CN116850447A - DBS surgical needle tract adjusting device and computer storage medium - Google Patents
DBS surgical needle tract adjusting device and computer storage medium Download PDFInfo
- Publication number
- CN116850447A CN116850447A CN202310815703.9A CN202310815703A CN116850447A CN 116850447 A CN116850447 A CN 116850447A CN 202310815703 A CN202310815703 A CN 202310815703A CN 116850447 A CN116850447 A CN 116850447A
- Authority
- CN
- China
- Prior art keywords
- needle track
- needle
- deviation
- physiological electric
- electric signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012549 training Methods 0.000 claims abstract description 57
- 239000013598 vector Substances 0.000 claims abstract description 52
- 238000004590 computer program Methods 0.000 claims abstract description 12
- 238000005457 optimization Methods 0.000 claims abstract description 12
- 238000013145 classification model Methods 0.000 claims description 20
- 238000001228 spectrum Methods 0.000 claims description 18
- 238000010801 machine learning Methods 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 11
- 238000000034 method Methods 0.000 description 27
- 238000005070 sampling Methods 0.000 description 6
- 230000003595 spectral effect Effects 0.000 description 6
- 238000012706 support-vector machine Methods 0.000 description 6
- 210000004556 brain Anatomy 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000001276 controlling effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 210000001905 globus pallidus Anatomy 0.000 description 4
- 230000000638 stimulation Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 238000002513 implantation Methods 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 210000005036 nerve Anatomy 0.000 description 2
- 208000012661 Dyskinesia Diseases 0.000 description 1
- 208000018737 Parkinson disease Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000001175 cerebrospinal fluid Anatomy 0.000 description 1
- 238000007428 craniotomy Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000007943 implant Substances 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004007 neuromodulation Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 230000001755 vocal effect Effects 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N1/00—Electrotherapy; Circuits therefor
- A61N1/02—Details
- A61N1/04—Electrodes
- A61N1/05—Electrodes for implantation or insertion into the body, e.g. heart electrode
- A61N1/0526—Head electrodes
- A61N1/0529—Electrodes for brain stimulation
- A61N1/0534—Electrodes for deep brain stimulation
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/388—Nerve conduction study, e.g. detecting action potential of peripheral nerves
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N1/00—Electrotherapy; Circuits therefor
- A61N1/18—Applying electric currents by contact electrodes
- A61N1/32—Applying electric currents by contact electrodes alternating or intermittent currents
- A61N1/36—Applying electric currents by contact electrodes alternating or intermittent currents for stimulation
- A61N1/3605—Implantable neurostimulators for stimulating central or peripheral nerve system
- A61N1/3606—Implantable neurostimulators for stimulating central or peripheral nerve system adapted for a particular treatment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Neurology (AREA)
- Neurosurgery (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Data Mining & Analysis (AREA)
- Medical Informatics (AREA)
- Physics & Mathematics (AREA)
- Radiology & Medical Imaging (AREA)
- Psychology (AREA)
- Theoretical Computer Science (AREA)
- Heart & Thoracic Surgery (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Databases & Information Systems (AREA)
- Primary Health Care (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Epidemiology (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Cardiology (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The application relates to the technical field of surgical needle tract adjustment, and discloses a DBS surgical needle tract adjustment device, which comprises a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the implementation steps are as follows: collecting physiological electric signals to obtain a physiological electric signal sample set, counting the time length and sequence of occurrence of various nuclear clusters in each section of physiological electric signals to obtain a first feature vector, marking a needle track deviation type label of each section of physiological electric signals, and combining the first feature vector and the needle track deviation type label to obtain a first sample training set; the first sample training set is imported into a first classifier to carry out training optimization, and a needle track deviation prediction model is obtained; and identifying the nucleus group type of the physiological electric signal to be detected, predicting the needle tract deviation azimuth of the physiological electric signal to be detected by combining the nucleus group type of the physiological electric signal to be detected and the needle tract deviation prediction model, and adjusting the needle tract according to the needle tract deviation azimuth. The application can automatically adjust the DBS operation needle track, and has high adjustment accuracy.
Description
The application relates to a method, a system and a device for regulating a DBS surgical needle tract, which are divisional applications of Chinese patent application with the application number of CN202010356723.0 and the application number of 29 of 04 month in 2020.
Technical Field
The invention relates to the technical field of surgical needle tract adjustment, in particular to a DBS surgical needle tract adjustment device and a computer storage medium.
Background
Deep brain stimulation (Deep Brain Stimulation, DBS) is an interventional neuromodulation technique that has brought back a quality of life to thousands of dyskinesia patients over the past 25 years. Particularly in the treatment of parkinson's disease, the motor score and the kinetic performance of the patient receiving the treatment can be improved by more than 80 percent. In order to provide an effective treatment to the patient, the physician needs to prepare an accurate surgical plan to ensure accurate implantation of the electrodes. However, during craniotomy, brain parenchyma can shift to a certain extent along with the partial loss of cerebrospinal fluid; the DBS operation requires MRI imaging of the brain prior to the operation and a process of fusion registration with CT during the operation, in which MRI is limited by its own accuracy and errors can occur during fusion with CT images, and in addition, the accuracy of stereotactic devices and errors introduced by manual operation can make the position of the target point planned prior to the operation inconsistent with the actual one.
For this problem, the doctor needs to make a temporary adjustment to the needle track. The process of needle tract adjustment often depends on the operation experience of a clinician, and to solve this problem, the clinician needs to record microelectrodes, puncture the microelectrodes from a position at a certain distance from the target point to the target point position according to a designed path, and record the neuron electrical activity of the passed position. The doctor judges the position of the current needle track relative to the actual target spot according to subjective experience by observing the record and correspondingly adjusts the current needle track so as to implant the puncture needle into the actual target spot.
The current process of microelectrode recording and needle tract adjustment by doctors is summarized as follows:
recording the physiological electrical signals at the needle tip from 10mm above the target until 5mm below the target, a set of physiological electrical signals at a series of points on the current actual needle track can be obtained (typically recorded every 1mm from 10 on the target to 5 on the target, recorded every 0.5mm from 5 on the target to 5 below the target, then signals of about 25 sites in total);
analyzing the series of collected signals by an operator, and preliminarily judging the offset direction and the offset of the actual position of the current electrode by comparing the actually recorded signals with the ideal signals and the characteristic signals of surrounding adjacent nuclear clusters;
the doctor gives temporary stimulation in operation, and changes different stimulation parameters, and simultaneously makes a verbal query with the patient, and whether the actual position of the current electrode deviates from the planned position or not is judged by the experience of the patient feedback, including the deviation direction and the deviation amount (the process has extremely high requirements on the subjective experience of the doctor);
the doctor withdraws the microelectrode recording needle from the middle channel, and changes one of the four other channels up, down, left and right according to the judgment result (namely the offset direction) of the microelectrode recording needle; and repeating the electrophysiological recording process, if the physiological electrical signal recording result is ideal, determining the current channel as the final electrode implanted channel if the characteristic signal recording length reaches the clinical requirement, otherwise, changing the channel again until the physiological electrical signal meets the requirement.
