CN114376615B - Mammary gland ultrasonic screening system and screening method based on artificial intelligence - Google Patents
Mammary gland ultrasonic screening system and screening method based on artificial intelligence Download PDFInfo
- Publication number
- CN114376615B CN114376615B CN202210205941.3A CN202210205941A CN114376615B CN 114376615 B CN114376615 B CN 114376615B CN 202210205941 A CN202210205941 A CN 202210205941A CN 114376615 B CN114376615 B CN 114376615B
- Authority
- CN
- China
- Prior art keywords
- probe
- focus
- scanning
- navigation
- breast
- 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.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/08—Detecting organic movements or changes, e.g. tumours, cysts, swellings
- A61B8/0825—Detecting organic movements or changes, e.g. tumours, cysts, swellings for diagnosis of the breast, e.g. mammography
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/44—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device
- A61B8/4444—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device related to the probe
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
-
- 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
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Radiology & Medical Imaging (AREA)
- Surgery (AREA)
- Molecular Biology (AREA)
- Animal Behavior & Ethology (AREA)
- Heart & Thoracic Surgery (AREA)
- Biophysics (AREA)
- Veterinary Medicine (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Quality & Reliability (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
The application discloses a breast ultrasonic screening system based on artificial intelligence, which belongs to the technical field of breast screening, and comprises a body surface micro-navigation subsystem, a control system and a control system, wherein the body surface micro-navigation subsystem is used for recording probe tracks, moving speed, coverage, probe pointing directions and inclined directions in the breast ultrasonic screening process and realizing standardized ultrasonic scanning tracks; the artificial intelligent focus identifying subsystem is used for focus detection, analysis and interpretation, and is combined with matching analysis of an artificial intelligent large database to make accurate diagnosis even higher than that of a common clinician, and the inspection report and focus display subsystem is used for generating a structured report, clearly describing and identifying the focus; according to the application, through standardized ultrasonic scanning track and standard image storage, automatic image storage, measurement and feature description of focus are completed by means of artificial intelligence, a structured report is generated, meanwhile, accurate diagnosis is made, a large amount of time can be saved, missed scanning and misdiagnosis can be prevented as much as possible, and large-scale screening quality and efficiency of mammary glands are remarkably improved.
Description
Technical Field
The application relates to the technical field of breast screening, in particular to a breast ultrasonic screening system and method based on artificial intelligence.
Background
According to reports, the incidence rate of breast cancer is high, and the breast cancer becomes malignant tumor with the highest incidence rate of females, thereby seriously threatening the body and physical and mental health of females. From the characteristics of the onset of the breast cancer, the breast cancer is slow to develop in the early stage, the screening time is sufficient, a small part of the breast cancer advances and spreads fast, but the majority of the breast cancer can reach a plurality of years, and the breast cancer can be found early only by ensuring that a woman performs breast cancer screening once a year. The early breast cancer belongs to in-situ cancer, does not need to carry out radiotherapy or chemotherapy, has very high success rate of direct intervention, and can reach more than 95% of survival rate of patients in 5 years.
Therefore, the mammary gland screening is particularly important, and the survival rate can be obviously improved by early diagnosis and early treatment. In the past, molybdenum target X-rays are mainly used as a main method, and Clinical Breast Examination (CBE), breast ultrasound US, breast magnetic resonance MRI and infrared rays are assisted, but the molybdenum target examination has radioactivity, and the accuracy rate of compact breast is obviously reduced.
It is common that asian females have mammary glands that are more suitable for ultrasound examination. Traditional breast ultrasound examination has advantages such as convenient, safety, but leak diagnosis rate is high, is applied to large-scale screening, can't ensure comprehensive scanning, and its reason is: unlike CT or MRI, the scanning is formed by a machine, so that a unified section standard can be formed, whether the scanning of each ultrasonic doctor is comprehensive or not can directly determine the diagnosis of a patient, otherwise, the diagnosis can be missed; that is, the existing ultrasonic breast screening method has the defects of high labor cost, high time cost, high requirement on the experience of doctors and the like, is difficult to use for large-scale screening, and particularly has the defects of poor medical conditions and low level of doctors in underdeveloped areas.
In order to solve the above-mentioned problems, a person skilled in the art has made many efforts, for example, chinese patent application with application number 202011603978.9 entitled "an artificial intelligent breast cancer ultrasonic screening system and screening method", discloses a breast ultrasonic screening method including a central processing module, a diagnosis module and a community service module, but the patent still cannot solve the problems of missed scan and misdiagnosis caused by the ultrasonic scanning track;
for example, a Chinese patent with application number 201911007859.4 and name of "breast ultrasound screening method, device and system" discloses a breast ultrasound method by means of an automation technology and an artificial intelligence technology, wherein a camera is used for shooting to obtain a depth image for model reconstruction, a three-dimensional structure model of a region to be scanned is obtained, and then a mechanical arm of a scanning mechanism is controlled in a remote transmission mode to drive an ultrasound probe for breast scanning. This patent requires extremely high equipment requirements, requires specialized camera equipment, mainframe, robotic arm and screening platform, is expensive to screen, and takes a long time to examine a patient, and is not suitable for large-scale examination such as breast screening.
