CN114862859B

CN114862859B - Image recognition method, device, system and computer readable storage medium

Info

Publication number: CN114862859B
Application number: CN202210801971.0A
Authority: CN
Inventors: 唐建东; 周国新
Original assignee: Sceneray Co Ltd
Current assignee: Jingyu Medical Technology Suzhou Co ltd
Priority date: 2022-07-08
Filing date: 2022-07-08
Publication date: 2022-09-27
Anticipated expiration: 2042-07-08
Also published as: CN114862859A

Abstract

The application provides an image identification method, an apparatus, a system and a computer-readable storage medium, which are used for identifying a slicing electrode implanted on an electrode lead in the intracranial of a patient, wherein the method comprises the following steps: acquiring medical image data of the patient; acquiring contour data of the electrode lead, contour data of image markers and contour data of the slice electrodes based on the medical image data; acquiring position information and posture information of the electrode lead based on the profile data of the electrode lead; acquiring position information of the image marker based on the outline data of the image marker; and respectively acquiring the outline data of each slice electrode from the outline data of a plurality of slice electrodes based on the position information of the image mark and the relative position relationship between the image mark and each slice electrode so as to obtain the identification result of each slice electrode. The slicing electrodes can be accurately positioned, so that accurate electrical stimulation treatment is realized.

Description

Image recognition method, device, system and computer readable storage medium

Technical Field

The present application relates to the field of implantable devices, image recognition, and deep learning technologies, and in particular, to an image recognition method, an apparatus, a system, and a computer-readable storage medium.

Background

In the implantable medical device industry, physicians continue to manually observe, identify implanted electrodes against images such as material Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and then infer the approximate location and orientation of the electrodes. The patient must carry and provide the medical image data of the affected part when seeing a doctor, which causes waste of time and resources. Physicians need to manually determine the position and orientation of the patch electrodes from medical images and precisely position and orient stimulation electrode leads within the patient (e.g., brain) that can deliver stimulation to the desired site and avoid side effects. However, scientific and rigorous data cannot be obtained in the mode so as to accurately position the slicing electrode, and the risk of errors caused by artificial judgment exists, so that economic loss is caused to a medical institution and the life safety of a patient is threatened; the accurate operation and subsequent guarantee of the implanted medical instrument determine the credit of medical institutions and medical personnel, and are social stable maintenance force. Therefore, accurate identification and positioning of the implanted click lead and the segmented electrodes based on the image identification method are necessary to obtain medical data which is convenient for a doctor to operate and has precision guarantee.

Patent CN202011444222.4 discloses a method for identifying the orientation of a slice electrode in brain in a medical image of brain: analyzing the medical image of the cranium and brain, acquiring the orientation coordinates of the slice electrodes in the brain according to the analysis result, and establishing an individual cranium and brain model according to the analysis result; calibrating the obtained position coordinates of the intracerebral segmented electrode on the individual craniocerebral model to form individualized intracerebral electrode positioning data; encoding the individualized intracerebral electrode positioning data to form an encoded data packet; and sending the coding data packet to a display device, and directly displaying the orientation of the intra-cerebral segmented electrode on a display screen of the display device by the display device according to the coding data packet. However, the patent can only identify the position coordinates of each segmented electrode, does not relate to direction identification, and is not beneficial for doctors to execute accurate electrical stimulation on each segmented electrode so as to achieve the purpose of performing accurate treatment on a specific target point.

Based on this, the present application provides an image recognition method, apparatus, system, and computer-readable storage medium to improve the prior art.

Disclosure of Invention

The application aims to identify and position a plurality of slicing electrodes and a related device, and aims to respectively identify and position and calculate the position and posture information of the slicing electrodes on an electrode lead implanted into the intracranial of a patient so as to facilitate a doctor to accurately execute electrostimulation treatment according to the position and posture information of each slicing electrode.

The purpose of the application is realized by adopting the following technical scheme:

in a first aspect, the present application provides an image recognition method for recognizing a segmented electrode implanted on an electrode lead in the intracranial of a patient, the electrode lead being provided with a plurality of segmented electrodes and image markers in the circumferential direction, the method comprising:

acquiring medical image data of the patient;

acquiring contour data of the electrode lead, contour data of the image marker and contour data of the slice electrodes based on the medical image data;

acquiring position information and posture information of the electrode lead based on the profile data of the electrode lead;

acquiring position information of the image marker based on the contour data of the image marker;

and respectively acquiring the outline data of each slice electrode from the outline data of a plurality of slice electrodes based on the position information of the image mark and the relative position relationship between the image mark and each slice electrode so as to obtain the identification result of each slice electrode.

The technical scheme has the beneficial effects that: firstly, medical image data (for example, CT data, MR data, etc.) of a patient is acquired, and recognition results of each sliced electrode in the intracranial of the patient are acquired (which is convenient for a doctor to perform accurate electrical stimulation treatment according to identification information of the sliced electrode). Specifically, the contour data of the electrode lead, the contour data of the image marker and the contour data of the plurality of segmented electrodes are acquired from the medical image data, and then the position information and the posture information of the electrode lead and the position information of the image marker are acquired, so that the contour data of each segmented electrode is acquired respectively, and the identification result of each segmented electrode is obtained. For example, the profile data of the intracranial electrode lead, the profile data of the image marker, and the profile data of the plurality of segment electrodes can be automatically acquired, and finally the individualized intracranial electrode lead profile data, the profile data of the image marker, and the profile data of the plurality of segment electrodes are programmed into a data packet format which can be directly displayed on a display device, so that a doctor can visually see the intracranial electrode lead profile, the profile of the image marker, and the profiles of the plurality of segment electrodes on the display device. Because the recognition results of all the segmented electrodes are automatically obtained, a doctor does not need to manually judge according to the images, the time for analyzing and judging the images when the doctor visits a doctor is saved, and the misjudgment risk caused by manual judgment is reduced; the judgment accuracy is improved, and the time consumed by manual judgment is saved; meanwhile, the patient does not need to carry medical images when in treatment, the treatment is convenient for the patient, a doctor can visually see the pose of the fragment electrode in the brain at any time on a display device during remote video inquiry, the remote program control is more convenient to carry out, the problem that scientific and rigorous data cannot be obtained is solved, the relative position of the fragment electrode in the cranium of the patient can be accurately positioned, the pose change of the fragment electrode is calculated, and accurate electrical stimulation treatment is realized.

In some optional embodiments, the acquiring the contour data of the electrode lead, the contour data of the image marker, and the contour data of the plurality of sliced electrodes based on the medical image data includes:

reconstructing a three-dimensional model of the patient's brain based on the medical image data;

and acquiring the contour data of the electrode lead, the contour data of the image mark and the contour data of the slice electrodes by using the three-dimensional model.