The method for manually judging and adjusting the DBS needle track has low adjusting efficiency, cannot guarantee the adjusting accuracy, is easy to cause errors due to manual misjudgment, has extremely high requirements on subjective experience of doctors, and prevents the development of DBS technology to a certain extent.
Disclosure of Invention
The invention aims to overcome the technical defects, and provides a DBS surgical needle tract adjusting device and a computer storage medium, which solve the technical problems of low adjusting efficiency, low adjusting accuracy and easy misjudgment when manually adjusting a DBS surgical needle tract in the prior art.
In order to achieve the above technical purpose, the technical scheme of the invention provides a DBS surgical needle tract adjusting device, which comprises a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the following steps are realized:
collecting physiological electric signals to obtain a physiological electric signal sample set, counting the time length and sequence of occurrence of various nuclear clusters in each section of the physiological electric signals to obtain a first feature vector, marking a needle track deviation type label of each section of the physiological electric signals, and combining the first feature vector and the needle track deviation type label to obtain a first sample training set; wherein, each section of the physiological electric signals is obtained by dividing the physiological electric signals in the physiological electric signal sample set into a plurality of sections according to the type of the nuclear cluster;
The first sample training set is led into a first classifier to carry out training optimization, and a needle track deviation prediction model is obtained;
and identifying the nucleus group type of the physiological electric signal to be detected, predicting the needle tract deviation azimuth of the physiological electric signal to be detected by combining the nucleus group type of the physiological electric signal to be detected and the needle tract deviation prediction model, and adjusting the needle tract according to the needle tract deviation azimuth.
Further, counting the time length and the sequence of occurrence of various nuclei in each section of the physiological electric signal to obtain a first feature vector, wherein the first feature vector specifically comprises:
counting the occurrence times of various kinds of nuclear clusters in the physiological electric signal, representing the occurrence time length of the various kinds of nuclear clusters according to the occurrence times of the various kinds of nuclear clusters, and arranging the occurrence times of the various kinds of nuclear clusters according to the occurrence sequence of the various kinds of nuclear clusters to obtain a first feature vector.
Further, the needle track deviation type label for marking each section of the physiological electric signal is specifically:
and respectively defining a deviation distance range between each needle track and the target needle track based on the deviation degree of each needle track in each needle track deviation direction to obtain a plurality of needle track deviation types, and marking the needle track deviation type label corresponding to each needle track deviation type label.
Further, before defining the range of the deviation distance from the target needle track based on the respective needle track deviation degree of each needle track deviation direction, respectively, a plurality of needle track deviation types are obtained, the method further comprises:
a plurality of lane departure directions are defined according to the target lane, and at least one lane departure degree is defined for each of the lane departure directions.
Further, the lane departure type tag corresponding to each of the lane departure type marks includes:
defining a label for each needle track deviation type, judging the needle track deviation type corresponding to each section of physiological electric signal according to the defined label, and marking the corresponding needle track deviation type label.
Further, defining a range of offset distances from the target lane based on the respective lane offset degrees of each lane offset direction, respectively, to obtain a plurality of lane offset types, further comprising:
controlling a puncture needle to enter from a middle needle track of a needle track group, collecting physiological electric signals corresponding to the middle needle track, and determining a middle needle track deviation direction and a middle needle track deviation degree based on the physiological electric signals corresponding to the middle needle track; the needle track group comprises a middle needle track and non-middle needle tracks which are positioned in four different directions around the middle needle track and are parallel to the middle needle track;
Determining whether the middle needle track is a target needle track based on the middle needle track deviation direction and the middle needle track deviation degree;
under the condition that the middle needle track is not the target needle track, controlling the puncture needle to enter from the non-middle needle track after being withdrawn from the middle needle track, collecting physiological electric signals corresponding to the non-middle needle track, and determining the deviation direction and the deviation degree of the non-middle needle track based on the physiological electric signals corresponding to the non-middle needle track;
a target stitch is determined from the non-intermediate stitch based on the non-intermediate stitch direction and the degree of non-intermediate stitch offset.
Further, the nucleus type of the physiological electric signal to be detected is identified, which is specifically:
inputting the physiological electric signal to be detected into a nucleus identification model to obtain the nucleus type.
Further, the nucleus identification model is obtained based on the following steps:
extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set to obtain a second characteristic vector, marking a nucleus group type label corresponding to each segment of the physiological electric signal, and combining the second characteristic vector and the nucleus group type label to obtain a second sample training set;
And importing the second sample training set into a second classifier to perform training optimization to obtain a nucleus identification model.
Further, extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set specifically includes: and extracting at least one of absolute medium bit difference, L2 norm, L infinity norm, fraction dimension, spectrum norm, power spectrum density and spectrum entropy of the physiological electric signal of each section.
Further, the first classifier is a machine learning method-based classification model or a neural network-based classification model, and the second classifier is a machine learning method-based classification model or a neural network-based classification model.
Further, the needle tract deviation azimuth of the physiological electric signal to be measured is predicted by combining the nucleus group type of the physiological electric signal to be measured and the needle tract deviation prediction model, and the method specifically comprises the following steps:
counting the time length and the sequence of occurrence of various nuclear clusters in the physiological electric signal to be detected to obtain a first feature vector of the physiological electric signal to be detected, and inputting the first feature vector of the physiological electric signal to be detected into the needle track deviation prediction model to obtain the needle track deviation azimuth.
The present invention also provides a computer storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of:
Collecting physiological electric signals to obtain a physiological electric signal sample set, counting the time length and sequence of occurrence of various nuclear clusters in each section of the physiological electric signals to obtain a first feature vector, marking a needle track deviation type label of each section of the physiological electric signals, and combining the first feature vector and the needle track deviation type label to obtain a first sample training set; wherein, each section of the physiological electric signals is obtained by dividing the physiological electric signals in the physiological electric signal sample set into a plurality of sections according to the type of the nuclear cluster;
the first sample training set is led into a first classifier to carry out training optimization, and a needle track deviation prediction model is obtained;
and identifying the nucleus group type of the physiological electric signal to be detected, predicting the needle tract deviation azimuth of the physiological electric signal to be detected by combining the nucleus group type of the physiological electric signal to be detected and the needle tract deviation prediction model, and adjusting the needle tract according to the needle tract deviation azimuth.