Therefore, there is a need in the art for a method of large-scale screening of breast that saves time, prevents missed scans, and misdiagnosis.
Disclosure of Invention
It is an object of the present application to provide an artificial intelligence based breast ultrasound screening system to solve the above-mentioned problems.
In order to achieve the above purpose, the technical scheme adopted by the application is as follows: an artificial intelligence based breast ultrasound screening system comprising:
the body surface micro-navigation subsystem is used for recording probe tracks, moving speed, coverage, probe pointing direction and inclination direction in the breast ultrasonic scanning process, prompting uncovered areas, displaying probe positions and directions in real time in body marking intention, and realizing standardized ultrasonic scanning tracks through artificial intelligence so as to facilitate quality control; the ultrasonic inspection process can be effectively monitored, so that the ultrasonic inspection quality is improved, and the aim of quality control is fulfilled;
the artificial intelligent focus recognition subsystem is used for recognizing, detecting and judging focuses, relies on the powerful image recognition and deep learning technology of artificial intelligence and the support of a large database, achieves the effect of being faster than a professional doctor in terms of detection efficiency and precision of images, and can reduce the misjudgment rate;
the inspection report and focus display subsystem is used for automatically storing images, measuring and describing characteristics of focuses, generating a structured report and generating a focus schematic diagram by combining body surface micro-navigation.
As a preferable technical scheme: the body surface micro-navigation subsystem comprises a magnetic navigator, an infrared scanner, a continuous scanning video recorder and other technologies capable of realizing track recording.
AI (Artificial Intelligence ) is the use of artificial intelligence software to identify ultrasound stored images, automatically identifying and diagnosing images as normal or focal. The artificial intelligence depends on powerful image recognition and deep learning technologies, can be faster than a professional doctor in terms of both detection efficiency and accuracy of images, and can also reduce the misjudgment rate of manual operation.
The application relates to an artificial intelligent focus recognition subsystem, which is a prior art in the field, and the basic technology and principle are applied to mammary gland ultrasonic diagnosis, so as to acquire the mammary gland image data of all quadrants of a checked person and synchronously and automatically acquire whether the image data has the characteristics of occupation, occupation size and occupation; based on the image data, an artificial intelligent mammary gland diagnosis and treatment algorithm model is constructed by using a deep learning algorithm, and a mammary gland focus screening diagnosis and treatment system is constructed based on the model so as to automatically analyze whether lesions exist at each part of the mammary gland of a checked person and obtain corresponding guidance and referral opinions. The breast diagnosis and treatment algorithm model is based on a huge breast ultrasonic image database, DICOM data image preprocessing is performed by utilizing the existing self-adaptive normalization technology, meanwhile, a regional convolution neural network and a multi-scale fine granularity classification model are used for detection, a three-dimensional depth reconstruction model of an attention neural network is matched and positioned, and a multi-view fusion network is combined with a migration learning technology to perform harmony, so that glandular parenchyma constitution, focus types and focus property analysis can be realized, and a graphic-text structured report is formed.
The application is by means of artificial intelligence technology and organically combines the artificial intelligence technology with the body surface micro-navigation subsystem, can more accurately and rapidly identify and diagnose the focus while preventing the leakage of the breast ultrasonic examination, thereby improving the efficiency and quality of the breast screening.
Compared with the prior art, such as the scanning track of the Chinese patent named breast ultrasonic screening method, device and system, the method and the device have the advantages that the optimal scanning track is designed through a computer after three-dimensional reconstruction, and the scanning track is the actual scanning track of an ultrasonic doctor, so that the method and the device can be used for quality control and review. Under the diagnosis experience of many years, an ultrasonic doctor forms a very perfect scanning thought, the scanning track is uniform and standard, excessive design is not needed, and only feedback evaluation of whether the scanning is complete is needed after the completion.
The application also aims at providing an artificial intelligence-based breast ultrasonic screening method, which adopts the technical scheme that the method comprises the following steps:
(1) An ultrasound examination, the ultrasound examination comprising the steps of:
(1.1) starting a body surface micro-navigation subsystem to finish initial calibration;
(1.2) starting an artificial intelligent focus recognition subsystem to assist focus detection;
(1.3) scanning bilateral breasts and armpits sequentially, wherein the scanning range covers armpits, quadrants of breasts and nipple areola areas, a body surface micro-navigation subsystem records scanning tracks, an artificial intelligent focus recognition subsystem continuously records scanning images, and real-time analysis and preservation of image numbers of focus detection are carried out;
(2) A normalized memory map, the normalized memory map comprising:
(2.1) conventional mapping;
(2.2) focus mapping;
(3) A structured inspection report is formed by the inspection report and the lesion display subsystem.
As a preferable technical scheme: in the step (1.1), a magnetic navigation method is adopted to carry out body surface micro navigation, and the specific method is as follows:
placing a magnetic force transmitting device below the examination bed, performing spatial navigation by utilizing magnetic force, attaching magnetic force patches on corresponding positions of the probe, performing probe marking on A, B, C pieces of the magnetic force patches, and recording the position and the direction of the probe in the examination process;
the patient is fully exposed, the fixed body position is not moved and then the calibration is carried out, and the probe is sequentially placed in the following points for calibration:
(1) the probe is arranged at the armpit at the left side of the patient, the long axis is parallel to the axillary midline, the mark of the A point at the upper edge of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(2) the probe is arranged at the nipple on the left side of the patient, the long axis is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface;
(3) the probe is arranged at the armpit on the right side of the patient, the long axis is parallel to the axillary midline, the mark of the A point on the upper edge of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(4) the probe is arranged at the nipple on the right side of the patient, the long axis is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface.
Then, the breast ultrasonic examination is sequentially carried out, the body surface micro-navigation system records the moving track and the moving speed of the probe, and the position and the direction of the probe can be automatically displayed on the body mark graph when the probe is static.
As a preferable technical scheme: in the step (1.1), the body surface micro navigation is carried out by adopting an infrared scanning method, and the specific method comprises the following steps:
an infrared scanning device is arranged above the examination bed, and the position of the probe is recorded by infrared scanning.
As a preferable technical scheme: in the step (1.1), a continuous scanning video recording method is adopted to carry out body surface micro navigation, and the specific method is as follows: and recording the scanned image of the probe by using the close-range camera, encrypting and storing the scanned image, and performing time axis matching with the imaging on the ultrasonic instrument.
As a preferable technical scheme: in step (1.3), the scanning mode includes, but is not limited to radial, reverse radial, rotary and parallel movement.
As a preferable technical scheme: the focus deposit map in the step (2.2) comprises a breast placeholder focus, which specifically comprises:
(1) the maximum diameter line section of the focus comprises a gray-scale static diagram and a longitudinal section, and automatic measurement is completed by artificial intelligence;
(2) the maximum section perpendicular to the step (1) comprises a gray-scale static diagram and a cross section, and the automatic measurement is completed by artificial intelligence;
(3) the most abundant section of blood flow in focus, namely color Doppler static diagram, the color sampling frame comprises focus and tissue of at least 1cm around the focus;
for the following nodules of BI-RADS 2 class, only the three sections are stored, and for the above nodules of BI-RADS 3 class, the following dynamic diagram is also stored, and meanwhile, the analysis function of the artificial intelligent focus recognition subsystem is started:
(4) slowly scanning the longitudinal section of the focus, and storing a gray level map from the outside of one side edge of the focus to the outside of the other side edge of the focus, wherein the focus is free-present-not present;
(5) slowly scanning the cross section of the focus, and storing the gray-scale images from the outside of one side edge of the focus to the outside of the other side edge of the focus.
As a preferable technical scheme: the focus map in the step (2.2) also comprises mammary duct expansion, which specifically comprises the following steps:
when the breast duct is found to expand in scanning, namely the inner diameter is more than or equal to 0.20cm, a gray-scale static image at the widest part of the breast duct is reserved;
observing whether there is a space-occupying focus in the catheter, if so, the time-to-live images are the same as the above;
the clinical examination or subjects complaining of nipple discharge and hemorrhage leave a gray-scale static image of the bilateral areola region.
As a preferable technical scheme: the focal map in step (2.2) also includes lymph node focal maps. The stored image comprises a lymph node maximum diameter line section, including a gray-scale static image and a longitudinal section, and the automatic measurement is completed by artificial intelligence; the maximum tangent plane perpendicular to the maximum radial tangent plane comprises a gray-scale static diagram and a cross section, and the automatic measurement is completed by artificial intelligence; the most abundant section of lymph node blood flow, i.e. the color Doppler static image, the color sampling frame contains the focus and its surrounding tissue of at least 1 cm.
Compared with the prior art, the application has the advantages that: according to the application, through standardized ultrasonic scanning tracks and standard image storage, automatic image storage, measurement and feature description of focuses are completed by means of an artificial intelligent focus recognition subsystem, a structural report is generated, meanwhile, accurate diagnosis is made, a large amount of time can be saved, missing scanning and misdiagnosis are prevented as much as possible, large-scale screening quality and efficiency of mammary glands are remarkably improved, the recognition accuracy of the focus recognition subsystem based on deep learning of the application on mammary glands can reach more than 90%, the auxiliary diagnosis effect is remarkable, and the screening efficiency is improved by more than 30% compared with the prior art.
Drawings
FIG. 1 is a schematic diagram of magnetic navigation according to an embodiment of the present application;
FIG. 2 is a schematic diagram of probe calibration according to an embodiment of the present application;
fig. 3 is a schematic diagram of a double milk nodule according to an embodiment of the present application: the boundaries between 1 and 12 represent the direction of the tumor, the circles with broken lines represent the distance from the nipple, the distance from the inner circle to the nipple is 2cm, the middle is 4cm, and the outer circle is 6cm; stars, triangles and circles all represent the location of the tumor; different shapes represent different classifications of mass: stars represent 4 or more classes; triangles represent class 3; circles represent class 2.
Detailed Description
The application will be further described with reference to the accompanying drawings.
Examples:
an artificial intelligence based breast ultrasonic screening system comprises a body surface micro-navigation subsystem, an artificial intelligence focus identification subsystem and an inspection report and focus display subsystem, wherein,
the body surface micro-navigation subsystem:
aiming at the problem that the breast ultrasound examination is easy to miss, the application provides a body surface micro-navigation design, namely, the probe track, the moving speed, the coverage area, the probe pointing direction, the inclined direction and the like in the breast ultrasound examination are recorded by means of science and technology, an uncovered area is prompted, the miss is avoided, the probe position and the probe direction are displayed in real time in the body marking intention, the body marking time can be saved for an examining doctor, the later quality control is convenient, and the ultrasound examination data is more convincing.
The supported technological means comprise: magnetic navigation, infrared scanning, continuous scanning video recording, and the like.
1.1 Magnetic navigation
And a magnetic force transmitting device is arranged below the examination bed, and space navigation is performed by utilizing magnetic force. In FIG. 1, magnetic patches (A, B, C pieces in total) are attached to corresponding positions of the probe to carry out probe marking, so as to record the position and the direction of the probe in the inspection process;
the patient is fully exposed, the fixed position is not moved, and then the calibration is carried out by sequentially placing the probe in the following points as shown in fig. 2:
(1) the probe is arranged at the armpit at the left side of the patient, the long axis is parallel to the axillary midline, the upper edge (marked by the point A) of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(2) the probe is arranged at the nipple on the left side of the patient, the long axis is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface;
(3) the probe is arranged at the armpit on the right side of the patient, the long axis is parallel to the axillary midline, the upper edge (marked by the point A) of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(4) the probe is arranged at the nipple on the right side of the patient, the long shaft is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface;
then, the breast ultrasonic examination is sequentially carried out, the body surface micro-navigation system records the moving track and the moving speed of the probe, and the position and the direction of the probe can be automatically displayed on the body mark graph when the probe is static.
1.2 Infrared scanning
An infrared scanning device is arranged above the examination bed, and the position of the probe is recorded by utilizing infrared scanning; if magnetic navigation is needed, infrared identification marks are attached to corresponding positions of the probe, and as the human body outline can be identified by infrared scanning, calibration can be omitted or only calibration can be carried out on nipple positions at two sides.
1.3 Continuous scanning video recording
And recording the scanned image of the probe by using the close-range camera, encrypting and storing the scanned image, and performing time axis matching with the imaging on the ultrasonic instrument.
The artificial intelligence focus recognition subsystem:
the large-scale screening workload is huge, the diagnosis level of first-line screening doctors such as communities is good and bad, a certain deviation is easy to generate on focus identification, the working pressure of the screening doctors can be reduced by utilizing the artificial intelligent focus identification subsystem for focus identification, the time is saved, the missed diagnosis rate and the misdiagnosis rate of breast ultrasonic examination can also be reduced, the AI system can be awakened and controlled by voice (the AI system is known by a person skilled in the art, and the working efficiency is further improved).
The inspection report and lesion display subsystem:
automatic image storage, measurement and feature description of the focus are completed by means of an AI focus recognition system, a structured report is generated, and a focus schematic diagram is generated by combining body surface micro navigation (figure 3). The examination report and the image are uploaded to the platform and pushed to the point, so that time is saved for doctors, and patients can conveniently acquire the report.
Methods of performing breast ultrasound screening that rely on the above-described systems are further described in more detail below and may include, but are not limited to, the following schemes:
1. propaganda and reservation
Community and family contractual doctor propaganda, public number article putting and community living committee warm reminding; weChat public number reserves time and place, and fills in the basic information table on line or assisted by community.
2. Sign-in and ordering
The WeChat public number reminding on the same day of inspection can be used for knowing the conditions of the candidates at different inspection sites through WeChat, and peak shifting inspection can be avoided, so that long-time waiting is avoided; and on-site scanning the identification card to obtain the number, and associating the related identification information and the past checking result from the big data center.
3. Before ultrasonic examination
3.1 The early information is input (preferably paperless, such as a mobile phone and a computer), and the submitting work number is preferably set, so that backtracking and calculation of workload are facilitated;
the staff checks whether the basic information table is completely filled in, and submits after the basic information table is qualified in filling;
inquiry and filling in "clinical information form", after filling in qualification "submit"
Part of patients are informed of the need of superior referral by the project participation informed consent
Whether ABVS quality control spot check informed consent is willing to participate in later quality control
3.2 Risk assessment
According to the contents submitted by project participation informed consent and ABVS quality control spot check informed consent, automatically completing risk assessment and labeling in the background, wherein the assessment basis is breast cancer risk assessment-T/CPMA 014 2020
3.3 Clinical palpation
The community doctor finishes palpation and fills in the form, preferably inputs voice, and the computer terminal operates;
3.4 Other examinations include, but are not limited to, blood drawing, molybdenum targets, gynecological examinations, and the like.
4. Ultrasonic inspection
4.1 Waiting area
Scrolling the announced video, wherein the video content can comprise how to cooperate with examination, how to acquire reports, breast cancer risk factors, how to review regularly, and the like;
4.2 Examination zone
4.2.1 Patient preparation
The patient fully exposes the breast and armpits (the two hands are lifted to abduct), and recommends to take a recumbent position (the larger breast can take half lateral positions or other special positions to be marked), (one exposure to the position is performed, the examination process is prevented from moving, and the trace record of the micro navigation system is prevented from being interfered);
4.2.2 Inspection flow
The linear array high-frequency probe is selected, the central frequency is more than or equal to 7.5MHz, and the frequency is improved as much as possible and can reach 15MHz at most on the premise of ensuring enough scanning depth. The depth of the image should envelop breast tissue and the pectoral large muscle behind, the maximum depth is suitable for displaying pleura, the image of a large number of lungs is not included, and the common depth is about 3-4 cm and can be adjusted according to specific conditions. The gray scale image should cause the subcutaneous fat lobules to appear as moderate echoes (a reasonable parameter range preset by the instrument engineer). The velocity scale of the color Doppler image is usually 3-5 cm/s, and the color Doppler image is suitable for preventing obvious color noise;
starting a body surface micro-navigation subsystem to finish initial calibration;
the detection function of the AI focus recognition system is started to assist focus detection.
Scanning the bilateral mammary glands and armpits sequentially, wherein the scanning mode comprises but is not limited to: radial, reverse radial, rotary, parallel movement and the like, the scanning range should be comprehensive, and the armpit, each quadrant of the mammary gland and the nipple areola area should be covered. The micro-navigation system will record the scan trajectory. The AI focus recognition system continuously records the scanned images, analyzes the scanned images in real time and reserves the image numbers of focus detection;
4.3 Image acquisition (standardized memory diagram)
4.3.1 conventional stock plot: leaving the gray-scale ultrasonic image (1 each on the left and right) of the section with the most glands displayed
4.3.2 lesion memory map:
4.3.2.1 mammary gland space occupying venereal disease range
(1) The maximum diameter line section (gray-scale static diagram, longitudinal section) of the focus, AI finishes automatic measurement;
(2) the maximum section (gray-scale static diagram, cross section) perpendicular to the step (1), AI completes automatic measurement;
(3) the most abundant section of blood flow in focus (color Doppler static diagram), the color sampling frame should contain focus and its peripheral at least 1cm tissue;
for the following nodules of BI-RADS 2 class, only the three sections are stored, and for the above nodules of BI-RADS 3 class, the following dynamic diagram is stored (meanwhile, the function of the AI focus recognition system is started to be "analyzed"):
(4) slowly scanning the longitudinal section of the focus, and storing a gray-scale map from the outside of one side edge of the focus to the outside of the other side edge (none-existence-non-existence);
(5) slowly scanning the cross section of the focus, and storing a gray-scale map from the outside of one side edge of the focus to the outside of the other side edge (none-existence-non-existence);
4.3.2.2 Mammary duct expansion
When the breast duct is found to expand (the inner diameter is more than or equal to 0.20 cm) in scanning, a gray-scale static image of the widest part of the breast duct is reserved;
observing whether there is a space-occupying focus in the catheter or not, if so, the time-to-live images are the same as above;
the examinee who clinically examines or complains of nipple discharge and hemorrhage should keep a gray-scale static image of the double-sided areola region, pay attention to the light scanning force, and do not need to manually press the catheter;
4.3.2.3 Lymph node lesions
The stored images are the same as those of breast placeholder lesions (1) (2) (3).
5. Report acquisition and reporting (hereinafter, X-class is classified according to BI-RADS, omitted for convenience of writing)
Class a.1: community doctors submit reports, patients inquire at mobile phone ends and public number reminds
b. Class 2, part 3: the benign lesions are found, and the c option standard is not met, community doctors submit reports, patients inquire at the mobile phone end, and public numbers are reminded
c. Class 0, part 3, class 4, and above: patient meeting 'first screening and second screening criteria', multiple persons double-blind interpretation after image uploading, and report submitting after examination by superior doctors (which can be compared with background searching past examination data):
c1. the method is characterized in that the method is good in possibility, the regular ultrasonic review is recommended, the patient inquires a report at a mobile phone end, and the public number is reminded;
c2. the method is characterized in that the possibility of malignancy, the ultrasound review of a superior hospital, a simple report, a public number prompt and a platform provide consultation are suggested;
the patient is subjected to ultrasonic review to a higher-level hospital;
c21. two screens for benign or low grade suspected malignancy: giving the patient a paper report and uploading the system;
special medical treatment of mammary gland, molybdenum target, MRI, elasticity imaging, radiography, ABVS and the like;
c22. highly suspected malignancy: giving the patient a paper report and uploading the system;
in time, the special department of mammary gland makes a doctor visit, molybdenum target, MRI, elastic presentation, radiography, ABVS, biopsy, operation and the like.
