CN111820950A - Personalized information determination device and ultrasonic training method - Google Patents
Personalized information determination device and ultrasonic training method Download PDFInfo
- Publication number
- CN111820950A CN111820950A CN202010583192.9A CN202010583192A CN111820950A CN 111820950 A CN111820950 A CN 111820950A CN 202010583192 A CN202010583192 A CN 202010583192A CN 111820950 A CN111820950 A CN 111820950A
- Authority
- CN
- China
- Prior art keywords
- information
- ultrasonic
- personalized
- training
- network model
- 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
- 238000012549 training Methods 0.000 title claims abstract description 125
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000015654 memory Effects 0.000 claims abstract description 34
- 238000005259 measurement Methods 0.000 claims abstract description 14
- 210000004204 blood vessel Anatomy 0.000 claims description 52
- 238000002604 ultrasonography Methods 0.000 claims description 38
- 210000002837 heart atrium Anatomy 0.000 claims description 15
- 210000003734 kidney Anatomy 0.000 claims description 15
- 210000004185 liver Anatomy 0.000 claims description 12
- 210000000481 breast Anatomy 0.000 claims description 11
- 210000001715 carotid artery Anatomy 0.000 claims description 10
- 210000001685 thyroid gland Anatomy 0.000 claims description 10
- 238000005452 bending Methods 0.000 claims description 7
- 238000012937 correction Methods 0.000 claims description 6
- 208000009453 Thyroid Nodule Diseases 0.000 claims description 5
- 208000024770 Thyroid neoplasm Diseases 0.000 claims description 5
- 208000019425 cirrhosis of liver Diseases 0.000 claims description 5
- 230000003902 lesion Effects 0.000 claims description 5
- 238000003384 imaging method Methods 0.000 claims description 3
- 238000001303 quality assessment method Methods 0.000 claims description 3
- 238000013441 quality evaluation Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 abstract description 28
- 238000012545 processing Methods 0.000 abstract description 10
- 230000000694 effects Effects 0.000 abstract description 4
- 238000004891 communication Methods 0.000 description 8
- 238000013528 artificial neural network Methods 0.000 description 7
- 210000003754 fetus Anatomy 0.000 description 6
- 210000004072 lung Anatomy 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000003211 malignant effect Effects 0.000 description 3
- 206010006187 Breast cancer Diseases 0.000 description 2
- 208000026310 Breast neoplasm Diseases 0.000 description 2
- 230000003187 abdominal effect Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 210000002758 humerus Anatomy 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- 210000000689 upper leg Anatomy 0.000 description 2
- 201000001320 Atherosclerosis Diseases 0.000 description 1
- 206010006272 Breast mass Diseases 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000747 cardiac effect Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Images
Classifications
-
- 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
- A61B8/5223—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for extracting a diagnostic or physiological parameter from medical diagnostic data
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B23/00—Models for scientific, medical, or mathematical purposes, e.g. full-sized devices for demonstration purposes
- G09B23/28—Models for scientific, medical, or mathematical purposes, e.g. full-sized devices for demonstration purposes for medicine
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Educational Technology (AREA)
- Chemical & Material Sciences (AREA)
- Algebra (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Business, Economics & Management (AREA)
- Educational Administration (AREA)
- Medicinal Chemistry (AREA)
- Theoretical Computer Science (AREA)
- Physiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Mathematics (AREA)
- Biophysics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Radiology & Medical Imaging (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a personalized information determining device and an ultrasonic training method. The device comprises a memory and a processor, wherein at least one program instruction is stored in the memory, and the processor loads and executes the at least one program instruction to realize the following steps: acquiring ultrasonic information, wherein the ultrasonic information is an ultrasonic image and/or an ultrasonic video; inputting the ultrasonic information into an individualized network model, wherein the output of the individualized network model is individualized information determined according to the ultrasonic information; the personalized network model is obtained by training according to sample information, the sample information comprises n pieces of ultrasonic sample information obtained by medical staff through historical drawing and diagnostic information or measurement information corresponding to each piece of ultrasonic sample information, and n is an integer larger than 1. The problems that manual diagnosis efficiency is low and misdiagnosis is possible in the existing scheme are solved; the effect of improving the diagnosis efficiency and accuracy is achieved.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a personalized information determining device and an ultrasonic training method.
Background
Ultrasound equipment is one of the commonly used auxiliary devices for medical diagnosis by virtue of being fast and painless and harmless to a subject.
In the existing scheme, after a medical staff examines an object, the medical staff judges whether the object has a lesion according to an obtained ultrasound image. However, due to the influences of the mood of the medical staff on the day and the busy degree of work, the judgment result of the medical staff has certain subjectivity, and misdiagnosis or efficiency influence may occur.
Disclosure of Invention
In view of this, the embodiment of the present invention provides a personalized information determining apparatus and an ultrasound training method, so as to solve the problems that misdiagnosis may occur and the efficiency is low in the existing scheme.
According to a first aspect, an embodiment of the present invention provides a personalized information determination apparatus, including a memory and a processor, where the memory stores at least one program instruction, and the processor loads and executes the at least one program instruction to implement the following steps:
acquiring ultrasonic information, wherein the ultrasonic information is an ultrasonic image and/or an ultrasonic video;
inputting the ultrasonic information into a personalized network model, wherein the output of the personalized network model is personalized information determined according to the ultrasonic information;
the personalized network model is obtained by training according to sample information, the sample information comprises n pieces of ultrasonic sample information obtained by medical staff through historical drawing and personalized information of diagnosis information or measurement information corresponding to each piece of ultrasonic sample information, and n is an integer larger than 1.
Optionally, the personalized information includes: a determination of a lesion in the ultrasound information and/or an image quality assessment of the ultrasound information.
Optionally, the ultrasound information is information obtained during ultrasound training;
the personalized network model is used for outputting the image quality evaluation of the ultrasonic information.
Optionally, the ultrasound information is information including a target object, and the target object includes at least one of a blood vessel, a liver, a kidney, a heart, a thyroid, a carotid artery, and a breast.
Optionally, the processor is further configured to implement the following steps:
receiving an adjusting instruction for adjusting the weight of each training parameter in the personalized network model; when the target object is a blood vessel, the training parameters comprise at least one of a blood vessel type, a blood vessel position, a blood vessel bending angle, an inner diameter of the blood vessel, a blood vessel plaque size, plaque elasticity and a new blood vessel in the plaque; when the target object is a liver, the training parameters comprise liver hardness and/or liver cirrhosis size; when the target object is a kidney, the training parameters comprise at least one of a long diameter, a wide diameter, a thick diameter, a resistance index and a pulsatility index of the kidney; when the target object is a heart, the training parameters include at least one of an inner diameter of an atrium, a thickness of an atrium wall, and a space between left and right atria; when the target object is a thyroid gland, the training parameters include a size and/or a shape of a thyroid nodule; when the target object is a carotid artery, the training parameters include the size and/or shape of plaque; when the target object is a breast, the training parameter comprises an aspect ratio;
and adjusting the weight of each training parameter in the personalized network model according to the adjusting instruction.
Optionally, the processor is further configured to implement the following steps:
receiving a correction instruction for correcting the personalized information output by the personalized network model;
and correcting the personalized information according to the correction instruction.
Optionally, the processor is further configured to implement the following steps:
after the personalized information is corrected, adding the ultrasonic information and the corrected personalized information to the sample information, and updating the personalized network model through the updated sample information;
alternatively, the first and second electrodes may be,
submitting the ultrasonic information and the corrected personalized information to a training server, adding the ultrasonic information and the corrected personalized information to the sample information by the training server, and updating the personalized network model through the updated sample information.
Optionally, the processor is further configured to implement the following steps:
acquiring the sample information;
and training the initialized network according to the sample information to obtain the personalized network model.
In a second aspect, there is provided a method of ultrasound training, the method comprising:
acquiring sample information, wherein the sample information comprises n pieces of ultrasonic sample information acquired by medical staff through historical imaging and diagnostic information or measurement information corresponding to each piece of ultrasonic sample information, and n is an integer greater than 1;
and training the initialized network according to the sample information to obtain a personalized network model, wherein the personalized network model is used for outputting personalized information corresponding to the ultrasonic information obtained by drawing by the medical staff.