Further, counting the time length and the sequence of occurrence of various nuclei in each section of the physiological electric signal to obtain a first feature vector, wherein the first feature vector specifically comprises:
counting the occurrence times of various kinds of nuclear clusters in the physiological electric signal, representing the occurrence time length of the various kinds of nuclear clusters according to the occurrence times of the various kinds of nuclear clusters, and arranging the occurrence times of the various kinds of nuclear clusters according to the occurrence sequence of the various kinds of nuclear clusters to obtain a first feature vector.
Further, the needle track deviation type label for marking each section of the physiological electric signal is specifically:
and respectively defining a deviation distance range between each needle track and the target needle track based on the deviation degree of each needle track in each needle track deviation direction to obtain a plurality of needle track deviation types, and marking the needle track deviation type label corresponding to each needle track deviation type label.
Further, before defining the range of the deviation distance from the target needle track based on the respective needle track deviation degree of each needle track deviation direction, respectively, a plurality of needle track deviation types are obtained, the method further comprises:
a plurality of lane departure directions are defined according to the target lane, and at least one lane departure degree is defined for each of the lane departure directions.
Further, the lane departure type tag corresponding to each of the lane departure type marks includes:
defining a label for each needle track deviation type, judging the needle track deviation type corresponding to each section of physiological electric signal according to the defined label, and marking the corresponding needle track deviation type label.
Further, defining a range of offset distances from the target lane based on the respective lane offset degrees of each lane offset direction, respectively, to obtain a plurality of lane offset types, further comprising:
Controlling a puncture needle to enter from a middle needle track of a needle track group, collecting physiological electric signals corresponding to the middle needle track, and determining a middle needle track deviation direction and a middle needle track deviation degree based on the physiological electric signals corresponding to the middle needle track; the needle track group comprises a middle needle track and non-middle needle tracks which are positioned in four different directions around the middle needle track and are parallel to the middle needle track;
determining whether the middle needle track is a target needle track based on the middle needle track deviation direction and the middle needle track deviation degree;
under the condition that the middle needle track is not the target needle track, controlling the puncture needle to enter from the non-middle needle track after being withdrawn from the middle needle track, collecting physiological electric signals corresponding to the non-middle needle track, and determining the deviation direction and the deviation degree of the non-middle needle track based on the physiological electric signals corresponding to the non-middle needle track;
a target stitch is determined from the non-intermediate stitch based on the non-intermediate stitch direction and the degree of non-intermediate stitch offset.
Further, the nucleus type of the physiological electric signal to be detected is identified, which is specifically:
inputting the physiological electric signal to be detected into a nucleus identification model to obtain the nucleus type.
Further, the nucleus identification model is obtained based on the following steps:
extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set to obtain a second characteristic vector, marking a nucleus group type label corresponding to each segment of the physiological electric signal, and combining the second characteristic vector and the nucleus group type label to obtain a second sample training set;
and importing the second sample training set into a second classifier to perform training optimization to obtain a nucleus identification model.
Further, extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set specifically includes: and extracting at least one of absolute medium bit difference, L2 norm, L infinity norm, fraction dimension, spectrum norm, power spectrum density and spectrum entropy of the physiological electric signal of each section.
Further, the first classifier is a machine learning method-based classification model or a neural network-based classification model, and the second classifier is a machine learning method-based classification model or a neural network-based classification model.
Further, the needle tract deviation azimuth of the physiological electric signal to be measured is predicted by combining the nucleus group type of the physiological electric signal to be measured and the needle tract deviation prediction model, and the method specifically comprises the following steps:
Counting the time length and the sequence of occurrence of various nuclear clusters in the physiological electric signal to be detected to obtain a first feature vector of the physiological electric signal to be detected, and inputting the first feature vector of the physiological electric signal to be detected into the needle track deviation prediction model to obtain the needle track deviation azimuth.
Compared with the prior art, the invention has the beneficial effects that: the invention provides a surgical needle tract adjusting and guiding device based on machine learning. The needle track deviation prediction model is obtained by training the first classifier, and the needle track deviation prediction model can output the direction of the current needle track, which is required to be adjusted, relative to the target needle track according to the occurrence sequence and the record length of various nuclear clusters which pass through the needle track successively. Therefore, the needle track deviation azimuth can be output only by inputting the physiological electric signal to be detected into the needle track deviation prediction model, so that a doctor can conveniently and correspondingly adjust the needle track, and guidance on needle track adjustment is realized. Solves the problem that the adjustment of the needle track of the DBS operation depends on the experience of doctors, particularly the surgeons with poor experience in the DBS operation can reasonably adjust the needle track of the puncture needle, reduce misjudgment caused by manual subjective judgment and improve the operation efficiency and success rate.
Drawings
FIG. 1 is a flow chart of an embodiment of a DBS surgical needle tract adjustment method provided by the present invention;
FIG. 2 is a schematic view of a track direction defining an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Example 1
As shown in fig. 1, embodiment 1 of the present invention provides a DBS surgical needle tract adjusting method, including the steps of:
s1, collecting physiological electric signals to obtain a physiological electric signal sample set, counting the occurrence time and sequence of various nuclear clusters in each physiological electric signal to obtain a first feature vector, marking a needle track deviation type label of each physiological electric signal, and combining the first feature vector and the needle track deviation type label to obtain a first sample training set;
s2, importing the first sample training set into a first classifier to perform training optimization to obtain a needle track deviation prediction model;
s3, identifying the nucleus group type of the physiological electric signal to be detected, predicting the needle tract deviation azimuth of the physiological electric signal to be detected by combining the nucleus group type of the physiological electric signal to be detected and the needle tract deviation prediction model, and adjusting the needle tract according to the needle tract deviation azimuth.
Specifically, each physiological electric signal refers to each segment of physiological electric signal, namely one physiological electric signal is divided into a plurality of segments according to the type of the nucleus, and then classification characteristic signals of each segment of physiological electric signal are respectively extracted, so that different characteristics of physiological electric signals corresponding to different types of the nucleus can be learned through training, and classification identification is realized.
Aiming at the current situation that the DBS operation needle tract adjustment has no effective guiding scheme, the embodiment of the invention provides a method for giving an adjusting direction for DBS needle tract adjustment by training a classifier, and the DBS needle tract is adjusted by assisting a computer. Specifically, training a first classifier, counting the sequence and time length of occurrence of the nuggets which pass through the needle track sequentially, taking the sequence and time length of occurrence of the nuggets as input, taking a needle track deviation type label as output, and performing training optimization on the first classifier to obtain a needle track deviation prediction model capable of predicting the deviation direction and degree of the current needle track relative to the target needle track according to the sequence and time of occurrence of the nuggets. And then, the nucleus type of the physiological electric signal to be detected is identified, and the needle track deviation direction is predicted by combining the nucleus type of the physiological electric signal to be detected and the needle track deviation prediction model, so that a neurosurgeon can be effectively guided, especially a surgeon with inexperienced experience in DBS operation can reasonably adjust the needle track of the puncture needle, misjudgment caused by manual subjective judgment is reduced, and the operation efficiency is improved. Based on the needle track deviation prediction model obtained through training, real-time and automatic prediction of the needle track deviation direction can be realized, and the manual judgment is not relied on.
