LU503428B1 - Method, Device and System for Diagnosing Pulmonary Embolism Based on non-contrast chest CT Images - Google Patents

Method, Device and System for Diagnosing Pulmonary Embolism Based on non-contrast chest CT Images Download PDF

Info

Publication number
LU503428B1
LU503428B1 LU503428A LU503428A LU503428B1 LU 503428 B1 LU503428 B1 LU 503428B1 LU 503428 A LU503428 A LU 503428A LU 503428 A LU503428 A LU 503428A LU 503428 B1 LU503428 B1 LU 503428B1
Authority
LU
Luxembourg
Prior art keywords
contrast
chest
pulmonary embolism
images
feature
Prior art date
Application number
LU503428A
Other languages
French (fr)
Inventor
Anqi Liu
Mei Deng
Min Liu
Haoyu Yang
Original Assignee
China Japan Friendship Hospital
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Japan Friendship Hospital filed Critical China Japan Friendship Hospital
Priority to LU503428A priority Critical patent/LU503428B1/en
Application granted granted Critical
Publication of LU503428B1 publication Critical patent/LU503428B1/en

Links

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
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/02—Devices for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03—Computerised tomographs
    • A61B6/032—Transmission computed tomography [CT]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/50—Clinical applications
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
    • A61B6/5211—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data
    • A61B6/5217—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data extracting a diagnostic or physiological parameter from medical diagnostic data
    • 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
    • G06T2207/20081—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

Abstract

The invention relates to a method, device and system for diagnosing pulmonary embolism based on non-contrast CT images. And the invention comprises: acquire a non-contrast CT image of the lung of the patient to be diagnosed; carry out three-dimensional reconstruction on the non- contrast CT image of the lung to obtain whole lung imaging; perform feature extraction on the whole lung imaging to obtain the feature vectors; input the feature vector into the trained machine learning model to obtain the pulmonary embolism diagnosis classification result of the patient to be diagnosed. And the invention provides a new noninvasive method for detecting acute pulmonary embolism for patients who do not have CTPA or have CTPA examination contraindications clinically, and has important clinical application value.

Description

DESCRIPTION 0503428
Method, Device and System for Diagnosing Pulmonary Embolism Based on non- contrast chest CT Images
TECHNICAL FIELD
The invention relates to the technical field of intelligent medical treatment, and relates to a method, device, system and storage medium for diagnosing pulmonary embolism based on non- contrast chest CT Images.
BACKGROUND
Acute pulmonary thromboembolism (APE) is a cardiovascular disease with high incidence.
Severe cases can lead to acute right heart failure and even sudden death. Because the clinical symptoms and signs are not specific, CT pulmonary angiography (CTPA) is a scheme for the diagnosis of APE, which is used to guide the diagnosis and evaluation of APE. Although CTPA
APE is the first-line non-invasive detection method, not everyone is suitable for CTPA examination, especially for high-risk patients with unstable hemodynamics, renal insufficiency or allergic to iodine contrast agent. CTPA cannot be used to diagnose APE, which leads to blindness in the treatment of APE. Compared with CTPA, non-contrast chest CT doesn't need to use the contrast agent, but generally clinicians believe that non-contrast chest CT can't effectively diagnose APE. At present, there is no good clinical solution for patients who need non-invasive detection of APE, such as the patients with clinical critical illness, contrast agent allergy and renal insufficiency.
SUMMARY
In order to solve the above problems, we re-adopted the non-contrast chest CT images which are discarded by clinicians as not having the diagnostic value of APE, and establish a method, device, system and storage medium for diagnosing pulmonary embolism based on non-contrast chest CT images, which provide a brand-new, rapid and non-invasive evaluation method for patients who don’t not have CTPA detection conditions or have CTPA imaging contraindications 903428 clinically.
The application discloses a method for diagnosing APE based on non-contrast chest CT data, which comprises the following steps:
Step 1, acquire a non-contrast chest CT image of the lung of the patient to be diagnosed,
Step 2, carry out three-dimensional reconstruction on the non-contrast chest CT image of the lung to obtain whole lung imaging. Preferably, each whole lung image is geometrically transformed into a map containing 36 sub-regions;
Step 3, perform feature extraction on the whole lung imaging to obtain the feature vectors;
Step 4, input the feature vector into the trained machine learning model to obtain the pulmonary embolism diagnosis classification result of the patient to be diagnosed.
Further, the method also includes ROI segmentation of the whole lung imaging to obtain the segmented whole lung imaging containing N sub-regions, and then preform feature extraction on the N sub-regions in the whole lung imaging to obtain the feature vectors;
Further, the feature extraction is to extract features of each sub-region in the whole lung imaging by using different filters to obtain different types of image omics features as feature vectors;
Further, the method also includes performing features fusion on the obtained feature vectors, and the features fusion is the features fusion of feature vectors of each feature vector in N sub- regions;
Further, step 3 also includes preforming the feature dimensionality reduction of the obtained feature vectors, that is, preforming the feature dimensionality reduction according to P value, correlation coefficient, minimum absolute contraction and selection operator LASSO regression;
Further, the lung non-contrast CT image adopts non-contrast CT to scan the whole chest of the patient to be diagnosed from the tip of the lung to the diaphragm;
Further, the machine learning model is selected from one or more of the following machine learning model algorithms: logistic regression algorithm, naive Bayes classification, support vector machine, k nearest neighbor, decision tree, random forest, xgboost, perceptron algorithm and logistic regression; Preferably, the decision tree is the gradient lifting decision tree.
The purpose of this application is to provide a system for diagnosing pulmonary embolish 09428 based on non-contrast CT images, and the system comprises: the non-contrast CT images acquisition unit, which is used for acquiring the non-contrast CT image of the lung of the patient to be diagnosed, the whole lung imaging unit, which is used for carrying out the three-dimensional reconstruction on the non-contrast CT image of the lung to obtain whole lung imaging; the feature extraction unit, which is used for performing feature extraction on the whole lung imaging to obtain feature vectors; and the APE diagnosis unit, which is used for inputting the feature vectors into the trained machine learning model to obtain the pulmonary embolism diagnosis classification result of the patient to be diagnosed.
The purpose of this application is to provide a device for diagnosing pulmonary embolism based on non-contrast CT images, and the device comprises: a memory and a processor; the memory is used for storing program instructions; and the processor is used for calling program instructions, and when the program instructions are executed, the pathological image classification method steps are realized.
A computer readable storage medium with a computer program stored on, and when the computer program is executed by the processor, the method steps of diagnosing pulmonary embolism based on the non-contrast chest CT images are realized.
Advantages of the application: 1. This application adopts the non-contrast chest CT images that clinicians think have no diagnostic value of APE, and establishes a method for diagnosing pulmonary embolism based on non-contrast chest CT images, combining the imaging characteristics of pulmonary images and machine learning algorithm modeling; 2. This application creatively provides a new noninvasive detection method for patients who do not have CTPA detection conditions clinically. This method can use non-contrast chest CT to assist in the diagnosis of APE in real emergencies. The research results show that this method is obviously superior to the visual evaluation of radiologists with different clinical experiences.
BRIEF DESCRIPTION OF THE FIGURES 0503428
Fig. 1 is a schematic flowchart of the method for diagnosing pulmonary embolism based on non-contrast chest CT images provided by the embodiment of the present invention;
Fig. 2 is a schematic diagram of the device for diagnosing pulmonary embolism based on non-contrast chest CT images provided by the embodiment of the present invention;
Fig. 3 is a schematic diagram of the system for diagnosing pulmonary embolism based on non-contrast chest CT images provided by the embodiment of the present invention;
Fig. 4 is a schematic diagram of whole lung imaging and sub-regions segmentation based on non-contrast chest CT images provided by the embodiment of the present invention;
Fig. 5 is a schematic diagram of features fusion provided by the embodiment of the present invention.
DETAILED DESCRIPTION OF THE INVENTION
Fig. 1 is a schematic flowchart of the method for diagnosing pulmonary embolism based on non-contrast chest CT images. Specifically, the method includes the following steps: 101. acquire a non-contrast chest CT image of the lung of the patient to be diagnosed;
In an embodiment, the non-contrast chest CT image adopts non-contrast chest CT to scan the whole chest of the patient to be diagnosed from the tip of the lung to the diaphragm 102. carry out three-dimensional reconstruction on non-contrast chest CT image of the lung to obtain whole lung imaging; 103. perform feature extraction on the whole lung imaging to obtain the feature vectors;
In an embodiment, the method also includes ROI segmentation of the whole lung imaging to obtain the segmented whole lung imaging containing N (N is a natural number, preferably 36) sub- regions, and then preform feature extraction on the N sub-regions in the whole lung imaging to obtain the feature vectors;
In an embodiment, the feature extraction is to extract features of each sub-region in the whole lung imaging by using different filters to obtain different types of image omics features as feature vectors;
In an embodiment, the method also includes performing features fusion on the obtained” 903428 feature vectors, and the features fusion is the features fusion of feature vectors of each feature vector in N sub-regions;
In an embodiment, see Fig. 5 for details. For whole lung imaging, each of 1004 image omics features is actually regarded as a feature vector.
In an embodiment, it also includes preforming the feature dimensionality reduction of the obtained feature vectors, that is, preforming the feature dimensionality reduction according to P value, correlation coefficient and LASSO regression. 104. input the feature vector into the trained machine learning model to obtain the pulmonary embolism diagnosis classification result of the patient to be diagnosed.
In an embodiment, the machine learning model is selected from one or more of the following machine learning model algorithms: logistic regression algorithm, naive Bayes classification, support vector machine, k nearest neighbor, decision tree, random forest, gradient lifting decision tree, perceptron algorithm and logistic regression; Preferably, the decision tree is the gradient lifting decision tree.
Fig. 2 is the device for diagnosing pulmonary embolism based on non-contrast chest CT images provided by the present embodiment, and the device comprises: a memory and a processor; the memory is used for storing program instructions; and the processor is used for calling program instructions, and when the program instructions are executed, the steps of the method for diagnosing pulmonary embolism based on non-contrast chest CT images are realized.
Fig. 3 is the system for diagnosing pulmonary embolism based on non-contrast chest CT images provided by the present embodiment, and the system comprises: the non-contrast chest CT images acquisition unit 301, which is used for acquiring the non- contrast chest CT images of the lung of the patient to be diagnosed, the whole lung imaging unit 302, which is used for carrying out the three-dimensional reconstruction on the non-contrast chest CT image of the lung to obtain whole lung imaging; the feature extraction unit 303, which is used for performing feature extraction on the whole lung imaging to obtain feature vectors;
and the pulmonary embolism diagnosis unit 304, which is used for inputting the feature 903428 vectors into the trained machine learning model to obtain the pulmonary embolism diagnosis classification result of the patient to be diagnosed.
One purpose of the present invention is to provide the computer readable storage medium with a computer program stored on, and when the computer program is executed by the processor, the method steps of diagnosing pulmonary embolism based on non-contrast chest CT images are realized.

Claims (10)

CLAIMS LU503428
1. A method for diagnosing pulmonary embolism based on non-contrast chest CT images is characterized in that the method comprises: step 1, acquire a non-contrast chest CT of the lung of the patient to be diagnosed; step 2, carry out three-dimensional reconstruction on the non-contrast chest CT image of the lung to obtain whole lung imaging; step 3, perform feature extraction on the whole lung imaging to obtain the feature vectors; step 4, input the feature vector into the trained machine learning model to obtain the pulmonary embolism diagnosis classification result of the patient to be diagnosed.
2. The method for diagnosing APE based on non-contrast chest CT images according to claim 1 is characterized in that the method also includes ROI segmentation of the whole lung imaging to obtain the segmented whole lung imaging containing N sub-regions, and the N is a natural number, and then preform feature extraction on the N sub-regions in the whole lung imaging to obtain the feature vectors.
3. The method for diagnosing pulmonary embolism based on non-contrast chest CTimages according to claim 2 is characterized in that the feature extraction is to extract features of each sub-region in the whole lung imaging by using different filters to obtain different types of image omics features as feature vectors.
4. The method for diagnosing APE based on non-contrast chest CT images according to claim 3 is characterized in that the method also includes performing features fusion on the obtained feature vectors, and the features fusion is the features fusion of feature vectors of each feature vector in N sub-regions.
5. The method for diagnosing pulmonary embolism based on the non-contrast chest CT images according to claim 1 is characterized in that step 3 also includes preforming the feature dimensionality reduction of the obtained feature vectors, that is, preforming the feature dimensionality reduction according to P value, correlation coefficient, minimum absolute contraction and LASSO regression.
6. The method for diagnosing pulmonary embolism based on the non-contrast chest CT images according to claim 1 is characterized in that tthe non-contrast chest CT images adopts the non-contrast chest CT to scan the whole chest of the patient to be diagnosed from the tip of the 903428 lung to the diaphragm.
7. The method for diagnosing pulmonary embolism based on the non-contrast chest CT images according to claim 1 is characterized in that the machine learning model is selected from one or more of the following machine learning model algorithms: logistic regression algorithm, naive Bayes classification, support vector machine, k nearest neighbor, decision tree, random forest, gradient lifting decision tree, perceptron algorithm and logistic regression.
8. A system for diagnosing pulmonary embolism based on the non-contrast chest CTimages is characterized in that the system comprises: the non-contrast chest CT images acquisition unit, which is used for acquiring the non- contrast CT image of the lung of the patient to be diagnosed, the whole lung imaging unit, which is used for carrying out the three-dimensional reconstruction on the non-contrast chest CT image of the lung to obtain whole lung imaging; the feature extraction unit, which is used for performing feature extraction on the whole lung imaging to obtain feature vectors; and the APE diagnosis unit, which is used for inputting the feature vectors into the trained machine learning model to obtain the APE diagnosis classification result of the patient to be diagnosed.
9. A device for diagnosing pulmonary embolism based on the non-contrast chest CT images is characterized in that the device comprises: a memory and a processor; the memory is used for storing program instructions; and the processor is used for calling program instructions, and when the program instructions are executed, the pathological image classification method steps according to any one of claims 1- 7 are realized.
10. A computer readable storage medium with a computer program stored on is characterized in that when the computer program is executed by the processor, the method steps of diagnosing APE based on the non-contrast chest CT images according to claim 1 are realized.
LU503428A 2023-02-02 2023-02-02 Method, Device and System for Diagnosing Pulmonary Embolism Based on non-contrast chest CT Images LU503428B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
LU503428A LU503428B1 (en) 2023-02-02 2023-02-02 Method, Device and System for Diagnosing Pulmonary Embolism Based on non-contrast chest CT Images

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
LU503428A LU503428B1 (en) 2023-02-02 2023-02-02 Method, Device and System for Diagnosing Pulmonary Embolism Based on non-contrast chest CT Images

Publications (1)

Publication Number Publication Date
LU503428B1 true LU503428B1 (en) 2023-08-02

Family

ID=87475022

Family Applications (1)

Application Number Title Priority Date Filing Date
LU503428A LU503428B1 (en) 2023-02-02 2023-02-02 Method, Device and System for Diagnosing Pulmonary Embolism Based on non-contrast chest CT Images

Country Status (1)

Country Link
LU (1) LU503428B1 (en)

Similar Documents

Publication Publication Date Title
US11576621B2 (en) Plaque vulnerability assessment in medical imaging
US20230104045A1 (en) System and method for ultrasound analysis
Jiang et al. Development and application of artificial intelligence in cardiac imaging
KR20190022216A (en) Eye image analysis method
US10991460B2 (en) Method and system for identification of cerebrovascular abnormalities
CN112184690B (en) Coronary vessel trend prediction method, prediction model training method and device
Luca et al. Impact of quality, type and volume of data used by deep learning models in the analysis of medical images
KR102071774B1 (en) Method for predicting cardio-cerebrovascular disease using eye image
JP2022515464A (en) Classification method and system of blood flow section based on artificial intelligence
CN111340794B (en) Quantification method and device for coronary artery stenosis
Ezquerra et al. PERFEX: an expert system for interpreting 3D myocardial perfusion
CN113470060A (en) Coronary artery multi-angle curved surface reconstruction visualization method based on CT image
LU503428B1 (en) Method, Device and System for Diagnosing Pulmonary Embolism Based on non-contrast chest CT Images
KR20200011530A (en) Method for predicting cardio-cerebrovascular disease using eye image
Wang et al. AVDNet: Joint coronary artery and vein segmentation with topological consistency
Lin et al. Echocardiography-based AI for detection and quantification of atrial septal defect
Chi et al. A composite of features for learning-based coronary artery segmentation on cardiac CT angiography
Chen et al. Spatio-temporal multi-task network cascade for accurate assessment of cardiac CT perfusion
Dey et al. Computed tomography and artificial intelligence
Preity et al. Automated computationally intelligent methods for ocular vessel segmentation and disease detection: a review
CN114155208B (en) Atrial fibrillation assessment method and device based on deep learning
US20230309940A1 (en) Explainable deep learning camera-agnostic diagnosis of obstructive coronary artery disease
Steeden et al. Artificial Intelligence-Based Evaluation of Congenital Heart Disease
de Moura et al. Fully automated identification and clinical classification of macular edema using optical coherence tomography images
Singh et al. Development and Validation of Artery-Vein Ratio Measurement Based on Deep Learning Architectures

Legal Events

Date Code Title Description
FG Patent granted

Effective date: 20230802