AU2021102129A4 - Automatic labeling method of emphysema in CT image based on image report - Google Patents
Automatic labeling method of emphysema in CT image based on image report Download PDFInfo
- Publication number
- AU2021102129A4 AU2021102129A4 AU2021102129A AU2021102129A AU2021102129A4 AU 2021102129 A4 AU2021102129 A4 AU 2021102129A4 AU 2021102129 A AU2021102129 A AU 2021102129A AU 2021102129 A AU2021102129 A AU 2021102129A AU 2021102129 A4 AU2021102129 A4 AU 2021102129A4
- Authority
- AU
- Australia
- Prior art keywords
- emphysema
- image
- clustering
- report
- data
- 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.)
- Ceased
Links
- 206010014561 Emphysema Diseases 0.000 title claims abstract description 54
- 238000002372 labelling Methods 0.000 title claims abstract description 20
- 238000000034 method Methods 0.000 claims abstract description 39
- 230000003902 lesion Effects 0.000 claims abstract description 16
- 238000000605 extraction Methods 0.000 claims abstract description 14
- 210000004072 lung Anatomy 0.000 claims abstract description 11
- 238000004364 calculation method Methods 0.000 claims abstract description 7
- 230000004199 lung function Effects 0.000 claims abstract description 4
- 239000000203 mixture Substances 0.000 claims description 14
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 7
- 230000005484 gravity Effects 0.000 claims description 6
- 230000004927 fusion Effects 0.000 claims description 4
- 238000000926 separation method Methods 0.000 claims description 4
- 238000004445 quantitative analysis Methods 0.000 claims description 2
- 238000007621 cluster analysis Methods 0.000 claims 1
- 238000003745 diagnosis Methods 0.000 abstract description 15
- 238000004458 analytical method Methods 0.000 abstract description 8
- 238000003384 imaging method Methods 0.000 abstract description 4
- 238000007781 pre-processing Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 4
- 208000006545 Chronic Obstructive Pulmonary Disease Diseases 0.000 description 3
- 238000011158 quantitative evaluation Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000009325 pulmonary function Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 1
- 238000013399 early diagnosis Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 208000023504 respiratory system disease Diseases 0.000 description 1
Classifications
-
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2576/00—Medical imaging apparatus involving image processing or analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/08—Detecting, measuring or recording devices for evaluating the respiratory organs
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
- A61B6/03—Computed tomography [CT]
- A61B6/032—Transmission computed tomography [CT]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/50—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20092—Interactive image processing based on input by user
- G06T2207/20104—Interactive definition of region of interest [ROI]
-
- 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
- G06T2207/30061—Lung
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
- G06V2201/031—Recognition of patterns in medical or anatomical images of internal organs
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention relates to a CT image emphysema automatic labeling method based on
image report, which comprises the following steps: (1) completing image image report, CT
image sequence input and image standardization preprocessing in the input module; (2)
extracting feature information in the image report, i.e., information about emphysema
description, according to the technique of lexical dictionary and rule pattern matching
technique in the speech semantic extraction module; (3) performing regional type division of
lung in the emphysema lesion extraction module first, performing clustering analysis to extract
emphysema lesions; (4) calculating the CT threshold value of each region according to the
calculation result of step 3, labeling each regional emphysema region in the output and display
module, display the emphysema region in the CT image to generarte the quantitative lung
function analysis report. Advantageous effects of the present invention: and simplifies the
process of emphysema diagnosis and treatment, and reduces the cost of patient visits.