Aiming at the pain point that the needle track adjustment is very dependent on the personal clinical experience of doctors, the invention relates to a mode of calculating a needle track adjustment scheme through a model obtained through training, thereby effectively improving the accuracy and repeatability of needle track adjustment, and simultaneously being very convenient for popularizing the scheme to hospitals in three-four-wire cities, so that the scheme is free from dependence on doctors with abundant experience, and the universality of operation is improved.
When training the first class, the samples need to be processed, and the time and the sequence of appearance of various nuclei in each physiological electric signal need to be counted first.
Preferably, counting the time and sequence of occurrence of each kind of nuclei in each physiological electric signal to obtain a first feature vector, specifically:
counting the occurrence times of various kinds of nuclear clusters in the physiological electric signal, representing the occurrence time of the various kinds of nuclear clusters according to the occurrence times of the various kinds of nuclear clusters, and arranging the occurrence times of the various kinds of nuclear clusters according to the occurrence sequence of the various kinds of nuclear clusters to obtain a first feature vector.
The puncture needle is advanced step by step to record the physiological electric signal after entering the target measuring range, the puncture needle depth corresponding to each step is the same, the physiological electric signal is sampled once by step, and then the step is continued, the step is completed manually by a surgeon, and the embodiment is to calculate one quantization of the appearance time of the puncture needle in each nucleus when the puncture needle is pushed at a constant speed, so that the appearance time of various nuclei can be replaced by the record times of the same physiological electric signal of the nuclei, and the record times multiplied by the step time interval is equal to the appearance time. For example, still taking GPi kernel as target kernel, a first eigenvector V is set 1 ＝(T GPi ,T GPe ,T imgip ,T tract ,T OT ) If physiological electric signals corresponding to GPi nucleus, GPe nucleus, imgip nucleus, track nucleus and OT nucleus recorded in the target spot measuring area are respectively 1, 2, 1, 0 and 0 times, a corresponding first eigenvector V can be constructed 1 ＝(1,2,1,0,0)。
Secondly, the needle track deviation type of the physiological electric signal needs to be divided and used as output.
Preferably, the trace deviation type label for marking each physiological electric signal is specifically:
defining a plurality of needle track deviation directions according to a target needle track, defining at least one needle track deviation degree for each needle track deviation direction, respectively defining a deviation distance range between each needle track deviation degree of each needle track deviation direction and the target needle track to obtain a plurality of needle track deviation types, defining a label for each needle track deviation type, judging the needle track deviation type corresponding to each physiological electric signal according to the definition, and marking the corresponding needle track deviation type label.
The definition of the needle track deviation direction can be set according to the requirement. Specifically, before defining the lane departure direction, a coordinate system is first defined, and then a plurality of lane departure directions are defined in the coordinate system. In this embodiment, four needle track deviation directions are defined, namely, front, back, inner and outer. As shown in fig. 2, a coordinate system is first defined, and the positive X-axis direction is defined as the direction from the Right side (Right) of the patient to the Left side (Left) of the patient; the positive Y-axis direction is the direction from the patient's Anterior chest (Anterior) to the patient's Posterior (Posterior); the positive Z-axis direction is directed from the sole (Foot) of the patient to the top of the Head (Head) of the patient. The target track is a track deployed in a preoperative plan for the purpose of inserting the puncture needle into the target nucleus, and in this embodiment, the target track is set to an X-axis positive direction, "forward" direction is defined as a Y-axis negative direction, "backward" direction is defined as a Y-axis positive direction, "in" direction is defined as a Z-axis negative direction, and "out" direction is defined as a Z-axis positive direction.
Each lane departure direction may define one or more lane departure levels, depending on the requirements, e.g., the higher the fine requirements for adjustment, the more lane departure levels are set. In this embodiment, two needle track deviation degrees, that is, a slight deviation and a severe deviation, are defined for each needle track deviation direction, respectively. The deviation distance range of each needle track deviation type is also set according to the requirement. In general, the shift of the nucleus is counted in millimeters, and thus, a slight deviation is defined in this embodiment as within 5mm of the target needle tract, and a severe deviation is defined as more than 5mm of the target needle tract. The four needle track deviation directions, two needle track deviation degrees of each needle track deviation direction, and simultaneously adding the target needle track, finally obtaining nine needle track deviation types, wherein the nine needle track deviation types are respectively represented as a target needle track, a slightly front needle track, a slightly rear needle track, a slightly inner needle track, a slightly outer needle track and a severely outer needle track. The needle tract is slightly forward of the target needle tract by within 5mm of the target needle tract, the needle tract is slightly forward of the target needle tract by more than 5mm of the target needle tract, the needle tract is slightly rearward of the target needle tract by more than 5mm of the target needle tract after the needle tract is slightly forward of the target needle tract, the needle tract is slightly inward of the target needle tract by more than 5mm of the target needle tract, the needle tract is slightly outward of the target needle tract by more than 5mm of the target needle tract, and the needle tract is slightly outward of the target needle tract by more than 5mm of the target needle tract.
The labels corresponding to each needle track deviation type can be freely set, and the labels are set to be mutually different. For example, in this embodiment, the target track label is "1", the track slightly forward label is "2", the track slightly forward label is "3", the track slightly backward label is "4", the track slightly backward label is "5", the track slightly inward label is "6", the track slightly inward label is "7", the track slightly outward label is "8", and the track slightly outward label is "9", that is, the classification label vector L of the track deviation type 1 ＝1,2,…,9。
In actual operation, the surgical instrument can simultaneously arrange the middle needle track and the front, back, left and right 4 needle tracks parallel to the middle needle track, the puncture needle firstly enters from the middle needle track, the physiological electric signals are collected to judge the deviation direction and degree, if the needle track is a target needle track, the current middle needle track is determined to be a final electrode implantation channel, otherwise, the needle track exits from the middle needle track, one of the 4 parallel needle tracks is selected to enter according to the judged direction and degree, or the 5 needle tracks are rearranged until the target needle track is found.
First eigenvector V extracted from each physiological electric signal 1 Marked lane departure type label L 1 Obtaining a first sample training set:
{(V 11 ,L 11 ),(V 12 ,L 12 ),…,(V 1n ,L 1n )}
wherein V is 1a A first eigenvector representing the a-th physiological electrical signal sample, L 1a A lane departure type label representing the a-th physiological electrical signal sample, a=1, 2, …, n, n being the number of training samples;
and importing the first sample training set into an SVM model, training and optimizing a first classifier, classifying the unknown sample by using the classifier, outputting a classification label, and predicting the direction of needle track deviation.