6. Quality control
6.1 Personnel qualification control trains an ultrasonic doctor (and a medical student), and an account number and a password are obtained after qualification audit and can participate in screening; and periodically online to participate in image secondary examination (comprising confirmed diagnosis images and images to be classified);
6.2 Scanning quality control: the micro navigation system records the scanning track, so that the missing scanning is avoided;
6.3 And (3) controlling the image quality: uploading all the reserved images; obtaining an account number and a password by an ultrasonic doctor after qualification auditing, logging in a double-blind interpretation image of an image system, and keeping an interpretation record in the background, wherein the suspicious image is interpreted repeatedly;
6.4 ABVS sampling quality control
According to the ABVS quality control sampling standard, an automatic mammary gland full volume imaging ultrasonic (ABVS) examination is carried out on an included patient, and the ABVS examination result is compared with a manual scanning result in a matching manner.
7. Follow-up visit
7.1 The examination result is a patient checked regularly, and the public number is reminded of expiration;
7.2 Suggesting patients with special doctor of mammary gland, combining other image examination and biopsy pathology, and reminding periodic review by BI-RADS classification marks;
7.3 The surgical patients are classified according to the postoperative pathological result marks and enter postoperative periodic follow-up
7.4 The mobile phone platform provides a consultation channel.
It should be noted that, the above-mentioned "basic information table", clinical information table ", project participation informed consent form", ABVS quality control spot check informed consent form ", breast cancer risk assessment-T/CPMA 014 2020", primary screening inclusion two-screen standard "and" ABVS quality control sampling standard "are all prior art, and those skilled in the art can select or modify as appropriate. The foregoing is the prior art known and understood by those skilled in the art, except as specifically illustrated and described herein, as well as where the present application is innovative.
The foregoing description of the preferred embodiments of the application is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the application.
Claims (9)
1. Mammary gland ultrasonic screening system based on artificial intelligence, characterized by comprising
The body surface micro-navigation subsystem is used for recording probe tracks, moving speed, coverage, probe pointing direction and inclination directions in the breast ultrasonic scanning process, prompting uncovered areas, displaying probe positions and directions in real time in body marking intention, and realizing standardized ultrasonic scanning tracks through artificial intelligence;
the artificial intelligent focus recognition subsystem is used for recognizing, detecting and judging focuses;
the inspection report and focus display subsystem is used for automatically storing images, measuring and describing characteristics of focuses, generating a structured report and generating a focus schematic diagram by combining body surface micro-navigation;
wherein: the body surface micro-navigation subsystem comprises a magnetic navigator, an infrared scanner or a continuous scanning video recorder;
when the magnetic navigation method is adopted for body surface micro navigation, the specific method is as follows:
placing a magnetic force transmitting device below the examination bed, performing spatial navigation by utilizing magnetic force, attaching magnetic force patches on corresponding positions of the probe, performing probe marking on A, B, C pieces of the magnetic force patches, and recording the position and the direction of the probe in the examination process; the patient is fully exposed, the fixed body position is not moved and then the calibration is carried out, and the probe is sequentially placed in the following points for calibration:
(1) the probe is arranged at the armpit at the left side of the patient, the long axis is parallel to the axillary midline, the mark of the A point at the upper edge of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(2) the probe is arranged at the nipple on the left side of the patient, the long axis is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface;
(3) the probe is arranged at the armpit on the right side of the patient, the long axis is parallel to the axillary midline, the mark of the A point on the upper edge of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(4) the probe is arranged at the nipple on the right side of the patient, the long shaft is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface;
then sequentially performing breast ultrasonic examination, and recording the moving track and moving speed of the probe by the body surface micro-navigation system, wherein the probe position and direction can be automatically displayed on the body marking intention when the probe is stationary;
when the infrared scanning method is adopted for body surface micro navigation, the specific method comprises the following steps:
an infrared scanning device is arranged above the examination bed, and the position of the probe is recorded by utilizing infrared scanning;
when the body surface micro-navigation is carried out by adopting a continuous scanning video recording method, the specific method comprises the following steps:
recording the scanned image of the probe by using a close-range camera, encrypting and storing the scanned image, and performing time axis matching with an image on an ultrasonic instrument;
the automatic image storage of the focus comprises a breast placeholder focus, and specifically comprises the following steps:
(1) the maximum diameter line section of the focus comprises a gray-scale static diagram and a longitudinal section, and automatic measurement is completed by artificial intelligence;
(2) the maximum section perpendicular to the step (1) comprises a gray-scale static diagram and a cross section, and the automatic measurement is completed by artificial intelligence;
(3) the most abundant section of blood flow in focus, namely color Doppler static diagram, the color sampling frame comprises focus and tissue of at least 1cm around the focus;
for the following nodules of BI-RADS 2 class, only the three sections are stored, and for the above nodules of BI-RADS 3 class, the following dynamic diagram is also stored, and meanwhile, the analysis function of the artificial intelligent focus recognition subsystem is started:
(4) slowly scanning the longitudinal section of the focus, and storing a gray level map from the outside of one side edge of the focus to the outside of the other side edge of the focus, wherein the focus is free-present-not present;
(5) slowly scanning the cross section of the focus, and storing the gray-scale images from the outside of one side edge of the focus to the outside of the other side edge of the focus.