The technical scheme has the beneficial effects that: the three-dimensional model of the brain of the patient is reconstructed by using the medical image data, so that the contour data of the electrode lead, the image mark and the plurality of sliced electrodes can be obtained in the three-dimensional model, the three-dimensional model can directly indicate the spatial position relation among the electrode lead, the image mark and the plurality of sliced electrodes, the three-dimensional model, the electrode lead, the image mark and the plurality of sliced electrodes can be displayed by using a display device subsequently, the visualization effect is good, a doctor can know the spatial position of each sliced electrode directly, and therefore the sliced electrode for releasing electrical stimulation energy can be selected accurately.

In some alternative embodiments, said reconstructing a three-dimensional model of the brain of the patient based on the medical image data comprises:

analyzing a DATA file of medical image DATA in a DICOM format, extracting pixel resolution in three directions of a coronal plane, a sagittal plane and a transverse section of a brain, and combining layer number and gray level image DATA in the transverse section direction into DATA in a RAW-DATA format;

according to the DATA in the RAW-DATA format, mapping the DATA of each layer into a two-dimensional texture image along the cross section direction, and drawing the two-dimensional texture image of each layer by using a gray value as an RGBA value of the two-dimensional texture image;

and (4) mapping by using the two-dimensional texture image of the corresponding layer so as to form a three-dimensional individual craniocerebral model.

The technical scheme has the beneficial effects that: when an individual craniocerebral model is established by using a surface drawing method, the outer surfaces of the brain gray matter, the brain white matter and the electrode lead are extracted by analyzing a data file in a DICOM format, and then the extracted structure is subjected to three-dimensional rendering by using a triangular surface to obtain a three-dimensional model.

The method can be used for modeling by using a surface rendering algorithm, the surface rendering algorithm is a method for three-dimensional modeling based on the surface of a constructed object, a gray-level isosurface of a medical image is extracted by using the surface rendering algorithm, the surface of the object is reconstructed, a series of two-dimensional slice data are regarded as a three-dimensional data field by using an MC (marching cubes) algorithm, a three-dimensional model is constructed by extracting the isosurface of the three-dimensional data, and a three-dimensional model is constructed by constructing a grid of the three-dimensional model. The corresponding output data (namely the gray level isosurface information) can be judged and obtained according to different gray level values of the medical image (namely the threshold value of the medical image), the application range is wide, and the algorithm accuracy is high.

By using a contour detection device (such as a CT device, an MR device, a PET device, an X-ray device, a PET-CT device, a PET-MR device, etc.), the contour data of the electrode lead, the contour data of the image marker, and the contour data of the plurality of slice electrodes (such as CT data, MR data, PET data, X-ray data, PET-CT data, PET-MR data, etc.) can be detected in a non-contact manner, so that the problem that the contour data of the electrode lead implanted in the cranium of the patient, the contour data of the image marker, and the contour data of the plurality of slice electrodes cannot be obtained by using a contact measurement method can be solved. The contour detection model can be obtained by training a large amount of training data, corresponding output data (namely contour data of an electrode lead, contour data of the image marker and contour data of the slice electrodes) can be obtained by predicting according to different input data (namely medical image data of a medical image), the application range is wide, and the intelligent level is high.

In some optional embodiments, the method further comprises:

acquiring target attitude information of the three-dimensional model;

generating a target two-dimensional image corresponding to the target posture information based on the medical image data and the target posture information;

and displaying the target two-dimensional image by using a display device and displaying the identification information of at least one slicing electrode on the target two-dimensional image.

The technical scheme has the beneficial effects that: the target two-dimensional image is displayed in a non-contact display mode by using a display device (such as a mobile phone, a tablet computer, a notebook computer, a desktop computer, a projector, a display and the like), and the identification information (such as one or more of Chinese, letters, numbers, symbols, shapes and colors to indicate the position information, the direction information, the quantity information and the like of the slicing electrode on the electrode lead or in the skull of the patient) of at least one slicing electrode is displayed on the target two-dimensional image.

In some optional embodiments, the obtaining target pose information of the three-dimensional model includes:

receiving a rotation operation aiming at the three-dimensional model by utilizing an interactive device, and determining target posture information of the three-dimensional model in response to the rotation operation.

The technical scheme has the beneficial effects that: the method comprises the steps of changing the angle and the direction of the three-dimensional model by utilizing interactive equipment (such as a mouse, a touch pad, a touch pen and the like) in a contact type dragging, rotating and zooming mode to obtain target posture information of different positions of the three-dimensional model, improving the efficiency of man-machine interaction by obtaining the target posture information in an all-around mode of 720 degrees, carrying out screenshot or projection operation by adopting a contact type dragging, rotating and clicking mode to generate a target two-dimensional image corresponding to the target posture information, and being more comprehensive in information obtaining, high in display flexibility and time cost saving.

generating a plurality of cross-sectional two-dimensional images of the brain of the patient based on the medical image data with the electrode lead, the image marker, and the plurality of slice electrodes as target objects, respectively;

detecting the contour of the target object from each two-dimensional image to obtain contour data of the target object in each two-dimensional image;

and acquiring the contour data of the target object based on the contour data of the target object in a plurality of two-dimensional images.

The technical scheme has the beneficial effects that: the method has the advantages that the two-dimensional images of the brain at the multiple cross sections are obtained through the imaging identification technology, the contour data in the two-dimensional images of the brain at the multiple cross sections are obtained through the contour detection technology, compared with the traditional method that the target image is identified and the data are obtained manually, the intelligent degree is high; for example, the trained imaging recognition model and contour detection model can be applied to the actual scene to acquire the contour data of the target object, and the recognition accuracy is high. The electrode lead, the image mark and the plurality of sliced electrodes are respectively used as target objects, namely, the electrode lead is respectively used as a target object, the image mark is used as a target object, and the plurality of sliced electrodes are used as target objects.

In some optional embodiments, the detecting a contour of the target object from each of the two-dimensional images to obtain contour data of the target object in each of the two-dimensional images includes:

detecting the contour of the target object from each two-dimensional image by using a contour detection model to obtain contour data of the target object in each two-dimensional image;

the training process of the contour detection model comprises the following steps:

acquiring a training set, wherein the training set comprises a plurality of training data, and each training data comprises a sample image and labeling data of contour data of a target object corresponding to the sample image;

for each training data in the training set, performing the following:

inputting a sample image in the training data into a preset deep learning model to obtain prediction data of the contour data of the target object corresponding to the sample image;

updating model parameters of the deep learning model based on the prediction data and the labeling data of the contour data of the target object corresponding to the sample image;

detecting whether a preset training end condition is met or not; if yes, taking the trained deep learning model as the contour detection model; if not, continuing to train the contour detection model by using the next training data.

The technical scheme has the beneficial effects that: the contour detection model can be obtained by training a large amount of training data, corresponding output data (namely contour data of electrode leads, contour data of the image markers and contour data of the plurality of segmented electrodes) can be obtained by predicting aiming at different input data (namely two-dimensional images), the trained contour test model has strong robustness and low overfitting risk, the calculation process is simple, the calculation speed is high, the calculation efficiency is high, and the consumed calculation resources are few. In practical application, the deep learning model is used for detecting the contours, and the purpose of detecting all contours can be achieved through configuration parameters, specifically, each contour can be classified into grades, such as: the outermost periphery, the first inner periphery, the second inner periphery and the like are respectively provided with a parent contour and an embedded contour index number of the current contour. All contours can be detected, the hierarchical relationship of the outer layer and the inner layer is established among the contours, and the information of all points on the contours is saved. The detection result can also be projected onto the corresponding position of the original image based on the contour deviation parameter.

In some alternative embodiments, the model parameters include at least one of: contour pixel coordinate vector, contour color, contour line width, line type, topology parameters, contour maximum level, and contour offset parameters.

The technical scheme has the beneficial effects that: various model parameters of the contour detection model are optimized to obtain a better model effect.

In some alternative embodiments, the medical image data includes one or more of CT data, MR data, PET data, X-ray data, PET-CT data, and PET-MR data;

the image marker comprises a plurality of feature angles; alternatively, the image mark includes a plurality of mark portions, each of which is provided on a different segment electrode.

The technical scheme has the beneficial effects that: because the image markers (such as circles, rectangles, triangles and the like) are directly arranged on the electrode leads or the sliced electrodes, the imaging identification can play a role in determining the orientation of the sliced electrodes, the pose information of the sliced electrodes can be further obtained by calculating the projection angle and the projection area of the image markers, and the image markers can be used for selecting the sliced electrodes for executing the stimulation task. When each marking part of the image mark is arranged on the slicing electrode, the area outside the slicing electrode does not need to be additionally provided with the marking part, so that the manufacturing cost of the stimulation electrode lead can be reduced, and the manufacturing difficulty of the stimulation electrode lead can be reduced.

In some optional embodiments, the obtaining, from the profile data of a plurality of sliced electrodes based on the position information of the image marker and the relative position relationship between the image marker and each sliced electrode, the profile data of each sliced electrode respectively includes:

acquiring the position information of each sliced electrode based on the position information of the image mark and the relative position relationship between the image mark and each sliced electrode;

and acquiring the contour data of each sliced electrode from the contour data of a plurality of sliced electrodes respectively based on the position information of each sliced electrode.

The technical scheme has the beneficial effects that: and identifying and positioning each sliced electrode by using the position information of the image mark and the relative position relationship between the image mark and each sliced electrode, and calculating the position information of each sliced electrode.

In some optional embodiments, the method further comprises:

and acquiring the posture information of each sliced electrode based on the contour data of each sliced electrode.

The technical scheme has the beneficial effects that: and identifying and positioning each sliced electrode by using the position information of the image mark and the relative position relationship between the image mark and each sliced electrode, and calculating the posture information of each sliced electrode.

I give here ] for example, it may be configured that a plurality of segment electrodes are arranged in a matrix in the circumferential direction of the electrode lead, a tangent plane of a center point of each of the segment electrodes is parallel to the axial direction of the electrode lead, an axial straight line passing through the center point of the electrode lead is located based on the position information of the electrode lead to obtain the location data of the axial straight line, and for each of the two-dimensional images (i.e., two-dimensional grayscale images), the area of each of the image markers in the two-dimensional image is calculated based on the profile data of each of the image markers in the two-dimensional image, and the position information of each of the segment electrodes is obtained. The projection planes corresponding to the plurality of two-dimensional images are all perpendicular to the cross section of the brain of the patient, and any two of the projection planes corresponding to the plurality of two-dimensional images are not parallel to each other. And acquiring the contour data of the electrode lead and the contour data of each image mark by using a contour detection model and a plurality of two-dimensional images. And carrying out binarization processing on the plurality of two-dimensional images, and carrying out contour detection on each two-dimensional image after binarization processing by using a contour detection model so as to obtain contour data of the electrode lead and contour data of each image mark. Based on the relative position information of the electrode lead in the patient's cranium, the relative position information of each piece of electrode in the patient's cranium is obtained, and then the posture information (such as three-dimensional coordinates, rotation angle, rotation direction, etc.) of the piece of electrode is obtained. By utilizing the position and area information of the image mark, the posture information of the slicing electrode is indirectly calculated, the intelligent degree is high, and the result accuracy is high.

The position information and pose information of each segmented electrode can be programmed into a data packet format that can be directly displayed on a display device, so that a doctor can visually identify each segmented electrode on an intracranial electrode lead on the display device and perform direct electrode stimulation treatment by observing the pose of each segmented electrode.

In a second aspect, the present application provides an image recognition apparatus for recognizing a segmented electrode implanted on an electrode lead in the intracranial of a patient, the electrode lead being provided with a plurality of segmented electrodes and image markers in the circumferential direction;

the apparatus comprises a processor configured to:

acquiring medical image data of the patient;

acquiring position information of the image marker based on the outline data of the image marker;

In some optional embodiments, the processor is further configured to acquire the contour data of the electrode lead, the contour data of the image marker, and the contour data of the plurality of sliced electrodes in the following manner:

In some optional embodiments, the processor is further configured to:

acquiring target attitude information of the three-dimensional model;

In some optional embodiments, the processor is further configured to obtain target pose information for the three-dimensional model by:

In some optional embodiments, the processor is further configured to acquire contour data of the target object in each of the two-dimensional images by:

for each training data in the training set, performing the following:

detecting whether a preset training end condition is met; if yes, taking the trained deep learning model as the contour detection model; if not, continuing to train the contour detection model by using the next training data.

the image marker comprises a plurality of characteristic angles; alternatively, the image mark includes a plurality of mark portions, each of which is provided on a different segment electrode.

In some optional embodiments, the processor is further configured to acquire contour data for each of the sliced electrodes by:

In some optional embodiments, the processor is further configured to:

In a third aspect, the present application provides an image recognition system, the system comprising:

any one of the above image recognition apparatuses;

a display device for providing a display function;

and the interaction equipment is used for providing an interaction function.

In a fourth aspect, the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of any of the methods described above or implements the functions of any of the apparatus described above.

Drawings

The present application is further described below with reference to the accompanying drawings and embodiments.

Fig. 1 shows a flowchart of an image recognition method provided in the present application.

Fig. 2 shows a schematic flowchart of a process for acquiring profile data provided in the present application.

Fig. 3 is a flowchart illustrating another image recognition method provided in the present application.

Fig. 4 shows a schematic flow chart of another method for acquiring profile data provided by the present application.

Fig. 5 shows a schematic flow chart for acquiring contour data of each sliced electrode provided in the present application.

Fig. 6 shows a block diagram of an image recognition apparatus provided in the present application.