Preferably, the nucleus type of the physiological electric signal to be detected is identified, specifically:
extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set to obtain a second characteristic vector, marking a nucleus group type label corresponding to each physiological electric signal, and combining the second characteristic vector and the nucleus group type label to obtain a second sample training set;
introducing the second sample training set into a second classifier for training and optimizing to obtain a nucleus identification model;
inputting the physiological electric signal to be detected into the nucleus identification model to obtain the nucleus type.
According to the method, an effective extraction scheme for the characteristics of the physiological electric signals is designed aiming at the mode that the physiological electric signals need to be manually distinguished based on the physiological electric signals in the current DBS operation, the second classifier is trained to classify the physiological electric signals, classification characteristic information of the physiological electric signals is extracted by analyzing the physiological electric signals recorded by the microelectrodes when the needle passes through, classification characteristic information is taken as input, a nucleus type label is taken as output, and the second classifier is trained and optimized to obtain a nucleus identification model capable of determining the corresponding nucleus of the physiological electric signals according to the classification characteristic information. The nucleus identification model obtained through the training of the second classifier can automatically judge the nerve nucleus, and the nerve nucleus sequence required by needle track deviation prediction is generated, so that the risk of incorrect identification of the nucleus type caused by manual misjudgment is reduced.
Preferably, extracting the plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set specifically includes: at least one of an absolute mid-range difference, an L2 norm, an L-infinity norm, a fractional dimension, a spectral norm, a power spectral density and a spectral entropy of each of the physiological electrical signals is extracted.
The embodiment researches the characteristics of the physiological electrical signals of the brain, and selects the seven physiological electrical signal characteristics which are effective in improving the classification performance at the same time, so as to classify the physiological electrical signals into corresponding nuclei.
Preferably, the extracting the absolute intermediate potential difference of the physiological electric signal specifically includes:
F1＝median(|X|-median(X))
wherein F1 is the absolute median difference of the physiological signal X, and medium () represents the median function;
the L2 norm of the physiological electric signal is extracted specifically as follows:
wherein F2 is L2 norm, x i The signal value of the physiological electric signal X at the sampling moment i is generated, and the sampling times of the physiological electric signal X are N;
the L-infinity norm of the physiological electric signal is extracted specifically as follows:
F3＝max|x i |
wherein F3 is L-infinity norm, x i I=1, 2, …, N is the signal value of the physiological electrical signal X at the i sampling time, and N is the sampling number of the physiological electrical signal X;
the fraction dimension of the physiological electric signal is extracted specifically as follows:
dividing the physiological electric signals X into k groups to obtain k groups of signal value vectors:
Wherein, the liquid crystal display device comprises a liquid crystal display device,for the m-th set of signal value vectors, int () represents a rounding function, N represents the number of samples of the physiological signal x, x [ m+jk ]]A signal value representing the physiological electrical signal X at the m+jk sampling instant, j=0, 1, …, int (N-m)/k; SLOPE { } is a SLOPE function of the fitted line of the point-to-group,
F4＝SLOPE{ln[L(k)],ln[L(1/k)]}
wherein F4 is the fractional dimension of physiological signal X, { ln [ L (k) ]],ln[L(1/k)]One set of point pairs, SLOPE { ln [ L (k)],ln[L(1/k)]The { ln [ L (k) } represents a point pair group],ln[L(1/k)]The slope of the fitted straight line of },
the extraction of the spectrum norms of the physiological electric signals is specifically as follows:
F5＝(maximum eigenvalue of X H X) 1/2
wherein F5 represents a spectrum norm, maximum eigenvalue of represents a maximum characteristic value, X H Representing the conjugate transpose of the physiological electrical signal X;
the power spectrum density of the physiological electric signal is extracted specifically as follows:
wherein F6 is the power spectral density, T represents the duration of the physiological electric signal x, and x (T) represents the signal value of the physiological electric signal x at the time of T sampling;
the spectral entropy of the physiological electric signal is extracted specifically as follows:
wherein F7 is spectral entropy, P (m) is the probability distribution value of physiological signal power spectrum on the frequency component of m,s (i) is a probability distribution value of a physiological electric signal power spectrum on a frequency component of i, i respectively takes each frequency component, S (m) is a physiological electric signal power spectrum, and S (m) = |X (m) | 2 X (m) is a discrete fourier transform of a discrete time signal of the physiological electrical signal X.
Extracting the seven classification characteristic information from each physiological electric signal sample to generate a second characteristic vector V 2 ＝(F1,F2,F3,F4,F5,F6,F7)。
Preferably, the labeling of the nucleolus type label corresponding to each physiological electric signal is specifically:
labeling a nucleolus type label corresponding to each nucleolus on a physiological electric signal generated by the nucleolus.
Marking a nucleus group type label corresponding to the physiological electric signal as L 2 =l, defining corresponding nucleolus type tags for the target nucleolus and its surrounding neighboring nucleoli according to the target nucleolus to which DBS needs to act. Taking the example of a general target nucleus globus pallidus inner nucleus GPi, nuclei that may appear in the vicinity thereof include: the outer nucleus GPe, the inner marrow plate imgip, the conducting bundle, the visual bundle OT of the globus pallidus can set the value range of the nucleus type label value as 1,2,3,4 and 5, which correspond to GPi, GPe, imgip, tract, OT five kinds of nucleus respectively. If the nucleus corresponding to the physiological electric signal is the globus pallidus inner side nucleus GPi, marking a nucleus type label '1' for the corresponding physiological electric signal; if the nucleus corresponding to the physiological electric signal is the globus pallidus outside nucleus GPe, marking a nucleus type label '2' for the corresponding physiological electric signal; if the nucleus corresponding to the physiological electric signal is an intramedullary plate imgip, marking a nucleus type label '3' for the corresponding physiological electric signal; if the nucleus corresponding to the physiological electric signal is a conductive bundle, marking a nucleus type label '4' for the corresponding physiological electric signal; if the nucleus corresponding to the physiological electric signal is the apparent beam OT, marking the corresponding physiological electric signal with a nucleus type label of '5'.
Second eigenvector V extracted from each physiological electric signal 2 Labeled nucleocapsid type tag L 2 Obtaining a second sample training set:
{(V 21 ,L 21 ),(V 22 ,L 22 ),…,(V 2n ,L 2n )}
wherein V is 2a A second eigenvector representing the a-th physiological electrical signal sample, L 2a A nucleus type label representing the a-th physiological electrical signal sample, a=1, 2, …, n, n being the number of training samples;
and importing the second sample training set into an SVM model, training and optimizing a second classifier, classifying the unknown physiological electric signals by using the classifier, and outputting a classified nucleus type label for identifying the nucleus corresponding to the physiological electric signals to be detected.
Preferably, the first classifier is a classification model based on a machine learning method or a classification model based on a neural network, and the second classifier is a classification model based on a machine learning method or a classification model based on a neural network.