2. An artificial intelligence based breast ultrasound screening method, characterized by using the system of claim 1, comprising the steps of:
(1) An ultrasound examination, the ultrasound examination comprising the steps of:
(1.1) starting a body surface micro-navigation subsystem to finish initial calibration;
(1.2) starting an artificial intelligent focus recognition subsystem to assist focus detection;
(1.3) scanning bilateral breasts and armpits sequentially, wherein the scanning range covers armpits, quadrants of breasts and nipple areola areas, a body surface micro-navigation subsystem records scanning tracks, an artificial intelligent focus recognition subsystem continuously records scanning images, and real-time analysis and preservation of image numbers of focus detection are carried out;
(2) A normalized memory map, the normalized memory map comprising:
(2.1) conventional mapping;
(2.2) focus mapping;
(3) A structured inspection report is formed by the inspection report and the lesion display subsystem.
3. The method according to claim 2, characterized in that: in the step (1.1), a magnetic navigation method is adopted to carry out body surface micro navigation, and the specific method is as follows:
placing a magnetic force transmitting device below the examination bed, performing spatial navigation by utilizing magnetic force, attaching magnetic force patches on corresponding positions of the probe, performing probe marking on A, B, C pieces of the magnetic force patches, and recording the position and the direction of the probe in the examination process; the patient is fully exposed, the fixed body position is not moved and then the calibration is carried out, and the probe is sequentially placed in the following points for calibration:
(1) the probe is arranged at the armpit at the left side of the patient, the long axis is parallel to the axillary midline, the mark of the A point at the upper edge of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(2) the probe is arranged at the nipple on the left side of the patient, the long axis is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface;
(3) the probe is arranged at the armpit on the right side of the patient, the long axis is parallel to the axillary midline, the mark of the A point on the upper edge of the probe is level with the collarbone, and the direction of the probe is parallel to the bed surface;
(4) the probe is arranged at the nipple on the right side of the patient, the long shaft is parallel to the collarbone, the nipple is positioned at the midpoint of the probe, and the direction of the probe is perpendicular to the bed surface;
then, the breast ultrasonic examination is sequentially carried out, the body surface micro-navigation system records the moving track and the moving speed of the probe, and the position and the direction of the probe can be automatically displayed on the body marking intention when the probe is static.
4. The method according to claim 2, characterized in that: in the step (1.1), the body surface micro navigation is carried out by adopting an infrared scanning method, and the specific method comprises the following steps:
an infrared scanning device is arranged above the examination bed, and the position of the probe is recorded by infrared scanning.
5. The method according to claim 2, characterized in that: in the step (1.1), a continuous scanning video recording method is adopted to carry out body surface micro navigation, and the specific method is as follows: and recording the scanned image of the probe by using the close-range camera, encrypting and storing the scanned image, and performing time axis matching with the imaging on the ultrasonic instrument.
6. The method according to claim 2, characterized in that: in step (1.3), the scanning mode includes radial, reverse radial, rotary or parallel movement.
7. The method according to claim 2, characterized in that: the focus deposit map in the step (2.2) comprises a breast placeholder focus, which specifically comprises:
(1) the maximum diameter line section of the focus comprises a gray-scale static diagram and a longitudinal section, and automatic measurement is completed by artificial intelligence;
(2) the maximum section perpendicular to the step (1) comprises a gray-scale static diagram and a cross section, and the automatic measurement is completed by artificial intelligence;
(3) the most abundant section of blood flow in focus, namely color Doppler static diagram, the color sampling frame comprises focus and tissue of at least 1cm around the focus;
for the following nodules of BI-RADS 2 class, only the three sections are stored, and for the above nodules of BI-RADS 3 class, the following dynamic diagram is also stored, and meanwhile, the analysis function of the artificial intelligent focus recognition subsystem is started:
(4) slowly scanning the longitudinal section of the focus, and storing a gray level map from the outside of one side edge of the focus to the outside of the other side edge of the focus, wherein the focus is free-present-not present;
(5) slowly scanning the cross section of the focus, and storing the gray-scale images from the outside of one side edge of the focus to the outside of the other side edge of the focus.
8. The method according to claim 2, characterized in that: the focus map in the step (2.2) also comprises mammary duct expansion, which specifically comprises the following steps:
when the breast duct is found to expand in scanning, namely the inner diameter is more than or equal to 0.20cm, a gray-scale static image at the widest part of the breast duct is reserved; observing whether there is a space-occupying focus in the catheter, if so, the time-to-live images are the same as the above; the clinical examination or subjects complaining of nipple discharge and hemorrhage leave a gray-scale static image of the bilateral areola region.
9. The method according to claim 2, characterized in that: the focal map in the step (2.2) also comprises a lymph node focal map; the lymph node focus map comprises:
(1) the maximum diameter line section of the lymph node comprises a gray-scale static image and a longitudinal section, and automatic measurement is completed by artificial intelligence;
(2) the maximum tangent plane perpendicular to the maximum radial tangent plane comprises a gray-scale static diagram and a cross section, and the automatic measurement is completed by artificial intelligence;
(3) the most abundant section of lymph node blood flow, i.e. the color Doppler static image, the color sampling frame contains the focus and its surrounding tissue of at least 1 cm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210205941.3A CN114376615B (en) | 2022-03-04 | 2022-03-04 | Mammary gland ultrasonic screening system and screening method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210205941.3A CN114376615B (en) | 2022-03-04 | 2022-03-04 | Mammary gland ultrasonic screening system and screening method based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114376615A CN114376615A (en) | 2022-04-22 |
| CN114376615B true CN114376615B (en) | 2023-09-15 |
Family
ID=81205981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210205941.3A Active CN114376615B (en) | 2022-03-04 | 2022-03-04 | Mammary gland ultrasonic screening system and screening method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114376615B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107809958A (en) * | 2015-06-30 | 2018-03-16 | 威斯康星校友研究基金会 | The operator that indiscipline or minimum level are trained carries out obstetrics' imaging at medical center |
| CN108665456A (en) * | 2018-05-15 | 2018-10-16 | 广州尚医网信息技术有限公司 | The method and system that breast ultrasound focal area based on artificial intelligence marks in real time |
| CN112508941A (en) * | 2020-12-25 | 2021-03-16 | 上海深博医疗器械有限公司 | Three-dimensional ultrasonic scanning integrity detection method and device |
| CN112603368A (en) * | 2020-12-25 | 2021-04-06 | 上海深博医疗器械有限公司 | Mammary gland ultrasonic navigation and diagnosis system and method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5210080B2 (en) * | 2008-07-31 | 2013-06-12 | 富士フイルム株式会社 | Medical imaging device |
-
2022
- 2022-03-04 CN CN202210205941.3A patent/CN114376615B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107809958A (en) * | 2015-06-30 | 2018-03-16 | 威斯康星校友研究基金会 | The operator that indiscipline or minimum level are trained carries out obstetrics' imaging at medical center |
| CN108665456A (en) * | 2018-05-15 | 2018-10-16 | 广州尚医网信息技术有限公司 | The method and system that breast ultrasound focal area based on artificial intelligence marks in real time |
| CN112508941A (en) * | 2020-12-25 | 2021-03-16 | 上海深博医疗器械有限公司 | Three-dimensional ultrasonic scanning integrity detection method and device |
| CN112603368A (en) * | 2020-12-25 | 2021-04-06 | 上海深博医疗器械有限公司 | Mammary gland ultrasonic navigation and diagnosis system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114376615A (en) | 2022-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wojcinski et al. | The automated breast volume scanner (ABVS): initial experiences in lesion detection compared with conventional handheld B-mode ultrasound: a pilot study of 50 cases | |
| JP5543444B2 (en) | Method and system for performing a biopsy | |
| Zeng et al. | Machine Learning-Based Medical Imaging Detection and Diagnostic Assistance | |
| CN114376615B (en) | Mammary gland ultrasonic screening system and screening method based on artificial intelligence | |
| CN115429327A (en) | Thyroid full-automatic ultrasonic intelligent scanning method and platform | |
| CN115399795A (en) | Nuclear medicine diagnosis device and diagnosis system used for same | |
| Li et al. | The concordance in lesion detection and characteristics between the Anatomical Intelligence and conventional breast ultrasound Scan method | |
| Turnbull et al. | Comparative effectiveness of magnetic resonance imaging in breast cancer (COMICE trial) | |
| O'Connor et al. | Complications of breast core biopsy | |
| Evans et al. | Sloane Project: a prospective audit of screen detected ductal carcinoma in situ | |
| UK Mammotome Trial Group et al. | UK National Health Service Breast Screening Programme (NHSBSP) multicentre image guided biopsy trial: an update | |
| Beral et al. | Breast cancer and hormone replacement therapy (HRT): collaborative reanalysis of data from 51 epidemiological studies of 52,705 women with breast cancer and 108,411 women without breast cancer | |
| Holland | Ductal carcinoma in situ (DCIS) | |
| Carder et al. | An audit of benign biopsies performed in patients assessed by the Leeds/Wakefield Breast Screening Unit | |
| Jensen et al. | Evaluation of Optical Sonography for Real-Time Breast Imaging and Biopsy Guidance | |
| Eve et al. | Changes in reporting practices with time when radiographers and radiologists double read screening mammograms | |
| Turnbull et al. | Predicting breast cancer response to chemotherapy using quantitative magnetic resonance (MR) imaging | |
| Whitehead | Hormone replacement therapy (HRT): indications and side effects | |
| Williams et al. | An evaluation of the current role of radiographer film readers in breast screening | |
| Boal et al. | Roadworks ahead: a collaborative approach to process redesign | |
| Lakhani | Molecular pathology of breast cancer | |
| Duffy | The effect of screening across the age spectrum | |
| Hendrick | Review of clinical trials | |
| Secretariat | University of York, 17–19 July 2002 | |
| Wivell et al. | Evaluation of the need for magnification views in the assessment of calcification in computerised mammography |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |