CN115841601A - Method, system, equipment and medium for identifying and judging head form of interventional catheter - Google Patents
Method, system, equipment and medium for identifying and judging head form of interventional catheter Download PDFInfo
- Publication number
- CN115841601A CN115841601A CN202211190946.XA CN202211190946A CN115841601A CN 115841601 A CN115841601 A CN 115841601A CN 202211190946 A CN202211190946 A CN 202211190946A CN 115841601 A CN115841601 A CN 115841601A
- Authority
- CN
- China
- Prior art keywords
- dsa
- catheter
- class
- representing
- situation
- 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
Images
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention provides a method, a system, equipment and a medium for identifying and judging the shape of a head end of an interventional catheter, which relate to the technical field of identification of medical instruments and comprise the following steps: image acquisition and processing steps: acquiring a data set from an interventional operation case database system, preprocessing the data set, defining the form category of an ROI (region of interest) picture of the head end of a catheter, and establishing a training set and a testing set; model training: building a training model, and training according to the training set; a real-time identification step: inputting the test set into the trained training model, and outputting the recognition and judgment result of the catheter shape. The invention can enhance the positioning area for displaying the head end of the catheter, assist in judging the pushing position of the catheter in the interventional operation and clearly determine the orientation of the head end of the catheter.
Description
Technical Field
The invention relates to the technical field of identification of medical instruments, in particular to a method, a system, equipment and a medium for identifying and judging the shape of a head end of an interventional catheter.
Background
Cardiovascular and cerebrovascular diseases have become a common disease at present. Because the interventional operation has the advantages of accuracy, rapidness, small harm to patients and the like, the interventional operation becomes a main means for treating cardiovascular and cerebrovascular diseases. During the interventional operation, the doctor extends one end of the catheter into the pathological part through the artery and the vein through a puncture way for delivering the contrast medium, so as to know the corresponding trend of the blood vessel and the morphological characteristics of the blood vessel, and provide important help for the judgment of the disease and the next treatment of the disease.
However, for the currently clinically mature radiography catheter, the operation difficulty of performing stretching or bending movement inside the human body is large, and the manual control of the catheter may cause injury to the patient, and the safety is low. In addition, the motion precision and the operation effect of the manual control catheter are determined by the experience and the capability of an operator, and the difference of the operation effect is large.
In conclusion, how to solve the problems of direction change, uncertainty and the like in the catheter pushing process has become a problem to be solved by the technical personnel in the field.
The invention patent with publication number CN110288693A discloses an intelligent micro-catheter shaping system for intracranial aneurysm interventional surgery, which comprises the steps of firstly obtaining an aneurysm image and related parameters, reconstructing a blood vessel three-dimensional model through a digital modeling module, calculating, segmenting the shape of the aneurysm and marking, calculating an optimal micro-catheter path, printing the shape of the micro-catheter path in equal proportion through a 3D printing module, shaping through an external detachable steam heating shaping device, and further applying to clinical treatment; the patented method has high cost and manufacturing cost, the realization of the technical route is complex, and the whole treatment time is prolonged; in addition, the method can not be applied to DSA images because the morphology is predicted in advance by CT images, and can not provide help in real-time surgery.
The invention patent with publication number CN111166329A provides a stretchable annular catheter shape determining method and a device, the patent method needs to be provided with an excitation signal emission source and a magnetic field signal emission source, a free shape plane of the annular catheter is determined according to coordinate data of a magnetic sensor after data collected by a catheter electrical impedance sensor and the magnetic sensor are amplified, and finally the catheter shape is corrected according to a calculation result. The method has high cost and high manufacturing cost, and the application scene of the operation is limited due to the introduction of the electromagnetic system. Meanwhile, this method cannot function on DSA images.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a method, a system, equipment and a medium for identifying and judging the shape of an interventional catheter tip.
According to the method, the system, the equipment and the medium for identifying and judging the shape of the head end of the interventional catheter, the scheme is as follows:
in a first aspect, a method for identifying and determining a morphology of an interventional catheter tip is provided, where the method includes:
image acquisition and processing steps: acquiring a data set from an interventional operation case database system, preprocessing the data set, defining the morphological category of an ROI (region of interest) picture of the head end of a catheter, and establishing a training set and a testing set;
model training: building a training model, and training according to the training set;
a real-time identification step: inputting the test set into the trained training model, and outputting the recognition and judgment result of the catheter shape.