Specifically, the machine learning method includes: support Vector Machines (SVM), adaboost, decision trees, random forests, nearest neighbor methods, K-means clustering. The deep learning method comprises the following steps: a fully connected classification network, a neural network. The classifier adopted in the embodiment is an existing relatively mature SVM classifier, and the SVM classifier has relatively excellent classifying generalization capability, so that the classifying precision and the classifying efficiency of unknown physiological electric signals can be effectively improved.
Preferably, the method predicts the needle tract deviation azimuth of the physiological electric signal to be measured by combining the nucleus group type of the physiological electric signal to be measured and the needle tract deviation prediction model, and specifically comprises the following steps:
counting the occurrence time and sequence of various nuclear clusters in the physiological electric signal to be detected to obtain a first feature vector of the physiological electric signal to be detected, and inputting the first feature vector of the physiological electric signal to be detected into the needle track deviation prediction model to obtain the needle track deviation azimuth.
The physiological electric signal to be measured refers to the physiological electric signal obtained by measuring the puncture needle in the measuring range of the target spot. Specifically, the puncture needle advances along the needle track planned before the operation, whether the puncture needle enters a measuring range near the target point is judged, if not, the puncture needle continues to advance along the needle track planned before the operation, if yes, the puncture needle is controlled to advance step by step, physiological electric signals of each step of the puncture needle are recorded, discharge nuclei corresponding to the recorded physiological electric signals are identified according to a training nuclei identification model, the sequence and duration of various discharge nuclei are recorded, whether the puncture needle enters a position outside the measuring range of the target point is judged, if yes, the sequence and duration of various discharge nuclei are counted, the needle track deviation direction of the puncture needle is predicted according to the needle track deviation prediction model, needle track adjustment suggestions are output, and doctors can adjust according to the needle track adjustment suggestions.
Example 2
The embodiment 2 of the invention provides a DBS surgical needle tract adjusting system, which comprises a signal acquisition and processing module, a first classification module and a prediction indication module;
the signal acquisition processing module is used for collecting physiological electric signals to obtain a physiological electric signal sample set, counting the occurrence time and sequence of various nuclear clusters in each physiological electric signal to obtain a first feature vector, marking a needle track deviation type label of each physiological electric signal, and combining the first feature vector and the needle track deviation type label to obtain a first sample training set;
the first classification module is used for guiding the first sample training set into a first classifier for training and optimizing to obtain a needle track deviation prediction model;
the prediction indication module is used for identifying the nucleus group type of the physiological electric signal to be detected, predicting the needle tract deviation azimuth of the physiological electric signal to be detected by combining the nucleus group type of the physiological electric signal to be detected and the needle tract deviation prediction model, and adjusting the needle tract according to the needle tract deviation azimuth.
Preferably, the system further comprises a second classification module;
the signal acquisition processing module is also used for extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set to obtain a second characteristic vector, marking a nucleus group type label corresponding to each physiological electric signal, and combining the second characteristic vector and the nucleus group type label to obtain a second sample training set;
The second classification module is used for guiding the second sample training set into a second classifier to perform training optimization so as to obtain a nucleus identification model;
the prediction indication module is also used for inputting the physiological electric signal to be detected into the nucleus identification model to obtain the nucleus type of the nucleus identification model.
The DBS surgical needle tract adjusting system provided by the embodiment of the invention is used for realizing the DBS surgical needle tract adjusting method, so that the DBS surgical needle tract adjusting system has the technical effects of the DBS surgical needle tract adjusting method, and is not described in detail herein.
Example 3
Embodiment 3 of the present invention provides a DBS surgical needle tract adjusting apparatus, including a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the DBS surgical needle tract adjusting method provided in embodiment 1 is implemented.
The DBS surgical needle tract adjusting device provided by the embodiment of the invention is used for realizing the DBS surgical needle tract adjusting method, so that the DBS surgical needle tract adjusting device has the technical effects of the DBS surgical needle tract adjusting method, and is not described in detail herein.
Example 4
Embodiment 4 of the present invention provides a computer storage medium having stored thereon a computer program which, when executed by a processor, implements the DBS surgical needle tract adjustment method provided in embodiment 1.
The computer storage medium provided by the embodiment of the invention is used for realizing the DBS operation needle track adjusting method, so that the technical effects of the DBS operation needle track adjusting method are achieved, and the computer storage medium is also provided and is not described herein.
The above-described embodiments of the present invention do not limit the scope of the present invention. Any other corresponding changes and modifications made in accordance with the technical idea of the present invention shall be included in the scope of the claims of the present invention.
Claims (22)
1. The DBS surgical needle tract adjusting device is characterized by comprising a processor and a memory, wherein the memory is stored with a computer program, and the computer program realizes the following steps when being executed by the processor:
collecting physiological electric signals to obtain a physiological electric signal sample set, counting the time length and sequence of occurrence of various nuclear clusters in each section of the physiological electric signals to obtain a first feature vector, marking a needle track deviation type label of each section of the physiological electric signals, and combining the first feature vector and the needle track deviation type label to obtain a first sample training set; wherein, each section of the physiological electric signals is obtained by dividing the physiological electric signals in the physiological electric signal sample set into a plurality of sections according to the type of the nuclear cluster;
The first sample training set is led into a first classifier to carry out training optimization, and a needle track deviation prediction model is obtained;
and identifying the nucleus group type of the physiological electric signal to be detected, predicting the needle tract deviation azimuth of the physiological electric signal to be detected by combining the nucleus group type of the physiological electric signal to be detected and the needle tract deviation prediction model, and adjusting the needle tract according to the needle tract deviation azimuth.
2. The DBS surgical needle tract adjusting device according to claim 1, wherein a first feature vector is obtained by counting a time length and a sequence of occurrence of each type of nuclei in each segment of the physiological electrical signal, specifically:
counting the occurrence times of various kinds of nuclear clusters in the physiological electric signal, representing the occurrence time length of the various kinds of nuclear clusters according to the occurrence times of the various kinds of nuclear clusters, and arranging the occurrence times of the various kinds of nuclear clusters according to the occurrence sequence of the various kinds of nuclear clusters to obtain a first feature vector.
3. The DBS surgical needle tract adjustment device according to claim 1, characterized by a needle tract departure type tag marking each segment of the physiological electrical signal, in particular:
and respectively defining a deviation distance range between each needle track and the target needle track based on the deviation degree of each needle track in each needle track deviation direction to obtain a plurality of needle track deviation types, and marking the needle track deviation type label corresponding to each needle track deviation type label.
4. The DBS surgical needle track adjusting apparatus according to claim 3, further comprising, before defining the range of the deviation distance from the target needle track based on the respective needle track deviation degree of each needle track deviation direction, respectively:
a plurality of lane departure directions are defined according to the target lane, and at least one lane departure degree is defined for each of the lane departure directions.