1/11
FIGURES
Start
Imaging o/
diagnosis Speech analysis of
reportand i mag ediag nosis
CT imag ing
datat
Extract that
characteristic
inrfo rmati on inth e
diagnosis report
No
are met ormot
Yes
Extracti ng Iles io ns
from CT image
data
Mark the les io n
an d s hade it
End
Figure1I
Description
1/11
Start
Imaging o/ diagnosis Speech analysis of reportand i mag ediag nosis CT imag ing datat
Extract that characteristic inrfo rmati on inth e diagnosis report
No
are met ormot
Yes
Extracti ng Iles io ns from CT image data
Mark the les io n an d s hade it
End
Figure1I
Automatic labeling method of emphysema in CT image based on image report
The invention relates to an automatic marking method of emphysema in CT images,
in particular to an automatic marking method of emphysema in CT images based on image
reports.
Chronic obstructive pulmonary disease (COPD) is a serious threat to human health.
At present, doctors' diagnosis of COPD mainly depends on pulmonary function
examination and CT image scanning. However, the image report does not indicate the
location information of emphysema on CT images, and the pulmonary function
examination cannot provide the location information of emphysema lesions, which leads
to the quantitative evaluation and severity grading of emphysema relying on subjective
experience, which affects the early diagnosis of patients, easily leads to missed diagnosis
and misdiagnosis, and reduces the effectiveness of diagnosis.
At present, some post-processing workstations equipped by CT manufacturers
provide manual quantitative evaluation tools, which can calculate possible emphysema
voxels and display them on images according to CT thresholds input by doctors. However,
the CT threshold needs to be set subjectively by doctors' experience, so there is an urgent
need for automatic detection and quantitative evaluation tools using machine learning or
deep learning technology. Machine learning or deep learning needs a certain number of
data set training models with labeling information, but it is unrealistic to rely on doctors to
manually label the location of emphysema in CT images to build standard data sets.
At present, there are a large number of image reports and CT image data for hospitals,
but the location and area size of emphysema are not marked on CT images.
A large amount of medical data: At present, there are CT images and corresponding
diagnosis reports of COPD patients in a certain scale in hospitals, but doctors do not mark
the location of emphysema on CT images, and there is no quantitative analysis result of
emphysema. There is only qualitative text description of emphysema in the image
diagnosis report, but no specific location information of lesions.
Extraction of phonetic semantics: At present, the application of machine learning and
deep thought learning in phonetic semantics has entered into daily life. Based on dictionary
technology and rule pattern matching technology, named entities and relationships in image
report text corpus are obtained. First of all, information extraction: in this stage, natural
language processing technology and statistical machine learning technology are mainly
used to identify, label and extract entity information in the text, and to mine the relationship
between entities, and to carry out tasks such as reference resolution and semantic
disambiguation, and to obtain the intermediate results of extraction or labeling, as well as
dictionary tools, vocabulary ontology and other resources in the extraction process.
In order to solve the above technical problems, the present invention provides an
automatic labeling method for emphysema in CT images based on image reports, aiming
at automatically labeling emphysema diagnosis results in image diagnosis reports in CT
images.
To achieve the above purpose, the automatic marking method of emphysema in CT
images based on image report of the present invention comprises the following steps:
(1) completing image image report, CT image sequence input and image
standardization preprocessing in the input module; (2) extracting feature information in the
image report, i.e., information about emphysema description, according to the technique of
lexical dictionary and rule pattern matching technique in the speech semantic extraction
module; (3) performing regional type division of lung in the emphysema lesion extraction
module first, performing clustering analysis to extract emphysema lesions; (4) calculating
the CT threshold value of each region according to the calculation result of step 3, labeling
each regional emphysema region in the output and display module, display the emphysema
region in the CT image to generarte the quantitative lung function analysis report.
In the step 2, the lesion information in the image report is extracted by text information
to obtain text features, then the text features are labeled by manual labeling and
crowdsourcing methods, and finally the labeled features are compared with the established
medical dictionary database to obtain lesion result information.
In step 3, the methods of mean clustering, Gaussian mixture model clustering, density
noise-based clustering and condensed hierarchical clustering are used to distinguish
healthy tissue from swollen tissue, and finally one of mean clustering, Gaussian mixture
model clustering, density noise-based clustering and condensed hierarchical clustering is
selected by voting method to distinguish healthy tissue from swollen tissue and determine
the location of swollen tissue.