Preferably, the image acquisition and processing step includes defining morphology categories of ROI pictures of the catheter tip, the morphology categories including:
class 0: representing the situation that the projection form in the DSA is approximately a straight line;
class 1: representing the situation that the DSA enters from the upper part of the visual field and the projection form is towards the right;
class 2: representing the situation that the DSA enters from the upper part of the visual field and the projection form is towards the left;
and 3 types: representing the situation that the DSA enters from the lower part of the visual field and the projection form is towards the left;
class 4: representing the situation that the projection mode enters from the lower part of the visual field and faces the right in the DSA;
class 5: representing the situation that the DSA enters from the left of the visual field and the projection form is upward;
class 6: representing the situation that the DSA enters from the left of the visual field and the projection form is downward;
7 types: representing the situation that the DSA enters from the right of the visual field and the projection form is upward;
class 8: this represents a case where the DSA is viewed from the right side and the projection mode is downward.
Preferably, the training model comprises: an input layer, a convolution layer, a pooling layer, a full-link layer and an output layer;
in the output layer, 84 neurons of the full connection layer are filled into a SoftMax function, a tensor with the output length equal to the number of categories is obtained, and the activated positions in the tensor represent the categories.
Preferably, the weight initialization of the interconnection among the model neurons in the training model adopts Gaussian distribution, the standard deviation of the distribution is (N/2) ^0.5, and N is the number of input nodes of each neuron.
Preferably, the type of the output result of the identification and determination result of the catheter shape comprises:
1) Displaying the ROI area selected by the frame, displaying the identification and judgment category of the head end form of the current leader, and additionally displaying a correct result/a statistical result;
2) Only the identification judgment category of the head end shape is displayed.
In a second aspect, there is provided a system for identifying and determining the morphology of an interventional catheter tip, the system comprising:
the image acquisition and processing module: collecting a data set from an interventional operation case database system, preprocessing the data set, defining morphological categories of ROI (region of interest) pictures of a catheter head end, and establishing a training set and a testing set;
a model training module: building a training model, and training according to the training set;
a real-time identification module: inputting the test set into the trained training model, and outputting the recognition and judgment result of the catheter shape.
Preferably, the image acquisition and processing module includes morphology categories defining ROI pictures of the catheter tip, the morphology categories including:
class 0: representing the situation that the projection form in the DSA is approximately a straight line;
class 1: representing the situation that the projection form enters from the upper part of the visual field and faces the right in the DSA;
class 2: representing the situation that the DSA enters from the upper part of the visual field and the projection form is towards the left;
class 3: representing the situation that the DSA enters from the lower part of the visual field and the projection form is towards the left;
class 4: representing the situation that the projection mode enters from the lower part of the visual field and faces the right in the DSA;
class 5: representing the situation that the DSA enters from the left of the visual field and the projection form is upward;
class 6: representing the situation that the DSA enters from the left of the visual field and the projection form is downward;
7 types: representing the situation that the DSA enters from the right of the visual field and the projection form is upward;
class 8: this represents a case where the DSA is viewed from the right side and the projection mode is downward.
Preferably, the training model comprises: an input layer, a convolution layer, a pooling layer, a full-link layer and an output layer;
in the output layer, 84 neurons of the full connection layer are filled into a SoftMax function to obtain a tensor with the output length equal to the number of categories, and the activated positions in the tensor represent the categories;
the weight initialization of the mutual connection between model neurons in the training model adopts Gaussian distribution, the standard deviation of the distribution is (N/2) ^0.5, and N is the number of input nodes of each neuron;
the type of the output result of the identification and judgment of the catheter form comprises the following steps:
1) Displaying the ROI area selected by the frame, displaying the identification and judgment category of the head end form of the current leader, and additionally displaying a correct result/a statistical result;
2) Only the identification judgment category of the head end shape is displayed.
In a third aspect, an apparatus is provided, the apparatus comprising:
one or more processors;
storage means for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement the steps in the method.
In a fourth aspect, a computer-readable storage medium is provided, in which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the method.
Compared with the prior art, the invention has the following beneficial effects:
1. according to the invention, DSA data is locally preprocessed, and a model is built according to Lenet5, so that the characteristics of the head end of the interventional catheter are identified in real time and classified and judged;
2. the invention greatly improves the effective rate of identifying the head end form of the interventional catheter by verifying on the DSA image data set.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is a schematic overall flow diagram of the present invention;
FIG. 2 is a definition of morphology categories of an ROI picture of a tip of an interventional catheter;
FIG. 3 is a schematic view of a model structure;
FIG. 4 is a schematic view showing a first type of the catheter tip shape determination result;
fig. 5 is a schematic view showing a type two of the result of the shape determination of the catheter tip.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications can be made by persons skilled in the art without departing from the concept of the invention. All falling within the scope of the present invention.