Optionally, the method further includes:
receiving an adjusting instruction for adjusting the weight of each training parameter in the personalized network model; when the target object in the ultrasonic information is a blood vessel, the training parameters comprise at least one of a blood vessel type, a blood vessel position, a blood vessel bending angle, an inner diameter of the blood vessel, a blood vessel plaque size, plaque elasticity and a new blood vessel in the plaque; when the target object is a liver, the training parameters comprise liver hardness and/or liver cirrhosis size; when the target object is a kidney, the training parameters comprise at least one of a long diameter, a wide diameter, a thick diameter, a resistance index and a pulsatility index of the kidney; when the target object is a heart, the training parameters include at least one of an inner diameter of an atrium, a thickness of an atrium wall, and a space between left and right atria; when the target object is a thyroid gland, the training parameters include a size and/or a shape of a thyroid nodule; when the target object is a carotid artery, the training parameters include the size and/or shape of plaque; when the target object is a breast, the training parameter comprises an aspect ratio;
and adjusting the weight of each training parameter in the personalized network model according to the adjusting instruction.
In a third aspect, an ultrasound processing apparatus is provided, the apparatus comprising a memory and a processor, the memory having stored therein at least one program instruction, the at least one program instruction being loaded and executed by the processor to implement the method of the first or second aspect.
In a fourth aspect, there is provided a computer storage medium having stored therein at least one program instruction which is loaded and executed by a processor to implement a method according to the first or second aspect.
After the ultrasonic information is obtained, outputting personalized information according to a personalized network model of the medical care personnel obtained through pre-training, wherein the personalized network model is a network obtained through training according to sample information of the medical care personnel; the problems that manual diagnosis efficiency is low and misdiagnosis is possible in the existing scheme are solved; the effect of automatically outputting the personalized information according to the record of the user so as to improve the diagnosis efficiency and accuracy is achieved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a method flow diagram of an ultrasound training method according to an embodiment of the present invention.
Fig. 2 is a flowchart of a method of determining personalized information according to an embodiment of the present invention.
Fig. 3 is a schematic hardware structure diagram of an ultrasound apparatus provided in an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, 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 some, but not all, embodiments of the present invention. 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.
Referring to fig. 1, a flowchart of a method of ultrasound training provided in an embodiment of the present application is shown, where as shown in fig. 1, the method includes:
After the medical staff makes a picture in history, the ultrasonic information obtained by drawing is judged to obtain a diagnosis result, and after the diagnosis result is determined to be accurate, the ultrasonic information and the corresponding diagnosis result can be added to a sample. Accordingly, the training server may obtain the sample.
During actual implementation, the training server can acquire the historical diagnosis record of a doctor from a working machine of medical personnel, or the working machine of the medical personnel can actively report the historical diagnosis record to the training server; and the training server takes the acquired historical diagnosis record as sample information. The historical diagnosis record comprises ultrasonic information and a diagnosis result corresponding to the ultrasonic information.
Generally, n is a large value, for example, n is a value greater than 1000, in order to ensure the training accuracy.
The ultrasound sample information may be information obtained by scanning a target object, and the target object may be an organ such as a blood vessel, a fetus, a heart, a lung, a thyroid, a carotid artery, a breast, and the like, which is not limited in this embodiment.
The diagnostic information includes: a determination of a lesion in the ultrasound information and/or an image quality assessment of the ultrasound information.
When the target object is a blood vessel, the diagnostic information may be a plaque to which the blood vessel belongs, such as a hard plaque and a soft plaque of atherosclerosis.
When the target object is a fetus, the diagnostic information may be the biperiate diameter, humerus length, femur length, abdominal circumference, head circumference, and the like of the fetus;
when the target object is a heart, the diagnostic information may be a cardiac cycle, a heart rate, or the like;
when the target object is a lung, the diagnosis information can be a B line, a pleural line, a lung effusion and the like;
when the target object is a thyroid gland, the diagnosis information is whether a nodule exists or not and the size of the nodule;
when the target object is a carotid artery, the diagnostic information is the nature of the carotid plaque, such as benign and malignant;
when the target object is a breast, the diagnostic information is the nature of the breast nodule, e.g., benign, malignant.
The above description is only given by taking the target object as the above examples, and in actual implementation, the target object may be another part or the diagnosis information may be other information when the target object is the part, which is not limited in this embodiment. In addition, the diagnostic information may be, in addition to the above, an image quality evaluation of the ultrasound image, for example, in a scene of training a new person, the diagnostic information may be an image quality score for mapping the new person, or whether the image quality is acceptable.
The measurement information is information obtained by measuring an object to be measured in the ultrasonic information, and the object to be measured may be a plaque, a fetus, or a kidney. For example, taking the target to be measured as a fetus as an example, the measurement information may be the size of the double apical diameter, the humerus length, the femur length, the abdominal circumference, and the head circumference. For another example, taking the object to be measured as a kidney as an example, the measurement information may be a long diameter, a wide diameter, a thick diameter, a resistance index and a pulsatility index of the kidney.
And 102, training the initialized network according to the sample information to obtain an individualized network model, wherein the individualized network model is used for outputting individualized information corresponding to the ultrasonic information.
And after the sample information is acquired, training the initialization network according to the sample information. Since the medical staff confirms the accurate historical diagnosis record to be used as the sample information for training, the diagnosis information or the measurement information output by the personalized network model obtained by training meets the diagnosis standard of the medical staff.
In practical implementation, the personalized network model is further used for outputting a personalized workflow according with the usage habit of the medical care personnel, which is not limited in this embodiment.
Optionally, when training the initialization network according to the sample information, the specific steps may include:
and extracting training parameters in the ultrasonic information, and training the initialization network according to each extracted training parameter.
When the target object in the ultrasonic information is a blood vessel, the training parameters comprise at least one of a blood vessel type, a blood vessel position, a blood vessel bending angle, an inner diameter of the blood vessel, a blood vessel plaque size, plaque elasticity and a new blood vessel in the plaque; when the target object is a liver, the training parameters comprise liver hardness and/or liver cirrhosis size; when the target object is a kidney, the training parameters comprise at least one of a long diameter, a wide diameter, a thick diameter, a resistance index and a pulsatility index of the kidney; when the target object is a heart, the training parameters include at least one of an inner diameter of an atrium, a thickness of an atrium wall, and a space between left and right atria; when the target object is a thyroid gland, the training parameters include a size and/or a shape of a thyroid nodule; when the target object is a carotid artery, the training parameters include the size and/or shape of plaque; when the target object is a breast, the training parameter comprises an aspect ratio;
taking the target object as a blood vessel as an example, each training parameter includes a score value as a weight to weight the diagnostic information, and for example, the training result of the blood vessel type 1 is that the weight of the angle of the bent blood vessel is 70%, the weight of the inner diameter of the blood vessel is 30%, the training result of the blood vessel type 2 is that the position of the blood vessel is 30%, the size of the blood vessel plaque is 20%, the elasticity of the plaque is 20%, and the new blood vessel in the plaque is 30%. Also, at this time, the diagnostic information may be a probability that the blood vessel has a lesion, for example, a probability of being a soft plaque is 80%.
When the target object is a breast, the training parameters include an aspect ratio, and when the aspect ratio is less than 1, the breast tumor is benign, and when the aspect ratio is greater than 1, the breast tumor is malignant.
In practical implementation, before the personalized network model is trained according to the training parameters, the training parameters may be scored, where the scoring includes the following types: scoring is carried out on the difference between each training parameter and a preset value, for example, scoring is carried out according to the difference between the bending angle of the blood vessel and a preset angle value, the larger the difference is, the lower the score is, the smaller the difference is, the higher the score is, similarly, scoring is carried out according to the difference between the inner diameter of the blood vessel and the preset inner diameter value, the smaller the difference is, the higher the score is, and the larger the difference is, the lower the score is; the score may also be a score of the degree of contribution of each training parameter to the diagnostic information, such as 70% contribution of the vessel bending angle to the diagnostic information and 30% contribution of the vessel inner diameter to the diagnostic information for a certain embodiment of the vessel.
The training parameters may be parameters defined by medical staff, such as referential indexes used by doctors to obtain diagnostic information, or parameters obtained by automatically recognizing ultrasound images, such as referential indexes of diagnostic information obtained by performing calculation with an algorithm, and the like, which is not limited in this embodiment. Taking training parameters as self-defined parameters of medical staff for example, in general, there are A, B, C, D and E five reference factors for determining the focus of a certain target object, and the factors concerned by different medical staff are different, for example, medical staff 1 sets A, B and C as training parameters, and medical staff 2 sets B, C, D and E as training parameters.
In one embodiment, the training parameters may be reference indicators that influence diagnostic information given by multiple physicians based on their own experience. Medical care personnel label each ultrasonic sample information in the sample information to form a data set which is formed by ultrasonic information, training parameter label and diagnosis information label. In one embodiment, the training parameters can extract indexes affecting diagnostic information from a diagnostic index list corresponding to a disease to form training parameters, and doctors label the training parameters and the diagnostic information to form the training parameters; in another embodiment, the training parameters may be obtained by processing input ultrasound information by algorithm 1, and labeled training parameters and diagnostic information are obtained by using data with ultrasound information using algorithm 2;
in another embodiment, the training parameter is composed of two parts, the training parameter set a is obtained by processing and learning the ultrasound information through an algorithm, and the training parameter set B is formed by marking the training parameter set diagnostic information by a doctor.
In practical implementation, the initialization network comprises a convolutional layer, a pooling layer, an activation function layer, a full connection layer, an embedding layer, a cyclic neural network layer, a long-term and short-term memory layer and the like. The initialization network mainly comprises two or more neural networks, and the first neural network generally comprises a convolution layer, a pooling layer, an activation function layer and a full-connection layer; processing the input ultrasonic information to obtain corresponding training parameters and scores, wherein the second neural network generally consists of an embedding layer, a full-connection layer, a circulating neural network layer and a long-short term memory layer, and is used for processing the specified training parameters and ultrasonic information and outputting diagnostic information; in another embodiment, the system further comprises a third neural network, the third neural network uses the ultrasonic information and the training parameters as input to regress personalized training parameter selection of a specific user or hospital with a specific personalized habit, and uses the training parameters as input to output ultrasonic diagnosis information. The initialization network is trained using the ultrasound information, labeled training parameters, and diagnostic information generated by the previous steps.
It should be noted that, before the training, each piece of ultrasound sample information may be preprocessed, and the initialization network may be trained according to the preprocessed ultrasound sample information. The preprocessing referred to herein may be normalization processing.
In summary, after the sample information is obtained, the initialized network is trained according to the sample information, so as to obtain a personalized network model, wherein the sample information includes n pieces of ultrasound sample information obtained by medical staff through historical imaging and diagnostic information or measurement information corresponding to each piece of ultrasound sample information; therefore, when the medical staff makes a picture again, the personalized information corresponding to the ultrasonic information can be directly output through the personalized network model; the problems that manual diagnosis efficiency is low and misdiagnosis is possible in the existing scheme are solved; the effect of automatically outputting the personalized information according to the record of the user so as to improve the diagnosis efficiency and accuracy is achieved.
It should be noted that, after the personalized network model is obtained through training, the medical staff may modify the weight of each training parameter of the personalized network model according to personal needs, that is, the method may further include:
firstly, receiving an adjusting instruction for adjusting the weight of each training parameter in the personalized network model;
secondly, the weight of each training parameter in the personalized network model is adjusted according to the adjusting instruction.
After receiving the adjustment instruction, the weight of the training parameter is adjusted. And then, the personalized network model after the weight is adjusted is used as the personalized network model obtained by final training, so that the accuracy of the personalized network model obtained by training is improved.
Referring to fig. 2, a flowchart of a method for determining personalized information according to an embodiment of the present application is shown, and as shown in fig. 2, the method includes:
the method comprises the steps of obtaining ultrasonic information obtained by medical staff by drawing, wherein the ultrasonic information can be information acquired by the medical staff in real time through ultrasonic equipment or information acquired by the ultrasonic equipment in advance. The ultrasonic information is an ultrasonic image and/or an ultrasonic video.
The ultrasound information may be information obtained by scanning a target object, and the target object may be an organ such as a blood vessel, a fetus, a heart, a lung, a thyroid, a carotid artery, a breast, and the like, which is not limited in this embodiment.
The personalized network model described in this embodiment is a model obtained by training in the embodiment shown in fig. 1.
The personalized network model is obtained by training according to sample information, the sample information comprises n pieces of ultrasonic sample information obtained by medical staff through historical drawing and diagnostic information or measurement information corresponding to each piece of ultrasonic sample information, and n is an integer larger than 1.
Optionally, after the personalized information is output by the personalized network model, the medical care personnel may subjectively determine whether the personalized information is accurate, and when the determination result is inaccurate, the medical care personnel may correct the personalized information, that is, the method may further include:
firstly, receiving a correction instruction for correcting the personalized information output by the personalized network model;
secondly, the personalized information is corrected according to the correction instruction.
The medical staff corrects the output of the personalized network model, which indicates that the personalized network model needs to be optimized, and at this time, in order to improve the accuracy of the personalized network model, the method may further include:
after the personalized information is corrected, adding the ultrasonic information and the corrected personalized information to the sample information, and updating the personalized network model through the updated sample information;
alternatively, the first and second electrodes may be,
submitting the ultrasonic information and the corrected personalized information to a training server, adding the ultrasonic information and the corrected personalized information to the sample information by the training server, and updating the personalized network model through the updated sample information.
By refreshing the sample information in time, the accuracy of the trained personalized network model is ensured, and the accuracy of the diagnosis information or the measurement information output by the personalized network model is further ensured.
In summary, after the ultrasound information is acquired, the diagnosis information or the measurement information is output according to the personalized network model of the medical care personnel obtained through pre-training, wherein the personalized network model is a network obtained through training according to the sample information of the medical care personnel; the problems that manual diagnosis efficiency is low and misdiagnosis is possible in the existing scheme are solved; the effect of automatically outputting the personalized information according to the record of the user so as to improve the diagnosis efficiency and accuracy is achieved.
The embodiment also discloses a personalized information determination device, which comprises a memory and a processor, wherein the memory stores at least one program instruction, and the processor executes the method in a mode of loading and executing the at least one program instruction.
The present embodiment also discloses a computer storage medium, in which at least one program instruction is stored, and the at least one program instruction is loaded by the processor and executes the method described above.
Referring to fig. 3, fig. 3 is a schematic structural diagram of an ultrasound apparatus according to an alternative embodiment of the present invention, and as shown in fig. 3, the ultrasound apparatus may include: at least one processor 61, such as a CPU (Central Processing Unit), at least one communication interface 63, memory 64, at least one communication bus 62. Wherein a communication bus 62 is used to enable the connection communication between these components. The communication interface 63 may include a Display (Display) and a Keyboard (Keyboard), and the optional communication interface 63 may also include a standard wired interface and a standard wireless interface. The Memory 64 may be a high-speed RAM Memory (volatile Random Access Memory) or a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. The memory 64 may optionally be at least one memory device located remotely from the processor 61. Wherein the processor 61 may be in connection with the apparatus described in fig. 3, an application program is stored in the memory 64, and the processor 61 calls the program code stored in the memory 64 for performing any of the above-mentioned method steps.
The communication bus 62 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The communication bus 62 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown in FIG. 3, but this does not mean only one bus or one type of bus.
The memory 64 may include a volatile memory (RAM), such as a random-access memory (RAM); the memory may also include a non-volatile memory (english: non-volatile memory), such as a flash memory (english: flash memory), a hard disk (english: hard disk drive, abbreviation: HDD), or a solid-state drive (english: SSD); the memory 64 may also comprise a combination of the above types of memory.
The processor 61 may be a Central Processing Unit (CPU), a Network Processor (NP), or a combination of CPU and NP.
The processor 61 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a Programmable Logic Device (PLD), or a combination thereof. The aforementioned PLD may be a Complex Programmable Logic Device (CPLD), a field-programmable gate array (FPGA), a General Array Logic (GAL), or any combination thereof.
Optionally, the memory 64 is also used to store program instructions. The processor 61 may call program instructions to implement the method as shown in the embodiments of fig. 1 and fig. 2 of the present application.
Embodiments of the present invention further provide a non-transitory computer storage medium, where computer-executable instructions are stored, and the computer-executable instructions may execute the method in any of the above method embodiments. The storage medium may be a magnetic Disk, an optical Disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a Flash Memory (Flash Memory), a Hard Disk (Hard Disk Drive, abbreviated as HDD), a Solid State Drive (SSD), or the like; the storage medium may also comprise a combination of memories of the kind described above.
Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations fall within the scope defined by the appended claims.
Claims (10)
1. An apparatus for personalized information determination, the apparatus comprising a memory having at least one program instruction stored therein and a processor that loads and executes the at least one program instruction to perform the steps of:
acquiring ultrasonic information, wherein the ultrasonic information is an ultrasonic image and/or an ultrasonic video;
inputting the ultrasonic information into a personalized network model, wherein the output of the personalized network model is personalized information determined according to the ultrasonic information;
the personalized network model is obtained by training according to sample information, the sample information comprises n pieces of ultrasonic sample information obtained by medical staff through historical drawing and diagnostic information or measurement information corresponding to each piece of ultrasonic sample information, and n is an integer larger than 1.
2. The apparatus of claim 1, wherein the personalization information comprises: a determination of a lesion in the ultrasound information and/or an image quality assessment of the ultrasound information.
3. The device of claim 1, wherein the ultrasound information is information obtained during ultrasound training;
the personalized network model is used for outputting the image quality evaluation of the ultrasonic information.
4. The apparatus of claim 1, wherein the ultrasound information is information including a target object including at least one of a blood vessel, a liver, a kidney, a heart, a thyroid, a carotid artery, and a breast.
5. The apparatus of claim 4, wherein the processor is further configured to implement the steps of:
receiving an adjusting instruction for adjusting the weight of each training parameter in the personalized network model; when the target object is a blood vessel, the training parameters comprise at least one of a blood vessel type, a blood vessel position, a blood vessel bending angle, an inner diameter of the blood vessel, a blood vessel plaque size, plaque elasticity and a new blood vessel in the plaque; when the target object is a liver, the training parameters comprise liver hardness and/or liver cirrhosis size; when the target object is a kidney, the training parameters comprise at least one of a long diameter, a wide diameter, a thick diameter, a resistance index and a pulsatility index of the kidney; when the target object is a heart, the training parameters include at least one of an inner diameter of an atrium, a thickness of an atrium wall, and a space between left and right atria; when the target object is a thyroid gland, the training parameters include a size and/or a shape of a thyroid nodule; when the target object is a carotid artery, the training parameters include the size and/or shape of plaque; when the target object is a breast, the training parameter comprises an aspect ratio;
and adjusting the weight of each training parameter in the personalized network model according to the adjusting instruction.
6. The apparatus of claim 1, wherein the processor is further configured to implement the steps of:
receiving a correction instruction for correcting the personalized information output by the personalized network model;
and correcting the personalized information according to the correction instruction.
7. The apparatus of claim 6, wherein the processor is further configured to implement the steps of:
after the personalized information is corrected, adding the ultrasonic information and the corrected personalized information to the sample information, and updating the personalized network model through the updated sample information;
alternatively, the first and second electrodes may be,
submitting the ultrasonic information and the corrected personalized information to a training server, adding the ultrasonic information and the corrected personalized information to the sample information by the training server, and updating the personalized network model through the updated sample information.
8. The apparatus of any of claims 1 to 6, wherein the processor is further configured to perform the steps of:
acquiring the sample information;
and training the initialized network according to the sample information to obtain the personalized network model.
9. A method of ultrasound training, the method comprising:
acquiring sample information, wherein the sample information comprises n pieces of ultrasonic sample information acquired by medical staff through historical imaging and diagnostic information or measurement information corresponding to each piece of ultrasonic sample information, and n is an integer greater than 1;
and training the initialized network according to the sample information to obtain a personalized network model, wherein the personalized network model is used for outputting personalized information corresponding to the ultrasonic information obtained by drawing by the medical staff.
10. The method of claim 9, further comprising:
receiving an adjusting instruction for adjusting the weight of each training parameter in the personalized network model; when the target object in the ultrasonic information is a blood vessel, the training parameters comprise at least one of a blood vessel type, a blood vessel position, a blood vessel bending angle, an inner diameter of the blood vessel, a blood vessel plaque size, plaque elasticity and a new blood vessel in the plaque; when the target object is a liver, the training parameters comprise liver hardness and/or liver cirrhosis size; when the target object is a kidney, the training parameters comprise at least one of a long diameter, a wide diameter, a thick diameter, a resistance index and a pulsatility index of the kidney; when the target object is a heart, the training parameters include at least one of an inner diameter of an atrium, a thickness of an atrium wall, and a space between left and right atria; when the target object is a thyroid gland, the training parameters include a size and/or a shape of a thyroid nodule; when the target object is a carotid artery, the training parameters include the size and/or shape of plaque; when the target object is a breast, the training parameter comprises an aspect ratio;
and adjusting the weight of each training parameter in the personalized network model according to the adjusting instruction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010583192.9A CN111820950A (en) | 2020-06-23 | 2020-06-23 | Personalized information determination device and ultrasonic training method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010583192.9A CN111820950A (en) | 2020-06-23 | 2020-06-23 | Personalized information determination device and ultrasonic training method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111820950A true CN111820950A (en) | 2020-10-27 |
Family
ID=72898069
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010583192.9A Pending CN111820950A (en) | 2020-06-23 | 2020-06-23 | Personalized information determination device and ultrasonic training method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111820950A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110490892A (en) * | 2019-07-03 | 2019-11-22 | 中山大学 | A kind of Thyroid ultrasound image tubercle automatic positioning recognition methods based on USFaster R-CNN |
| CN110751179A (en) * | 2019-09-18 | 2020-02-04 | 无锡祥生医疗科技股份有限公司 | Focus information acquisition method, focus prediction model training method and ultrasonic equipment |
| CN111179227A (en) * | 2019-12-16 | 2020-05-19 | 西北工业大学 | Mammary gland ultrasonic image quality evaluation method based on auxiliary diagnosis and subjective aesthetics |
| CN111242921A (en) * | 2020-01-12 | 2020-06-05 | 刘涛 | Method and system for automatically updating medical ultrasonic image auxiliary diagnosis system |
| CN111310851A (en) * | 2020-03-03 | 2020-06-19 | 四川大学华西第二医院 | Artificial intelligence ultrasonic auxiliary system and application thereof |
-
2020
- 2020-06-23 CN CN202010583192.9A patent/CN111820950A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110490892A (en) * | 2019-07-03 | 2019-11-22 | 中山大学 | A kind of Thyroid ultrasound image tubercle automatic positioning recognition methods based on USFaster R-CNN |
| CN110751179A (en) * | 2019-09-18 | 2020-02-04 | 无锡祥生医疗科技股份有限公司 | Focus information acquisition method, focus prediction model training method and ultrasonic equipment |
| CN111179227A (en) * | 2019-12-16 | 2020-05-19 | 西北工业大学 | Mammary gland ultrasonic image quality evaluation method based on auxiliary diagnosis and subjective aesthetics |
| CN111242921A (en) * | 2020-01-12 | 2020-06-05 | 刘涛 | Method and system for automatically updating medical ultrasonic image auxiliary diagnosis system |
| CN111310851A (en) * | 2020-03-03 | 2020-06-19 | 四川大学华西第二医院 | Artificial intelligence ultrasonic auxiliary system and application thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7086336B2 (en) | Tissue nodule detection method and its model Training method, equipment, equipment, system, and its computer program | |
| CN108701354B (en) | Method and system for identifying contour of interest region in ultrasonic image | |
| EP2365471B1 (en) | Diagnosis assisting apparatus, coronary artery analyzing method and recording medium having a coronary artery analzying program stored therein | |
| JP2017515609A (en) | System and method for quantifying valves | |
| CN111325759B (en) | Vessel segmentation method, apparatus, computer device, and readable storage medium | |
| CN110458837B (en) | Image post-processing method and device, electronic equipment and storage medium | |
| CN112052896B (en) | Image processing method and device, and classification model training method and device | |
| JP6957214B2 (en) | Information processing equipment, information processing system, information processing method and program | |
| US11996182B2 (en) | Apparatus and method for medical image reading assistant providing representative image based on medical use artificial neural network | |
| JPWO2020027228A1 (en) | Diagnostic support system and diagnostic support method | |
| CN111568451A (en) | Exposure dose adjusting method and system | |
| CN112529919A (en) | System and method for generating a bull's eye map generation of a subject's heart | |
| CN113989407A (en) | Training method and system for limb part recognition model in CT image | |
| CN116130090A (en) | Ejection fraction measuring method and device, electronic device, and storage medium | |
| CN111863204A (en) | Mammary gland disease AI auxiliary diagnosis method and system based on molybdenum target X-ray photographic examination | |
| CN111860636A (en) | Measurement information prompting method and ultrasonic training method | |
| JP2020151450A (en) | Image identification method and image identification device | |
| WO2020007026A1 (en) | Segmentation model training method and apparatus, and computer-readable storage medium | |
| CN110992312B (en) | Medical image processing method, medical image processing device, storage medium and computer equipment | |
| CN111383236B (en) | Method, apparatus and computer-readable storage medium for labeling regions of interest | |
| CN110739050B (en) | Left ventricle full-parameter and confidence coefficient quantification method | |
| JP2023504026A (en) | Automatic protocol specification in medical imaging systems | |
| US20230222668A1 (en) | Image processing apparatus, image processing method, and recording medium | |
| CN111820950A (en) | Personalized information determination device and ultrasonic training method | |
| CN116050691A (en) | Medical image report evaluation method, device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201027 |
|
| RJ01 | Rejection of invention patent application after publication |