CN110633385B - Medical image retrieval and compression method - Google Patents
Medical image retrieval and compression method Download PDFInfo
- Publication number
- CN110633385B CN110633385B CN201910907305.3A CN201910907305A CN110633385B CN 110633385 B CN110633385 B CN 110633385B CN 201910907305 A CN201910907305 A CN 201910907305A CN 110633385 B CN110633385 B CN 110633385B
- Authority
- CN
- China
- Prior art keywords
- image
- densenet
- medical
- medical image
- searched
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000006835 compression Effects 0.000 title claims abstract description 19
- 238000007906 compression Methods 0.000 title claims abstract description 17
- 238000005516 engineering process Methods 0.000 claims abstract description 13
- 238000004891 communication Methods 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 17
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000003708 edge detection Methods 0.000 claims description 10
- 230000009467 reduction Effects 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 7
- 238000000354 decomposition reaction Methods 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 abstract description 6
- 238000004458 analytical method Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/136—Incoming video signal characteristics or properties
- H04N19/14—Coding unit complexity, e.g. amount of activity or edge presence estimation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/63—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding using sub-band based transform, e.g. wavelets
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/85—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/14—Picture signal circuitry for video frequency region
- H04N5/142—Edging; Contouring
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Health & Medical Sciences (AREA)
- Radiology & Medical Imaging (AREA)
- Medical Informatics (AREA)
- General Engineering & Computer Science (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Physics & Mathematics (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Compression Of Band Width Or Redundancy In Fax (AREA)
Abstract
The invention discloses a medical image retrieval and compression method, which comprises the following steps of 101, retrieving a medical image; step 201, compressing each image obtained in step 101; step 301, the programs in step 101 and step 201 are imported into RAM of a single-chip microcomputer, a chip of 5G network communication technology is inserted into the single-chip microcomputer to transmit by means of a 5G network, and the single-chip microcomputer is provided with a USB interface to be inserted into a host computer of a hospital computer, so that medical image retrieval and compression and resource sharing among different hospitals are achieved. The method can realize the retrieval and compression of medical images. The medical image retrieval and compression software is packaged through the singlechip, and the medical image information transmission among different hospitals is realized by adopting a 5G network communication technology for transmission, so that medical resource sharing is realized.
Description
Technical Field
The invention relates to the technical field of medical image retrieval and compression, and a singlechip technology and a 5G network communication technology, which can be used for medical image retrieval and compression and realize medical resource sharing among different hospitals.
Background
With the rapid development of medical imaging technology, hospitals generate a large amount of medical image data every day, and how to effectively and rapidly retrieve and compress the medical image data in a large amount of medical image data has become a problem to be solved. The medical image can be searched to play a role in assisting a diagnosis doctor, so that the working efficiency of the doctor can be improved to a certain extent; the medical image is compressed, so that the storage space of medical image data can be effectively reduced, medical image data transmission can be better performed, and the medical image compression method can be applied to medical information sharing and telemedicine.
2014 Xia et al propose a convolutional neural network based hash algorithm (Convolution Neural Network Hashing, CNNH) which is a completely new attempt to combine CNN with hash (see, xia R, pan Y, lai H, et al, supervised hashing for image retrieval via image representation learning// Proceedings of the 23rd International Joint Conference on Artificial Intelligence.Quebec City,Canada,2014:2156-2162.). The dense convolutional neural networks (Dense Convolutional Network, denseNet) were proposed by Gao Huang, zhuang Liu at the university of Qinghua and Laurensvan der Maaten at Facebook AI Research of Kannel, 2016 (see Huang G, liu Z, kilian Q W, et al Densey connected convolutional networks [ EB/OL ]. ArXiv preprint arXiv,1608.06993,2016.), and image retrieval methods based on DenseNet and supervised hashing were employed in the present invention. The basic principle of the method is as follows: firstly, extracting high-level semantic features of an image by using a trained and optimized DenseNet model, and secondly, carrying out hash coding on the extracted features by using an improved supervised hash code, thereby realizing retrieval.
