CN111311589A - Pulmonary nodule detection and property judgment system and method - Google Patents
Pulmonary nodule detection and property judgment system and method Download PDFInfo
- Publication number
- CN111311589A CN111311589A CN202010148254.3A CN202010148254A CN111311589A CN 111311589 A CN111311589 A CN 111311589A CN 202010148254 A CN202010148254 A CN 202010148254A CN 111311589 A CN111311589 A CN 111311589A
- Authority
- CN
- China
- Prior art keywords
- nodule
- model
- pulmonary
- lung
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000002685 pulmonary effect Effects 0.000 title claims abstract description 125
- 238000001514 detection method Methods 0.000 title claims abstract description 86
- 238000000034 method Methods 0.000 title claims abstract description 13
- 210000004072 lung Anatomy 0.000 claims abstract description 69
- 238000003325 tomography Methods 0.000 claims abstract description 54
- 238000013145 classification model Methods 0.000 claims abstract description 34
- 206010056342 Pulmonary mass Diseases 0.000 claims abstract description 21
- 238000002372 labelling Methods 0.000 claims abstract description 19
- 238000007781 pre-processing Methods 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000013500 data storage Methods 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 7
- 238000010276 construction Methods 0.000 claims abstract description 6
- 230000003211 malignant effect Effects 0.000 claims description 15
- 238000000605 extraction Methods 0.000 claims description 10
- 230000011218 segmentation Effects 0.000 claims description 8
- 230000002308 calcification Effects 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 238000013528 artificial neural network Methods 0.000 abstract description 2
- 206010054107 Nodule Diseases 0.000 description 23
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 206010058467 Lung neoplasm malignant Diseases 0.000 description 1
- 206010028980 Neoplasm Diseases 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 201000005202 lung cancer Diseases 0.000 description 1
- 208000020816 lung neoplasm Diseases 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000036210 malignancy Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 210000000115 thoracic cavity Anatomy 0.000 description 1
- 238000011282 treatment Methods 0.000 description 1
- 238000011269 treatment regimen Methods 0.000 description 1
Images
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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
-
- 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
- G06T2207/20081—Training; 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30061—Lung
- G06T2207/30064—Lung nodule
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a pulmonary nodule detection and property judgment system and method, and relates to a neural network. The first acquisition module acquires a lung tomography image; a second acquisition module to acquire clinical data; the preprocessing module is used for obtaining a lung parenchyma tomography image; the data storage module is used for storing the nodule detection model and the nodule classification model; the nodule detection module includes: the detection unit is used for detecting and obtaining the position of a pulmonary nodule candidate region; the classification unit is used for processing the positions of the candidate areas to obtain real nodule areas; the labeling module is used for manually labeling each high-dimensional feature and clinical data to obtain a labeled image; the training module is used for training to obtain a pulmonary nodule property judgment model; the model construction module is used for constructing a pulmonary nodule identification model; and the prediction module is used for inputting the lung tomography image of the patient to be detected into the lung nodule identification model to obtain the prediction probability. Has the following beneficial effects: the accuracy of real nodule property judgment is improved.
Description
Technical Field
The invention relates to the field of neural networks, in particular to a pulmonary nodule detection and property judgment system and method.
Background
With the development of machine learning algorithms and computer hardware, deep learning has begun to be widely applied to the processing of medical images. In the field of medical thoracic imaging, the prior art focuses on detection of pulmonary nodules and improvement of corresponding algorithms, and mainly includes identification, positioning, identification and the like of the nodules on a pulmonary tomography image. However, the false positive rate of the detection result of the existing computer aided detection system is higher, which is not only beneficial for the clinician to improve the efficiency of clinical work, but also increases the workload of the clinician, and fails to further classify the nature of the nodule, thus having little clinical significance.
In addition, for the detected pulmonary nodules, accurate judgment of benign and malignant nodules can help clinicians to make follow-up and treatment strategies, but the existing few nodule property judgment systems only rely on the characteristic information of images and do not fully utilize the information of clinical and laboratory related examinations of patients. Therefore, a complete system for detecting the nodules and determining the quality and the malignancy of the nodules by combining with multi-aspect information such as clinical data has important clinical significance, and is expected to greatly improve the efficiency of clinical diagnosis and reduce the burden of imaging doctors in the future.