The embodiment of the invention provides an identification and judgment method of the shape of a head end of an interventional catheter, which is used for identifying and judging the shape of the head end of the catheter in a vascular interventional operation in a real-time DSA image to serve for an automatic operation. Specifically, the identification method mainly comprises three parts of image acquisition and processing, model training and real-time identification. In the DSA radiography process, an original DSA image is preprocessed, case data are made by adopting a similar Mnist hand-written data set, then a model trained based on Lenet5 is input to learn the characteristics of the pre-marked head end shape of the interventional catheter, and the characteristic knowledge is stored in a weight value which continuously changes along with gradient reduction, so that training is completed, and a final model for prediction is obtained; finally, the shape of the head end of the interventional catheter is identified; meanwhile, the invention can also enhance the positioning area for displaying the head end of the catheter, can help doctors, computers or surgical robots to judge the pushing position of the catheter in the interventional operation, helps doctors, computers or surgical robots to clearly and correctly orient the head end of the catheter, and is favorable for giving more accurate operation decision. Referring to fig. 1, the method specifically includes the following steps:
image acquisition and processing steps: data sets are collected from an interventional surgical case database system and preprocessed, morphological categories of ROI pictures of a catheter tip are defined, and a training set and a testing set are established.
Specifically, all data sets in this step are from a standard slave interventional surgical case database system. Then locally preprocessed to determine ROI (region of interest) pictures of the catheter tip by extracting the catheter tip position. Limiting the pixel size of the ROI picture to the size required by a model input layer, and making case data by adopting a similar Mnist handwriting data set; and finally, carrying out category marking on all collected pictures, and dividing the pictures into a training set and a testing set according to a universal proportion.
Classification criteria are defined in the process data set. Catheter tip morphology was defined as 9 categories:
class 0: representing the situation that the projection form in the DSA is approximately a straight line;
class 1: this represents a case where the DSA enters from above the field of view and the projection mode is directed to the right.
Class 2: this represents the case where the DSA is viewed from above and the projection pattern is oriented to the left.
And 3 types: this represents a case where the DSA enters from below the visual field and the projection pattern is left.
Class 4: this represents a case where the DSA enters from below the visual field and the projection mode is directed to the right.
Class 5: this represents the case where the DSA is viewed from the left and the projection is directed upward.
Class 6: this represents the case where the DSA is viewed from the left and the projection mode is oriented downward.
7 types: this represents a case where the DSA is viewed from the right side and the projection mode is directed upward.
Class 8: this represents a case where the DSA is viewed from the right side and the projection mode is downward.
Model training: and building a training model and training according to the training set.
Specifically, the steps include: the training model is built by following the Lenet5 model, and during the training process, when the nonlinear layer, sigmoid, is passed, the distance from the center is far (the derivative is close to 0), so that the gradient disappearance occurs prematurely. Therefore, the weight initialization of the interconnection among the model neurons adopts Gaussian distribution, the standard deviation of the distribution is (N/2) ^0.5, and N is the number of input nodes of each neuron. In the final output layer, 84 neurons of the second fully-connected layer are filled into a SoftMax function to obtain a tensor with the output length equal to the number of categories, and the activated positions in the tensor represent the categories. (for example, a tensor of [0,0,0,1, 0] is activated at index =3, and thus the image represented by the tensor belongs to the third category, and belongs to the contrast catheter with the starting point down and the direction to the left).
A real-time identification step: and inputting the test set into the trained training model, and outputting a recognition judgment result of the catheter shape.
Specifically, the recognition result is displayed in real time from a DSA image display of the automatic surgical robot system, and the display strictly conforms to the DICOM Part 14 standard, can accurately present the fine shadow Part of the medical image and further conforms to the diagnosis requirement of a doctor. The DICOM calibrated display can ensure that the display consistency is presented for a long time at different display terminals. In the DSA display, the vacant position of an original DSA image is reasonably utilized to display the identification and judgment result of the catheter shape, the head end region is selected in a frame, and the prediction result and the statistical result are displayed in the lower right corner in a unified mode.
During real-time identification, the shape judgment result of the catheter tip shows that two outputs are allowed: 1. and displaying the ROI area selected by the frame, displaying the identification and judgment category of the head end shape of the current front conduit, and displaying a correct result/statistical result in the lower right corner. 2. Only the type of identification and judgment of the shape of the tip is shown, for example, type 2, the catheter starting point is above, and the catheter tip is oriented to the left.
Next, the present invention will be described in more detail.
1. Image acquisition and processing:
the method comprises the steps of collecting required case data from an interventional operation case database system to manufacture an applicable data set, screening out a DSA image of catheter radiography, and enabling all the radiography catheters to enter from a single direction and output from the single direction. All data sets were set as single-channel grayscale threshold pictures. Then locally preprocessed to determine ROI (region of interest) pictures of the catheter tip by extracting the catheter tip position. Limiting the pixel size of the ROI picture to the size required by a model input layer, and making case data by adopting a similar Mnist handwriting data set; and finally, carrying out category marking on all collected pictures, and dividing the pictures into a training set and a testing set according to a universal proportion.
Unlike tomography, DSA images are the dose of x-rays transmitted through the radiation field to the imaging field, and the catheter is presented in DSA renderings as a 2-dimensional projection of the radiation direction, as seen by the physician, as is the catheter tip morphology. Therefore, the catheter tip configurations are defined as 9 categories, wherein 0 category represents the case where the projection configurations are approximately a straight line in the DSA, 1 category 2 category represents the case where the DSA enters from above the visual field, the projection configurations are right and left, 3 category 4 category represents the case where the DSA enters from below the visual field, the projection configurations are left and right, 5 category 6 category represents the case where the DSA enters from left of the visual field, the projection configurations are up and down, 7 category 8 category represents the case where the DSA enters from right of the visual field, and the projection configurations are up and down, respectively, as shown in fig. 2.
2. Model training:
the training model is built according to a Lenet5 model, and comprises an input layer, a convolution layer, a pooling layer and a full-connection layer. The model structure is as shown in fig. 3, in the final output layer, 84 neurons of the second fully-connected layer are filled into a SoftMax function, so as to obtain a tensor with the output length equal to the number of categories, and the activated positions in the tensor represent the categories. (for example, a tensor of [0,0,0,1, 0] is activated at index =3, and thus the image represented by the tensor belongs to the third category, and belongs to the contrast catheter with the starting point down and the direction to the left). Inputting the training set in 1 into a model for training, wherein the model updates the weight in the training process and each weight parameter needs to have a corresponding initial value. The initialization of the model weight is important for the training of the network, poor initialization parameters can cause the problem of gradient propagation, and the training speed is reduced; going through a nonlinear layer such as sigmoid is farther from the center (derivative is close to 0), so that the gradient vanishing occurs prematurely. Therefore, the weight initialization of the interconnection among the model neurons adopts Gaussian distribution, the standard deviation of the distribution is (N/2) ^0.5, and N is the number of input nodes of each neuron.
3. Real-time identification:
the identification result is displayed in real time from a DSA image display of the full-automatic surgical robot system, the display strictly conforms to the DICOM Part 14 standard, the fine shadow Part of the medical image can be accurately presented, and the diagnosis requirement of a doctor is met. The DICOM calibrated display can ensure that the display consistency is presented for a long time at different display terminals.
In a DSA display, the vacant space of an original DSA image is reasonably utilized, the recognition and judgment result of the catheter shape is displayed, the head end region is selected, two result type output modes are set, and the prediction result and the statistical result are displayed at the lower right corner. As shown in fig. 4 and 5.
Fig. 4 is a schematic diagram showing a first type of the catheter tip morphology determination result, in which a rectangular frame is used to select the tip ROI region. The current conduit head end shape recognition is class 0, the statistical result is 70 predicted, and 51 correct. Fig. 5 shows a schematic view of the type two of the catheter tip shape determination results, and optionally only the tip shape determination is displayed, as identified in the figure as type 2 (catheter start point is above, catheter tip is oriented to the left).
The present invention further provides a system for identifying and determining the shape of the tip of the interventional catheter, and those skilled in the art can understand the method for identifying and determining the shape of the tip of the interventional catheter provided by the present invention as a specific implementation manner of the system for identifying and determining the shape of the tip of the interventional catheter, that is, the system for identifying and determining the shape of the tip of the interventional catheter can be implemented by executing the steps of the method for identifying and determining the shape of the tip of the interventional catheter.
The method, the system, the equipment and the medium for identifying and judging the shape of the head end of the interventional catheter provided by the embodiment of the invention realize real-time identification of the characteristics of the head end of the interventional catheter and carry out classification judgment by locally preprocessing DSA data and building a model according to Lenet 5. The identification of the morphology of the tip of the interventional catheter was finally demonstrated to be effective by validation on the DSA image dataset.
Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, units provided by the present invention as pure computer readable program code, the system and its various devices, modules, units provided by the present invention can be fully implemented by logically programming method steps in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system and various devices, modules and units thereof provided by the invention can be regarded as a hardware component, and the devices, modules and units included in the system for realizing various functions can also be regarded as structures in the hardware component; means, modules, units for realizing various functions can also be regarded as structures in both software modules and hardware components for realizing the methods.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (10)
1. A method for identifying and judging the shape of the head end of an interventional catheter is characterized by comprising the following steps:
image acquisition and processing steps: acquiring a data set from an interventional operation case database system, preprocessing the data set, defining the form category of an ROI (region of interest) picture of the head end of a catheter, and establishing a training set and a testing set;
model training: building a training model, and training according to the training set;
a real-time identification step: inputting the test set into the trained training model, and outputting the recognition and judgment result of the catheter shape.
2. The method as claimed in claim 1, wherein the image obtaining and processing step includes defining a morphology category of an ROI image of the catheter tip, the morphology category includes:
class 0: representing the situation that the projection form in the DSA is approximately a straight line;
class 1: representing the situation that the projection form enters from the upper part of the visual field and faces the right in the DSA;
class 2: representing the situation that the DSA enters from the upper part of the visual field and the projection form is towards the left;
class 3: representing the situation that the DSA enters from the lower part of the visual field and the projection form is towards the left;
class 4: representing the situation that the projection mode enters from the lower part of the visual field and faces the right in the DSA;
class 5: representing the situation that the DSA enters from the left of the visual field and the projection form is upward;
class 6: representing the situation that the DSA enters from the left of the visual field and the projection form is downward;
7 types: representing the situation that the DSA enters from the right of the visual field and the projection form is upward;
class 8: this represents a case where the DSA is viewed from the right side and the projection mode is downward.
3. The method for recognizing and determining the morphology of an interventional catheter tip as set forth in claim 1, wherein the training model comprises: an input layer, a convolution layer, a pooling layer, a full-link layer and an output layer;
in the output layer, a plurality of neurons of the full connection layer are filled into a SoftMax function, tensor with the output length equal to the number of categories is obtained, and the activated positions in the tensor represent the categories.
4. The method of claim 1, wherein the initialization of the weights of the interconnections between the neurons in the training model is Gaussian distribution with a standard deviation of (N/2) ^0.5, and N is the number of input nodes per neuron.
5. The method for identifying and determining the morphology of the interventional catheter tip as set forth in claim 1, wherein the output of the identification and determination result of the morphology of the catheter includes two types:
1) Displaying the ROI area selected by the frame, displaying the identification and judgment category of the head end form of the current leader pipe, and additionally displaying a correct result/a statistical result;
2) Only the identification judgment category of the head end shape is displayed.
6. A system for identifying and determining the morphology of an interventional catheter tip, comprising:
the image acquisition and processing module: acquiring a data set from an interventional operation case database system, preprocessing the data set, defining the form category of an ROI (region of interest) picture of the head end of a catheter, and establishing a training set and a testing set;
a model training module: building a training model, and training according to the training set;
a real-time identification module: inputting the test set into the trained training model, and outputting the recognition and judgment result of the catheter shape.
7. The system for recognizing and determining the morphology of an interventional catheter tip as set forth in claim 6, wherein the image acquisition and processing module includes morphology categories defining an ROI picture of the catheter tip, the morphology categories including:
class 0: representing the situation that the projection form in the DSA is approximately a straight line;
class 1: representing the situation that the projection form enters from the upper part of the visual field and faces the right in the DSA;
class 2: representing the situation that the DSA enters from the upper part of the visual field and the projection form is towards the left;
class 3: representing the situation that the DSA enters from the lower part of the visual field and the projection form is towards the left;
class 4: representing the situation that the projection mode enters from the lower part of the visual field and faces the right in the DSA;
class 5: representing the situation that the DSA enters from the left of the visual field and the projection mode is upward;
class 6: representing the situation that the DSA enters from the left of the visual field and the projection form is downward;
7 types: representing the situation that the DSA enters from the right of the visual field and the projection form is upward;
class 8: this represents a case where the DSA is viewed from the right side and the projection mode is downward.
8. The system of claim 6, wherein the training model comprises: an input layer, a convolution layer, a pooling layer, a full-link layer and an output layer;
in the output layer, filling a plurality of neurons of the full connection layer into a SoftMax function to obtain a tensor with the output length equal to the number of categories, wherein the activated positions in the tensor represent the categories;
the weight initialization of the mutual connection between model neurons in the training model adopts Gaussian distribution, the standard deviation of the distribution is (N/2) ^0.5, and N is the number of input nodes of each neuron;
the output of the identification and judgment result of the catheter form comprises the following two types:
1) Displaying the ROI area selected by the frame, displaying the identification and judgment category of the head end form of the current leader pipe, and additionally displaying a correct result/a statistical result;
2) Only the identification judgment category of the head end shape is displayed.
9. An apparatus, characterized in that the apparatus comprises:
one or more processors;
storage means for storing one or more programs;
the one or more programs, when executed by the one or more processors, cause the one or more processors to perform the steps of the method recited in any of claims 1-5.
10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211190946.XA CN115841601A (en) | 2022-09-28 | 2022-09-28 | Method, system, equipment and medium for identifying and judging head form of interventional catheter |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211190946.XA CN115841601A (en) | 2022-09-28 | 2022-09-28 | Method, system, equipment and medium for identifying and judging head form of interventional catheter |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115841601A true CN115841601A (en) | 2023-03-24 |
Family
ID=85575461
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211190946.XA Pending CN115841601A (en) | 2022-09-28 | 2022-09-28 | Method, system, equipment and medium for identifying and judging head form of interventional catheter |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115841601A (en) |
-
2022
- 2022-09-28 CN CN202211190946.XA patent/CN115841601A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2015163089A1 (en) | Medical image information system, medical image information processing method, and program | |
Girum et al. | Learning with context feedback loop for robust medical image segmentation | |
JP2017174039A (en) | Image classification device, method, and program | |
Han et al. | Automatic segmentation of human placenta images with U-Net | |
CN111369528A (en) | Coronary artery angiography image stenosis region marking method based on deep convolutional network | |
US20210059758A1 (en) | System and Method for Identification, Labeling, and Tracking of a Medical Instrument | |
CN110310280A (en) | Hepatic duct and the image-recognizing method of calculus, system, equipment and storage medium | |
US20240078676A1 (en) | Interactive coronary labeling using interventional x-ray images and deep learning | |
JP2022531989A (en) | Automatic coronary angiography analysis | |
WO2021117043A1 (en) | Automatic stenosis detection | |
Lee et al. | Online learning for the hyoid bone tracking during swallowing with neck movement adjustment using semantic segmentation | |
da Cruz et al. | Enabling autonomous medical image data annotation: A human-in-the-loop reinforcement learning approach | |
KR102103281B1 (en) | Ai based assistance diagnosis system for diagnosing cerebrovascular disease | |
US20230222668A1 (en) | Image processing apparatus, image processing method, and recording medium | |
CN115841601A (en) | Method, system, equipment and medium for identifying and judging head form of interventional catheter | |
CN114359208B (en) | Head and neck blood vessel segmentation method and device, readable storage medium and electronic equipment | |
WO2022223042A1 (en) | Surgical path processing system, method, apparatus and device, and storage medium | |
CN111144163A (en) | Vein and artery identification system based on neural network | |
CN114782443A (en) | Device and storage medium for data-based enhanced aneurysm risk assessment | |
KR102095731B1 (en) | Mra image learning method and assistance diagnosis method for blood vessel lesion of deep learning based assistance diagnosis system | |
Miao et al. | CoWRadar: Visual Quantification of the Circle of Willis in Stroke Patients. | |
WO2024156255A1 (en) | Target recognition method based on image, and neural network model processing method | |
CN112950582B (en) | 3D lung focus segmentation method and device based on deep learning | |
Yao | Machine Learning and Image Processing for Clinical Outcome Prediction: Applications in Medical Data from Patients with Traumatic Brain Injury, Ulcerative Colitis, and Heart Failure | |
Zamani et al. | DeepNav: Joint View Learning for Direct Optimal Path Perception in Cochlear Surgical Platform Navigation |
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 |