5. The DBS surgical needle track adjustment device according to claim 3, wherein the needle track deviation type label corresponding to each of the needle track deviation type marks includes:
defining a label for each needle track deviation type, judging the needle track deviation type corresponding to each section of physiological electric signal according to the defined label, and marking the corresponding needle track deviation type label.
6. The DBS surgical needle track adjusting apparatus according to claim 3, wherein a range of deviation distances from the target needle track is defined in each needle track deviation degree based on each needle track deviation direction, respectively, to obtain a plurality of needle track deviation types, further comprising:
controlling a puncture needle to enter from a middle needle track of a needle track group, collecting physiological electric signals corresponding to the middle needle track, and determining a middle needle track deviation direction and a middle needle track deviation degree based on the physiological electric signals corresponding to the middle needle track; the needle track group comprises a middle needle track and non-middle needle tracks which are positioned in four different directions around the middle needle track and are parallel to the middle needle track;
Determining whether the middle needle track is a target needle track based on the middle needle track deviation direction and the middle needle track deviation degree;
under the condition that the middle needle track is not the target needle track, controlling the puncture needle to enter from the non-middle needle track after being withdrawn from the middle needle track, collecting physiological electric signals corresponding to the non-middle needle track, and determining the deviation direction and the deviation degree of the non-middle needle track based on the physiological electric signals corresponding to the non-middle needle track;
a target stitch is determined from the non-intermediate stitch based on the non-intermediate stitch direction and the degree of non-intermediate stitch offset.
7. The DBS surgical needle tract adjusting device according to claim 1, wherein the identifying of the nucleus type of the physiological electrical signal to be measured is specifically:
inputting the physiological electric signal to be detected into a nucleus identification model to obtain the nucleus type.
8. The DBS surgical needle tract adjustment device according to claim 7, wherein the nucleus identification model is derived based on the steps of:
extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set to obtain a second characteristic vector, marking a nucleus group type label corresponding to each segment of the physiological electric signal, and combining the second characteristic vector and the nucleus group type label to obtain a second sample training set;
And importing the second sample training set into a second classifier to perform training optimization to obtain a nucleus identification model.
9. The DBS surgical needle tract adjustment device according to claim 8, wherein extracting a plurality of classification characteristic information of each physiological electrical signal in the physiological electrical signal sample set, specifically comprises: and extracting at least one of absolute medium bit difference, L2 norm, L infinity norm, fraction dimension, spectrum norm, power spectrum density and spectrum entropy of the physiological electric signal of each section.
10. The DBS surgical needle tract adjustment device according to claim 8, wherein the first classifier is a machine learning method based classification model or a neural network based classification model, and the second classifier is a machine learning method based classification model or a neural network based classification model.
11. The DBS surgical needle tract adjusting apparatus according to claim 1, wherein the needle tract deviation azimuth of the physiological electric signal to be measured is predicted by combining the nucleus group type of the physiological electric signal to be measured and the needle tract deviation prediction model, specifically:
counting the time length and the sequence of occurrence of various nuclear clusters in the physiological electric signal to be detected to obtain a first feature vector of the physiological electric signal to be detected, and inputting the first feature vector of the physiological electric signal to be detected into the needle track deviation prediction model to obtain the needle track deviation azimuth.
12. A computer storage medium having a computer program stored thereon, the computer program, when executed by a processor, performing the steps of:
collecting physiological electric signals to obtain a physiological electric signal sample set, counting the time length and sequence of occurrence of various nuclear clusters in each section of the physiological electric signals to obtain a first feature vector, marking a needle track deviation type label of each section of the physiological electric signals, and combining the first feature vector and the needle track deviation type label to obtain a first sample training set; wherein, each section of the physiological electric signals is obtained by dividing the physiological electric signals in the physiological electric signal sample set into a plurality of sections according to the type of the nuclear cluster;
the first sample training set is led into a first classifier to carry out training optimization, and a needle track deviation prediction model is obtained;
and identifying the nucleus group type of the physiological electric signal to be detected, predicting the needle tract deviation azimuth of the physiological electric signal to be detected by combining the nucleus group type of the physiological electric signal to be detected and the needle tract deviation prediction model, and adjusting the needle tract according to the needle tract deviation azimuth.
13. The computer storage medium according to claim 12, wherein the time length and the sequence of occurrence of each type of nuclei in each segment of the physiological electrical signal are counted to obtain a first feature vector, specifically:
Counting the occurrence times of various kinds of nuclear clusters in the physiological electric signal, representing the occurrence time length of the various kinds of nuclear clusters according to the occurrence times of the various kinds of nuclear clusters, and arranging the occurrence times of the various kinds of nuclear clusters according to the occurrence sequence of the various kinds of nuclear clusters to obtain a first feature vector.
14. The computer storage medium of claim 12, wherein the lane departure type tag marking each of the segments of the physiological electrical signal is in particular:
and respectively defining a deviation distance range between each needle track and the target needle track based on the deviation degree of each needle track in each needle track deviation direction to obtain a plurality of needle track deviation types, and marking the needle track deviation type label corresponding to each needle track deviation type label.
15. The computer storage medium of claim 14, further comprising, before defining the range of departure distances from the target lane based on the respective lane departure degree for each lane departure direction, respectively:
a plurality of lane departure directions are defined according to the target lane, and at least one lane departure degree is defined for each of the lane departure directions.
16. The computer storage medium of claim 14, wherein the lane departure type label corresponding to each of the lane departure type flags comprises:
Defining a label for each needle track deviation type, judging the needle track deviation type corresponding to each section of physiological electric signal according to the defined label, and marking the corresponding needle track deviation type label.
17. The computer storage medium of claim 14, wherein a plurality of lane departure types are obtained in a range of departure distances from the target lane defined based on respective degrees of lane departure for each lane departure direction, respectively, further comprising:
controlling a puncture needle to enter from a middle needle track of a needle track group, collecting physiological electric signals corresponding to the middle needle track, and determining a middle needle track deviation direction and a middle needle track deviation degree based on the physiological electric signals corresponding to the middle needle track; the needle track group comprises a middle needle track and non-middle needle tracks which are positioned in four different directions around the middle needle track and are parallel to the middle needle track;
determining whether the middle needle track is a target needle track based on the middle needle track deviation direction and the middle needle track deviation degree;
under the condition that the middle needle track is not the target needle track, controlling the puncture needle to enter from the non-middle needle track after being withdrawn from the middle needle track, collecting physiological electric signals corresponding to the non-middle needle track, and determining the deviation direction and the deviation degree of the non-middle needle track based on the physiological electric signals corresponding to the non-middle needle track;
A target stitch is determined from the non-intermediate stitch based on the non-intermediate stitch direction and the degree of non-intermediate stitch offset.