The invention provides a pulmonary nodule detection and property judgment system and method, which are used for screening early lung cancer in real nodules by adopting a deep learning algorithm to construct a pulmonary nodule model and identifying authenticity and properties of the pulmonary nodules so as to realize early intervention or treatment.
Disclosure of Invention
In order to solve the above problems, the present invention provides a pulmonary nodule detecting and property determining system, including:
the first acquisition module is used for acquiring tomographic images of the lungs of a plurality of patients;
the second acquisition module is used for acquiring clinical data related to each lung tomography image;
the preprocessing module is connected with the first acquisition module and used for preprocessing images of the lung tomography scans to obtain corresponding lung parenchyma tomography images;
the data storage module is used for storing a node detection model and a node classification model which are generated in advance;
a nodule detection module connected to the preprocessing module and the data storage module, respectively, the nodule detection module comprising:
the detection unit is used for respectively carrying out nodule detection on each lung parenchyma tomography image according to the nodule detection model to obtain a lung nodule candidate region position;
the classification unit is connected with the detection unit and used for processing the positions of the pulmonary nodule candidate regions according to the nodule classification model to obtain a real nodule region;
the labeling module is respectively connected with the second acquisition module and the nodule detection module and is used for respectively extracting high-dimensional features of pulmonary nodules contained in each real nodule region and performing artificial labeling according to the extracted high-dimensional features and the clinical data to obtain a labeled image containing the nodule properties of the pulmonary nodules;
the nodule property is a true probability that the pulmonary nodule is a malignant nodule;
the training module is connected with the labeling module and used for obtaining a pulmonary nodule property judgment model according to the labeled image training;
a model construction module, respectively connected to the data storage module and the training module, configured to construct a pulmonary nodule identification model according to the nodule detection model, the nodule classification model, and the pulmonary nodule property determination model, where an input of the pulmonary nodule identification model is an input of the nodule detection model, an output of the nodule detection model is an input of the nodule classification model, an output of the nodule classification model is an input of the pulmonary nodule property determination model, and an output of the pulmonary nodule property determination model is an output of the pulmonary nodule identification model;
and the prediction module is connected with the model construction module and used for inputting the pulmonary tomography image of the patient to be detected into the pulmonary nodule identification model to obtain the prediction probability representing that the pulmonary nodule of the patient to be detected is a malignant nodule, and a doctor gives a clinical guidance suggestion of the patient to be detected according to the prediction probability.
Preferably, the preprocessing module comprises:
the segmentation unit is used for segmenting the lung tomography image by using 3D-Slicer software to obtain a plurality of segmented images;
and the extraction unit is connected with the segmentation unit and is used for respectively carrying out lung parenchyma extraction on each segmentation image to obtain a lung parenchyma tomography image.
Preferably, the nodule detection model is a Faster R-CNN pulmonary nodule detection model.
Preferably, the pulmonary nodule property determination model is a 3D MS-DenseNet pulmonary nodule property determination model.
Preferably, the high-dimensional features include nodule density, and/or nodule contour definition, and/or nodule calcification, and/or nodule congestion.
Preferably, the true probability is 0 or 1.
A pulmonary nodule detection and property determination method is applied to a pulmonary nodule detection and property determination system and comprises the following steps:
step S1, acquiring lung tomography images of a plurality of patients;
step S2, acquiring clinical data related to each lung tomography image;
step S3, preprocessing each lung tomography to obtain corresponding lung parenchyma tomography images;
step S4, respectively carrying out nodule detection on each lung parenchyma tomography image according to a pre-generated nodule detection model to obtain lung nodule candidate region positions;
step S5, processing the positions of the candidate regions of pulmonary nodules according to a pre-generated nodule classification model to obtain a real nodule region;
step S6, respectively extracting high-dimensional features of pulmonary nodules contained in each real nodule region, and performing artificial labeling according to the extracted high-dimensional features and the clinical data to obtain labeled images containing the nodule properties of the pulmonary nodules;
the nodule property is a true probability that the pulmonary nodule is a malignant nodule;
step S7, training according to the labeling image to obtain a lung nodule property judgment model;
step S8, constructing a pulmonary nodule identification model according to the nodule detection model, the nodule classification model and the pulmonary nodule property determination model, wherein the input of the pulmonary nodule identification model is the input of the nodule detection model, the output of the nodule detection model is the input of the nodule classification model, the output of the nodule classification model is the input of the pulmonary nodule property determination model, and the output of the pulmonary nodule property determination model is the output of the pulmonary nodule identification model;
step S9, inputting the lung tomography image of the patient to be detected into the lung nodule identification model to obtain the prediction probability representing that the lung nodule of the patient to be detected is a malignant nodule, and giving out the clinical guidance suggestion of the patient to be detected by the doctor according to the prediction probability.
Preferably, the step S3 includes:
step S31, segmenting the lung tomography image by using 3D-Slicer software to obtain a plurality of segmented images;
in step S32, lung parenchyma extraction is performed on each of the divided images to obtain a lung parenchyma tomographic image.
Has the following beneficial effects:
according to the method, authenticity of the pulmonary nodules is detected, and meanwhile, the authenticity nodules are subjected to property judgment and are provided for reference for doctors, so that the burden of the doctors is relieved, and the accuracy of the authenticity nodule property judgment is improved.
Drawings
FIG. 1 is a schematic diagram of a pulmonary nodule detection and property determination system according to a preferred embodiment of the present invention;
FIG. 2 is a flow chart illustrating a method for pulmonary nodule detection and property determination according to a preferred embodiment of the present invention;
fig. 3 is a schematic flow chart of acquiring a tomographic image of lung parenchyma according to a preferred embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the embodiments and features of the embodiments may be combined with each other without conflict.
The invention is further described with reference to the following drawings and specific examples, which are not intended to be limiting.
In order to solve the above problems, the present invention provides a pulmonary nodule detecting and property determining system, as shown in fig. 1, including:
the system comprises a first acquisition module 1, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring lung tomography images of a plurality of patients;
the second acquisition module 2 is used for acquiring clinical data related to each lung tomography image;
the preprocessing module 3 is connected with the first acquisition module 1 and is used for preprocessing images of the lung tomography scans to obtain corresponding lung parenchyma tomography images;
the data storage module 4 is used for storing a node detection model and a node classification model which are generated in advance;
the nodule detection module 5 is respectively connected with the preprocessing module 3 and the data storage module 4, and the nodule detection module 5 comprises:
a detecting unit 51, configured to perform nodule detection on each pulmonary parenchymal tomography image according to a nodule detection model, respectively, to obtain positions of nodule candidate regions in the lung;
the classification unit 52 is connected to the detection unit 51, and is configured to process the positions of the candidate regions of pulmonary nodules according to the nodule classification model to obtain a real nodule region;
the labeling module 6 is respectively connected with the second acquiring module 2 and the nodule detecting module 5, and is used for respectively extracting high-dimensional features of pulmonary nodules contained in each real nodule region, and performing artificial labeling according to the extracted high-dimensional features and clinical data to obtain a labeled image containing the properties of the pulmonary nodules;
the nodule property is the true probability that a pulmonary nodule is a malignant nodule;
the training module 7 is connected with the labeling module 6 and used for obtaining a pulmonary nodule property judgment model according to the labeled image training;
the model building module 8 is respectively connected with the data storage module 4 and the training module 7 and is used for building a pulmonary nodule identification model according to the nodule detection model, the nodule classification model and the pulmonary nodule property judgment model, the input of the pulmonary nodule identification model is the input of the nodule detection model, the output of the nodule detection model is the input of the nodule classification model, the output of the nodule classification model is the input of the pulmonary nodule property judgment model, and the output of the pulmonary nodule property judgment model is the output of the pulmonary nodule identification model;
and the prediction module 9 is connected with the model construction module 8 and is used for inputting the pulmonary tomography image of the patient to be detected into the pulmonary nodule identification model to obtain the prediction probability representing that the pulmonary nodule of the patient to be detected is a malignant nodule, and the doctor gives out the clinical guidance suggestion of the patient to be detected according to the prediction probability.
Specifically, in this embodiment, first, the first obtaining module 1 obtains lung tomographic images of a plurality of patients, and preferably, the lung tomographic images are called from a hospital database and used as one of key data for constructing the system; then, clinical data related to each lung tomography image is acquired through a second acquisition module 2, and the acquired clinical data related to the lung tomography image is called to complete data of an image matched with the clinical data for later-stage system learning samples; then, image preprocessing is carried out on each lung tomography through a preprocessing module 3 to obtain corresponding lung parenchyma tomography images, and the lung parenchyma tomography images are obtained from the lung tomography images to obtain images which really need to be learned; then, a nodule detection model and a nodule classification model which are generated in advance by the data storage module 4 can be used for detecting and classifying nodules; a nodule detection module 5, the nodule detection module 5 comprising: the detection unit 51 is used for respectively carrying out nodule detection on each lung parenchyma tomography image according to the nodule detection model to obtain the position of a candidate region of a pulmonary nodule, and the detection unit 51 can be used for obtaining a pulmonary nodule region which comprises a real region and a suspicious region of the nodule; a classification unit 52, which processes the positions of the candidate regions of pulmonary nodules according to the nodule classification model obtained by the detection unit 51 to obtain real nodule regions, and removes suspicious regions to obtain real nodule regions; and the labeling module 6 is used for manually labeling the high-dimensional features and the clinical data in the real nodule region respectively to obtain a labeled image containing the nodule property of the pulmonary nodule, wherein the nodule property is the real probability that the pulmonary nodule is a malignant nodule, all training materials required by subsequent system training are really completed at the moment, and the system can be trained.
Then the training module 7 trains according to the labeled image to obtain a pulmonary nodule property judgment model; the model building module 8 is used for building a pulmonary nodule identification model according to the nodule detection model, the nodule classification model and the pulmonary nodule property judgment model, wherein the input of the pulmonary nodule identification model is the input of the nodule detection model, the output of the nodule detection model is the input of the nodule classification model, the output of the nodule classification model is the input of the pulmonary nodule property judgment model, and the output of the pulmonary nodule property judgment model is the output of the pulmonary nodule identification model; firstly, a lung parenchymal tomography image is used as input of a lung nodule identification model and input into a nodule detection model to detect the position of a lung nodule candidate region, then the position of the lung nodule candidate region is used for obtaining a real nodule region through a nodule classification model, then the real nodule region is used for obtaining the prediction probability that a lung nodule is a malignant nodule through a lung nodule property judgment model, and the identification and the output of the lung nodule identification model are completed.
In a preferred embodiment of the present invention, the preprocessing module 3 includes:
the segmentation unit 31 is configured to segment the lung tomography image by using 3D-Slicer software to obtain a plurality of segmented images;
and an extraction unit 32 connected to the segmentation unit 31 for performing lung parenchyma extraction on each segmented image to obtain a lung parenchyma tomographic image.
Specifically, in this embodiment, the 3D-Slicer software is used to segment the lung tomographic image to obtain segmented images, which facilitates the input of the segmented images, and the lung parenchyma is extracted from each segmented image to obtain the lung parenchyma tomographic image, which facilitates the extraction of the lung parenchyma from the segmented images.
In a preferred embodiment of the present invention, the nodule detection model is the Faster R-CNN pulmonary nodule detection model.
In a preferred embodiment of the present invention, the pulmonary nodule property determination model is a 3D MS-DenseNet pulmonary nodule property determination model.
In a preferred embodiment of the invention, the high-dimensional features include nodule density, and/or nodule contour definition, and/or nodule calcification, and/or nodule congestion.
In the preferred embodiment of the present invention, the true probability is 0 or 1.