In 1996, said et al proposed an SPIHT image compression method (see Said a, pearlman W A.A New, fast, and Efficient Image Code Based on Set Partitioning in Hierarchical Trees [ J ]. IEEE Transactions on Circuits and Systems for Video Technology,1996,6 (3): 243-250.), but this method would lose high frequency information such as texture and contour of an image, and the present invention adopts an image compression method combining Canny (see j.canny.a computational approach to edge detection.ieee Transactions on Pattern Analysis and Machine Intelligence,1986,8 (6): 679-698) and SPIHT in consideration of importance of the high frequency information for medical image diagnosis. The basic principle of the method is as follows: firstly, carrying out Canny edge detection on an image, and carrying out Huffman coding and decoding on the extracted edge image to obtain an edge reconstruction image; secondly, encoding the image by using a SPIHT algorithm, and performing Huffman encoding and decoding on the encoded code stream, and obtaining a pair of reconstruction images after decoding by using the SPIHT algorithm and wavelet inverse transformation; and finally, adding the two obtained reconstructed images to restore the original image. The method has the advantages that the high-frequency information of the reconstructed image is well reserved, but a part of high-frequency information redundancy also has the defect that the visual effect of the reconstructed image is reduced.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and complete medical image retrieval, compression and transmission by means of a singlechip technology and a 5G network communication technology, thereby realizing resource sharing among different hospitals.
The invention relates to a method for searching and compressing medical images, which comprises the following steps:
step 101, retrieving a medical image:
step one, preprocessing each medical image in an image dataset formed by a plurality of medical images, wherein the steps are as follows:
first, converting a medical image in a Dicom format into a jpg format and converting a medical image resolution size into 224x224 by a computer program;
secondly, converting the single-channel medical image into a three-channel image through a computer vision library Opencv;
step two, processing the preprocessed image data set to construct an image feature library, which comprises the following specific steps:
firstly, constructing a DenseNet model in a Spyder editor of Python 3.6;
dividing the preprocessed image data set into a training set and a testing set, wherein 70% of three-channel image data are used as the training set, then performing normalization processing on neurons input into the DenseNet network model by using a BN algorithm to serve as neurons input into the DenseNet network model, performing training optimization on the DenseNet model by using the training set, and performing testing on the DenseNet network model by using the testing set; when training the DenseNet network model, optimizing a loss function of the DenseNet network model by adopting a RMSProp algorithm;
thirdly, inputting each image in the image dataset into a DenseNet model after training and optimization, and then extracting the output of the last pooling layer of the DenseNet model as the characteristic of each image to construct an image characteristic library for forming the image dataset;
step three, performing dimension reduction processing on the data in the image feature library by using a KPCA projection method;
step four, respectively carrying out KSH coding on the image feature data of each image after KPCA projection processing to obtain a hash code base of the image features;
inputting an image to be searched in the optimized DenseNet model, extracting the output of the last pooling layer of the DenseNet model as the characteristic of the image to be searched, performing dimension reduction processing on the characteristic data of the image to be searched by using a KPCA projection method, performing KSH coding on the characteristic data of the image to be searched after the dimension reduction processing to obtain a hash code of the image to be searched, comparing the Hamming distance between the hash code of the image to be searched and a Hamming code library, performing similarity measurement, and arranging images in the Hamming code library with the Hamming distance within a set range of the image to be searched according to the sequence of sequentially increasing distances and storing the images as search results;
step 201, compressing each image obtained in step 101, including the following steps:
firstly, carrying out Canny edge detection on each retrieved image, and extracting high-frequency information of the image;
secondly, performing Huffman coding on the image high-frequency information extracted by Canny edge detection, and performing Huffman decoding on the obtained code stream to obtain an edge reconstruction image;
performing wavelet decomposition on the retrieved image, wherein the decomposition level is 5, performing SPIHT coding on wavelet coefficients, performing Huffman coding on the code stream to obtain an optimized compressed code stream, and sequentially performing Huffman decoding, SPIHT decoding and wavelet inverse transformation on the optimized compressed code stream to obtain a reconstructed image with high-frequency information loss;
step four, adding the edge reconstruction image obtained in the step two in the step 201 and the reconstruction image with loss of high-frequency information obtained in the step three, and recovering to obtain an original image;
step 301, the programs in step 101 and step 201 are imported into RAM of a single-chip microcomputer, a chip of 5G network communication technology is inserted into the single-chip microcomputer to transmit by means of a 5G network, and the single-chip microcomputer is provided with a USB interface to be inserted into a host computer of a hospital computer, so that medical image retrieval and compression and resource sharing among different hospitals are achieved.
The beneficial effects of the invention are as follows:
(1) The retrieval and compression of medical images can be realized. The method can effectively assist the diagnostician by searching in a large number of medical images, thereby improving the working efficiency. The medical image compression method can well reserve the high-frequency information of the medical image, thereby ensuring that important diagnostic information is not lost.
(2) The medical image retrieval and compression software is packaged through the singlechip, and the medical image information transmission among different hospitals is realized by adopting a 5G network communication technology for transmission, so that medical resource sharing is realized.
Drawings
FIG. 1 is a flowchart of medical image retrieval;
fig. 2 is a flow chart of image compression.
Detailed Description
The invention will now be described in detail with reference to the drawings and to specific embodiments.
The invention relates to a medical image searching and compressing method as shown in the attached drawings, which comprises the following steps:
step 101, retrieving a medical image:
in order to make the medical image meet the input requirement of the DenseNet model, each medical image in the image dataset formed by a plurality of medical images needs to be preprocessed, and the steps are as follows:
firstly, converting a medical image in a Dicom format into a jpg format through a computer program, wherein the medical image resolution in the Dicom format is mostly 512x512, and the medical image resolution is required to be converted into 224x224 in order to meet the input requirement of a model;
secondly, converting the single-channel medical image into a three-channel image through a computer vision library Opencv;
step two, processing the preprocessed image data set to construct an image feature library, which comprises the following specific steps:
firstly, constructing a DenseNet model in a Spyder editor of Python 3.6;
dividing the preprocessed image data set into a training set and a test set, wherein 70% of three-channel image data are used as the training set, then using BN algorithm (see Ioffe S, szegedy C.batch normalization: accelerating deep network training by reducing internal covariate shift [ C ]. Proceedings of 32nd International Conference on Machine Learning,Piscataway,NJ:IEEE,2015:148-156.-Ioffe S, szegedy C. Bulk normalization: accelerating deep network training [ C ]. 32nd machine learning international conference discussion set by reducing internal coordination variable movements, piscataway, NJ: IEEE,2015: 148-156.). Normalization processing is performed on neurons input to the DenseNet network model as neurons input to the DenseNet network model, training optimization is performed on the DenseNet model by using the training set, and testing of the DenseNet network model is performed by using the test set; when training the DenseNet network model, the loss function of the DenseNet network model is optimized by adopting an RMSProp algorithm (see Tielemant, hinton G.RMSProp: divide the gradient by a running average of its recent magnitude [ R ]. COURSERA: neural Networks for Machine Learning, 2012-Tielemant, hinton G.RMSProp: dividing the gradient by the moving average value of the nearest gradient [ R ]. COURERA: machine-learned neural network, 2012.), and adopting an RMSProp algorithm to optimize the loss function so as to solve the problem of large swing amplitude of the DenseNet network model in the updating process and accelerate the convergence speed of the DenseNet network model, thereby realizing model optimization and increasing the accuracy and robustness of the model.
The BN algorithm is used in this step to well address the effects of shifting and increasing the input data.
As one embodiment of the present invention, the parameters of the DenseNet model at training are set as follows: the learning rate (learning rate) was 1e-6, the batch size (batch_size) was 32, and the number of rounds (epochs) was 70.
Thirdly, inputting each image in the image dataset into a DenseNet model after training and optimization, and then extracting the output of the last pooling layer of the DenseNet model as the characteristic of each image to construct an image characteristic library for forming the image dataset.
And thirdly, performing dimension reduction treatment on the data in the image feature library by using a KPCA projection method, fully mining nonlinear features contained in the data through KPCA projection (see H.Hotelling. Analysis of a complex of statistical variables into principal components [ J ]. Journal of educational psychology,1993,24 (6): 417-H.Hotelling. Analyzing complex statistical variables into main components [ J ]. Education psychology journal, 1993,24 (6): 417), and reducing projection errors.
And fourthly, respectively carrying out KSH coding on the image characteristic data of each image after KPCA projection processing (see Liu Wei, wang Jun, ji Rongrong, et al, superviced hashing with kernels [ C ]// Proc of IEEE Conference on Computer Vision and Pattern Recognition.Piscataway, NJ:IEEE Press,2012:2074-2081.— Liu Wei, wang Jun, ji Rongrong, and the like).
Inputting an image to be searched in the optimized DenseNet model, extracting the output of the last pooling layer of the DenseNet model as the characteristic of the image to be searched, performing dimension reduction processing on the characteristic data of the image to be searched by using a KPCA projection method, performing KSH coding on the characteristic data of the image to be searched after the dimension reduction processing to obtain a hash code of the image to be searched, comparing the Hamming distance between the hash code of the image to be searched and a Hamming code library, performing similarity measurement, and arranging images (usually only taking 20 images) in the Hamming code library with the Hamming distance within a set range of the image to be searched according to the sequence of sequentially increasing distances and storing the images as search results.
Step 201, compressing each image obtained in step 101, including the following steps:
firstly, carrying out Canny edge detection on each retrieved image (see J.canny.A computational approach to edge detection [ J ]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,8 (6): 679-698.—a calculation method of J.canny. Edge detection [ J ]. IEEE mode analysis and machine intelligence journal, 1986,8 (6): 679-698.) and extracting high-frequency information of the image.
And secondly, performing Huffman coding on the image high-frequency information extracted by Canny edge detection, and performing Huffman decoding on the obtained code stream to obtain an edge reconstruction image.
Performing wavelet decomposition on the retrieved image, wherein the decomposition level is 5, performing SPIHT coding on the wavelet coefficient (see Said A, pearlman W A.A New, fast, and Efficient Image Code Based on Set Partitioning in Hierarchical Trees [ J ] IEEE Transactions on Circuits and Systems for Video Technology,1996,6 (3): 243-250.—Said A, pearlman W A. A novel, rapid and efficient image coding based on multi-level tree set splitting [ J ]. IEEE video technical circuit and system academic, 1996,6 (3): 243-250), performing Huffman coding on the code stream to optimize 01 space of the obtained code stream after the wavelet coefficient is coded by SPIHT, obtaining an optimized compressed code stream, and sequentially performing Huffman decoding, SPIHT decoding and wavelet inverse transformation on the optimized compressed code stream to obtain a reconstructed image losing high-frequency information.
And step four, adding the edge reconstruction image obtained in the step two in the step 201 and the reconstruction image with loss of high-frequency information obtained in the step three, and recovering to obtain an original image.
Step 301, the programs in step 101 and step 201 are imported into RAM of a single-chip microcomputer, a chip of 5G network communication technology is inserted into the single-chip microcomputer to transmit by means of a 5G network, and the single-chip microcomputer is provided with a USB interface so as to be inserted into a host computer of a hospital computer, thereby realizing retrieval and compression of medical images and resource sharing among different hospitals.
Claims (1)
1. The medical image searching and compressing method is characterized by comprising the following steps:
step 101, retrieving a medical image:
step one, preprocessing each medical image in an image dataset formed by a plurality of medical images, wherein the steps are as follows:
first, converting a medical image in a Dicom format into a jpg format and converting a medical image resolution size into 224x224 by a computer program;
secondly, converting the single-channel medical image into a three-channel image through a computer vision library Opencv;
step two, processing the preprocessed image data set to construct an image feature library, which comprises the following specific steps:
firstly, constructing a DenseNet model in a Spyder editor of Python 3.6;
dividing the preprocessed image data set into a training set and a testing set, wherein 70% of three-channel image data are used as the training set, then performing normalization processing on neurons input into the DenseNet network model by using a BN algorithm to serve as neurons input into the DenseNet network model, performing training optimization on the DenseNet model by using the training set, and performing testing on the DenseNet network model by using the testing set; when training the DenseNet network model, optimizing a loss function of the DenseNet network model by adopting a RMSProp algorithm;
thirdly, inputting each image in the image dataset into a DenseNet model after training and optimization, and then extracting the output of the last pooling layer of the DenseNet model as the characteristic of each image to construct an image characteristic library for forming the image dataset;
step three, performing dimension reduction processing on the data in the image feature library by using a KPCA projection method;
step four, respectively carrying out KSH coding on the image feature data of each image after KPCA projection processing to obtain a hash code base of the image features;
inputting an image to be searched in the optimized DenseNet model, extracting the output of the last pooling layer of the DenseNet model as the characteristic of the image to be searched, performing dimension reduction processing on the characteristic data of the image to be searched by using a KPCA projection method, performing KSH coding on the characteristic data of the image to be searched after the dimension reduction processing to obtain a hash code of the image to be searched, comparing the Hamming distance between the hash code of the image to be searched and a Hamming code library, performing similarity measurement, and arranging images in the Hamming code library with the Hamming distance within a set range of the image to be searched according to the sequence of sequentially increasing distances and storing the images as search results;
step 201, compressing each image obtained in step 101, including the following steps:
firstly, carrying out Canny edge detection on each retrieved image, and extracting high-frequency information of the image;
secondly, performing Huffman coding on the image high-frequency information extracted by Canny edge detection, and performing Huffman decoding on the obtained code stream to obtain an edge reconstruction image;
performing wavelet decomposition on the retrieved image, wherein the decomposition level is 5, performing SPIHT coding on wavelet coefficients, performing Huffman coding on the code stream to obtain an optimized compressed code stream, and sequentially performing Huffman decoding, SPIHT decoding and wavelet inverse transformation on the optimized compressed code stream to obtain a reconstructed image with high-frequency information loss;
step four, adding the edge reconstruction image obtained in the step two in the step 201 and the reconstruction image with loss of high-frequency information obtained in the step three, and recovering to obtain an original image;
step 301, the programs in step 101 and step 201 are imported into RAM of a single-chip microcomputer, a chip of 5G network communication technology is inserted into the single-chip microcomputer to transmit by means of a 5G network, and the single-chip microcomputer is provided with a USB interface to be inserted into a host computer of a hospital computer, so that medical image retrieval and compression and resource sharing among different hospitals are achieved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910907305.3A CN110633385B (en) | 2019-09-24 | 2019-09-24 | Medical image retrieval and compression method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910907305.3A CN110633385B (en) | 2019-09-24 | 2019-09-24 | Medical image retrieval and compression method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110633385A CN110633385A (en) | 2019-12-31 |
| CN110633385B true CN110633385B (en) | 2023-05-12 |
Family
ID=68973655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910907305.3A Expired - Fee Related CN110633385B (en) | 2019-09-24 | 2019-09-24 | Medical image retrieval and compression method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110633385B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111883234A (en) * | 2020-08-05 | 2020-11-03 | 杜翠 | Medical information system and image retrieval method thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7085379B1 (en) * | 1999-11-26 | 2006-08-01 | Sharp Kabushiki Kaisha | Image compression device and image decompressing device, and computer-readable recorded medium on which program for allowing computer to execute image compressing method and image decompressing method |
| CN108898175A (en) * | 2018-06-26 | 2018-11-27 | 北京工业大学 | Area of computer aided model building method based on deep learning gastric cancer pathological section |
| CN109461495A (en) * | 2018-11-01 | 2019-03-12 | 腾讯科技（深圳）有限公司 | A kind of recognition methods of medical image, model training method and server |
| CN109741265A (en) * | 2018-11-20 | 2019-05-10 | 哈尔滨理工大学 | A kind of spine CT image bilateral filtering denoising method based on feature selecting |
| CN110069644A (en) * | 2019-04-24 | 2019-07-30 | 南京邮电大学 | A kind of compression domain large-scale image search method based on deep learning |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4073942B2 (en) * | 2006-12-01 | 2008-04-09 | シャープ株式会社 | Image compression apparatus and image expansion apparatus, and computer-readable recording medium recording a program for causing a computer to execute the image compression method and the image expansion method |
| US20150003511A1 (en) * | 2010-11-26 | 2015-01-01 | Christopher Carmichael | WEAV Video Super Compression System |
| CN107220373B (en) * | 2017-06-19 | 2019-09-17 | 太原理工大学 | A kind of Lung neoplasm CT image Hash search method based on medicine sign and convolutional neural networks |
| CN109840290B (en) * | 2019-01-23 | 2020-09-22 | 北京航空航天大学 | End-to-end depth hash-based dermoscope image retrieval method |
-
2019
- 2019-09-24 CN CN201910907305.3A patent/CN110633385B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7085379B1 (en) * | 1999-11-26 | 2006-08-01 | Sharp Kabushiki Kaisha | Image compression device and image decompressing device, and computer-readable recorded medium on which program for allowing computer to execute image compressing method and image decompressing method |
| CN108898175A (en) * | 2018-06-26 | 2018-11-27 | 北京工业大学 | Area of computer aided model building method based on deep learning gastric cancer pathological section |
| CN109461495A (en) * | 2018-11-01 | 2019-03-12 | 腾讯科技（深圳）有限公司 | A kind of recognition methods of medical image, model training method and server |
| CN109741265A (en) * | 2018-11-20 | 2019-05-10 | 哈尔滨理工大学 | A kind of spine CT image bilateral filtering denoising method based on feature selecting |
| CN110069644A (en) * | 2019-04-24 | 2019-07-30 | 南京邮电大学 | A kind of compression domain large-scale image search method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110633385A (en) | 2019-12-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhao et al. | Learning to forecast and refine residual motion for image-to-video generation | |
| Dasan et al. | A novel dimensionality reduction approach for ECG signal via convolutional denoising autoencoder with LSTM | |
| CN111428073B (en) | Image retrieval method of depth supervision quantization hash | |
| CN111260653B (en) | Image segmentation method and device, storage medium and electronic equipment | |
| WO2024104035A1 (en) | Long short-term memory self-attention model-based three-dimensional medical image segmentation method and system | |
| CN103023510B (en) | A kind of movement data compression method based on sparse expression | |
| CN116110597B (en) | Digital twinning-based intelligent analysis method and device for patient disease categories | |
| CN107292821B (en) | A kind of super-resolution image reconstruction method and system | |
| Liu et al. | Multi-modality empowered network for facial action unit detection | |
| Ji et al. | Lung nodule detection in medical images based on improved YOLOv5s | |
| CN117974693B (en) | Image segmentation method, device, computer equipment and storage medium | |
| CN110633385B (en) | Medical image retrieval and compression method | |
| CN114886404A (en) | Heart rhythm data classification method and device, electronic equipment and storage medium | |
| CN117238034A (en) | Human body posture estimation method based on space-time transducer | |
| Chen et al. | A novel imbalanced dataset mitigation method and ECG classification model based on combined 1D_CBAM-autoencoder and lightweight CNN model | |
| CN117011650A (en) | Method and related device for determining image encoder | |
| Valenti et al. | ROS-neuro integration of deep convolutional autoencoders for EEG signal compression in real-time BCIs | |
| Hudagi et al. | Bayes-probabilistic-based fusion method for image inpainting | |
| Yang et al. | Superimposed semantic communication for iot-based real-time ecg monitoring | |
| CN114120245B (en) | Crowd image analysis method, device and equipment based on deep neural network | |
| Ma | Summary of research on application of deep learning in image recognition | |
| WO2023134087A1 (en) | Method and apparatus for generating inquiry template, electronic device, and storage medium | |
| CN116095183A (en) | Data compression method and related equipment | |
| Sinha et al. | Data deduplication applications in cognitive science and computer vision research | |
| CN113723519A (en) | Electrocardio data processing method and device based on contrast learning and storage medium |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20230512 |