Fig. 7 shows a schematic structural diagram of an image recognition system provided by the present application.

Fig. 8 shows a schematic structural diagram of a program product provided in the present application.

Detailed Description

The technical solutions in the present application will be described below with reference to the drawings and the detailed description of the present application, and it should be noted that, in the case of no conflict, any combination between the embodiments or technical features described below may form a new embodiment.

In the present application, "at least one" means one or more, "a plurality" means two or more. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone, wherein A and B can be singular or plural. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. "at least one of the following" or similar expressions refer to any combination of these items, including any combination of the singular or plural items. For example, at least one (one) of a, b, or c, may represent: a, b, c, a and b, a and c, b and c, a and b and c, wherein a, b and c can be single or multiple. It is to be noted that "at least one item" may also be interpreted as "one or more item(s)".

It is also noted that the terms "exemplary" or "such as" and the like are used herein to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "such as" is not necessarily to be construed as preferred or advantageous over other embodiments or designs. Rather, use of the word "exemplary" or "such as" is intended to present concepts related in a concrete fashion.

In the following, a brief description of one of the application areas of the present application, i.e. an implantable device, is first given.

An implantable neurostimulation system (an implantable medical system) generally includes a stimulator implanted in a patient and a programming device disposed outside the patient. The existing nerve regulation and control technology is mainly characterized in that an electrode is implanted in a specific structure (namely a target spot) in a body through a three-dimensional operation, and a stimulator implanted in the body of a patient sends electric pulses to the target spot through the electrode to regulate and control the electric activity and the function of a corresponding nerve structure and network, so that symptoms are improved, and pain is relieved. The stimulator may be any one of an Implantable nerve electrical stimulation device, an Implantable cardiac electrical stimulation System (also called a cardiac pacemaker), an Implantable Drug Delivery System (I DDS for short), and a lead switching device. Examples of the implantable neural electrical Stimulation device include Deep Brain Stimulation (DBS), Cortical Brain Stimulation (CNS), Spinal Cord Stimulation (SCS), Sacral Nerve Stimulation (SNS), and Vagal Nerve Stimulation (VNS).

The stimulator may include an IPG (implantable pulse generator) disposed in the patient's body, an extension lead and an electrode lead, and supplies controllable electrical stimulation energy to the body tissue by means of a sealed battery and circuit, and delivers one or two controllable specific electrical stimulations to specific regions of the body tissue through the implanted extension lead and electrode lead. The extension lead is used in cooperation with the IPG and is used as a transmission medium of the electrical stimulation signal to transmit the electrical stimulation signal generated by the IPG to the electrode lead. The electrode leads deliver electrical stimulation to specific areas of tissue within the body through a plurality of electrode contacts. The stimulator is provided with one or more paths of electrode leads on one side or two sides, a plurality of electrode contacts are arranged on the electrode leads, and the electrode contacts can be uniformly arranged or non-uniformly arranged on the circumference of the electrode leads. As an example, the electrode contacts may be arranged in an array of 4 rows and 3 columns (12 electrode contacts in total) in the circumferential direction of the electrode lead. The electrode contacts may include stimulation electrode contacts and/or collection electrode contacts. The electrode contact may have a sheet shape, an annular shape, a dot shape, or the like.

In some possible implementations, the stimulated in vivo tissue may be brain tissue of the patient, and the stimulated site may be a specific site of the brain tissue. The sites stimulated are generally different when the patient's disease type is different, as are the number of stimulation contacts (single or multiple) used, the application of one or more (single or multiple) specific electrical stimulation signals, and stimulation parameter data. The type of disease to which the present application is applicable is not limited, and may be the type of disease to which Deep Brain Stimulation (DBS), Spinal Cord Stimulation (SCS), pelvic stimulation, gastric stimulation, peripheral nerve stimulation, functional electrical stimulation are applicable. Among the types of diseases that DBS may be used for treatment or management include, but are not limited to: convulsive disorders (e.g., epilepsy), pain, migraine, psychiatric disorders (e.g., Major Depressive Disorder (MDD)), manic depression, anxiety, post-traumatic stress disorder, depression, Obsessive Compulsive Disorder (OCD), behavioral disorders, mood disorders, memory disorders, mental state disorders, movement disorders (e.g., essential tremor or parkinson's disease), huntington's disease, alzheimer's disease, drug addiction, autism, or other neurological or psychiatric diseases and injuries. When the DBS is used for treating drug addiction patients, the DBS can help drug addicts to abstain drugs and improve the happiness and the life quality of the drug addicts.

In the application, when the program control device is connected with the stimulator in a program control manner, the program control device can be used for adjusting stimulation parameters of the stimulator (different electrical stimulation signals corresponding to different stimulation parameters are different), the stimulator can sense bioelectricity activity of deep brain of a patient to acquire electrophysiological signals, and the stimulation parameters of the electrical stimulation signals of the stimulator can be continuously adjusted through the acquired electrophysiological signals.

The programming device may be a physician programmer (i.e., a programming device used by a physician) or a patient programmer (i.e., a programming device used by a patient). The program control device may be, for example, a tablet computer, a notebook computer, a desktop computer, a mobile phone, or other intelligent terminal devices.

The data interaction of this application to doctor's program controller and stimulator does not restrict each other, when doctor long-range programme-controlled (doctor is in the hospital this moment, and the patient is at home), doctor's program controller can carry out data interaction through server, patient program controller and stimulator. When the doctor is off-line and the patient is in face-to-face program control, the doctor program controller can perform data interaction with the stimulator through the patient program controller, and the doctor program controller can also perform data interaction with the stimulator directly.

The patient programmer may include a master (in communication with the server) and a slave (in communication with the stimulator), with the master and slave being communicatively coupled. The doctor program controller can perform data interaction with the server through a 3G/4G/5G network, the server can perform data interaction with the host through the 3G/4G/5G network, the host can perform data interaction with the submachine through a Bluetooth protocol/WIFI protocol/USB protocol, the submachine can perform data interaction with the stimulator through a 401MHz-406MHz working frequency band/2.4 GHz-2.48GHz working frequency band, and the doctor program controller can perform data interaction with the stimulator directly through the 401MHz-406MHz working frequency band/2.4 GHz-2.48GHz working frequency band.

Besides the application field of the implanted device, the implantable medical device can also be applied to the technical field of other medical devices and even non-medical devices, and the implantable medical device is not limited by the application and can be applied to occasions related to image recognition.

Method embodiment

Referring to fig. 1, fig. 1 shows a schematic flow chart of an image recognition method provided in the present application.

The application provides an image identification method, which is used for identifying a slice electrode implanted on an electrode lead in the intracranial of a patient, wherein a plurality of slice electrodes and image marks are arranged on the circumferential direction of the electrode lead, and the method comprises the following steps:

step S101: acquiring medical image data of the patient;

step S102: acquiring contour data of the electrode lead, contour data of the image marker and contour data of the slice electrodes based on the medical image data;

step S103: acquiring position information and posture information of the electrode lead based on the profile data of the electrode lead;

step S104: acquiring position information of the image marker based on the outline data of the image marker;

step S105: and respectively acquiring the outline data of each slice electrode from the outline data of a plurality of slice electrodes based on the position information of the image mark and the relative position relationship between the image mark and each slice electrode so as to obtain the identification result of each slice electrode.

Therefore, medical image data (for example, CT data, MR data, etc.) of the patient is first acquired, and the recognition results of the slicing electrodes in the intracranial of the patient are acquired (which is convenient for a doctor to perform accurate electrical stimulation treatment according to the identification information of the slicing electrodes).

Specifically, the contour data of the electrode lead, the contour data of the image marker and the contour data of the plurality of segmented electrodes are acquired from the medical image data, and then the position information and the posture information of the electrode lead and the position information of the image marker are acquired, so that the contour data of each segmented electrode is acquired respectively, and the identification result of each segmented electrode is obtained.

For example, the profile data of the intracranial electrode lead, the profile data of the image marker, and the profile data of the plurality of segment electrodes can be automatically acquired, and finally the individualized intracranial electrode lead profile data, the profile data of the image marker, and the profile data of the plurality of segment electrodes are programmed into a data packet format which can be directly displayed on a display device, so that a doctor can visually see the intracranial electrode lead profile, the profile of the image marker, and the profiles of the plurality of segment electrodes on the display device.

Because the recognition results of all the segmented electrodes are automatically obtained, the doctor does not need to manually judge the images, the time for analyzing the images and judging the images when the doctor visits the doctor is saved, and the misjudgment risk caused by manual judgment is reduced; the judgment accuracy is improved, and the time consumed by manual judgment is saved; meanwhile, the patient does not need to carry medical images when in treatment, the treatment is convenient for the patient, a doctor can visually see the pose of the fragment electrode in the brain at any time on a display device during remote video inquiry, the remote program control is more convenient to carry out, the problem that scientific and rigorous data cannot be obtained is solved, the relative position of the fragment electrode in the cranium of the patient can be accurately positioned, the pose change of the fragment electrode is calculated, and accurate electrical stimulation treatment is realized.

The number of electrode leads in the patient's cranium is not limited in this application and can be, for example, 1, 2, 3, 4, 6, 8, etc.

The implantation position of the electrode lead is not limited in the present application, and may be, for example, distributed entirely in the left and right brains of the patient, or may be distributed in the left and right brains of the patient, respectively.

The number of the segmented electrodes on the electrode lead is not limited in the present application, and may be, for example, 4, 8, 12, 24, etc.

The arrangement of the segmented electrodes on the electrode lead is not limited in the present application, and may be arranged in a uniform arrangement manner, such as a matrix (multiple rows and multiple columns), a ring, a diamond, or a non-uniform arrangement manner.

Medical image information in the present application may include, for example, CT data, MR data, PET data, X-ray data, PET-CT data, PET-MR data, and the like. Accordingly, the data acquisition device used can be, for example, a CT device, an MR device, a PET device, an X-ray device, a PET-CT device, a PET-MR device, etc. Among them, ct (computed tomography) is computed tomography, mr (magnetic resonance) is magnetic resonance, and pet (positron Emission tomography) is positron Emission tomography.

The contour data is not limited in the present application, and may be, for example, point cloud data of contour points, positioning data of contour lines, positioning data of contour surfaces (for example, each curved surface may be regarded as a combination of a plurality of triangular surface patches), and the like.

In the present application, the position information of the electrode lead, the image mark, and the segmented electrode may be represented by three-dimensional coordinate data of a central point, for example, wherein the position information of the electrode lead may also be represented by a linear equation. The attitude information of the electrode lead, the image mark and the sliced electrode can be represented by three attitude angles, namely a pitch angle, an inclination angle and a roll angle.

In the application, the electrode lead is fixedly provided with the slice electrodes and the image marks, so that the relative position relationship between each slice electrode and the image marks is fixed and can be obtained in advance.

Based on the relative position relationship between the image mark and each slice electrode, the contour data of the slice electrodes can be divided to obtain the contour data of each slice electrode.

Referring to fig. 2, fig. 2 is a schematic diagram illustrating a flow chart for acquiring profile data according to the present application.

In some optional embodiments, the step S102 may include:

step S201: reconstructing a three-dimensional model of the patient's brain based on the medical image data;

step S202: and acquiring the contour data of the electrode lead, the contour data of the image mark and the contour data of the slice electrodes by using the three-dimensional model.

Therefore, the three-dimensional model of the brain of the patient is reconstructed by using the medical image data, so that the contour data of the electrode lead, the image mark and the plurality of sliced electrodes can be obtained in the three-dimensional model, the three-dimensional model can directly indicate the spatial position relation among the electrode lead, the image mark and the plurality of sliced electrodes, the three-dimensional model, the electrode lead, the image mark and the plurality of sliced electrodes can be displayed by using the display equipment subsequently, the visualization effect is good, a doctor can intuitively know the spatial position of each sliced electrode conveniently, and the sliced electrode for releasing the electrical stimulation energy can be accurately selected.

In some optional embodiments, the step S201 may include:

according to the DATA in the RAW-DATA format, mapping the DATA of each layer into a two-dimensional texture image along the transverse direction, and drawing the two-dimensional texture image of each layer by using a gray value as an RGBA value of the two-dimensional texture image;

mapping is performed using the two-dimensional texture images of the corresponding layers, thereby forming a three-dimensional individual craniocerebral model (i.e., a three-dimensional model).

Therefore, when the individual craniocerebral model is established by using the surface rendering method, the data file in the DICOM format is analyzed to extract the outer surfaces of the grey brain matter, the white brain matter and the electrode lead, and also can extract structures such as epidermis, bones and the like, and then the extracted surface is subjected to three-dimensional rendering by using a triangular surface to obtain the three-dimensional model.

The method can be used for modeling by using a surface rendering algorithm, the surface rendering algorithm is a method for three-dimensional modeling based on the surface of a constructed object, a gray-scale isosurface of a medical image is extracted by using the surface rendering algorithm, the surface of the object is reconstructed, a series of two-dimensional slice data are regarded as a three-dimensional data field by using an MC (marching cubes) algorithm, and a surface grid of a three-dimensional model is constructed by extracting the isosurface of the three-dimensional data, so that the three-dimensional model is constructed.

The corresponding output data (namely the gray level isosurface information) can be judged and obtained according to different gray level values of the medical image (namely the threshold value of the medical image), the application range is wide, and the algorithm accuracy is high.

By using a contour detection device (such as a CT device, an MR device, a PET device, an X-ray device, a PET-CT device, a PET-MR device, etc.), the contour data of the electrode lead, the contour data of the image marker, and the contour data of the plurality of slice electrodes (such as CT data, MR data, PET data, X-ray data, PET-CT data, PET-MR data, etc.) can be detected in a non-contact manner, so that the problem that the contour data of the electrode lead implanted in the cranium of the patient, the contour data of the image marker, and the contour data of the plurality of slice electrodes cannot be obtained by using a contact measurement method can be solved.

The contour detection model can be obtained by training a large amount of training data, corresponding output data (namely contour data of an electrode lead, contour data of the image marker and contour data of the slice electrodes) can be obtained by predicting according to different input data (namely medical image data of a medical image), the application range is wide, and the intelligent level is high.

The method can also be used for modeling by utilizing a volume rendering algorithm, the volume rendering can provide the rendering effect which is closest to the vision of human eyes, and a Ray casting algorithm (Ray-casting) is utilized, so that each pixel of the medical image emits a Ray along the sight line direction, the Ray penetrates through a volume data set, the sampling is carried out according to a certain step length, the gray value and the opacity of each sampling point are calculated, and then the accumulated gray value and the opacity value are calculated point by point from the front to the back or from the back to the front. The volume rendering algorithm is favorable for keeping the details of the medical image, is suitable for rendering three-dimensional models with fuzzy regional characteristics and high correlation with voxel characteristics, and directly converts discrete data of a three-dimensional space into a three-dimensional model and considers the transmission, emission and reflection effects of each voxel on light. Therefore, the volume rendering can better show the spatial volume details of the three-dimensional model.

In other alternative embodiments, the step S102 may include:

and carrying out three-dimensional contour detection on the medical image data to obtain contour data of the electrode lead, contour data of the image marker and contour data of the slice electrodes.

In some optional embodiments, the present application may utilize a preset deep learning network to train to obtain a three-dimensional contour segmentation model, and perform three-dimensional contour detection and segmentation on medical image data by using the three-dimensional contour segmentation model to obtain the contour data. The training process of the three-dimensional contour segmentation model is not limited, and may be, for example, a training mode of supervised learning, or a training mode of semi-supervised learning, or a training mode of unsupervised learning.

In other alternative embodiments, the present application may adopt other three-dimensional contour detection methods, such as the contour detection method in the patent CN106898012A-CT image chest contour automatic detection method.

Referring to fig. 3, fig. 3 is a schematic flow chart of another image recognition method provided in the present application.

In some optional embodiments, the method may further comprise:

step S106: acquiring target posture information of the three-dimensional model;

step S107: generating a target two-dimensional image corresponding to the target posture information based on the medical image data and the target posture information;

step S108: and displaying the target two-dimensional image by using a display device and displaying the identification information of at least one slicing electrode on the target two-dimensional image.

Therefore, by using a display device (such as a mobile phone, a tablet computer, a notebook computer, a desktop computer, a projector, a display and the like), displaying the target two-dimensional image in a non-contact display mode and displaying identification information (such as one or more of Chinese, letters, numbers, symbols, shapes and colors to indicate position information, direction information, quantity information and the like of the slicing electrodes on the electrode lead or in the cranium of the patient) of at least one slicing electrode on the target two-dimensional image, the identification information of the slicing electrodes can be intuitively acquired in a mode of displaying images and marking images, the display efficiency is high, and the intelligentization level is high.

In the present application, the target attitude information may be represented by three attitude angles.

The number of the slice electrodes for displaying the identification information is not limited, and for example, the identification information of the slice electrodes with the exposed parts larger than the preset ratio can be displayed. Generally, when 3, 4 or even more segment electrodes are arranged in the same circumferential row, only the identification information of one or two segment electrodes is displayed (because most of the rest segment electrodes are not exposed).

In some optional embodiments, the step S106 may include:

Therefore, by utilizing interactive equipment (such as a mouse, a touch pad, a touch pen and the like), the angle and the direction of the three-dimensional model are changed in a contact type dragging, rotating and zooming mode to obtain target posture information of different positions of the three-dimensional model, the man-machine interaction efficiency is improved in a mode of obtaining the target posture information in 720 degrees in all directions, screenshot or projection operation is carried out in a contact type dragging, rotating and clicking mode to generate a target two-dimensional image corresponding to the target posture information, the obtained information is more comprehensive, the display activity is high, and the time cost is saved.

The interactive device is not limited in the present application, and may be, for example, a mobile phone, a tablet computer, a notebook computer, a desktop computer, an intelligent wearable device, or an intelligent terminal device with a mouse, a touch pad, and a touch pen, or the interactive device may be a workstation or a console.

The manner in which the various (manual) operations are received by the interactive device is not limited in this application. The operations are divided according to input modes, and may include, for example, a text input operation, an audio input operation, a video input operation, a key operation, a mouse operation, a keyboard operation, an intelligent stylus operation, and the like.

Referring to fig. 4, fig. 4 is a schematic flow chart illustrating another method for acquiring profile data provided by the present application.

In some optional embodiments, the step S102 may include:

step S301: generating two-dimensional images of a plurality of cross sections of the brain of the patient based on the medical image data with the electrode lead, the image marker and the plurality of slice electrodes as target objects, respectively;

step S302: detecting the contour of the target object from each two-dimensional image to obtain contour data of the target object in each two-dimensional image;

step S303: and acquiring the contour data of the target object based on the contour data of the target object in a plurality of two-dimensional images.

Therefore, two-dimensional images of a plurality of cross sections (also called horizontal planes and medical anatomy concepts) of the brain are obtained through an imaging identification technology, and contour data in the two-dimensional images of the cross sections of the brain are obtained through a contour detection technology, so that compared with the traditional method of manually identifying and obtaining target images, the intelligent degree is high; for example, the trained imaging recognition model and contour detection model can be applied to the actual scene to acquire the contour data of the target object, and the recognition accuracy is high.

The electrode lead, the image mark and the plurality of sliced electrodes are respectively used as target objects, namely, the electrode lead is respectively used as a target object, the image mark is used as a target object, and the plurality of sliced electrodes are used as target objects.

In some optional embodiments, the step S302 may include:

for each training data in the training set, performing the following:

Therefore, the contour detection model can be obtained by training a large amount of training data, corresponding output data (namely contour data of the electrode lead, contour data of the image marker and contour data of the plurality of segmented electrodes) can be obtained by predicting aiming at different input data (namely two-dimensional images), the trained contour test model has strong robustness and low overfitting risk, the calculation process is simple, the calculation speed is high, the calculation efficiency is high, and the consumed calculation resources are few.

Through design, a proper amount of neuron calculation nodes and a multilayer operation hierarchical structure are established, a proper input layer and a proper output layer are selected, a preset deep learning model can be obtained, through learning and tuning of the preset deep learning model, a function relation from input to output is established, although the function relation between the input and the output cannot be found 100%, the function relation can be close to a real association relation as much as possible, the contour detection model obtained through training can be used for respectively obtaining contour data of a target object based on each two-dimensional image, and the accuracy and the reliability of calculation results are high.

In practical application, the deep learning model is used for detecting the contours, and the purpose of detecting all contours can be achieved through configuration parameters, specifically, each contour can be classified into grades, such as: the outermost periphery, the first inner periphery, the second inner periphery and the like are respectively provided with a parent contour and an embedded contour index number of the current contour.

All contours can be detected, the hierarchical relationship between the outer layer and the inner layer is established between the contours, and the information of all points on the contours is saved.

The detection result can also be projected to the corresponding position of the original image based on the contour deviation parameter.

In some alternative embodiments, the present application may use the above-mentioned training process to train to obtain the contour detection model, and in other alternative embodiments, the present application may use a pre-trained contour detection model.

The method for acquiring the annotation data is not limited in the present application, and for example, a manual annotation method, an automatic annotation method, or a semi-automatic annotation method may be adopted.

The training process of the contour detection model is not limited in the present application, and may be, for example, the above-mentioned supervised learning training mode, or may be a semi-supervised learning training mode, or may be an unsupervised learning training mode.

The preset training end condition is not limited in the present application, and may be, for example, that the training frequency reaches the preset frequency (the preset frequency is, for example, 1 time, 3 times, 10 times, 100 times, 1000 times, 10000 times, etc.), or may be that training data in a training set all complete one or more times of training, or may be that a total loss value obtained by this training is not greater than a preset loss value.

Therefore, various model parameters of the contour detection model are optimized to obtain a better model effect.

In some alternative embodiments, the medical image data includes one or more of CT data, MR data, PET data, X-ray data, PET-CT data, and PET-MR data.

In some alternative embodiments, the image marker comprises a plurality of characteristic angles; alternatively, the image mark includes a plurality of mark portions, each of which is provided on a different segment electrode.

Therefore, since the image markers (such as circles, rectangles, triangles and the like) are directly arranged on the electrode leads or the slice electrodes, the imaging identification can be used for determining the orientation of the slice electrodes, the pose information of the slice electrodes can be further obtained by calculating the projection angles (projection angles corresponding to characteristic angles) and the projection areas of the image markers, and the image markers can be used for selecting the slice electrodes for executing the stimulation task. When each marking part of the image mark is arranged on the split electrode, the area outside the split electrode does not need to be additionally provided with the marking part, so that the manufacturing cost of the electrode lead can be reduced, and the manufacturing difficulty of the electrode lead can be reduced.

In other alternative embodiments, the image mark includes a characteristic angle, as long as each sliced electrode can be identified, and the number of characteristic angles in the image mark is not limited in the present application.

In addition to the characteristic corners, the image markers may be provided with rounded corners, chamfers, arcs, broken lines, etc., wherein the arcs may be wavy, and the broken lines may be jagged, etc.

Referring to fig. 5, fig. 5 shows a schematic flowchart for acquiring the outline data of each sliced electrode provided in the present application.

In some optional embodiments, the step S105 may include:

step S401: acquiring the position information of each sliced electrode based on the position information of the image mark and the relative position relationship between the image mark and each sliced electrode;

step S402: and acquiring the outline data of each sliced electrode from the outline data of a plurality of sliced electrodes respectively based on the position information of each sliced electrode.

Therefore, each slice electrode can be identified and positioned by utilizing the position information of the image mark and the relative position relation between the image mark and each slice electrode, and the position information of each slice electrode is calculated.

In some optional embodiments, the method may further comprise:

Therefore, each slice electrode can be identified and positioned by utilizing the position information of the image mark and the relative position relation between the image mark and each slice electrode, and the posture information of each slice electrode is calculated.

For example, one or more image markers may be arranged regularly or irregularly in the circumferential direction of the electrode lead (i.e., one or more image markers may be disposed on the electrode lead), a plurality of segment electrodes may be arranged in a matrix in the circumferential direction of the electrode lead, a tangent plane of a center point of each segment electrode is parallel to an axial direction of the electrode lead, an axial straight line passing through the center point of the electrode lead is located based on the position information of the electrode lead to obtain location data of the axial straight line, and for each two-dimensional image (i.e., a two-dimensional grayscale image), an area of each image marker in the two-dimensional image is calculated based on profile data of each image marker in the two-dimensional image, so as to obtain position information of each segment electrode.

The projection planes corresponding to the plurality of two-dimensional images are all perpendicular to the cross section of the brain of the patient, and any two of the projection planes corresponding to the plurality of two-dimensional images are not parallel to each other.

And acquiring the contour data of the electrode lead and the contour data of each image mark by using a contour detection model and a plurality of two-dimensional images.

And carrying out binarization processing on the two-dimensional images, and carrying out contour detection on each two-dimensional image after binarization processing by using a contour detection model so as to obtain contour data of the electrode lead and contour data of each image mark.

Based on the relative position information of the electrode lead in the patient's cranium, the relative position information of each piece of electrode in the patient's cranium is obtained, and then the posture information (such as three-dimensional coordinates, rotation angle, rotation direction, etc.) of the piece of electrode is obtained.

By utilizing the position and area information of the image mark, the posture information of the slicing electrode is indirectly calculated, the intelligent degree is high, and the result accuracy is high.

Device embodiments

The application also provides an image recognition device, which is used for recognizing the segmented electrodes implanted on the electrode lead in the intracranial of a patient, wherein a plurality of the segmented electrodes and image marks are arranged on the circumferential direction of the electrode lead;

the apparatus comprises a processor configured to:

acquiring medical image data of the patient;

In some optional embodiments, the processor may be further configured to acquire the contour data of the electrode lead, the contour data of the image marker, and the contour data of the plurality of sliced electrodes in the following manner:

In some optional embodiments, the processor may be further configured to:

acquiring target attitude information of the three-dimensional model;

In some optional embodiments, the processor may be further configured to obtain target pose information for the three-dimensional model by:

In some alternative embodiments, the processor may be further configured to obtain the contour data of the electrode lead, the contour data of the image marker, and the contour data of the plurality of sliced electrodes in the following manner:

In some optional embodiments, the processor may be further configured to acquire the contour data of the target object in each of the two-dimensional images in the following manner:

for each training data in the training set, performing the following:

updating model parameters of the deep learning model based on the prediction data and the annotation data of the contour data of the target object corresponding to the sample image;

In some alternative embodiments, the model parameters may include at least one of: contour pixel coordinate vector, contour color, contour line width, line type, topology parameters, contour maximum level, and contour offset parameters.

In some alternative embodiments, the medical image data may include one or more of CT data, MR data, PET data, X-ray data, PET-CT data, and PET-MR data;

the image marker may comprise a plurality of feature angles; alternatively, the image marker may include a plurality of marker portions each of which is provided on a different patch electrode.

In some optional embodiments, the processor may be further configured to acquire the contour data of each of the sliced electrodes by:

In some optional embodiments, the processor may be further configured to:

Referring to fig. 6, fig. 6 is a block diagram illustrating a structure of an image recognition apparatus 200 provided in the present application.

The image recognition apparatus 200 may include, for example, at least one memory 210, at least one processor 220, and a bus 230 connecting different platform systems.

The memory 210 may include readable media in the form of volatile memory, such as Random Access Memory (RAM)211 and/or cache memory 212, and may further include Read Only Memory (ROM) 213.

The memory 210 further stores a computer program, and the computer program can be executed by the processor 220, so that the processor 220 implements the steps of any one of the methods, and the specific implementation manner of the method is consistent with the implementation manner and the achieved technical effect described in the implementation manner of the method, and some contents are not described again.

Memory 210 may also include a utility 214 having at least one program module 215, such program modules 215 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Accordingly, the processor 220 may execute the computer programs described above, and may execute the utility 214.

The processor 220 may employ one or more Application Specific Integrated Circuits (ASICs), DSPs, Programmable Logic Devices (PLDs), Complex Programmable Logic Devices (CPLDs), Field-Programmable Gate arrays (FPGAs), or other electronic components.

Bus 230 may be one or more of any of several types of bus structures including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a local bus using any of a variety of bus architectures.

The image recognition apparatus 200 may also communicate with one or more external devices 240, such as a keyboard, pointing device, bluetooth device, etc., and may also communicate with one or more devices capable of interacting with the image recognition apparatus 200, and/or with any device (e.g., router, modem, etc.) that enables the image recognition apparatus 200 to communicate with one or more other computing devices. Such communication may be through input-output interface 250. Also, the image recognition apparatus 200 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network such as the internet) through the network adapter 260. The network adapter 260 may communicate with other modules of the image recognition apparatus 200 via the bus 230. It should be appreciated that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the image recognition apparatus 200, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, to name a few.

System implementation

Referring to fig. 7, fig. 7 is a schematic structural diagram illustrating an image recognition system provided in the present application.

The present application provides an image recognition system, the system comprising:

any of the above image recognition apparatuses 10;

a display device 20 for providing a display function;

an interaction device 30 for providing interaction functionality.

Media embodiments

The present application further provides a computer-readable storage medium, where a computer program is stored, and when the computer program is executed by a processor, the steps of any one of the methods are implemented, and a specific implementation manner of the method is consistent with the implementation manner and the achieved technical effect described in the implementation manner of the method, and some contents are not described again.

Referring to fig. 8, fig. 8 shows a schematic structural diagram of a program product provided in the present application.

The program product is for implementing any of the methods described above. The program product may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited in this respect, and in this application, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

A computer readable storage medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable storage medium may also be any readable medium that can communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing. Program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the C language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

While the present application is described in terms of various aspects, including exemplary embodiments, the principles of the invention should not be limited to the disclosed embodiments, but are also intended to cover various modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.

Claims

1. An image identification method is characterized in that the method is used for identifying a segmented electrode implanted on an electrode lead in the intracranial of a patient, and a plurality of segmented electrodes and image markers are arranged on the circumferential direction of the electrode lead;

the method comprises the following steps:

acquiring medical image data of the patient;

respectively acquiring contour data of each slice electrode from contour data of a plurality of slice electrodes based on the position information of the image mark and the relative position relationship between the image mark and each slice electrode to obtain the identification result of each slice electrode;

the acquiring of the contour data of the electrode lead, the contour data of the image marker and the contour data of the plurality of sliced electrodes based on the medical image data comprises:

and acquiring the contour data of the electrode lead, the contour data of the image mark and the contour data of the plurality of sliced electrodes by utilizing the three-dimensional model.

2. The image recognition method of claim 1, further comprising:

acquiring target attitude information of the three-dimensional model;

3. The image recognition method according to claim 2, wherein the obtaining of the target pose information of the three-dimensional model includes:

4. The image recognition method according to claim 1, wherein the acquiring contour data of the electrode lead, contour data of the image marker, and contour data of the plurality of sliced electrodes based on the medical image data comprises:

5. The image recognition method according to claim 4, wherein the detecting the contour of the target object from each of the two-dimensional images to obtain contour data of the target object in each of the two-dimensional images comprises:

for each training data in the training set, performing the following:

6. The image recognition method of claim 5, wherein the model parameters comprise at least one of: contour pixel coordinate vector, contour color, contour line width, line type, topology parameters, contour maximum level, and contour offset parameters.

7. The image recognition method of claim 1, wherein the medical image data includes one or more of CT data, MR data, PET data, X-ray data, PET-CT data, and PET-MR data;

8. The image recognition method according to claim 1, wherein the obtaining of the contour data of each of the patch electrodes from the contour data of the plurality of patch electrodes based on the position information of the image marker and the relative positional relationship between the image marker and each of the patch electrodes comprises:

9. The image recognition method of claim 8, further comprising:

10. An image recognition device is used for recognizing a slice electrode implanted on an electrode lead in the intracranial of a patient, wherein a plurality of slice electrodes and image markers are arranged on the circumference of the electrode lead;

the apparatus comprises a processor configured to:

acquiring medical image data of the patient;

the processor is further configured to acquire profile data of the electrode lead, profile data of the image marker, and profile data of the plurality of sliced electrodes by:

11. The image recognition device of claim 10, wherein the processor is further configured to:

acquiring target attitude information of the three-dimensional model;

12. The image recognition device of claim 11, wherein the processor is further configured to obtain target pose information for the three-dimensional model by:

13. The image recognition device of claim 10, wherein the processor is further configured to obtain the profile data of the electrode lead, the image marker, and the plurality of sliced electrodes by:

14. The image recognition device of claim 13, wherein the processor is further configured to obtain the contour data of the target object in each of the two-dimensional images by:

for each training data in the training set, performing the following:

15. The image recognition device of claim 14, wherein the model parameters comprise at least one of: contour pixel coordinate vector, contour color, contour line width, line type, topology parameters, contour maximum level, and contour offset parameters.

16. The image recognition device of claim 10, wherein the medical image data includes one or more of CT data, MR data, PET data, X-ray data, PET-CT data, and PET-MR data;

17. The image recognition device of claim 10, wherein the processor is further configured to obtain the contour data for each sliced electrode by:

18. The image recognition device of claim 17, wherein the processor is further configured to:

19. An image recognition system, the system comprising:

the image recognition device of any one of claims 10-18;

a display device for providing a display function;

and the interaction equipment is used for providing an interaction function.

20. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which, when executed by a processor, implements the steps of the method of any of claims 1-9 or implements the functionality of the apparatus of any of claims 10-18.