The mean clustering algorithm is as follows: CT image sequence, extracting input data
point set; Selecting k data randomly from n data as clustering centroid; Measure the
distance from each remaining data to each centroid, and classify it into the nearest centroid
class; Recalculate the obtained centroid of each class; Iterate 2 ~ 4 steps until the new centroid is equal to or less than the specified threshold, and the algorithm ends; Wherein that formula in the data distribution process.
Si= {xp: I Ix-i j I I Ix-j12Vi, j,1 i, i : k} Si is the ith data point set, Xp is the data point to be determined, and mi and mj are
the ith and j gravity points of kmeans.
Data update process:
-i 1 Si IZ (t)Xj
When the calculation reaches the t-th time, the center of gravity of the m-th time is
the average value of all Euclidean distances xi belonging to the current point set Si, and the
new center point is calculated by summation until the center point no longer changes or
reaches the maximum number of cycles.
For the Gaussian mixture model, the number of classifications is the number of
Gaussian models, and the single Gaussian probability density function is defined as
follows:
#(|6=e (1)
Wherein 0= (p,a), p, G2 represent sample mean and variance respectively, and y
represents Gaussian density of data; Multiple Gaussian mixture model combinations are
defined as follows:
Multiple Gaussian mixture model combinations are defined as follows:
P(y19)= (2) $ICg(yIk)
In which, ak is the probability that k class in the sample set is selected: ai = p (z =k 0),
wherein z = k means that the sample belongs to k class, then <p (y 10 k) = p (y z =k, 0),
obviously *, ,and y is observed values;
The distribution of multiple classes can be obtained by calculating p and (2
The density-based noise applies a spatial clustering algorithm: starting from any data
that has not been accessed, calculating the number of all points in the neighborhood (e) of
the point, if there are enough points, starting clustering, otherwise marking the changed
points as noise; For the first point in the new cluster, the points in its domain (e) will also
become a part of the same cluster; This process is that all the points in the neighborhood
(e) belong to the same class, and the above process is repeated until all the points are
marked.
The condensed hierarchical clustering is to decompose a given data set hierarchically
until a certain condition is met; After the distance value has been obtained, the elements
are connected and a structure is constructed by separation and fusion; Firstly, each object
is regarded as a cluster, and then these clusters are merged into larger and larger clusters
until a certain termination condition is met.
The advantageous effect of the present invention: the information such as the location
and severity of emphysema reported by the image is displayed visually on the CT image,
then not only can save the time of doctors' diagnosis and improve the efficiency of
diagnosis, but also can meet the demand of automatic labeling of emphysema lesions and
significantly reduce the workload of doctors in labeling image data, and the method is of
great significance in building a standard image data set for respiratory system diseases. It also simplifies the process of emphysema diagnosis and treatment, and reduces the cost of patient care.
Figure 1 is a workflow diagram of the present invention.
Figure 2 is a workflow diagram of steps.
Figure 3 is a workflow diagram of mean clustering.
Figure 4 is the CT thresholds of interest for each of the lungs derived according to
Figure 3.
Figure 5 is the workflow diagram of Gaussian mixture model.
Figure 6 is the CT thresholds of interest for each of the lungs derived from Figure 5.
Figure 7 workflow diagram of density-based noise applied to spatial clustering.
Figure 8 is the CT thresholds of interest for each of the lungs derived in Figure 7.
Figure 9 is the workflow diagram of coalescent hierarchical clustering.
Figure 10 shows the CT thresholds of interest for each of the lungs derived in Figure
9.
Figure 11 shows the source data.
Figure 12 shows the results of the clustering of the source data of Figure 11.
The present invention will be further explained with reference to the accompanying
drawings. It should be understood that the preferred embodiments described herein are only
used to illustrate and explain the present invention, and are not used to limit the present
invention.
