment of the present invention is applied to electronic equipment 100, and electronic equipment 100 can be with
It is, but is not limited to smart phone, tablet computer, personal computer, vehicle-mounted computer, personal digital assistant (personal
Digital assistant, PDA) etc..Referring to Fig. 1, Fig. 1 shows the portion of electronic equipment provided in an embodiment of the present invention
Separation structure schematic diagram, electronic equipment 100 include processor 101, memory 102, bus 103 and communication interface 104.Processor
101, memory 102 and communication interface 104 are connected by bus 103, and processor 101 is used to executing to be stored in memory 102
Executable module, such as computer program.
Processor 101 may be a kind of IC chip, the processing capacity with signal.During realization, hemotoncus
Each step for expanding prediction technique can pass through the integrated logic circuit of the hardware in processor 101 or the instruction of software form
It completes.Above-mentioned processor 101 can be general processor 101, including central processing unit (Central Processing
Unit, abbreviation CPU), network processing unit (Network Processor, abbreviation NP) etc.;It can also be digital signal processor
(Digital Signal Processor, abbreviation DSP), specific integrated circuit (Application Specific
Integrated Circuit, abbreviation ASIC), ready-made programmable gate array (Field-Programmable Gate Array,
Abbreviation FPGA) either other programmable logic device, discrete gate or transistor logic, discrete hardware components.
Memory 102 may include high-speed random access memory (RAM:Random Access Memory), it is also possible to
It further include non-labile memory (non-volatile memory), for example, at least a magnetic disk storage.Memory 102
It may be, but not limited to, random access memory (Random Access Memory, RAM), read-only memory (Read Only
Memory, ROM), programmable read only memory (Programmable Read-Only Memory, PROM) is erasable read-only to deposit
Reservoir (Erasable Programmable Read-Only Memory, EPROM), electricallyerasable ROM (EEROM)
(Electric Erasable Programmable Read-Only Memory, EEPROM) etc..
Bus 103 can be ISA (Industry Standard Architecture) bus, PCI (Peripheral
Component Interconnect) bus or EISA (Extended Industry Standard Architecture) be total
Line etc..It is only indicated with a four-headed arrow in Fig. 1, it is not intended that an only bus or a type of bus.
Electronic equipment 100 realizes the electronic equipment by least one communication interface 104 (can be wired or wireless)
Communication connection between 100 and external equipment.Memory 102 is for storing program, such as hematoma Enlargement in Spontaneous prediction meanss 200.Blood
The swollen prediction meanss 200 that expand include that at least one can be stored in the memory in the form of software or firmware (firmware)
In 102 or the software function module that is solidificated in the operating system (operating system, OS) of electronic equipment 100.It is described
Processor 101 executes described program after receiving and executing instruction to realize hematoma Enlargement in Spontaneous prediction technique.
It should be understood that structure shown in FIG. 1 is only the structure application schematic diagram of electronic equipment 100, electronic equipment 100
It may also include than shown in Fig. 1 more perhaps less component or with the configuration different from shown in Fig. 1.Shown in Fig. 1
Each component can be realized using hardware, software, or its combination.
Based on above-mentioned electronic equipment 100, a kind of possible implementation of hematoma Enlargement in Spontaneous prediction technique, the party is given below
The executing subject of method can be above-mentioned electronic equipment 100, referring to Fig. 2, Fig. 2 shows a kind of blood provided in an embodiment of the present invention
The swollen flow chart for expanding prediction technique.Hematoma Enlargement in Spontaneous prediction technique the following steps are included:
S1 obtains clinical examination data and the CT image comprising hemotoncus.
In embodiments of the present invention, clinical examination data may be, but not limited to, the systolic pressure of patient, diastolic pressure, blood glucose,
Whether renal function Glasgow coma score, Antiplatelet therapy situation, smokes, falls ill to time difference between baseline CT etc.,
CT image, which can be, shoots patient's brain, the obtained CT image comprising hemotoncus.CT image includes multiple faultage images,
And every tension fault image has its corresponding image thicknesses.It should be noted that clinical examination data and CT image belong to together
One patient.
CT image is inputted preset automatic parted pattern and is split, obtains segmentation result, and according to segmentation result by S2
Calculate volume of hematoma.
In embodiments of the present invention, automatic parted pattern can be comprising mixing the U-shaped network of empty convolution sum (referred to as, U-
Net divide network in mixing cavity), wherein U-Net is an image segmentation network.Referring to Fig. 3, mixing cavity segmentation net
Network may include 5 pond layers, include altogether 9 convolution blocks, referring to Fig. 4, the convolution of each convolutional layer in each convolution block
Core is made of tetra- groups of cavity convolution of 3*3 that voidage is respectively 1,2,3,5.