A pulmonary nodule detecting and property determining method applied to a pulmonary nodule detecting and property determining system as shown in fig. 2, comprising the following steps:
step S1, acquiring lung tomography images of a plurality of patients;
step S2, acquiring clinical data related to each lung tomography image;
step S3, preprocessing each lung tomography image to obtain corresponding lung parenchyma tomography images;
step S4, respectively carrying out nodule detection on each lung parenchyma tomography image according to a pre-generated nodule detection model to obtain lung nodule candidate region positions;
step S5, processing the positions of the candidate regions of pulmonary nodules according to a pre-generated nodule classification model to obtain real nodule regions;
step S6, respectively extracting high-dimensional features of pulmonary nodules contained in each real nodule region, and performing artificial labeling according to the extracted high-dimensional features and clinical data to obtain labeled images containing the nodule properties of the pulmonary nodules;
the nodule property is the true probability that a pulmonary nodule is a malignant nodule;
step S7, training according to the labeling image to obtain a lung nodule property judgment model;
step S8, constructing a pulmonary nodule identification model according to the nodule detection model, the nodule classification model and the pulmonary nodule property judgment model, wherein the input of the pulmonary nodule identification model is the input of the nodule detection model, the output of the nodule detection model is the input of the nodule classification model, the output of the nodule classification model is the input of the pulmonary nodule property judgment model, and the output of the pulmonary nodule property judgment model is the output of the pulmonary nodule identification model;
step S9, inputting the lung tomography image of the patient to be detected into the lung nodule identification model to obtain the prediction probability representing that the lung nodule of the patient to be detected is a malignant nodule, and the doctor gives out the clinical guidance suggestion of the patient to be detected according to the prediction probability.
In a preferred embodiment of the present invention, as shown in fig. 3, step S3 includes:
step S31, segmenting the lung tomography image by using 3D-Slicer software to obtain a plurality of segmented images;
in step S32, lung parenchyma extraction is performed on each of the divided images to obtain a lung parenchyma tomographic image.
While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention.
Claims (8)
1. A pulmonary nodule detection and property determination system, comprising:
the first acquisition module is used for acquiring tomographic images of the lungs of a plurality of patients;
the second acquisition module is used for acquiring clinical data related to each lung tomography image;
the preprocessing module is connected with the first acquisition module and used for preprocessing images of the lung tomography scans to obtain corresponding lung parenchyma tomography images;
the data storage module is used for storing a node detection model and a node classification model which are generated in advance;
a nodule detection module connected to the preprocessing module and the data storage module, respectively, the nodule detection module comprising:
the detection unit is used for respectively carrying out nodule detection on each lung parenchyma tomography image according to the nodule detection model to obtain a lung nodule candidate region position;
the classification unit is connected with the detection unit and used for processing the positions of the pulmonary nodule candidate regions according to the nodule classification model to obtain a real nodule region;
the labeling module is respectively connected with the second acquisition module and the nodule detection module and is used for respectively extracting high-dimensional features of pulmonary nodules contained in each real nodule region and performing artificial labeling according to the extracted high-dimensional features and the clinical data to obtain a labeled image containing the nodule properties of the pulmonary nodules;
the nodule property is a true probability that the pulmonary nodule is a malignant nodule;
the training module is connected with the labeling module and used for obtaining a pulmonary nodule property judgment model according to the labeled image training;
a model construction module, respectively connected to the data storage module and the training module, configured to construct a pulmonary nodule identification model according to the nodule detection model, the nodule classification model, and the pulmonary nodule property determination model, where an input of the pulmonary nodule identification model is an input of the nodule detection model, an output of the nodule detection model is an input of the nodule classification model, an output of the nodule classification model is an input of the pulmonary nodule property determination model, and an output of the pulmonary nodule property determination model is an output of the pulmonary nodule identification model;
and the prediction module is connected with the model construction module and used for inputting the pulmonary tomography image of the patient to be detected into the pulmonary nodule identification model to obtain the prediction probability representing that the pulmonary nodule of the patient to be detected is a malignant nodule, and a doctor gives a clinical guidance suggestion of the patient to be detected according to the prediction probability.
2. The pulmonary nodule detection and property determination system of claim 1 wherein the preprocessing module comprises:
the segmentation unit is used for segmenting the lung tomography image by using 3D-Slicer software to obtain a plurality of segmented images;
and the extraction unit is connected with the segmentation unit and is used for respectively carrying out lung parenchyma extraction on each segmentation image to obtain a lung parenchyma tomography image.
3. The pulmonary nodule detection and property determination system of claim 1 wherein the nodule detection model is the fast R-CNN pulmonary nodule detection model.
4. The pulmonary nodule detection and property determination system of claim 1 wherein the pulmonary nodule property determination model is a 3D MS-DenseNet pulmonary nodule property determination model.
5. The pulmonary nodule detection and property determination system of claim 1 wherein the high dimensional features include nodule density, and/or nodule contour definition, and/or nodule calcification, and/or nodule congestion.
6. The pulmonary nodule detection and property determination system of claim 1 wherein the true probability is 0 or 1.
7. A pulmonary nodule detecting and property determining method applied to the pulmonary nodule detecting and property determining system according to any one of claims 1 to 6, comprising the steps of:
step S1, acquiring lung tomography images of a plurality of patients;
step S2, acquiring clinical data related to each lung tomography image;
step S3, preprocessing each lung tomography to obtain corresponding lung parenchyma tomography images;
step S4, respectively carrying out nodule detection on each lung parenchyma tomography image according to a pre-generated nodule detection model to obtain lung nodule candidate region positions;
step S5, processing the positions of the candidate regions of pulmonary nodules according to a pre-generated nodule classification model to obtain a real nodule region;
step S6, respectively extracting high-dimensional features of pulmonary nodules contained in each real nodule region, and performing artificial labeling according to the extracted high-dimensional features and the clinical data to obtain labeled images containing the nodule properties of the pulmonary nodules;
the nodule property is a true probability that the pulmonary nodule is a malignant nodule;
step S7, training according to the labeling image to obtain a lung nodule property judgment model;
step S8, constructing a pulmonary nodule identification model according to the nodule detection model, the nodule classification model and the pulmonary nodule property determination model, wherein the input of the pulmonary nodule identification model is the input of the nodule detection model, the output of the nodule detection model is the input of the nodule classification model, the output of the nodule classification model is the input of the pulmonary nodule property determination model, and the output of the pulmonary nodule property determination model is the output of the pulmonary nodule identification model;
step S9, inputting the lung tomography image of the patient to be detected into the lung nodule identification model to obtain the prediction probability representing that the lung nodule of the patient to be detected is a malignant nodule, and giving out the clinical guidance suggestion of the patient to be detected by the doctor according to the prediction probability.
8. The pulmonary nodule detection and property determination method of claim 7, wherein the step S3 includes:
step S31, segmenting the lung tomography image by using 3D-Slicer software to obtain a plurality of segmented images;
in step S32, lung parenchyma extraction is performed on each of the divided images to obtain a lung parenchyma tomographic image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010148254.3A CN111311589A (en) | 2020-03-05 | 2020-03-05 | Pulmonary nodule detection and property judgment system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010148254.3A CN111311589A (en) | 2020-03-05 | 2020-03-05 | Pulmonary nodule detection and property judgment system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111311589A true CN111311589A (en) | 2020-06-19 |
Family
ID=71160472
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010148254.3A Pending CN111311589A (en) | 2020-03-05 | 2020-03-05 | Pulmonary nodule detection and property judgment system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111311589A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112116603A (en) * | 2020-09-14 | 2020-12-22 | 中国科学院大学宁波华美医院 | Pulmonary nodule false positive screening method based on multitask learning |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108670285A (en) * | 2018-06-05 | 2018-10-19 | 胡晓云 | A kind of CT pulmonary tuberculosis detection artificial intelligence diagnosis and therapy system |
| CN108805209A (en) * | 2018-06-14 | 2018-11-13 | 清华大学深圳研究生院 | A kind of Lung neoplasm screening method based on deep learning |
| CN108986067A (en) * | 2018-05-25 | 2018-12-11 | 上海交通大学 | Pulmonary nodule detection method based on cross-module state |
| 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 |
| CN110458801A (en) * | 2019-06-24 | 2019-11-15 | 深圳市未来媒体技术研究院 | A kind of 3D dual path neural network and the pulmonary nodule detection method based on the network |
| CN110534192A (en) * | 2019-07-24 | 2019-12-03 | 大连理工大学 | A kind of good pernicious recognition methods of Lung neoplasm based on deep learning |
| CN110807764A (en) * | 2019-09-20 | 2020-02-18 | 成都智能迭迦科技合伙企业(有限合伙) | Lung cancer screening method based on neural network |
-
2020
- 2020-03-05 CN CN202010148254.3A patent/CN111311589A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108986067A (en) * | 2018-05-25 | 2018-12-11 | 上海交通大学 | Pulmonary nodule detection method based on cross-module state |
| CN108670285A (en) * | 2018-06-05 | 2018-10-19 | 胡晓云 | A kind of CT pulmonary tuberculosis detection artificial intelligence diagnosis and therapy system |
| CN108805209A (en) * | 2018-06-14 | 2018-11-13 | 清华大学深圳研究生院 | A kind of Lung neoplasm screening method based on deep learning |
| 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 |
| CN110458801A (en) * | 2019-06-24 | 2019-11-15 | 深圳市未来媒体技术研究院 | A kind of 3D dual path neural network and the pulmonary nodule detection method based on the network |
| CN110534192A (en) * | 2019-07-24 | 2019-12-03 | 大连理工大学 | A kind of good pernicious recognition methods of Lung neoplasm based on deep learning |
| CN110807764A (en) * | 2019-09-20 | 2020-02-18 | 成都智能迭迦科技合伙企业(有限合伙) | Lung cancer screening method based on neural network |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112116603A (en) * | 2020-09-14 | 2020-12-22 | 中国科学院大学宁波华美医院 | Pulmonary nodule false positive screening method based on multitask learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109583440B (en) | Medical image auxiliary diagnosis method and system combining image recognition and report editing | |
| KR102645120B1 (en) | System and method for integrating tomographic image reconstruction and radiomics using neural networks | |
| CN107103187B (en) | Lung nodule detection grading and management method and system based on deep learning | |
| CN109087703B (en) | Peritoneal transfer marking method of abdominal cavity CT image based on deep convolutional neural network | |
| US10706534B2 (en) | Method and apparatus for classifying a data point in imaging data | |
| JP4640845B2 (en) | Image processing apparatus and program thereof | |
| Jaffar et al. | Fuzzy entropy based optimization of clusters for the segmentation of lungs in CT scanned images | |
| Xu et al. | Using transfer learning on whole slide images to predict tumor mutational burden in bladder cancer patients | |
| CN113261012B (en) | Method, device and system for processing image | |
| CN111369573A (en) | Auxiliary identification system and method for lung adenocarcinoma subtypes | |
| KR102258756B1 (en) | Determination method for stage of cancer based on medical image and analyzing apparatus for medical image | |
| Manikandan et al. | Segmentation and Detection of Pneumothorax using Deep Learning | |
| Cifci | SegChaNet: a novel model for lung cancer segmentation in CT scans | |
| Peña-Solórzano et al. | Findings from machine learning in clinical medical imaging applications–Lessons for translation to the forensic setting | |
| Tian et al. | Radiomics and its clinical application: artificial intelligence and medical big data | |
| CN112529900B (en) | Method, device, terminal and storage medium for matching ROI in mammary gland image | |
| US20050002548A1 (en) | Automatic detection of growing nodules | |
| CN113298773A (en) | Heart view identification and left ventricle detection device and system based on deep learning | |
| CN111311589A (en) | Pulmonary nodule detection and property judgment system and method | |
| RU2668699C1 (en) | Intellectual method of diagnostics and detection of neoplasms in lungs | |
| Dhalia Sweetlin et al. | Patient-Specific Model Based Segmentation of Lung Computed Tomographic Images. | |
| Grace John et al. | Extreme learning machine algorithm‐based model for lung cancer classification from histopathological real‐time images | |
| Widodo et al. | Texture feature extraction to improve accuracy of malignant and benign cancer detection on ct-scan images | |
| CN111265234A (en) | Method and system for judging properties of lung mediastinal lymph nodes | |
| Hinisha et al. | Lung Nodule Identification |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200619 |