As shown in the figures, the automatic marking method of emphysema in CT images
based on image report of the present invention comprises the following steps:
(2) completing image image report, CT image sequence input and image
standardization preprocessing in the input module; (2) extracting feature information in the
image report, i.e., information about emphysema description, according to the technique of
lexical dictionary and rule pattern matching technique in the speech semantic extraction
module; (3) performing regional type division of lung in the emphysema lesion extraction
module first, performing clustering analysis to extract emphysema lesions; (4) calculating
the CT threshold value of each region according to the calculation result of step 3, labeling
each regional emphysema region in the output and display module, display the emphysema
region in the CT image to generarte the quantitative lung function analysis report.
In the step 2, the lesion information in the image report is extracted by text information
to obtain text features, then the text features are labeled by manual labeling and
crowdsourcing methods, and finally the labeled features are compared with the established
medical dictionary database to obtain lesion result information.
In step 3, the methods of mean clustering, Gaussian mixture model clustering, density
noise-based clustering and condensed hierarchical clustering are used to distinguish
healthy tissue from swollen tissue, and finally one of mean clustering, Gaussian mixture
model clustering, density noise-based clustering and condensed hierarchical clustering is
selected by voting method to distinguish healthy tissue from swollen tissue and determine
the location of swollen tissue.
The mean clustering algorithm is as follows: CT image sequence, extracting input data
point set; Selecting k data randomly from n data as clustering centroid; Measure the distance from each remaining data to each centroid, and classify it into the nearest centroid class; Recalculate the obtained centroid of each class; Iterate 2 ~ 4 steps until the new centroid is equal to or less than the specified threshold, and the algorithm ends; Wherein that formula in the data distribution process.
S,= {xp: 1,-1i [MIG I [! xI--rij I[2Vi, j,1 5i, j :5 k} Si is the ith data point set, Xp is the data point to be determined, and mi and mj are
the ith and j gravity points of kmeans.
Data update process:
- Q+) (t)X 1
When the calculation reaches the t-th time, the center of gravity of the m-th time is
the average value of all Euclidean distances xi belonging to the current point set Si, and the
new center point is calculated by summation until the center point no longer changes or
reaches the maximum number of cycles.
For the Gaussian mixture model, the number of classifications is the number of
Gaussian models, and the single Gaussian probability density function is defined as
follows:
O( 1( (9 c r2__ Z2 (1 Wherein 0= (pt,2, 2 represent sample mean and variance respectively, and y
represents Gaussian density of data; Multiple Gaussian mixture model combinations are
defined as follows:
Multiple Gaussian mixture model combinations are defined as follows:
P(y|)= 1 ag y I9k) (2)
In which, ak is the probability that k class in the sample set is selected: ai= p (z =k 0),
wherein z = k means that the sample belongs to k class, then p (y 10 k) = p (y z =k, 0),
obviously 0 ,and y is observed values;
The distribution of multiple classes can be obtained by calculating p and
. The density-based noise applies a spatial clustering algorithm: starting from any data
that has not been accessed, calculating the number of all points in the neighborhood (e) of
the point, if there are enough points, starting clustering, otherwise marking the changed
points as noise; For the first point in the new cluster, the points in its domain (e) will also
become a part of the same cluster; This process is that all the points in the neighborhood
(e) belong to the same class, and the above process is repeated until all the points are
marked.
The condensed hierarchical clustering is to decompose a given data set hierarchically
until a certain condition is met; After the distance value has been obtained, the elements
are connected and a structure is constructed by separation and fusion; Firstly, each object
is regarded as a cluster, and then these clusters are merged into larger and larger clusters
until a certain termination condition is met.
The condensed hierarchical clustering is to decompose a given data set hierarchically
until a certain condition is met; After the distance value has been obtained, the elements
are connected and a structure is constructed by separation and fusion; First, each object is
regarded as a cluster, and then these clusters are merged into larger and larger clusters until
a certain termination condition is met.