18. The computer storage medium according to claim 12, wherein the identification of the type of nuclei of the physiological electrical signal to be measured is in particular:
inputting the physiological electric signal to be detected into a nucleus identification model to obtain the nucleus type.
19. The computer storage medium of claim 18, wherein the clique recognition model is derived based on the steps of:
extracting a plurality of classification characteristic information of each physiological electric signal in the physiological electric signal sample set to obtain a second characteristic vector, marking a nucleus group type label corresponding to each segment of the physiological electric signal, and combining the second characteristic vector and the nucleus group type label to obtain a second sample training set;
and importing the second sample training set into a second classifier to perform training optimization to obtain a nucleus identification model.
20. The computer storage medium of claim 19, wherein extracting the plurality of classification characteristic information for each physiological electrical signal in the physiological electrical signal sample set comprises: and extracting at least one of absolute medium bit difference, L2 norm, L infinity norm, fraction dimension, spectrum norm, power spectrum density and spectrum entropy of the physiological electric signal of each section.
21. The computer storage medium of claim 19, wherein the first classifier is a machine learning method based classification model or a neural network based classification model, and the second classifier is a machine learning method based classification model or a neural network based classification model.
22. The computer storage medium according to claim 12, wherein the needle tract deviation azimuth of the physiological electric signal to be measured is predicted by combining the nucleus group type of the physiological electric signal to be measured and the needle tract deviation prediction model, specifically:
counting the time length and the sequence of occurrence of various nuclear clusters in the physiological electric signal to be detected to obtain a first feature vector of the physiological electric signal to be detected, and inputting the first feature vector of the physiological electric signal to be detected into the needle track deviation prediction model to obtain the needle track deviation azimuth.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310815703.9A CN116850447A (en) | 2020-04-29 | 2020-04-29 | DBS surgical needle tract adjusting device and computer storage medium |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310815703.9A CN116850447A (en) | 2020-04-29 | 2020-04-29 | DBS surgical needle tract adjusting device and computer storage medium |
| CN202010356723.0A CN111569251B (en) | 2020-04-29 | 2020-04-29 | DBS surgical needle tract adjusting method, system and device |
Related Parent Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010356723.0A Division CN111569251B (en) | 2020-04-29 | 2020-04-29 | DBS surgical needle tract adjusting method, system and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116850447A true CN116850447A (en) | 2023-10-10 |
Family
ID=72115023
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010356723.0A Active CN111569251B (en) | 2020-04-29 | 2020-04-29 | DBS surgical needle tract adjusting method, system and device |
| CN202310815703.9A Pending CN116850447A (en) | 2020-04-29 | 2020-04-29 | DBS surgical needle tract adjusting device and computer storage medium |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010356723.0A Active CN111569251B (en) | 2020-04-29 | 2020-04-29 | DBS surgical needle tract adjusting method, system and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN111569251B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113797440B (en) * | 2021-09-27 | 2022-05-10 | 首都医科大学附属北京天坛医院 | Automatic implantation system of deep brain electrode based on image and electrophysiological real-time positioning |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010048613A2 (en) * | 2008-10-24 | 2010-04-29 | The General Hospital Corporation | System and method for dynamically configurable deep brain stimulation |
| US9008414B2 (en) * | 2011-10-04 | 2015-04-14 | Medtronic Navigation, Inc. | Method and apparatus for assisted trajectory planning |
| US9247895B2 (en) * | 2014-03-21 | 2016-02-02 | Pacesetter, Inc. | Systems and methods for performing deep brain stimulation |
| CN107847138B (en) * | 2015-05-10 | 2021-08-24 | 阿尔法欧米伽神经科技有限公司 | Automatic brain detector guiding system |
| CN109585018A (en) * | 2018-11-09 | 2019-04-05 | 青岛歌尔微电子研究院有限公司 | Information processing method, information processing unit and physiological detection equipment |
| CN110135454A (en) * | 2019-04-02 | 2019-08-16 | 成都真实维度科技有限公司 | A kind of deep learning model training method based on 3D faulted scanning pattern data set |
| US20210343017A1 (en) * | 2020-04-30 | 2021-11-04 | Medtronic, Inc. | Post operative implantation site monitoring and medical device performance |
-
2020
- 2020-04-29 CN CN202010356723.0A patent/CN111569251B/en active Active
- 2020-04-29 CN CN202310815703.9A patent/CN116850447A/en active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| CN111569251A (en) | 2020-08-25 |
| CN111569251B (en) | 2023-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bao et al. | A map of object space in primate inferotemporal cortex | |
| Jog et al. | Tetrode technology: advances in implantable hardware, neuroimaging, and data analysis techniques | |
| US9730642B2 (en) | Neuropsychological spatiotemporal pattern recognition | |
| Lewicki | A review of methods for spike sorting: the detection and classification of neural action potentials | |
| US11980427B2 (en) | Brain navigation methods and device | |
| Liu et al. | Functional organization of speed tuned neurons in visual area MT | |
| US8849727B2 (en) | Method and system for classifying brain signals in a BCI using a subject-specific model | |
| US10039507B2 (en) | Advanced intraoperative neural targeting system and method | |
| US20220151537A1 (en) | Brain navigation lead | |
| WO2016028888A1 (en) | System and method for annotating and analyzing eeg waveforms | |
| CN107669244A (en) | Epileptic abnormal discharge site positioning method and system based on EEG-fMRI | |
| Chelaru et al. | Spike source localization with tetrodes | |
| Khosravi et al. | Intraoperative localization of STN during DBS surgery using a data-driven model | |
| CN106485039A (en) | The construction method of Butut distinguished in a kind of Chinese brain language | |
| CN111569251B (en) | DBS surgical needle tract adjusting method, system and device | |
| EP3744242A1 (en) | Brain navigation lead | |
| CN110584600B (en) | High-frequency oscillation rhythm detection method based on intelligent algorithm optimized fuzzy clustering | |
| Cardona et al. | NEUROZONE: On-line recognition of brain structures in stereotactic surgery-application to Parkinson's disease | |
| Taghva | An automated navigation system for deep brain stimulator placement using hidden Markov models | |
| Xiao et al. | Automatic identification of sweet spots from MERs for electrodes implantation in STN-DBS | |
| Lu et al. | A Systematic Review of Neurophysiology-Based Localization Techniques Used in Deep Brain Stimulation Surgery of the Subthalamic Nucleus | |
| CN114748162A (en) | Path planning method and readable storage medium | |
| US20070167856A1 (en) | Microelectrode recording analysis and visualization for improved target localization | |
| CN111436923B (en) | Puncture path determination device, surgical navigation system, surgical navigation apparatus, and storage medium | |
| Lima et al. | Analysis and classification of microelectrode recordings in deep brain stimulation surgery |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |