Imaging recognition method and related device for stimulating electrode lead

Abstract

The present application provides an imaging recognition method and related device for a stimulating electrode lead. The device includes: an image acquisition module for using an image acquisition device to collect a target image of the stimulating electrode lead in real time; a detection result module for obtaining a marker detection result corresponding to the target image, where the marker detection result includes one or more tags and their confidence levels and position information; a confidence level judgment module for judging whether the confidence level of each tag meets a preset condition; when the confidence level of at least one tag does not meet the preset condition, the image acquisition module is called again; when the confidence levels of all tags meet the preset condition, the result output module is called; and a result output module for outputting the marker detection result to a preset user device. Through the imaging recognition device, a more accurate recognition result of the stimulating electrode lead can be obtained.

Description

Technical Field

[0001] This application relates to the technical field of implantable medical devices, and in particular, to an imaging recognition method and related device for a stimulating electrode lead wire. Background Art

[0002] In the prior art, for deep brain electrical stimulation therapy (DBS, Deep Brain Stimulation), it involves delivering electrical stimulation to nerve structures in specific regions of the brain to stimulate or inhibit cell activities, and can effectively treat movement disorders such as chronic pain, Parkinson's disease, essential tremor, epilepsy, as well as mental diseases such as depression and obsessive-compulsive disorder. Specifically, the stimulating electrode for applying electrical stimulation acts on the patient's head and stimulates the designated part of the brain, playing a therapeutic role in the patient's brain injury. At the same time, the other end of the stimulating electrode is connected to a nerve stimulator through a stimulating electrode lead wire. Currently, in order to meet the expectation of accurately implanting the stimulating electrode lead wire at the desired site in the brain and avoiding side effects on other parts of the brain, various imaging techniques are usually used to assist the stimulating electrode lead wire to be relatively accurately implanted at the desired site in the brain. Such imaging techniques include magnetic resonance imaging (MRI, Magnetic Resonance Imaging), computed tomography (CT, Computed Tomography), X-ray, fluorescence imaging, and stereoscopic imaging.

[0003] In specific applications, doctors expect to accurately place and orient the stimulating electrode lead wire that can release stimulation in a patient (such as the brain) to deliver electrical stimulation to the intended site and avoid side effects. For example, it is expected to deliver the stimulation of the stimulating electrode lead wire to a very small target point so as not to stimulate other adjacent brain tissues; if the stimulation is not accurately delivered to the desired target point, the curative effect will be reduced, and adjacent brain tissues will receive unnecessary excessive stimulation, causing pain to the patient.

[0004] Therefore, there is an urgent need to design a new imaging recognition device for a stimulating electrode lead wire to assist doctors in improving the accuracy of placing the stimulating electrode lead wire into a patient's body. Summary of the Invention

[0005] The purpose of this application is to provide an imaging recognition method and related device for a stimulating electrode lead wire. By directly recognizing the marks set on the electrode sheet through imaging techniques and judging the confidence level of each label, the problem of low recognition accuracy of the stimulating electrode lead wire is solved.

[0006] The purpose of this application is achieved by adopting the following technical solutions:

[0007] In a first aspect, the present application provides an imaging recognition device for a stimulating electrode lead. A plurality of electrode pads are arranged on the outer peripheral surface of the stimulating electrode lead, and marks are respectively arranged on at least some of the electrode pads for identifying the electrode pads during imaging. The device includes: an image acquisition module for using an image acquisition device to collect a target image of the stimulating electrode lead in real time; a detection result module for obtaining a mark detection result corresponding to the target image, the mark detection result including one or more tags and their confidence levels and position information; a confidence level judgment module for judging whether the confidence level of each tag meets the preset condition; when the confidence level of at least one tag does not meet the preset condition, the image acquisition module is called again; when the confidence levels of all tags meet the preset condition, the result output module is called; and a result output module for outputting the mark detection result to a preset user device.

[0008] The beneficial effect of this technical solution is that detection is performed based on the target image collected in real time, and the confidence level of the tags in the detection result is judged through preset conditions (the tags are used to indicate the identification of the electrode pads). When the confidence level of each tag does not meet the preset condition, the image acquisition module is called again for image acquisition, detection, and judgment until a mark detection result that meets the confidence level condition is obtained. Compared with the prior art, a more accurate recognition result of the stimulating electrode lead is obtained. Doctors refer to the recognition result of the stimulating electrode lead without having to make complex logical judgments, with a high degree of intelligence. Even doctors without rich experience can accurately deliver the stimulation to the desired target point, shortening the time for doctors to place and orient the stimulating electrode lead, improving the efficiency of doctors to accurately place and orient the stimulating electrode lead, reducing the pain of patients during the process of doctors placing and orienting the stimulating electrode lead, and thus improving the therapeutic effect of electrical stimulation treatment on patients.

[0009] At the same time, since the marks are directly arranged on the electrode pads, the electrode pads with marks can play a role in determining the electrode orientation through imaging recognition and can also be used to generate stimulation signals. There is no need to additionally preset marking components in non-electrode pad areas, which can reduce the manufacturing cost of the stimulating electrode lead and the manufacturing difficulty of the stimulating electrode lead.

[0010] In some optional embodiments, the detection result module includes: an imaging recognition unit for using an imaging recognition model to perform imaging recognition on the target image to obtain a mark detection result corresponding to the target image; wherein, the training process of the imaging recognition model is as follows: obtaining a first training set, the first training set including a plurality of first training images and their corresponding annotation data of mark detection results; and using the first training set to train a preset first deep learning model to obtain the imaging recognition model.

[0011] The beneficial effects of this technical solution are as follows. Compared with the traditional manual recognition of target images, the imaging recognition model has a high degree of intelligence. Applying the trained imaging recognition model to the imaging recognition of the stimulation electrode lead in the actual scenario has a high recognition accuracy.

[0012] In some alternative embodiments, during the training process of the imaging recognition model, the training of the preset first deep learning model using the first training set includes: for each first training image in the first training set, inputting the first training image into the preset first deep learning model to obtain prediction data of the marked detection result corresponding to the first training image; based on the prediction data of the marked detection result corresponding to the first training image and the annotation data, updating the model parameters of the preset first deep learning model; detecting whether the preset first training end condition is satisfied. If so, stop the training and use the trained preset first deep learning model as the imaging recognition model. If not, continue to train the preset first deep learning model using the next training data.

[0013] The beneficial effects of this technical solution are as follows. Compared with the traditional recognition system, which often only analyzes and compares with existing images and their marked detection results, this application uses the first training set to train the first deep learning model, making the recognition effect of the finally formed imaging recognition model more matched with the actual imaging result, and the user obtains a more satisfactory imaging recognition result of the stimulation electrode lead, improving the user experience.

[0014] In some alternative embodiments, the detection result module includes: a target detection unit for performing target detection on the target image to obtain one or more sub-images and their corresponding position information, each sub-image corresponding to a mark; a sub-image classification unit for classifying the marks of each sub-image to obtain the label and its confidence level corresponding to each sub-image; a marking result unit for obtaining the marked detection result corresponding to the target image based on the label, its confidence level, and the position information corresponding to each sub-image, where the marked detection result includes one or more labels, their confidence levels, and position information.

[0015] The beneficial effects of this technical solution are as follows. Based on the target detection unit, the sub-image classification unit, and the marking result unit, the marks in the target image are classified by each sub-image to obtain the label and its confidence level corresponding to each sub-image, and then the marked detection result corresponding to the target image including all labels, their confidence levels, and position information is obtained, with a high degree of intelligence.

[0016] In some alternative embodiments, the sub - graph classification unit includes: a sub - graph classification sub - unit, configured to perform label classification on each sub - graph by using a label classification model to obtain a label classification result corresponding to each sub - graph; wherein, the training process of the label classification model is as follows: obtain a second training set, the second training set includes a plurality of second training images and annotation data of their corresponding label classification results; use the second training set to train a preset second deep - learning model to obtain the label classification model.

[0017] The beneficial effect of this technical solution is that by performing label classification on each sub - graph through the sub - graph classification sub - unit to obtain the label classification result corresponding to each sub - graph for training the imaging recognition model, the robustness of the imaging recognition model can be improved and its fitting risk can be effectively reduced.

[0018] In some alternative embodiments, during the training process of the label classification model, the step of using the second training set to train a preset second deep - learning model includes: for each second training image in the second training set, input the second training image into the preset second deep - learning model to obtain prediction data of the label detection result corresponding to the second training image; based on the prediction data of the label detection result corresponding to the second training image and the annotation data, update the model parameters of the preset second deep - learning model; detect whether a preset second training end condition is met, if so, stop training and use the trained preset second deep - learning model as the label classification model, if not, continue to train the preset second deep - learning model with the next training data.

[0019] The beneficial effect of this technical solution is that the second training end condition for ending training can be configured based on actual needs, and the trained label classification model has strong robustness and low over - fitting risk.

[0020] In some alternative embodiments, the device further includes: a result display module, configured to display the target image and its corresponding label detection result by using the user device.

[0021] The beneficial effect of this technical solution is that through the setting of the result display module, the target image of the stimulating electrode lead, the label of the electrode patch in the stimulating electrode lead, its confidence level and position information collected in real - time by the image acquisition device can be intuitively displayed on the display module. On the one hand, it makes the real - time information of the stimulating electrode lead more convenient for doctors to refer to; on the other hand, it enables patients or their families to intuitively understand the progress of diagnosis and treatment, alleviates the tension of patients and their families, and promotes the relationship of trust and understanding between doctors and patients.

[0022] Second aspect, the present application provides an imaging recognition method for a stimulating electrode lead. A plurality of electrode pads are arranged on the outer peripheral surface of the stimulating electrode lead, and markers are respectively arranged on at least some of the electrode pads. The markers are used to identify the electrode pads during imaging. The method includes: S101: Use an image acquisition device to collect a target image of the stimulating electrode lead in real time; S102: Obtain a marker detection result corresponding to the target image. The marker detection result includes one or more labels and their confidence levels and position information; S103: Determine whether the confidence level of each label meets the preset condition; when the confidence level of at least one label does not meet the preset condition, re-execute step S101 to obtain a new target image; when the confidence levels of all labels meet the preset condition, execute step S104; S104: Output the marker detection result to a preset user device.

[0023] In some alternative embodiments, step S102 includes: performing imaging recognition on the target image by using an imaging recognition model to obtain a marker detection result corresponding to the target image; wherein, the training process of the imaging recognition model is as follows: Obtain a first training set, the first training set includes a plurality of first training images and their corresponding annotation data of marker detection results; use the first training set to train a preset first deep learning model to obtain the imaging recognition model.

[0024] In some alternative embodiments, during the training process of the imaging recognition model, the use of the first training set to train a preset first deep learning model includes: for each first training image in the first training set, input the first training image into the preset first deep learning model to obtain prediction data of a marker detection result corresponding to the first training image; based on the prediction data of the marker detection result corresponding to the first training image and the annotation data, update the model parameters of the preset first deep learning model; detect whether a preset first training end condition is met. If so, stop training and use the trained preset first deep learning model as the imaging recognition model. If not, continue to train the preset first deep learning model with the next training data.

[0025] In some alternative embodiments, obtaining the marker detection result corresponding to the target image includes: performing target detection on the target image to obtain one or more sub-images and their corresponding position information, and each sub-image corresponds to a marker; performing marker classification on each sub-image to obtain the label and its confidence level corresponding to each sub-image; based on the label, confidence level, and position information corresponding to each sub-image, obtain the marker detection result corresponding to the target image. The marker detection result includes one or more labels and their confidence levels and position information.

[0026] In some alternative embodiments, step S202 includes: classifying each sub-graph using a label classification model to obtain a label classification result corresponding to each sub-graph; wherein, the training process of the label classification model is as follows: obtaining a second training set, the second training set including a plurality of second training images and annotation data of their corresponding label classification results; training a preset second deep learning model using the second training set to obtain the label classification model.

[0027] In some alternative embodiments, during the training process of the label classification model, the training of the preset second deep learning model using the second training set includes: for each second training image in the second training set, inputting the second training image into the preset second deep learning model to obtain prediction data of the label detection result corresponding to the second training image; updating the model parameters of the preset second deep learning model based on the prediction data of the label detection result corresponding to the second training image and the annotation data; detecting whether a preset second training end condition is satisfied, if so, stopping the training and using the trained preset second deep learning model as the label classification model, if not, continuing to train the preset second deep learning model using the next training data.

[0028] In some alternative embodiments, the method further includes step S105: displaying the target image and its corresponding label detection result using the user device.

[0029] In a third aspect, the present application provides an electronic device for imaging and identifying a stimulation electrode lead, with electrode pads disposed on the outer peripheral surface of the stimulation electrode lead, and markers are respectively disposed on at least some of the electrode pads for identifying the electrode pads during imaging; the electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the steps of the above method are implemented.

[0030] In a fourth aspect, the present application provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present application will be further described below in conjunction with the drawings and embodiments.

[0032] Figure 1 is a schematic structural diagram of an imaging and identification device for a stimulation electrode lead provided by an embodiment of the present application;

[0033] Figure 2It is a partial perspective view of a stimulating electrode lead provided by an embodiment of the present application;

[0034] Figure 3 It is a partial structural view of a stimulating electrode lead in a planarized state provided by an embodiment of the present application;

[0035] Figure 4 It is another partial structural view of a stimulating electrode lead in a planarized state provided by an embodiment of the present application;

[0036] Figure 5 It is another partial structural view of a stimulating electrode lead in a planarized state provided by an embodiment of the present application;

[0037] Figure 6 It is another partial structural view of a stimulating electrode lead in a planarized state provided by an embodiment of the present application;

[0038] Figure 7 It is another partial structural view of a stimulating electrode lead in a planarized state provided by an embodiment of the present application;

[0039] Figure 8 It is a structural view of a detection result module provided by an embodiment of the present application;

[0040] Figure 9 It is a structural view of an imaging recognition device for a stimulating electrode lead provided by an embodiment of the present application;

[0041] Figure 10 It is a flowchart of an imaging recognition method for a stimulating electrode lead provided by an embodiment of the present application;

[0042] Figure 11 It is a flowchart of a process for obtaining a marker detection result provided by an embodiment of the present application;

[0043] Figure 12 It is another flowchart of an imaging recognition method for a stimulating electrode lead provided by an embodiment of the present application;

[0044] Figure 13 It is a structural view of an electronic device provided by an embodiment of the present application;

[0045] Figure 14 It is a structural view of a computer-readable storage medium provided by an embodiment of the present application. Detailed implementation manners

[0046] Next, in combination with the accompanying drawings and specific implementation manners, the present application will be further described. It should be noted that, on the premise of no conflict, any combination of the following-described embodiments or technical features can form a new embodiment.

[0047] In the description and claims of this application and the above-mentioned drawings, the terms "first", "second", "third", "fourth", etc. (if any) are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of this application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "corresponding to" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device comprising a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0048] The Chinese patent with the publication number CN112604159A discloses a segmented electrode. The electrode orientation can be identified by additionally setting marks, and the position and direction of the electrode can be determined by the correspondence between the predefined mark direction and the electrode stimulation sheet. However, this method requires doctors to have strong logical judgment ability, and errors are likely to occur during the judgment process, bringing unnecessary pain to patients.

[0049] See Figure 1 , an imaging recognition device for a stimulation electrode lead is provided in an embodiment of this application. A plurality of electrode sheets are arranged on the outer peripheral surface of the stimulation electrode lead, and marks are respectively arranged on at least some of the electrode sheets for identifying the electrode sheets during imaging.

[0050] The above device can implement a recognition method different from the prior art. (Using a manufacturing process such as a flexible thin-film circuit) directly utilize the structure of the marks arranged on the electrode sheets, and identify the positions and directions of the electrode sheets with different marks in the stimulation electrode lead through the images obtained by the imaging technology. That is to say, compared with the additional marks in the prior art, the marks respectively arranged on the electrode sheets can be used to identify the orientation during imaging.

[0051] See Figures 3 to 7 , which is a schematic diagram of the stimulation electrode lead in a flattened state. The marks arranged on the electrode sheets can be used to distinguish and identify each electrode sheet. For example, the shape difference of each electrode sheet in the electrode sheets is used as a mark, or the different positions of the connection points on the electrode sheet are used as marks for distinguishing different electrode sheets ( Figure 3 , Figure 4 , Figure 6 and Figure 7 ), or the different shapes of the connection points on the electrode sheet are used as marks for distinguishing different electrode sheets ( Figure 5) or a combination of different marking settings of the above electrode patches. A doctor can implant a stimulating electrode wire in the deep brain region of a patient. The surface of the stimulating electrode wire can be provided with a plurality of electrode patches arranged in a regular matrix, and a plurality of electrode patches arranged on the stimulating electrode wire can release stimulation through a nerve stimulator.

[0052] When the electrode patches are regularly arranged, different columns of electrodes can be marked in different rows to identify the electrode patches. In one example, there are 4 rows of electrode patches, with 3 columns in each row (12 output labels, i.e., electrode patch No. 1 to electrode patch No. 12). In one embodiment, the electrode patch in the first row and the first column and the electrode patch in the second row and the second column can be marked; in another embodiment, the electrode patch in the second row and the third column and the electrode patch in the fourth row and the second column can be marked; in yet another embodiment, the electrode patch in the first row and the third column and the electrode patch in the third row and the first column can be marked; in yet another embodiment, the electrode patch in the first row and the second column and the electrode patch in the fourth row and the third column can be marked.

[0053] In another example, there are 5 rows of electrode patches, with 4 columns in each row (20 output labels, i.e., electrode patch No. 1 to electrode patch No. 20). In one embodiment, the electrode patch in the first row and the first column and the electrode patch in the second row and the second column can be marked; in another embodiment, the electrode patch in the second row and the third column and the electrode patch in the fourth row and the second column can be marked; in yet another embodiment, the electrode patch in the first row and the third column and the electrode patch in the third row and the first column can be marked; in yet another embodiment, the electrode patch in the first row and the second column and the electrode patch in the fifth row and the third column can be marked.

[0054] In yet another example, there are 5 rows of electrode patches, with 4 columns in each row (20 output labels, i.e., electrode patch No. 1 to electrode patch No. 20). In one embodiment, the electrode patch in the first row and the first column, the electrode patch in the second row and the second column, and the electrode patch in the third row and the third column can be marked; in another embodiment, the electrode patch in the second row and the third column, the electrode patch in the fourth row and the second column, and the electrode patch in the fifth row and the fourth column can be marked; in yet another embodiment, the electrode patch in the first row and the third column, the electrode patch in the third row and the first column, and the electrode patch in the fourth row and the second column can be marked; in yet another embodiment, the electrode patch in the first row and the second column, the electrode patch in the fourth row and the third column, and the electrode patch in the fifth row and the first column can be marked.

[0055] Among them, the objects of identifying the stimulating electrode wire through the imaging recognition device can be the first diagnosing doctor of the patient, the consultation expert, etc., who treat the patient. The patients in the embodiments of the present application can be Parkinson's patients, or mental illness patients such as depression patients and obsessive-compulsive disorder patients, and can also be drug addiction patients or drug addicts in rehabilitation.

[0056] Through the stimulating electrode lead, the electrical stimulation of the stimulator can be delivered to a specific area of the human body to apply stimulating treatment. In this embodiment, the stimulating electrode lead can release electrical stimulation to the neural structure of the brain to stimulate or inhibit cell activities, and can effectively treat, for example, spastic diseases (such as epilepsy), pain, migraine, mental diseases (such as major depressive disorder (MDD)), bipolar disorder, anxiety disorder, post-traumatic stress disorder, dysthymia, obsessive-compulsive disorder (OCD), behavioral disorders, mood disorders, memory disorders, mental state disorders, movement disorders (such as essential tremor or Parkinson's disease), Huntington's disease, Alzheimer's disease, drug addiction, autism or other neurological or psychiatric diseases and impairments.

[0057] The stimulator can be any one of an implantable neuroelectrical stimulation device, an implantable cardiac electrical stimulation system (also known as a cardiac pacemaker), an implantable drug delivery device (Implantable Drug Delivery System, abbreviated as IDDS), and a lead adapter device. When the stimulator is an implantable neuroelectrical stimulation device, the implantable neuroelectrical stimulation device is, for example, a deep brain stimulation system (Deep Brain Stimulation, abbreviated as DBS), an implantable cortical nerve stimulation system (Cortical Nerve Stimulation, abbreviated as CNS), an implantable spinal cord stimulation system (Spinal Cord Stimulation, abbreviated as SCS), an implantable sacral nerve stimulation system (Sacral Nerve Stimulation, abbreviated as SNS), an implantable vagus nerve stimulation system (Vagus Nerve Stimulation, abbreviated as VNS), etc.

[0058] The device includes an image acquisition module 101, a detection result module 102, a confidence judgment module 103, and a result output module 104.

[0059] The image acquisition module 101 is configured to use an image acquisition device to collect a target image of the stimulating electrode lead in real time. The image acquisition device can include an imaging device capable of implementing imaging technologies such as magnetic resonance imaging (MRI), computed tomography (CT), X-ray, fluorescence imaging, and stereoscopic imaging. See Figure 2 , which is a partial perspective schematic diagram of a stimulating electrode lead obtained by X-ray.

[0060] The detection result module 102 is configured to obtain a marker detection result corresponding to the target image, and the marker detection result includes one or more labels, their confidence levels, and position information. The position information can be the coordinate values corresponding to the marker, and the position of the corresponding marker on the target image can be accurately obtained through the position information.

[0061] The confidence judgment module 103 is used to judge whether the confidence of each label meets the preset condition; when the confidence of at least one label does not meet the preset condition, the image acquisition module is called again; when the confidence of all labels meets the preset condition, the result output module is called.

[0062] The result output module 104 is used to output the marker detection result to a preset user device. Among them, the user device for receiving the marker detection result can adopt a program controller existing in the prior art, that is, the user device can be a separate hardware device, which is an electronic device capable of performing data interaction with the stimulator through a wireless network or a wired network, such as a tablet computer, a computer, a mobile phone or a smart wearable device, etc. The user can use such a program control device to receive the marker detection result. Generally speaking, a computer program (i.e., software) is installed in the user device, and when the computer program is executed by a processor, it can realize the function of receiving the marker detection result in the embodiment of the present application.

[0063] The present application does not limit the preset condition. The preset condition is, for example, a numerical range preset condition. In one embodiment, the preset condition is that the confidence is not less than a preset confidence. The preset confidence is, for example, 0.95, 0.97, 0.94, 0.98.

[0064] The present application can set the same or different preset conditions for different patients. In a specific application, the same preset condition is set for different patients, that is, the confidence is not less than 0.96.

[0065] In another specific application, differential preset conditions can be set according to the different diseases or treatment stages of the patients to achieve personalized and customized diagnosis and treatment for the patients. For example, a doctor judges the positions and directions of the electrode patches of the stimulating electrode wires implanted in the bodies of patients Zhang San, Li Si, and Wang Wu through the above-mentioned imaging recognition device for the stimulating electrode wires. Refer to Table 1 below for the specific judgment situation.

[0066] Table 1

[0067]

[0068]

[0069] Generally speaking, the consultation time of doctors and the treatment time of patients are both very precious. Therefore, doctors and patients are more eager to locate the stimulating electrode lead to the effective expected target point to reduce the need for re-diagnosis and treatment caused by poor positioning. Thus, by judging whether the confidence level of each label meets the preset condition, even when the confidence level of a label does not meet the preset condition, the image acquisition module will be called again, saving more treatment time for doctors to deliver stimulation to very small target points through the stimulating electrode lead to the patient during treatment without stimulating adjacent brain tissues, and thereby reducing the discomfort of the patient during treatment.

[0070] Therefore, based on the target image collected in real time for detection, the confidence level of the label in the detection result is judged through the preset condition. The label is used to indicate the identification of the electrode patch. When the confidence level of each label does not meet the preset condition, the image acquisition module 101 is called again for image acquisition, detection and judgment until a marker detection result that meets the confidence level condition is obtained. Compared with the prior art, a more accurate identification result of the stimulating electrode lead can be obtained, and the degree of intelligence is high. Doctors refer to the identification result of the stimulating electrode lead without the need for complex logical judgment by doctors. Even doctors who are not experienced can accurately deliver the stimulation to the expected target point, shortening the time for doctors to place and orient the stimulating electrode lead, improving the efficiency of doctors to accurately place and orient the stimulating electrode lead, reducing the pain of the patient during the period when doctors place and orient the stimulating electrode lead, and improving the therapeutic effect of electrical stimulation on the patient.

[0071] At the same time, since the marker is directly set on the electrode patch, the electrode patch with the marker can play a role in determining the electrode orientation through imaging recognition, and can also be used to generate stimulation signals. There is no need to additionally preset a marker component in the non-electrode patch area, which can reduce the manufacturing cost of the stimulating electrode lead and the manufacturing difficulty of the stimulating electrode lead.

[0072] In some optional embodiments, any two of the multiple electrode patches are insulated from each other, and the multiple electrode patches include multiple stimulating electrode patches and multiple acquisition electrode patches. At this time, the stimulating electrode lead can not only be used to release electrical stimulation energy, but also be used to collect bioelectrical signals of tissues in the organism.

[0073] See Figure 9 , the device may further include a result display module 105, and the result display module 105 is used to display the target image and its corresponding marker detection result by using the user equipment. Among them, the result display module 105 may include device modules such as a display and a projector that provide a display function.

[0074] Among them, the marker detection result and the target image are displayed on the display module, which can be understood as displaying the target image on the interface of the display module. The position information is used to correspond the marker detection result and the image displayed on the target image. On the electrode pads displayed on the target image, the corresponding labels, confidence levels, etc. can be displayed. Among them, the displayed labels can be Electrode Pad 1, Electrode Pad 2... Electrode Pad N, etc., and the marked confidence levels can be 0.91, 0.94, 0.98, etc.

[0075] Thus, through the setting of the result display module 105, the target image of the stimulation electrode lead, the labels and confidence levels of the markers of the electrode pads in the stimulation electrode lead, and the position information collected in real time by the image acquisition device can be intuitively displayed on the display module. On the one hand, it makes the real-time information of the stimulation electrode lead more convenient for doctors to refer to; on the other hand, it enables patients or their families to intuitively understand the diagnosis and treatment process, relieve the tension of patients and their families, and promote the relationship of trust and understanding between doctors and patients.

[0076] In some embodiments, the detection result module may include an imaging recognition unit. The imaging recognition unit may be used to perform imaging recognition on the target image using an imaging recognition model to obtain the marker detection result corresponding to the target image.

[0077] Among them, the training process of the imaging recognition model is as follows:

[0078] Obtain a first training set, where the first training set includes a plurality of first training images and the annotation data of the corresponding marker detection results;

[0079] Use the first training set to train a preset first deep learning model to obtain the imaging recognition model.

[0080] Thus, through the imaging recognition model, compared with the traditional manual recognition of the target image, the degree of intelligence is high; applying the trained imaging recognition model to the imaging recognition of the stimulation electrode lead in the actual scenario, the recognition accuracy is high.

[0081] Specifically, in the training process of the imaging recognition model, the step of using the first training set to train a preset first deep learning model may include:

[0082] For each first training image in the first training set, input the first training image into a preset first deep learning model to obtain prediction data of the marker detection result corresponding to the first training image; update the model parameters of the preset first deep learning model based on the prediction data of the marker detection result corresponding to the first training image and the annotation data; detect whether a preset first training end condition is satisfied. If so, stop the training and use the trained preset first deep learning model as the imaging recognition model. If not, continue to train the preset first deep learning model with the next training data.

[0083] Training a preset first deep learning model using a first training set can obtain a trained imaging recognition model. The imaging recognition model can be obtained by training with a large amount of training data, can predict corresponding marker detection results for various input data, has a wide range of applications, and a high level of intelligence. By design, establishing an appropriate number of neuron computing nodes and a multi-layer operation hierarchy, and selecting appropriate input and output layers, a preset first deep learning model can be obtained. Through the learning and optimization of the preset first deep learning model, a functional relationship from input to output is established. Although the functional relationship between input and output cannot be found 100%, it can approximate the real correlation relationship as much as possible. The imaging recognition model trained thereby can realize the self-diagnosis function of imaging recognition, and the diagnosis result has high reliability.

[0084] Therefore, compared with traditional recognition systems, which often only analyze and compare with existing images and their marker detection results, this application trains a first deep learning model using a first training set, making the recognition effect of the finally formed imaging recognition model more matched with the actual imaging result, and the user obtains a more satisfactory imaging recognition result of the stimulating electrode lead, improving the user experience.

[0085] See Figure 8 , in some embodiments, the detection result module may further include a target detection unit 201, a sub-image classification unit 202, and a marker result unit 203.

[0086] The target detection unit 201 is configured to perform target detection on the target image to obtain one or more sub-images and their corresponding position information, and each sub-image corresponds to a marker respectively. The main attributes of the target image can be reflected through the sub-images, and processing such as compression and denoising of the image data can be realized.

[0087] The sub-image classification unit 202 is configured to perform marker classification on each sub-image to obtain the label and its confidence corresponding to each sub-image.

[0088] The labeled result unit 203 is configured to obtain the labeled detection result corresponding to the target image based on the labels corresponding to each sub - graph, their confidence levels, and location information. The labeled detection result includes one or more labels, their confidence levels, and location information.

[0089] Thus, based on the object detection unit, the sub - graph classification unit, and the labeled result unit, the labels in the target image are classified by each sub - graph to obtain the labels corresponding to each sub - graph and their confidence levels, and further obtain the labeled detection result corresponding to the target image, including all labels, their confidence levels, and location information, with a high degree of intelligence.

[0090] In some embodiments, the sub - graph classification unit may include a sub - graph classification sub - unit, which is configured to use a label classification model to classify the labels of each sub - graph and obtain the label classification result corresponding to each sub - graph.

[0091] Among them, the training process of the label classification model is as follows:

[0092] Obtain a second training set, which includes a plurality of second training images and the annotation data of their corresponding label classification results;

[0093] Use the second training set to train a preset second deep - learning model to obtain the label classification model.

[0094] Thus, by using the sub - graph classification sub - unit to classify the labels of each sub - graph and obtain the label classification result corresponding to each sub - graph for training the label classification model, the robustness of the label classification model can be improved, and its fitting risk can be effectively reduced.

[0095] Specifically, during the training process of the label classification model, the step of using the second training set to train a preset second deep - learning model may include the following steps:

[0096] For each second training image in the second training set, input the second training image into the preset second deep - learning model to obtain the predicted data of the labeled detection result corresponding to the second training image;

[0097] Based on the predicted data of the labeled detection result corresponding to the second training image and the annotation data, update the model parameters of the preset second deep - learning model;

[0098] Detect whether the preset second training end condition is met. If so, stop training and use the trained preset second deep - learning model as the label classification model. If not, continue to train the preset second deep - learning model with the next training data.

[0099] Therefore, the second training end condition for the end of training can be configured based on actual requirements, and the trained marker classification model has strong robustness and a low risk of overfitting.

[0100] Training the preset second deep learning model using the second training set can obtain a trained marker classification model. The marker classification model can be trained with a large amount of training data, can predict corresponding marker detection results for various input data, has a wide range of applications, and a high level of intelligence. By designing and establishing an appropriate number of neuron computing nodes and a multi-layer operation hierarchy, and selecting appropriate input and output layers, the preset second deep learning model can be obtained. Through the learning and optimization of the preset second deep learning model, a functional relationship from input to output can be established. Although the functional relationship between input and output cannot be found 100%, it can approximate the real correlation relationship as much as possible. The marker classification model trained thereby can realize the self-diagnosis function for imaging recognition, and the diagnosis result has high reliability.

[0101] See Figure 10 , this application embodiment also provides an imaging recognition method for a stimulating electrode lead. Since the role of the imaging recognition method for the stimulating electrode lead is the same as or similar to that of the above-mentioned imaging recognition device for the stimulating electrode lead, it will not be elaborated here.

[0102] Among them, a plurality of electrode pads are arranged on the outer peripheral surface of the stimulating electrode lead, and markers are respectively arranged on at least some of the electrode pads, and the markers are used to identify the electrode pads during imaging.

[0103] The method includes steps S101 to S104.

[0104] Step S101: Use an image acquisition device to collect a target image of the stimulating electrode lead in real time.

[0105] Step S102: Obtain the marker detection result corresponding to the target image, and the marker detection result includes one or more labels and their confidence levels and position information.

[0106] Step S103: Determine whether the confidence level of each label meets the preset condition; when the confidence level of at least one label does not meet the preset condition, re-execute step S101 to obtain a new target image; when the confidence levels of all labels meet the preset condition, execute step S104.

[0107] Step S104: Output the marker detection result to a preset user device.

[0108] In some embodiments, step S102 may include: performing imaging recognition on the target image using an imaging recognition model to obtain the marker detection result corresponding to the target image.

[0109] Among them, the training process of the imaging recognition model is as follows: Obtain a first training set, where the first training set includes a plurality of first training images and annotation data of their corresponding marked detection results; Use the first training set to train a preset first deep learning model to obtain the imaging recognition model.

[0110] In some embodiments, during the training process of the imaging recognition model, the using the first training set to train a preset first deep learning model may include:

[0111] For each first training image in the first training set, input the first training image into the preset first deep learning model to obtain prediction data of the marked detection result corresponding to the first training image; Based on the prediction data of the marked detection result corresponding to the first training image and the annotation data, update the model parameters of the preset first deep learning model; Detect whether a preset first training end condition is satisfied. If so, stop training and use the trained preset first deep learning model as the imaging recognition model. If not, continue to train the preset first deep learning model with the next training data.

[0112] See Figure 11 , in some embodiments, obtaining the marked detection result corresponding to the target image may include step S201 to step S203.

[0113] Step S201: Perform object detection on the target image to obtain one or more sub-images and their corresponding position information, and each sub-image corresponds to a mark respectively.

[0114] Step S202: Perform mark classification on each sub-image to obtain the label and its confidence level corresponding to each sub-image.

[0115] Step S203: Based on the label, its confidence level, and the position information corresponding to each sub-image, obtain the marked detection result corresponding to the target image, where the marked detection result includes one or more labels, their confidence levels, and position information.

[0116] In some embodiments, step S202 may include:

[0117] Use a mark classification model to perform mark classification on each sub-image to obtain the mark classification result corresponding to each sub-image.

[0118] Among them, the training process of the mark classification model is as follows:

[0119] Obtain a second training set, where the second training set includes a plurality of second training images and annotation data of their corresponding labeled classification results;

[0120] Use the second training set to train a preset second deep learning model to obtain the labeled classification model.

[0121] In some embodiments, during the training process of the labeled classification model, the using the second training set to train a preset second deep learning model may include:

[0122] For each second training image in the second training set, input the second training image into a preset second deep learning model to obtain prediction data of the labeled detection result corresponding to the second training image;

[0123] Based on the prediction data of the labeled detection result corresponding to the second training image and the annotation data, update the model parameters of the preset second deep learning model;

[0124] Detect whether a preset second training end condition is satisfied. If so, stop training and use the trained preset second deep learning model as the labeled classification model. If not, continue to train the preset second deep learning model with the next training data.

[0125] See Figure 12 , in some embodiments, the method may further include step S105.

[0126] Step S105: Use the user device to display the target image and its corresponding labeled detection result.

[0127] See Figure 13 , an embodiment of the present application further provides an electronic device 200 for imaging and identifying a stimulating electrode lead. The stimulating electrode lead is disposed on the outer peripheral surface of the stimulating electrode lead, and at least some of the electrode patches are respectively provided with labels for identifying the electrode patches during imaging.

[0128] The electronic device 200 includes at least one memory 210, at least one processor 220, and a bus 230 connecting different platform systems.

[0129] The memory 210 may include a readable medium in the form of a volatile memory, such as a random access memory (RAM) 211 and / or a cache memory 212, and may further include a read-only memory (ROM) 213.

[0130] Among them, the memory 210 also stores a computer program, which can be executed by the processor 220, so that the processor 220 executes the steps of the above method in the embodiments of the present application. The specific implementation manner is the same as the implementation manner and the achieved technical effects described in the above method embodiments, and some contents will not be elaborated herein.

[0131] The memory 210 may further include a utility 214 having at least one program module 215. Such program modules 215 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment.

[0132] Correspondingly, the processor 220 can execute the above computer program and can also execute the utility 214.

[0133] The bus 230 may represent one or more of several types of bus structures, including a memory bus or a memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any bus structure in a variety of bus structures.

[0134] The electronic device 200 can also communicate with one or more external devices 240, such as a keyboard, a pointing device, a Bluetooth device, etc., and can also communicate with one or more devices capable of interacting with the electronic device 200, and / or communicate with any device (such as a router, a modem, etc.) that enables the electronic device 200 to communicate with one or more other computing devices. Such communication can be carried out through the input / output interface 250. Moreover, the electronic device 200 can also communicate with one or more networks (such as 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 can communicate with other modules of the electronic device 200 through the bus 230. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in combination with the electronic device 200, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, etc.

[0135] The embodiments of the present application also provide a computer-readable storage medium. The specific implementation manner is the same as the implementation manner and the achieved technical effects described in the above method embodiments, and some contents will not be elaborated herein.

[0136] The computer-readable storage medium is used to store a computer program; when the computer program is executed, the steps of the above method in the embodiments of the present application are implemented.

[0137] Figure 14Fig. 0 shows the program product 300 provided by this embodiment for implementing the above method. It can be a portable compact disc read-only memory (CD-ROM) and includes program code, and can run on a terminal device, such as a personal computer. However, the program product 300 of the present invention is not limited thereto. In this document, a readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in conjunction with an instruction execution system, apparatus, or device. The program product 300 can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (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 above.

[0138] A computer-readable storage medium can include a data signal propagated in a baseband or as part of a carrier wave, in which the readable program code is carried. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than the readable storage medium, and this readable medium can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above. The program code for performing the operations of the present invention can be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as an independent software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, by using an Internet service provider to connect through the Internet).

[0139] This application is described from the perspectives of purpose of use, efficacy, progress, and novelty, and has met the functional enhancement and usage requirements emphasized by the Patent Law. The above description and accompanying drawings of this application are only preferred embodiments of this application, and do not limit this application thereby. Therefore, all those that are similar or identical to the structure, device, features, etc. of this application, that is, all equivalent substitutions or modifications made according to the scope of the patent application of this application, shall fall within the scope of patent application protection of this application.

Claims

1. An imaging recognition device for a stimulating electrode lead, characterized in that, A plurality of electrode patches are arranged on the outer peripheral surface of the stimulation electrode wire, and marks are respectively arranged on at least some of the electrode patches, and the marks are used to identify the electrode patches during imaging; The device includes: An image acquisition module, configured to use an image acquisition device to acquire a target image of the stimulation electrode wire in real time; A detection result module, configured to obtain a mark detection result corresponding to the target image, where the mark detection result includes one or more tags and their confidence levels and position information; A confidence level judgment module, configured to judge whether the confidence level of each tag meets a preset condition; when the confidence level of at least one tag does not meet the preset condition, the image acquisition module is called again; when the confidence levels of all tags meet the preset condition, the result output module is called; A result output module, configured to output the mark detection result to a preset user device.

2. The imaging and recognition device for the stimulating electrode lead according to claim 1, wherein, The detection result module includes: An imaging recognition unit, configured to perform imaging recognition on the target image by using an imaging recognition model to obtain a mark detection result corresponding to the target image; Wherein, the training process of the imaging recognition model is as follows: Obtain a first training set, where the first training set includes a plurality of first training images and annotation data of their corresponding mark detection results; Use the first training set to train a preset first deep learning model to obtain the imaging recognition model.

3. The imaging and recognition device for the stimulating electrode lead according to claim 2, characterized in that, During the training process of the imaging recognition model, the using the first training set to train a preset first deep learning model includes: For each first training image in the first training set, input the first training image into a preset first deep learning model to obtain prediction data of a mark detection result corresponding to the first training image; Based on the prediction data of the mark detection result corresponding to the first training image and the annotation data, update the model parameters of the preset first deep learning model; Detect whether a preset first training end condition is met. If so, stop training and use the trained preset first deep learning model as the imaging recognition model. If not, continue to train the preset first deep learning model with the next training data.

4. The imaging and recognition device for the stimulating electrode lead according to claim 1, wherein, The detection result module includes: A target detection unit, configured to perform target detection on the target image to obtain one or more sub-images and their corresponding position information, and each sub-image corresponds to a mark; A sub-image classification unit, configured to perform mark classification on each sub-image to obtain the tag and its confidence level corresponding to each sub-image; A mark result unit, configured to obtain a mark detection result corresponding to the target image based on the tag, its confidence level, and position information corresponding to each sub-image, where the mark detection result includes one or more tags and their confidence levels and position information.

5. The imaging and recognition device for the stimulating electrode lead according to claim 4, wherein The sub-image classification unit includes: A sub-image classification subunit, configured to perform mark classification on each sub-image by using a mark classification model to obtain a mark classification result corresponding to each sub-image; Wherein, the training process of the mark classification model is as follows: Obtain a second training set, where the second training set includes a plurality of second training images and annotation data of their corresponding mark classification results; Train a preset second deep learning model using the second training set to obtain the marker classification model.

6. The imaging and recognition device for the stimulating electrode lead according to claim 5, wherein During the training process of the marker classification model, the step of training the preset second deep learning model using the second training set includes: For each second training image in the second training set, input the second training image into the preset second deep learning model to obtain prediction data of the marker detection result corresponding to the second training image; Update the model parameters of the preset second deep learning model based on the prediction data of the marker detection result corresponding to the second training image and the annotation data; Detect whether a preset second training end condition is satisfied. If so, stop the training and use the trained preset second deep learning model as the marker classification model. If not, continue to train the preset second deep learning model using the next training data.

7. The imaging and recognition device for the stimulating electrode lead according to claim 1, characterized in that, The device further includes: A result display module for displaying the target image and its corresponding marker detection result using the user device.

8. An imaging recognition method for a stimulating electrode lead, characterized in that, A plurality of electrode patches are arranged on the outer peripheral surface of the stimulation electrode wire, and markers are respectively arranged on at least some of the electrode patches for identifying the electrode patches during imaging; The method includes: S101: Use an image acquisition device to collect a target image of the stimulation electrode wire in real time; S102: Obtain a marker detection result corresponding to the target image, where the marker detection result includes one or more labels and their confidence levels and position information; S103: Determine whether the confidence level of each label meets a preset condition; when the confidence level of at least one label does not meet the preset condition, re-execute step S101 to obtain a new target image; when the confidence levels of all labels meet the preset condition, execute step S104; S104: Output the marker detection result to a preset user device.

9. An electronic device, characterized in that, For imaging and identifying a stimulation electrode wire, a plurality of electrode patches are arranged on the outer peripheral surface of the stimulation electrode wire, and markers are respectively arranged on at least some of the electrode patches for identifying the electrode patches during imaging; The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method according to claim 8 are implemented.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to claim 8 are implemented.

Images