Claims (7)
1. A CT image emphysema automatic labeling method based on image report is
characterized by comprising the following steps: (1) Completing image report, CT image
sequence input and image standardization pretreatment in an input module; (2) In the
speech semantic extraction module, the feature information in the image report is extracted
according to the dictionary technology and the rule pattern matching technology, that is,
the information about emphysema description; (3) The lungs are divided into regions in the
emphysema focus extraction module; The right lung has the same width from top to
bottom, and the left lung will be divided into two regions in the middle; Then cluster
analysis is carried out in the lung area to extract emphysema lesions; (4) Calculating that
CT threshold value of each region according to the calculation result of step 3, labeling the
emphysema region of each region in the output and display module, then shading the
emphysema and healthy region of each region according to the CT threshold value of each
region, while accurately locate the focus position of pulmonary emphysema in each region,
then displaying the emphysema region in the CT image, and generate a quantitative
analysis report of lung function.
2. The CT image emphysema automatic annotation method based on image report
according to claim 1, which is characterized in that the extraction of lesion information in
the image report in step 2 obtains text features by text information extraction, and then
labels text features by manual annotation and crowdsourcing method, and finally compares
the annotated features with the established medical dictionary database to obtain lesion
result information.
3. According to claim 1, the CT image emphysema automatic annotation method based on
image report is characterized in that in step 3, mean clustering, Gaussian mixture model
clustering, density noise based clustering and agglomerative hierarchical clustering are
used to distinguish healthy tissue and emphysema tissue respectively; Finally, the voting
method is used to select one of mean clustering, Gaussian mixture model clustering, density
noise based clustering and agglomerative hierarchical clustering to distinguish healthy
tissue and emphysema tissue, and determine the location of emphysema.
4. The method for automatically labeling emphysema in CT images based on image report
according to claim 3, which is characterized in that the mean clustering algorithm is: CT
image sequence, extracting input data point set; Selecting k data randomly from n data as
clustering centroid; Measure the distance from each remaining data to each centroid, and
classify it into the nearest centroid class; Recalculate the obtained centroid of each class;
Iterate 2 ~ 4 steps until the new centroid is equal to or less than the specified threshold, and
the algorithm ends; The formula in the process of data distribution:
Si {x, :Ix, - x, -- vi, j, 5 i, j iJk1 Si is the ith data point set, X is the data point to be determined, and mi and mj are the ith
and j gravity points of kmeans.
Data update process:
At the time of calculation to the t-th time, the center of gravity of the mi-th is the average
value of all Euclidean distances xi belonging to the point set Si, and the new center point
is calculated by summation until the center point no longer changes or reaches the
maximum number of cycles.
5. The method for automatically labeling emphysema in CT images based on image report
according to claim 3, which is characterized in that for Gaussian mixture models, the
number of classifications is the number of Gaussian models, and the single Gaussian
probability density function is defined as follows:
2 2 In the formula,O=(p,y ),a represent the sample mean and variance respectively, and y
represents the Gaussian density of the data;
In which, ak is the probability that k class in the sample set is selected: ak= p (z =k 0),
wherein z = k means that the sample belongs to k class, then p(y 10 k) = p (y z =k, 0),
obviously y ,and y is observed values;
The distribution of multiple classes can be obtained by calculating p and a
6. The method for automatically labeling emphysema in CT images based on image report
according to claim 3, which is characterized by applying spatial clustering algorithm based
on density noise: starting from any data that has not been accessed, calculating the number
of all points in the neighborhood (e) of the point, if there are enough points, starting
clustering, otherwise marking the changed points as noise; For the first point in the new
cluster, the points in its domain (e) will also become a part of the same cluster; This process
is that all the points in the neighborhood (e) belong to the same class, and the above process
is repeated continuously until all the points are marked.
7. The method for automatically labeling emphysema in CT images based on image report
according to claim 3, which is characterized in that the hierarchical clustering is to
decompose a given data set hierarchically until a certain condition is met; After the distance value has been obtained, the elements are connected and a structure is constructed by separation and fusion; Firstly, each object is regarded as a cluster, and then these clusters are merged into larger and larger clusters until a certain termination condition is met.
FIGURES 1/11
Figure 1
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2021102129A AU2021102129A4 (en) | 2021-04-22 | 2021-04-22 | Automatic labeling method of emphysema in CT image based on image report |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2021102129A AU2021102129A4 (en) | 2021-04-22 | 2021-04-22 | Automatic labeling method of emphysema in CT image based on image report |
Publications (1)
Publication Number | Publication Date |
---|---|
AU2021102129A4 true AU2021102129A4 (en) | 2021-06-24 |
Family
ID=76442159
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
AU2021102129A Ceased AU2021102129A4 (en) | 2021-04-22 | 2021-04-22 | Automatic labeling method of emphysema in CT image based on image report |
Country Status (1)
Country | Link |
---|---|
AU (1) | AU2021102129A4 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115050442A (en) * | 2022-08-17 | 2022-09-13 | 深圳市指南针医疗科技有限公司 | Disease category data reporting method and device based on mining clustering algorithm and storage medium |
-
2021
- 2021-04-22 AU AU2021102129A patent/AU2021102129A4/en not_active Ceased
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115050442A (en) * | 2022-08-17 | 2022-09-13 | 深圳市指南针医疗科技有限公司 | Disease category data reporting method and device based on mining clustering algorithm and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110993064B (en) | Deep learning-oriented medical image labeling method and device | |
CN111933251B (en) | Medical image labeling method and system | |
EP3567605A1 (en) | Structured report data from a medical text report | |
CN107563383A (en) | A kind of medical image auxiliary diagnosis and semi-supervised sample generation system | |
US10755413B1 (en) | Method and system for medical imaging evaluation | |
CN109949938B (en) | Method and device for standardizing medical non-standard names | |
CN111949759A (en) | Method and system for retrieving medical record text similarity and computer equipment | |
CN110033859A (en) | Assess method, system, program and the storage medium of the medical findings of patient | |
EP3654343A1 (en) | Application of deep learning for medical imaging evaluation | |
CN106383853A (en) | Realization method and system for electronic medical record post-structuring and auxiliary diagnosis | |
US20220301670A1 (en) | Automated information extraction and enrichment in pathology report using natural language processing | |
CN109448822A (en) | CT image pulmonary emphysema automatic marking method based on image report | |
CN115062165B (en) | Medical image diagnosis method and device based on film reading knowledge graph | |
US20230117134A1 (en) | A method of and system for calcium scoring of coronary arteries | |
CN113662664B (en) | Instrument tracking-based objective and automatic evaluation method for surgical operation quality | |
CN114419314A (en) | System and method for automatically labeling medical images | |
Sahin et al. | Detection and classification of COVID-19 by using faster R-CNN and mask R-CNN on CT images | |
Wang et al. | Cataract detection based on ocular B-ultrasound images by collaborative monitoring deep learning | |
AU2021102129A4 (en) | Automatic labeling method of emphysema in CT image based on image report | |
CN114191665A (en) | Method and device for classifying man-machine asynchronous phenomena in mechanical ventilation process | |
Chen et al. | Visual-linguistic causal intervention for radiology report generation | |
CN115206478A (en) | Medical report generation method and device, electronic equipment and readable storage medium | |
CN117237351B (en) | Ultrasonic image analysis method and related device | |
Du et al. | A novel emotion-aware method based on the fusion of textual description of speech, body movements, and facial expressions | |
CN113658690A (en) | Intelligent medical guide method and device, storage medium and electronic equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
FGI | Letters patent sealed or granted (innovation patent) | ||
MK22 | Patent ceased section 143a(d), or expired - non payment of renewal fee or expiry |