Please continue to refer to Fig. 3, for each convolutional layer in convolution block 1-1, convolution kernel is respectively 1 by 8 voidages, 2,3,
The mixing cavity convolution of the 3 × 3 of 5 forms, totally 32 convolution kernels, to be cut out after avoiding convolution to characteristic pattern, uses outer lining
Value outer lining operation identical with voidage is to keep characteristic pattern size consistent.After convolution operation, batch rule are carried out to characteristic pattern
Generalized then carries out nonlinear activation using line rectification function.In the contraction phase, the one convolution block operation of every process is then right
Characteristic pattern carries out 2 × 2 pondization operation.In the expansionary phase, the one convolution block operation of every process then carries out 2 to primitive character figure
× 2 transposition convolution, and folded operation is carried out with the high-resolution feature of same single order, for example, the input feature vector figure of convolution block 4-2 is total
512 are contained, wherein 256 characteristic pattern transposition convolution by low resolution are come another 256 by directly replicating volume
The high-resolution characteristic pattern of block 4-1 output.
The mixing cavity segmentation network that the embodiment of the present invention proposes has the advantages that following two points: (1) convolution nuclear volume is whole
Body becomes original half, and two layers of convolution operation of each convolution block has been become three-layer coil product operation, so that model is narrower
It is deeper, studies have shown that narrow and deep network effect may be more preferable.(2) neuron number is less, introduces and mixes empty convolution,
Receptive field is promoted, and the feature of multiple dimensioned contextual information can be captured in conjunction with U-Net, so that the network can adapt to difference
The hemotoncus of size, while use outer lining value corresponding with voidage, avoid and are cut out to characteristic pattern.
By using mixing cavity segmentation network, the receptive field range of the neuron of same layer can be expanded, so that
The segmentation network can adapt to various sizes of hemotoncus, and then promote hemotoncus segmentation performance.
Segmentation result can be the hemotoncus region and non-hemotoncus region carry out area in every tension fault image in CT image
After point, obtained result.Volume of hematoma calculation formula is previously stored in electronic equipment 100, mixing cavity segmentation network is defeated
Segmentation result out brings volume into and calculates network, can calculate volume of hematoma.
Referring to Fig. 5, step S2 can specifically include following sub-step:
Multiple faultage images are inputted mixing cavity segmentation network and carry out image segmentation, obtain every tension fault image pair by S21
The segmented image comprising hemotoncus region and non-hemotoncus region answered constitutes segmentation result.
In embodiments of the present invention, segmented image can be faultage image and carry out image segmentation, and what is obtained includes hemotoncus area
The image in domain and the non-hemotoncus region in addition to hemotoncus region, segmentation result can be the corresponding segmentation figure of all faultage images
The summation of picture.Referring to Fig. 6, white portion is hemotoncus region, black portions are non-hemotoncus region.
Multiple faultage images are inputted into mixing cavity segmentation network and carry out image segmentation, it is corresponding to obtain every tension fault image
Segmented image comprising hemotoncus region and non-hemotoncus region, the step of obtaining segmentation result, it can be understood as, by every tension fault figure
As input mixing cavity segmentation network obtains the tomograph so that segmentation network in mixing cavity is split the faultage image
As the corresponding segmented image comprising hemotoncus region and non-hemotoncus region, identical processing is carried out to every tension fault image, it can
To obtain the corresponding segmented image of every tension fault image, the corresponding segmented image of all faultage images together constitutes segmentation knot
Fruit.
S22 counts the hemotoncus pixel quantity in the corresponding hemotoncus region of every tension fault image according to segmentation result.
In embodiments of the present invention, hemotoncus pixel quantity can be the pixel in hemotoncus region in a segmented image
Quantity.Since every tension fault image passes through image segmentation, its corresponding segmented image can be obtained, then described according to segmentation knot
Fruit counts the hemotoncus pixel quantity step in the corresponding hemotoncus region of every tension fault image, it can be understood as, it counts each and breaks
The quantity that the pixel of hemotoncus region (white area in Fig. 6) is characterized in the corresponding segmented image of tomographic image, can be obtained every
The corresponding hemotoncus pixel quantity of the tomography image.
S23 obtains the image thicknesses and hemotoncus pixel point areas of every tension fault image.
In embodiments of the present invention, image thicknesses can be the thickness of scanning slice.For example, 5mm, 9mm, 10mm.Hemotoncus picture
Vegetarian refreshments area can be understood as the area of hemotoncus pixel, the i.e. length of hemotoncus pixel and wide product.Each tension fault image
There are its corresponding image thicknesses and hemotoncus pixel point areas, obtains all faultage images in the shooting for carrying out a CT
Image thicknesses and the hemotoncus pixel point areas of hemotoncus pixel are generally consistent.Obtain every tension fault image image thicknesses and
The step of hemotoncus pixel point areas, it can be understood as, CT file is exported from CT equipment, includes every tension fault figure in CT file
The image thicknesses of picture and the length and width of hemotoncus pixel, according to the length and width of hemotoncus pixel, available hemotoncus pixel
Area.
It should be noted that the execution sequence of step S23 and step S22 can be handed in other embodiments of the invention
It changes, also may be performed simultaneously step S23 and step S22, be not limited thereto.
S24, by the image thicknesses and hemotoncus pixel point areas of all hemotoncus pixel quantity and every tension fault image
It brings volume of hematoma calculation formula into and carries out volume calculating, obtain volume of hematoma.
In embodiments of the present invention, it is previously stored with volume of hematoma calculation formula in electronic equipment 100, step S22 is obtained
To the corresponding hemotoncus region of every tension fault image hemotoncus pixel quantity and the obtained every tension fault image of step S23
Image thicknesses and hemotoncus pixel point areas bring into volume of hematoma calculation formula, the volume of hematoma of patient can be obtained.Hemotoncus
It is as follows that volume calculates formula expression:
Wherein, pixelsiFor the corresponding hemotoncus pixel quantity of the i-th tension fault image, areaiFor the i-th tension fault image
Hemotoncus pixel point areas, thicknessiFor the image thicknesses of the i-th tension fault image, V is volume of hematoma, and n is the interruption of CT image
The quantity of tomographic image.
As an implementation, when image thicknesses are 6mm, hemotoncus pixel point areas is 25mm2, include altogether in CT image
5 tension fault images: the hemotoncus pixel quantity of the first tension fault image is 234;The hemotoncus pixel number of second tension fault image
Amount is 124;The hemotoncus pixel quantity of third tension fault image is 218;The hemotoncus pixel quantity of 4th tension fault image is
327;When the hemotoncus pixel quantity of 5th tension fault image is 116, volume of hematoma V=234*25*6+124*25*6+218*
25*6+327*25*6+116*25*6=152850mm3。
Segmentation result, volume of hematoma and clinical examination data are formed test sample by S3.
In embodiments of the present invention, face what the obtained segmentation result of step S2 and volume of hematoma and step S1 obtained
Bed checks that data are combined, and constitutes the test sample of the patient.
S4 calculates the matching degree in preset multiple support samples between each support sample and test sample, and foundation
Whether multiple matching degrees, prediction hemotoncus can expand.
In embodiments of the present invention, support sample can be segmentation result, the blood of pre-stored non-hematoma Enlargement in Spontaneous patient
Swollen volume and clinical examination data, can also be segmentation result, volume of hematoma and the clinic of pre-stored hematoma Enlargement in Spontaneous patient
Check data.It should be noted that all support samples should belong to same type in the process for executing a step S4
, or be the corresponding support sample of non-hematoma Enlargement in Spontaneous patient, or be the corresponding support sample of hematoma Enlargement in Spontaneous patient.
The mode for obtaining support sample can be consistent with the mode of forecast sample is obtained, and carries out to support sample and test sample identical
Processing.
It is previously stored with multiple same type of support samples in electronic equipment 100, calculates multiple same type of supports
Each of sample supports the matching degree between sample and test sample, obtains multiple matching degrees, and count multiple matching degrees
In be greater than the number of matches of preset matching degree, and preset whether hemotoncus can expand according to the number of matches.
Referring to Fig. 7, step S4 may include following sub-step when supporting sample is the sample of non-hematoma Enlargement in Spontaneous patient
It is rapid:
S41 calculates the first matching degree in preset multiple support samples between each support sample and test sample.
In embodiments of the present invention, it is the corresponding support sample of non-hematoma Enlargement in Spontaneous patient that the first matching degree, which can be support sample,
Matching degree between sheet and test sample.It calculates in preset multiple support samples between each support sample and test sample
The step of first matching degree, it can be understood as, test sample and a support sample are inputted into the preset twin network of three-dimensional, i.e.,
The first matching degree between the support sample and test sample can be calculated, in a manner mentioned above, in multiple support samples
Each support sample standard deviation carry out identical processing, each of multiple support samples support sample can be obtained and test
The first matching degree between sample.
Referring to Fig. 8, step S41 can specifically include following sub-step:
S411 carries out feature extraction to the segmentation result in test sample, obtains image feature vector.
In embodiments of the present invention, referring to Fig. 9, three-dimensional twin network includes that feature extraction network and matching degree calculate net
Network, feature extraction network calculate the calculating that network is used to carry out matching degree for carrying out feature extraction, matching degree.Feature extraction net
Network may include two identical subcharacters and extract network, be respectively used to carry out feature extraction to the segmentation result in test sample
Feature extraction is carried out with to segmentation result in support sample.
The parameter sharing of two sub- feature extraction networks, subcharacter extract network by four convolutional layers and a full articulamentum
Composition, it is every to pass through a convolution operation in convolution block 1-1,1-2 and 1-3, then carry out primary batch of standardization, nonlinear activation
It is operated with maximum pondization, maximum pond ruler of the maximum pond in convolution block 1-1,1-3 having a size of 2 × 2 × 1, in convolution block 1-2
Very little is 2 × 2 × 2, and after convolution block 1-3 operation, characteristic pattern is flattened, and becomes one-dimensional characteristic pattern, and pass through full articulamentum
Feature is become the one-dimensional characteristic vector of 64 length, as image feature vector by 1-1.
S412 splices image feature vector and volume of hematoma, clinical examination data, obtains hemotoncus feature vector.
In embodiments of the present invention, hemotoncus feature vector can be image feature vector and volume of hematoma, clinical examination
After data are spliced, obtained one-dimensional characteristic vector.Image feature vector and volume of hematoma, clinical examination data are spelled
The step of connecing, obtaining hemotoncus feature vector, it can be understood as, firstly, by volume of hematoma and clinical examination number in test sample
According to being spliced, statistical nature 1 is obtained, then, statistical nature 1 is spliced with image feature vector, hemotoncus can be obtained
Feature vector.
S413 obtains the corresponding default feature vector of each support sample.
In embodiments of the present invention, support sample can be segmentation result, the blood of pre-stored non-hematoma Enlargement in Spontaneous patient
Swollen volume and clinical examination data, default feature vector can be the corresponding hemotoncus feature vector of support sample.
The step of obtaining each support sample corresponding default feature vector, it can be understood as, firstly, utilizing feature extraction
Subcharacter in network extracts network and carries out feature extraction, the corresponding spy of the sample that is supported to the segmentation result in support sample
Levy vector;Then, the volume of hematoma in sample will be supported to splice with clinical examination data, obtains statistical nature 2;Finally,
The feature vector corresponding with support sample of statistical nature 2 is spliced, default feature vector can be obtained.Wherein, spy is utilized
Sign extracts the subcharacter in network and extracts network to the segmentation result progress feature extraction in support sample, and be supported sample pair
The specific process for the feature vector answered is consistent with to the segmentation result progress characteristic extraction procedure in test sample, herein no longer
It repeats.
It should be noted that it is corresponding to be required to acquisition support sample due to for the test sample that each is inputted
Default feature vector, then, default feature vector corresponding for the support sample obtained can store and set in electronics
In standby 100 memory 102, when obtaining the corresponding default feature vector of support sample again, direct use, without carrying out
It computes repeatedly.Only for the support sample newly increased, above-mentioned step is just executed.
S414 calculates the matching degree between hemotoncus feature vector and each default feature vector.
In embodiments of the present invention, matching degree is inputted after hemotoncus feature vector and default feature vector being spliced to calculate
Network, matching degree calculate network and can be made of three full articulamentum blocks, by the hemotoncus feature vector after splicing and preset spy
Levy vector and input full link block 2-1, obtain the feature vector that a length is 128, then, by the length be 128 feature to
Amount inputs full link block 2-2, obtains the feature vector that a length is 64, finally, the feature vector that the length is 64 is inputted
One 1 × 1 full articulamentum obtains the matching degree between hemotoncus feature vector and default feature vector, as the first matching degree.
Above-mentioned identical processing is carried out to the corresponding default feature vector of each support sample, hemotoncus feature can be obtained
The first matching degree between vector and each default feature vector.
S42 compares the first matching degree of each of multiple first matching degrees with preset matching value.
S43 counts the first number of matches for being greater than preset matching value in multiple first matching degrees.
In embodiments of the present invention, the first number of matches is first for being greater than preset matching value in multiple first matching degrees
Quantity with degree.By the first matching degree of each of multiple first matching degrees obtained in step S41 with preset matching value
(for example, 60%) is compared, and the first matching degree that will be greater than preset matching value sets 1, will be less than first of preset matching value
0 is set with degree, the quantity that the first matching degree is 1 is then counted, the first number of matches can be obtained.
For example, preset matching value is 60% when multiple first matching degrees are respectively 80%, 30%, 25%, 65%, 73%
When, the first matching degree 80%, 65%, 73% is all larger than preset matching value 60%, by the first matching degree 80%, 65%, 73%
It is set to 1, the first matching degree 30%, 25% is respectively less than preset matching value 60%, the first matching degree 30%, 25% is set to 0, most
Afterwards, the quantity for counting 1 is 3, and the first number of matches is 3.
S44, when the first number of matches is greater than preset quantity, prediction hemotoncus will not expand.
In embodiments of the present invention, the first number of matches and preset quantity (for example, 60) are compared, when the first matching
When quantity is greater than preset quantity, prediction hemotoncus will not expand.
S45, when the first number of matches is less than or equal to preset quantity, prediction hemotoncus can expand.
In embodiments of the present invention, the first number of matches and preset quantity (for example, 60) are compared, when the first matching
When quantity is less than or equal to preset quantity, prediction hemotoncus can expand.
It should be noted that predicted quantity is related to the support quantity of sample, the 60% of support sample size can be,
That is predicted quantity can be 60 when supporting the quantity of sample to be 100.
Referring to Fig. 10, step S4 can also include following sub-step when supporting sample is the sample of hematoma Enlargement in Spontaneous patient
It is rapid:
S46 calculates the second matching degree in preset multiple support samples between each support sample and test sample.
In embodiments of the present invention, it is the corresponding support sample of hematoma Enlargement in Spontaneous patient that the second matching degree, which can be support sample,
Matching degree between test sample.Calculate the in preset multiple support samples between each support sample and test sample
The step of two matching degrees, it can be understood as, test sample and a support sample are inputted into the preset twin network of three-dimensional
The second matching degree between the support sample and test sample is calculated, in a manner mentioned above, in multiple support samples
Each support sample standard deviation carries out identical processing, and each of multiple support samples support sample and test specimens can be obtained
The second matching degree between this.
The step of calculating the second matching degree of each of the multiple support samples between support sample and test sample, can
With referring particularly to sub-step S411-S414, details are not described herein.
S47 compares the second matching degree of each of multiple second matching degrees with preset matching value.
S48 counts the second number of matches for being greater than preset matching value in multiple second matching degrees.
In embodiments of the present invention, the second number of matches is second for being greater than preset matching value in multiple second matching degrees
Quantity with degree.By the second matching degree of each of multiple second matching degrees obtained in step S46 with preset matching value
(for example, 60%) is compared, and the second matching degree that will be greater than preset matching value sets 1, will be less than second of preset matching value
0 is set with degree, the quantity that the second matching degree is 1 is then counted, the second number of matches can be obtained.
For example, preset matching value is 60% when multiple second matching degrees are respectively 80%, 40%, 15%, 65%, 83%
When, the second matching degree 80%, 65%, 83% is all larger than preset matching value 60%, by the second matching degree 80%, 65%, 83%
It is set to 1, the second matching degree 40%, 15% is respectively less than preset matching value 60%, the second matching degree 40%, 15% is set to 0, most
Afterwards, the quantity for counting 1 is 3, and the second number of matches is 3.
S49, when the second number of matches is greater than preset quantity, prediction hemotoncus can expand.
In embodiments of the present invention, the second number of matches and preset quantity (for example, 60) are compared, when the second matching
When quantity is greater than preset quantity, prediction hemotoncus can expand.
S410, when the second number of matches is less than or equal to preset quantity, prediction hemotoncus will not expand.
In embodiments of the present invention, the second number of matches and preset quantity (for example, 60) are compared, when the second matching
When quantity is less than or equal to preset quantity, prediction hemotoncus will not expand.
Compared with prior art, the embodiment of the present invention has the advantage that
Firstly, characteristics of image is automatically extracted using neural network, so that the characteristics of image extracted is not by the subjectivity of people
It influences, and can effectively promote diagosis efficiency, and effectively reduce cost.
Secondly, having comprehensively considered characteristics of image and statistical nature to predict whether hemotoncus can expand, hematoma Enlargement in Spontaneous is improved
The accuracy rate of prediction.
Finally, the receptive field range of the neuron of same layer can be expanded using mixing cavity segmentation network, so that
Segmentation network can adapt to various sizes of hemotoncus, and then promote hemotoncus segmentation performance.
For the method flow of above-mentioned Fig. 2,5,7,8 and Figure 10, be given below a kind of hematoma Enlargement in Spontaneous prediction meanss 200 can
The implementation of energy, the hematoma Enlargement in Spontaneous prediction meanss 200 can be using the device architectures of the electronic equipment 100 in above-described embodiment
It realizes, or the processor 101 in the electronic equipment 100 is realized, please refers to Figure 11, Figure 11 shows the embodiment of the present invention
The block diagram of the hematoma Enlargement in Spontaneous prediction meanss of offer.Hematoma Enlargement in Spontaneous prediction meanss 200 include obtaining module 201 and processing mould
Block 202.
Obtain module 201, the CT image for obtaining clinical examination data and comprising hemotoncus;
Processing module 202 is split for CT image to be inputted preset automatic parted pattern, obtains segmentation result,
And volume of hematoma is calculated according to segmentation result;Segmentation result, volume of hematoma and clinical examination data are formed into test sample;Meter
The matching degree in preset multiple support samples between each support sample and test sample is calculated, and according to multiple matching degrees, in advance
Survey whether hemotoncus can expand.
In embodiments of the present invention, when supporting sample is the sample of non-hematoma Enlargement in Spontaneous patient, processing module 202 executes meter
The matching degree in preset multiple support samples between each support sample and test sample is calculated, and according to multiple matching degrees, in advance
The step of whether hemotoncus can expand is surveyed, is specifically used for: calculating each support sample and test specimens in preset multiple support samples
The first matching degree between this;The first matching degree of each of multiple first matching degrees is compared with preset matching value;
Count the first number of matches for being greater than preset matching value in multiple first matching degrees;When the first number of matches is greater than preset quantity
When, prediction hemotoncus will not expand;When the first number of matches is less than or equal to preset quantity, prediction hemotoncus can expand.
In embodiments of the present invention, when supporting sample is the sample of hematoma Enlargement in Spontaneous patient, processing module 202 executes calculating
Matching degree in preset multiple support samples between each support sample and test sample, and according to multiple matching degrees, prediction
The step of whether hemotoncus can expand, is specifically used for: calculating each support sample and test sample in preset multiple support samples
Between the second matching degree;The second matching degree of each of multiple second matching degrees is compared with preset matching value;System
Count the second number of matches for being greater than preset matching value in multiple second matching degrees;When the second number of matches is greater than preset quantity,
Prediction hemotoncus can expand;When the second number of matches is less than or equal to preset quantity, prediction hemotoncus will not expand.
In embodiments of the present invention, automatic parted pattern includes mixing cavity segmentation network, and CT image includes multiple tomographies
Image, processing module 202, which is executed, to be inputted preset automatic parted pattern for CT image and is split, and obtains segmentation result, and according to
The step of calculating volume of hematoma according to segmentation result is specifically used for: by multiple faultage images input mixing cavity segmentation network into
Row image segmentation obtains the corresponding segmented image comprising hemotoncus region and non-hemotoncus region of every tension fault image, constitutes segmentation
As a result;According to segmentation result, the hemotoncus pixel quantity in the corresponding hemotoncus region of every tension fault image is counted;Obtain every tension fault
The image thicknesses and hemotoncus pixel point areas of image;By the image of all hemotoncus pixel quantity and every tension fault image
Thickness and hemotoncus pixel point areas bring volume of hematoma calculation formula into and carry out volume calculating, obtain volume of hematoma.
In embodiments of the present invention, volume of hematoma calculation formula expression formula is as follows:
Wherein, pixelsiFor the corresponding hemotoncus pixel quantity of the i-th tension fault image, areaiFor the i-th tension fault image
Hemotoncus pixel point areas, thicknessiFor the image thicknesses of the i-th tension fault image, n is the quantity that CT image interrupts tomographic image,
V is volume of hematoma.
In embodiments of the present invention, mixing cavity segmentation network includes 5 pond layers, altogether includes 9 convolution blocks, Mei Gejuan
The convolution kernel of each convolutional layer in block is made of tetra- groups of cavity convolution of 3*3 that voidage is respectively 1,2,3,5.
In embodiments of the present invention, processing module 202, which executes, calculates each support sample in preset multiple support samples
The step of matching degree between test sample, it is specifically used for: feature extraction is carried out to the segmentation result in test sample, is obtained
Image feature vector;Image feature vector and volume of hematoma, clinical examination data are spliced, hemotoncus feature vector is obtained;
Obtain the corresponding default feature vector of each support sample;Calculate between hemotoncus feature vector and each default feature vector
With degree.
It is apparent to those skilled in the art that for convenience and simplicity of description, the hemotoncus of foregoing description
The specific work process for expanding prediction meanss 200, can refer to corresponding processes in the foregoing method embodiment, no longer superfluous herein
It states.
In conclusion the embodiment of the present invention provides a kind of hematoma Enlargement in Spontaneous prediction technique and device, which comprises obtain
Clinical examination data and CT image comprising hemotoncus;CT image is inputted preset automatic parted pattern to be split, is divided
It cuts as a result, and calculating volume of hematoma according to segmentation result;Segmentation result, volume of hematoma and clinical examination data are formed and are tested
Sample;The matching degree in preset multiple support samples between each support sample and test sample is calculated, and according to multiple
With degree, predict whether hemotoncus can expand.Compared with prior art, present invention has the advantage that according to existing image data
It predicts with clinical examination data whether hemotoncus can expand, whether can occur hematoma Enlargement in Spontaneous, and carry out if can predict in advance
Corresponding processing, avoids and determines hematoma Enlargement in Spontaneous not in time, and leading to some patientss, sb.'s illness took a turn for the worse during this period or even dead feelings
Condition.
In several embodiments provided herein, it should be understood that disclosed device and method can also pass through
Other modes are realized.The apparatus embodiments described above are merely exemplary, for example, flow chart and block diagram in attached drawing
Show the device of multiple embodiments according to the present invention, the architectural framework in the cards of method and computer program product,
Function and operation.In this regard, each box in flowchart or block diagram can represent the one of a module, section or code
Part, a part of the module, section or code, which includes that one or more is for implementing the specified logical function, to be held
Row instruction.It should also be noted that function marked in the box can also be to be different from some implementations as replacement
The sequence marked in attached drawing occurs.For example, two continuous boxes can actually be basically executed in parallel, they are sometimes
It can execute in the opposite order, this depends on the function involved.It is also noted that every in block diagram and or flow chart
The combination of box in a box and block diagram and or flow chart can use the dedicated base for executing defined function or movement
It realizes, or can realize using a combination of dedicated hardware and computer instructions in the system of hardware.
In addition, each functional module in each embodiment of the present invention can integrate one independent portion of formation together
Point, it is also possible to modules individualism, an independent part can also be integrated to form with two or more modules.
It, can be with if the function is realized and when sold or used as an independent product in the form of software function module
It is stored in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially in other words
The part of the part that contributes to existing technology or the technical solution can be embodied in the form of software products, the meter
Calculation machine software product is stored in a storage medium, including some instructions are used so that a computer equipment (can be a
People's computer, server or network equipment etc.) it performs all or part of the steps of the method described in the various embodiments of the present invention.
And storage medium above-mentioned includes: that USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), arbitrary access are deposited
The various media that can store program code such as reservoir (RAM, Random Access Memory), magnetic or disk.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.It should also be noted that similar label and letter exist
Similar terms are indicated in following attached drawing, therefore, once being defined in a certain Xiang Yi attached drawing, are then not required in subsequent attached drawing
It is further defined and explained.
Claims (10)
1. a kind of hematoma Enlargement in Spontaneous prediction technique, which is characterized in that the described method includes:
Obtain clinical examination data and the CT scan CT image comprising hemotoncus;
The CT image is inputted preset automatic parted pattern to be split, obtains segmentation result, and tie according to the segmentation
Fruit calculates volume of hematoma;
The segmentation result, volume of hematoma and the clinical examination data are formed into test sample;
Calculate the matching degree in preset multiple support samples between each support sample and the test sample, and foundation
Multiple matching degrees, predict whether the hemotoncus can expand.
2. the method according to claim 1, wherein when the sample that the support sample is non-hematoma Enlargement in Spontaneous patient
When, the matching degree calculated in preset multiple support samples between each support sample and the test sample, and
According to multiple matching degrees, the step of whether hemotoncus can expand predicted, comprising:
Calculate the first matching degree in preset multiple support samples between each support sample and the test sample;
The first matching degree of each of multiple first matching degrees is compared with preset matching value;
Count the first number of matches for being greater than the preset matching value in the multiple first matching degree;
When first number of matches is greater than preset quantity, prediction hemotoncus will not expand;
When first number of matches is less than or equal to preset quantity, prediction hemotoncus can expand.
3. the method according to claim 1, wherein when the sample that the support sample is hematoma Enlargement in Spontaneous patient
When, the matching degree calculated in preset multiple support samples between each support sample and the test sample, and
According to multiple matching degrees, the step of whether hemotoncus can expand predicted, comprising:
Calculate the second matching degree in preset multiple support samples between each support sample and the test sample;
The second matching degree of each of multiple second matching degrees is compared with preset matching value;
Count the second number of matches for being greater than the preset matching value in the multiple second matching degree;
When second number of matches is greater than preset quantity, prediction hemotoncus can expand;
When second number of matches is less than or equal to preset quantity, prediction hemotoncus will not expand.
4. the method according to claim 1, wherein the automatic parted pattern includes mixing cavity segmentation net
Network, the CT image include multiple faultage images, described to be split the preset automatic parted pattern of CT image input,
Obtain segmentation result, and the step of calculating volume of hematoma according to the segmentation result, comprising:
Multiple described faultage images are inputted into the mixing cavity segmentation network and carry out image segmentation, obtain every tension fault image pair
The segmented image comprising hemotoncus region and non-hemotoncus region answered constitutes segmentation result;
According to segmentation result, the hemotoncus pixel quantity in the corresponding hemotoncus region of every tension fault image is counted;
Obtain the image thicknesses and hemotoncus pixel point areas of every tension fault image;
The image thicknesses and hemotoncus pixel point areas of all hemotoncus pixel quantity and every tension fault image are brought into
Volume of hematoma calculation formula carries out volume calculating, obtains volume of hematoma.
5. according to the method described in claim 4, it is characterized in that, the volume of hematoma calculation formula expression formula is as follows:
Wherein, pixelsiFor the corresponding hemotoncus pixel quantity of the i-th tension fault image, areaiFor the hemotoncus of the i-th tension fault image
Pixel point areas, thicknessiFor the image thicknesses of the i-th tension fault image, n is the quantity that CT image interrupts tomographic image, and V is
Volume of hematoma.
6. according to the method described in claim 4, it is characterized in that, segmentation network in mixing cavity includes 5 pond layers, altogether
Comprising 9 convolution blocks, the convolution kernel of each convolutional layer in each convolution block is respectively 1,2,3,5 3*3 tetra- by voidage
The empty convolution composition of group.
7. method according to claim 1-3, which is characterized in that described to calculate in preset multiple support samples
The step of each matching degree supported between sample and the test sample, comprising:
Feature extraction is carried out to the segmentation result in the test sample, obtains image feature vector;
Described image feature vector and the volume of hematoma, clinical examination data are spliced, hemotoncus feature vector is obtained;
Obtain the corresponding default feature vector of each support sample;
Calculate the matching degree between the hemotoncus feature vector and each default feature vector.
8. a kind of hematoma Enlargement in Spontaneous prediction meanss, which is characterized in that described device includes:
Obtain module, the CT scan CT image for obtaining clinical examination data and comprising hemotoncus;
Processing module is split for the CT image to be inputted preset automatic parted pattern, obtains segmentation result, and according to
Volume of hematoma is calculated according to the segmentation result;The segmentation result, volume of hematoma and the clinical examination data are formed and are surveyed
Sample sheet;The matching degree in preset multiple support samples between each support sample and the test sample is calculated, and
According to multiple matching degrees, predict whether the hemotoncus can expand.
9. device according to claim 8, which is characterized in that when the support sample is the sample that hemotoncus does not expand patient
When, the processing module is specifically used for:
Calculate the first matching degree in preset multiple support samples between each support sample and the test sample;
The first matching degree of each of multiple first matching degrees is compared with preset matching value;
Count the first number of matches for being greater than the preset matching value in the multiple first matching degree;
When first number of matches is greater than preset quantity, prediction hemotoncus will not expand;
When first number of matches is less than or equal to preset quantity, prediction hemotoncus can expand.
10. device according to claim 8, which is characterized in that when the support sample is that volume of hematoma expands patient's
When sample, the processing module is specifically used for:
Calculate the second matching degree in preset multiple support samples between each support sample and the test sample;
The second matching degree of each of multiple second matching degrees is compared with preset matching value;
Count the second number of matches for being greater than the preset matching value in the multiple second matching degree;
When second number of matches is greater than preset quantity, prediction hemotoncus can expand;
When second number of matches is less than or equal to preset quantity, prediction hemotoncus will not expand.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910583573.4A CN110288589B (en) | 2019-06-28 | 2019-06-28 | Hematoma expansion prediction method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910583573.4A CN110288589B (en) | 2019-06-28 | 2019-06-28 | Hematoma expansion prediction method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110288589A true CN110288589A (en) | 2019-09-27 |
CN110288589B CN110288589B (en) | 2021-07-02 |
Family
ID=68020860
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910583573.4A Active CN110288589B (en) | 2019-06-28 | 2019-06-28 | Hematoma expansion prediction method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110288589B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111079671A (en) * | 2019-12-20 | 2020-04-28 | 深圳集智数字科技有限公司 | Method and device for detecting abnormal articles in scene |
CN111723817A (en) * | 2020-06-30 | 2020-09-29 | 重庆大学 | Pulmonary nodule auxiliary detection method |
CN113349810A (en) * | 2021-05-27 | 2021-09-07 | 北京安德医智科技有限公司 | Cerebral hemorrhage focus identification and hematoma expansion prediction method and device |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016205811A1 (en) * | 2015-06-19 | 2016-12-22 | New York University | Determination of accelerated brain atrophy using computed tomography |
CN107506579A (en) * | 2017-08-14 | 2017-12-22 | 西南大学 | Cerebral hemorrhage forecast model method for building up and system based on integrated study |
CN109190690A (en) * | 2018-08-17 | 2019-01-11 | 东北大学 | The Cerebral microbleeds point detection recognition method of SWI image based on machine learning |
-
2019
- 2019-06-28 CN CN201910583573.4A patent/CN110288589B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016205811A1 (en) * | 2015-06-19 | 2016-12-22 | New York University | Determination of accelerated brain atrophy using computed tomography |
CN107506579A (en) * | 2017-08-14 | 2017-12-22 | 西南大学 | Cerebral hemorrhage forecast model method for building up and system based on integrated study |
CN109190690A (en) * | 2018-08-17 | 2019-01-11 | 东北大学 | The Cerebral microbleeds point detection recognition method of SWI image based on machine learning |
Non-Patent Citations (2)
Title |
---|
WU GANG ETC.: ""A New Method for Predicting Intracerebral Hematoma Expansion:Volume Difference Measured by ABC/2 and Slicer Software"", 《第十三届中国医师协会神经外科医师年会摘要集》 * |
张华博: ""基于深度学习的图像分割研究与应用"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111079671A (en) * | 2019-12-20 | 2020-04-28 | 深圳集智数字科技有限公司 | Method and device for detecting abnormal articles in scene |
CN111723817A (en) * | 2020-06-30 | 2020-09-29 | 重庆大学 | Pulmonary nodule auxiliary detection method |
CN111723817B (en) * | 2020-06-30 | 2023-09-29 | 重庆大学 | Auxiliary detection method for pulmonary nodules |
CN113349810A (en) * | 2021-05-27 | 2021-09-07 | 北京安德医智科技有限公司 | Cerebral hemorrhage focus identification and hematoma expansion prediction method and device |
CN113349810B (en) * | 2021-05-27 | 2022-03-01 | 北京安德医智科技有限公司 | Cerebral hemorrhage focus identification and hematoma expansion prediction system and device |
Also Published As
Publication number | Publication date |
---|---|
CN110288589B (en) | 2021-07-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110288589A (en) | Hematoma Enlargement in Spontaneous prediction technique and device | |
Kramer et al. | Aerobic exercise for women during pregnancy | |
CN110197724A (en) | Predict the method, apparatus and computer equipment in diabetes illness stage | |
JP4879368B2 (en) | Medical work support device | |
CN111710420A (en) | Complication morbidity risk prediction method, system, terminal and storage medium based on electronic medical record big data | |
JP6634163B2 (en) | Method and apparatus for extracting diagnostic objects from medical documents | |
Appannagari et al. | Risk factors for inadequate colonoscopy bowel preparations in African Americans and whites at an urban medical center | |
Jiang et al. | Magnetic resonance imaging (mri) brain tumor image classification based on five machine learning algorithms | |
Beaulieu et al. | Endoscopy reporting standards | |
CN112634231A (en) | Image classification method and device, terminal equipment and storage medium | |
JP2014067344A (en) | Graph creation device, graph creation method, and graph creation program | |
EP3805807A1 (en) | Tomographic image prediction device and tomographic image prediction method | |
CN112017784B (en) | Coronary heart disease risk prediction method based on multi-modal data and related equipment | |
Jusuf et al. | Assessing acne severity: teledermatology versus face to face consultations during the COVID-19 pandemic | |
JP5846925B2 (en) | Medical image display device | |
KR101472709B1 (en) | Health Medical Examination Method including Detailed Examination Information | |
CN112990339A (en) | Method and device for classifying stomach pathological section images and storage medium | |
KR20210145359A (en) | Method for providing information on diagnosing renal failure and device using the same | |
JP2011113428A (en) | Medical information processing apparatus and program | |
JP5796319B2 (en) | Inspection result management apparatus, inspection result management program, and inspection result management method | |
CN111128330A (en) | Automatic entry method and device for electronic case report table and related equipment | |
CN111599480A (en) | Method, device, terminal and readable medium for evaluating adverse drug reactions | |
Pionteck et al. | Intracranial aneurysm wall displacement predicts instability | |
Suwandi et al. | A Systematic Literature Review: Diabetic Retinopathy Detection Using Deep Learning | |
CN116580841B (en) | Disease diagnosis device, device and storage medium based on multiple groups of study data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |