Neurosurgery data processing method and system

Abstract

The invention provides a neurosurgery data processing method and system, and relates to a data processing technology, by constructing a medical image database, recorded patient data can be marked and classified in combination with indexes of multiple dimensions, and the classification basis can comprise indexes such as disease types, lesion positions, shapes, volumes and the like. Therefore, the data in the medical image database can be mapped to the virtual medical auxiliary tool, such as a virtual operation guide plate, the operation data matched with the lesion information of the current patient is recommended to a doctor through the mapped data, more accurate operation reference is provided, and the scientificity of operation decision making is improved.

Description

Technical Field

[0001] The present invention relates to data processing technology, and in particular to a neurosurgery data processing method and system. Background Art

[0002] In the field of neurosurgery, surgical precision plays a decisive role in the patient's treatment outcome and prognosis. With the rapid development of medical technology, medical imaging technologies such as CT and MRI provide doctors with a wealth of patient information. However, how to efficiently utilize this massive amount of data to provide accurate and reliable guidance for surgical operations has become a key issue.

[0003] While some methods exist for processing patient medical imaging data and clinical information, they still have numerous shortcomings. For one thing, the classification and management of patient data is crude, often relying solely on simple classification based on a single dimension. This makes it difficult to comprehensively and accurately label and categorize patient data, leading to inefficiencies and poor accuracy when searching for similar cases and referencing previous treatment experience. Furthermore, when using patient data for surgical guidance, doctors often rely solely on the current patient's data and their own experience, lacking comprehensive and accurate references.

[0004] Therefore, how to combine medical imaging data to provide comprehensive and accurate operational references has become an urgent problem that needs to be solved today. Summary of the Invention

[0005] The present invention provides a neurosurgery data processing method and system, which can provide comprehensive and accurate operation reference in combination with medical imaging data.

[0006] A first aspect of the present invention provides a neurosurgery data processing method, comprising: Build a medical imaging database, enter patient data, and label and classify it; Screening similar feature data according to the classification results, and determining evaluation data of the similar feature data according to the comparison results of the initial data and the treatment data of the similar feature data; Obtain spatial dimension variables of similar feature data corresponding to the current patient data, map the spatial dimension variables and evaluation data to the virtual medical assistive device, and generate operation reference data.

[0007] Optionally, in a possible implementation of the first aspect, classifying the patient data according to the following steps includes: Retrieve the classification list, determine the classification indicators in the classification list as target indicators in turn, and divide the patient data with the same target indicators into the same group according to the target indicators to obtain multiple data groups.

[0008] Optionally, in a possible implementation of the first aspect, screening similar feature data according to the classification result, and determining evaluation data of the similar feature data according to a comparison result between initial data and treatment data of the similar feature data includes: Determining that patient data corresponding to the same data group are similar feature data; The clearance dose is obtained by subtracting the residual dose from the initial dose, and the evaluation data of the corresponding patient data is obtained according to the ratio of the clearance dose to the initial dose, wherein the initial data includes the initial dose and the treatment data includes the residual dose.

[0009] Optionally, in a possible implementation of the first aspect, obtaining spatial dimension variables of similar feature data corresponding to current patient data, mapping the spatial dimension variables and evaluation data to a virtual medical assistive device, and generating operation reference data includes: Mapping the operation position and operation direction obtained by parsing the spatial dimension variables to the virtual medical assistive device, and binding the operation position and operation direction to corresponding evaluation data; Dividing the virtual medical assistive device into multiple reference areas based on the regional accuracy level, selecting the reference area with the largest point density as the target area, and determining its corresponding single attribute; Obtain the high-frequency position area with the largest point frequency and the high-frequency angle area with the largest angle frequency in the target area, determine the mean of the evaluation data corresponding to the high-frequency position area and the high-frequency angle area as the reference coefficient respectively, and obtain the operation reference data according to the high-frequency position area and the high-frequency angle area and their corresponding reference coefficients.

[0010] Optionally, in a possible implementation of the first aspect, obtaining a high-frequency location area with the maximum frequency of points within a target area of a single attribute by the following steps includes: Calculate the average distance of each operating position in the target area, and offset the average distance according to a preset ratio to obtain a judgment threshold; The operation positions with a partition spacing less than the judgment threshold are grouped into the same partition set, and the corresponding point frequency is obtained based on the ratio of the number of points in each partition set to the total number; The partition set with the largest point frequency is selected, and the center point corresponding to the partition set is determined as the center of the circle. The distance corresponding to the operation position farthest from the center of the circle is the radius length to generate a high-frequency position area.

[0011] Optionally, in a possible implementation of the first aspect, obtaining a high-frequency angle zone with a maximum angle frequency within a target zone of a single attribute by the following steps includes: Divide the angle range into multiple zones according to the angle accuracy level, count the number of points where the operation direction falls into each angle range, and calculate the angle frequency corresponding to each angle range based on the ratio of the number of points to the total number of points; The angle range with the largest angle frequency is determined as the high-frequency angle range.

[0012] Optionally, in a possible implementation of the first aspect, after selecting the reference area with the largest point density as the target area, the method further includes: Calculate the average evaluation difference between the reference area and the target area adjacent to the target area, merge the reference area with the target area whose average evaluation difference is less than the evaluation difference threshold, obtain the updated target area, and determine its corresponding merged attributes; For the target area with merged attributes, the operation positions and operation directions that meet the outlier conditions are screened out, and the center position of the remaining operation positions is determined as the reference position, and the center angle of the remaining operation directions is determined as the reference angle. The operation reference data is obtained according to the reference position and reference angle.

[0013] Optionally, in a possible implementation manner of the first aspect, screening out operation positions that meet an outlier condition includes: Dividing the target area of the merged attributes into multiple screening areas according to the outlier screening accuracy, and calculating the first quantity mean of the points in each of the screening areas; A first quantity difference between the number of points in each screening area and the first quantity mean is calculated, and the screening area whose first quantity difference is greater than the first threshold is determined to meet the outlier condition, and the corresponding operation position is screened out.

[0014] Optionally, in a possible implementation of the first aspect, screening out operation directions that meet an outlier condition includes: Divide the screening angle intervals into multiple intervals according to the outlier screening accuracy, and calculate the second quantity means corresponding to all the screening angle intervals; Calculate the second quantity difference between the number of points corresponding to each screening angle interval and the second quantity mean, determine that the screening angle interval whose second quantity difference is greater than the second threshold meets the outlier condition, and screen out its corresponding operation direction.

[0015] Optionally, in a possible implementation of the first aspect, the method further includes: Counting the history areas of all virtual medical assistive devices in the current patient data history, and overwriting the history areas to the current virtual medical assistive device; Determine the current area of the virtual medical assistive device, compare the historical area with the current area to obtain overlapping areas and non-overlapping areas, and extract the time tags of the historical areas corresponding to the overlapping areas; Operation reference data is generated based on the overlapping area, the non-overlapping area, and the time tag.

[0016] A second aspect of the present invention provides a neurosurgery data processing system, comprising: The data module is used to build a medical imaging database, enter patient data, and perform labeling and classification; An evaluation module, configured to screen similar feature data according to the classification results, and determine evaluation data of the similar feature data according to a comparison result between the initial data and the treatment data of the similar feature data; The reference module is used to obtain spatial dimension variables of similar feature data corresponding to the current patient data, map the spatial dimension variables and evaluation data to the virtual medical assistive device, and generate operation reference data.

[0017] The beneficial effects of the present invention are as follows: The present invention integrates the multi-dimensional data of patients by constructing a medical imaging-related database, realizes the systematic management and classification of neurosurgery-related data, provides comprehensive data support for surgical decision-making, and improves the accuracy and scientificity of surgery.

[0018] The present invention can objectively evaluate the effects of different surgical positions and directions, providing a quantitative basis for doctors to choose the best surgical plan, helping doctors to make more scientific and reasonable clinical decisions and improve the pertinence and effectiveness of treatment.

[0019] The present invention is based on data matching and guide plate mapping. By intuitively displaying evaluation data at different positions and directions on the guide plate, doctors can quickly obtain effective information and reduce surgical risks.

[0020] The present invention provides doctors with convenient surgical planning assistance through virtual surgical guides, improves the efficiency of surgical planning, and also facilitates doctors to adjust personalized surgical plans to meet the needs of different patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of the present invention; Figure 2 is a flowchart of a neurosurgery data processing method provided by an embodiment of the present invention; Figure 3 Schematic diagram of the structure of a neurosurgery data processing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0023] See also Figure 1 , is a schematic diagram of an application scenario provided by an embodiment of the present invention. By constructing a medical imaging database, the entered patient data can be labeled and classified using indicators from multiple dimensions. The classification criteria can include indicators such as disease type, lesion location, shape, and volume. This allows the data in the medical imaging database to be mapped to virtual medical aids, such as a virtual surgical guide. This mapped data can then be used to recommend surgical procedures appropriate for the patient's lesion information, providing doctors with more accurate operational references and improving the scientific nature of surgical decisions.

[0024] The medical imaging database is a database that stores various types of medical imaging data (such as CT, MRI, PET, and other different modalities) and related patient clinical information (such as medical records, examination reports, surgical records, etc.). The virtual medical aid is a virtual surgical guide.

[0025] See also Figure 2 , is a flow chart of a neurosurgery data processing method provided by an embodiment of the present invention, Figure 2 The execution subject of the method shown may be a software and / or hardware device. The execution subject of the present application may include but is not limited to at least one of the following: user equipment, network equipment, etc. Among them, the user equipment may include but is not limited to computers, smart phones, personal digital assistants (PDAs) and the electronic devices mentioned above. Network equipment may include but is not limited to a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers, wherein cloud computing is a type of distributed computing, a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. It includes steps S101 to S103, as follows: S101, build a medical imaging database, enter patient data and perform labeling and classification.

[0026] It is understandable that building a comprehensive and structured medical data storage and management platform can efficiently retrieve, analyze and utilize neurosurgery patient data, providing a solid data foundation for clinical decision-making and medical research.

[0027] When building a medical imaging database, you can record the patient's identity information (such as name, ID number, medical record number), demographic characteristics (age, gender, race), and basic health status (past medical history, allergy history, etc.); various attributes of medical images, including imaging modality (CT, MRI, PET, etc.), acquisition time, imaging device model, image resolution, and storage path of image files; as well as disease diagnosis results (such as brain tumor type and grade, location and amount of cerebral hemorrhage, etc.), diagnosis time, examination items and indicators based on which the diagnosis is based, etc., and you can set associated fields (such as patient ID) to ensure that the data can be accurately associated and integrated.

[0028] When entering patient data, both medical imaging data and clinical information can be entered. For medical imaging data, a data transmission interface can be used to directly import image files from imaging devices (such as CT machines and MRI scanners) into a designated storage location in the database, and the image attribute information can be automatically or manually filled in the image data table. Clinical information can be manually entered by medical staff or through data integration with other systems (such as electronic medical records (EMRs)). Patient records, examination reports, test results, and other information can be extracted and populated into corresponding database table fields.

[0029] When labeling patient data, neurosurgeons can use professional medical image annotation tools (such as 3D Slicer and ITK-SNAP) to accurately annotate lesions in medical images. For example, when annotating an MRI image of a brain tumor, the tumor's contours can be carefully delineated, differentiating its internal tissue components (such as solid tumor tissue, necrotic areas, and cystic areas), and potentially marking the relationship between the tumor and surrounding important neurovascular structures. For patient clinical information, key information can be annotated, such as disease severity, the presence of complications, and key features of the treatment plan (such as radiotherapy dose).

[0030] When classifying patient data, it can be classified by combining data from multiple dimensions, such as disease type, lesion location, shape, volume, etc. Through multi-level and multi-dimensional classification, it is possible to quickly locate data subsets with similar characteristics, and provide more accurate data support by referring to past treatment experience and effects.

[0031] In some embodiments, patient data may be categorized by: Retrieve the classification list, determine the classification indicators in the classification list as target indicators in turn, and divide the patient data with the same target indicators into the same group according to the target indicators to obtain multiple data groups.

[0032] A classification list is a predefined set of various classification metrics. Select a classification metric from the list one by one and set it as the key basis for data segmentation, or the target metric. This process is performed one by one, with data segmentation performed on one classification metric at a time. For example, if the classification list contains multiple classification metrics such as disease type, lesion location, shape, and volume, you can first select "Disease Type" as the target metric, and then repeat the same process with the other metrics.

[0033] Based on the determined target metric, all patient data are traversed and compared. Patient data with the same target metric value are grouped together. For example, if the target metric is "Disease Type," all patient data with "Cerebral Hemorrhage" are grouped together, all patient data with "Brain Tumor" are grouped together, and so on. This process is repeated until all categorical metrics in the classification list have been used as target metrics to complete the data partitioning.

[0034] As different classification metrics are applied sequentially, patient data is gradually broken down into distinct groups. Each group contains patient data with the same or similar characteristics. For example, after using classification metrics like "disease type," "lesion location," "lesion shape," and "lesion volume," multiple groups with similar characteristics may be generated.

[0035] By dividing according to different classification indicators, patient data with similar characteristics are concentrated in the same data group, so that when conducting treatment effect evaluation, surgical plan formulation and other analyses, the data of patients with similar conditions can be compared and referenced more accurately.

[0036] S102 , screening similar feature data according to the classification result, and determining evaluation data of the similar feature data according to a comparison result between the initial data and the treatment data of the similar feature data.

[0037] Based on the data classification results of step S101, patient data with similar characteristics (i.e., similar feature data) can be screened. For example, patient data with similar disease type, lesion location, shape, and volume can be grouped into the same data subset as similar feature data. By evaluating similar feature data, patient data with similar conditions can be analyzed together, and treatment effects can be quantified. This allows doctors to intuitively compare treatment outcomes for different patients with similar conditions, providing a clear quantitative basis for adjusting and optimizing treatment plans.

[0038] Initial data primarily refers to information about the patient's condition before treatment, such as the initial hematoma volume for patients with intracerebral hemorrhage, tumor size and location for patients with tumors, and the patient's initial symptoms and signs. Treatment data refers to information about the patient during and after treatment, such as the residual hematoma volume after treatment for patients with intracerebral hemorrhage, postoperative pathology reports for patients with tumors, the patient's recovery status, and the occurrence of complications. Evaluation data are numerical values used to reflect the effectiveness of treatment.

[0039] By comparing the initial data and treatment data, we can know the initial state of the disease and the changes after treatment, quantify the treatment effect of patients with similar conditions, and provide a rich data basis for evaluating treatment effects.

[0040] Based on the above embodiment, the specific implementation of step S102 may be: Determine that patient data corresponding to the same data group are similar feature data; subtract the residual dose from the initial dose to obtain a clearance dose, and obtain evaluation data of the corresponding patient data according to the ratio of the clearance dose to the initial dose, wherein the initial data includes the initial dose and the treatment data includes the residual dose.

[0041] It can be understood that the patient data in the same data group have similar features, and thus can be regarded as similar feature data.

[0042] The initial volume refers to the pre-treatment measurement of the patient's hematoma volume, typically determined through imaging studies such as CT scans, such as volume or weight. The residual volume is the remaining hematoma volume after treatment, as measured by imaging studies. The clearance volume is the initial volume minus the residual volume, reflecting the amount of hematoma removed or reduced after treatment. The ratio of the clearance volume to the initial volume provides a direct indicator of the treatment's hematoma-clearing effect, clearly demonstrating the therapeutic effect from a quantitative perspective.

[0043] S103, obtaining spatial dimension variables of similar feature data corresponding to the current patient data, mapping the spatial dimension variables and evaluation data to a virtual medical assistive device, and generating operation reference data.

[0044] Spatial variables primarily include position and orientation information during surgical procedures. For example, during brain tumor resection, preoperative CT imaging data is used to generate a 3D model of the brain using a 3D reconstruction algorithm. The tumor's location is accurately marked on the model, thereby determining the optimal position coordinates for surgical instruments to enter the brain. The operating direction, such as the angle between the puncture needle and a reference plane (such as the sagittal, coronal, or transverse plane) or reference axis (such as the central axis) during puncture surgery, can be derived through processing and analysis of medical images.

[0045] After obtaining the position and direction information of the surgical operation, a positioning structure that matches the operation position and direction can be precisely designed on the virtual surgical guide model, such as constructing a virtual positioning hole at a specific position and angle. At the same time, the evaluation data previously calculated by comparing the initial data and treatment data with similar feature data is bound to the corresponding position and direction on the virtual guide. In other words, different areas of the virtual guide are assigned corresponding treatment effect evaluation values based on the corresponding operation position and direction. These evaluation values can be identified with numbers or other visual methods, so that doctors can intuitively understand the possible treatment effects of performing operations in different positions and directions.

[0046] When generating reference data, the data mapped onto the virtual surgical guide can be combined to identify optimal surgical data (such as position and orientation data during the procedure) to provide recommendations to physicians. This helps doctors better understand key surgical areas and parameters, reducing subjectivity and variability in surgical procedures. Reference data refers to data that combines multiple aspects of surgical position, angle, and assessment data to provide detailed reference for surgical procedures.

[0047] Based on the above embodiment, the specific implementation of step S103 may be: The operation position and operation direction obtained by analyzing the spatial dimension variables are mapped to the virtual medical assistive device, and the operation position and operation direction are bound to corresponding evaluation data.

[0048] By integrating the spatial information of surgical operations and treatment effect evaluation data into virtual medical aids, it can provide intuitive and quantitative references to help doctors plan the best surgical path in advance and choose the appropriate operating angle and position, so as to improve the accuracy and success rate of actual surgery and reduce surgical risks.

[0049] The virtual medical assistive device is divided into a plurality of reference areas based on the regional accuracy level, the reference area with the largest point density is selected as the target area, and its corresponding single attribute is determined.

[0050] It is understandable that the reason for dividing multiple reference areas is to reasonably divide the virtual medical aids and screen out areas where surgical operations are more concentrated and may have better effects as key objects (i.e., target areas), so as to analyze the characteristics and laws of the area more deeply and provide more targeted guidance reference for surgical operations.

[0051] The regional accuracy level can be determined based on factors such as the surgical precision requirements, the design features of the virtual medical device, and clinical experience. The virtual medical device model is then divided into multiple reference zones based on this level. These reference zones can be small areas divided according to spatial position, angular range, and other factors. For example, a high-precision neurosurgery virtual guide may be divided into smaller, more precise reference zones to meet the high-precision requirements of the surgical position and direction; whereas for surgeries with relatively low precision requirements, the reference zones may be larger. In practical applications, the reference zones can be rectangular.

[0052] By counting the number of simulated surgical points within different reference areas (i.e., point density), the reference area with the highest point density is identified and designated as the target area. Point density reflects the frequency of operations performed on each reference area. Areas with high point density typically indicate that doctors have preferred these areas in past surgeries, potentially indicating better surgical outcomes or feasibility.

[0053] A corresponding single attribute is determined for the target area, which means that the target area has the characteristics of being independent and clear and not fused or merged with other areas.

[0054] Obtain the high-frequency position area with the largest point frequency and the high-frequency angle area with the largest angle frequency in the target area, determine the mean of the evaluation data corresponding to the high-frequency position area and the high-frequency angle area as the reference coefficient respectively, and obtain the operation reference data according to the high-frequency position area and the high-frequency angle area and their corresponding reference coefficients.

[0055] For a single-attribute target area, obtaining high-frequency position and angle areas can accurately determine the critical ranges of surgical position and angle during the surgical procedure, providing more accurate reference data. The high-frequency position area represents the area within the target area where surgical operation positions are most concentrated. The frequency of surgical operations within this area is high, and this may be a more ideal surgical operation position range. The high-frequency angle area represents the angle range within the target area where surgical operation directions are most concentrated. The likelihood of surgical operations within this angle range is high, and this may be a more appropriate surgical operation angle range.

[0056] Point frequency refers to how often an operating point appears within a specific location range within the target area. Angle frequency refers to how often an operating direction appears within a specific angle range within the target area.

[0057] By calculating the mean of the evaluation data, two reference coefficients are generated, reflecting the average level of expected treatment efficacy when performing procedures in the high-frequency position and angle zones, respectively. Based on the high-frequency position and angle zones and their corresponding reference coefficients, operation reference data is generated. This operation reference data integrates key surgical locations and angles, along with corresponding treatment efficacy references, providing comprehensive and specific references and guidance for physicians' surgical procedures.

[0058] In some embodiments, the high-frequency location area with the largest point frequency within a target area with a single attribute may be obtained by the following steps: Calculate the mean distance of each operating position in the target area, and offset the distance mean according to the preset ratio to obtain the judgment threshold; divide the operating positions with a spacing less than the judgment threshold into the same partition set, and obtain the corresponding point frequency according to the ratio of the number of points in each partition set to the total number; select the partition set with the largest point frequency, and determine the center point corresponding to the partition set as the center of the circle, and the spacing corresponding to the operating position farthest from the center of the circle as the radius length to generate a high-frequency position area.

[0059] Within the target area, there are multiple surgical locations, each of which can be represented by spatial coordinates. To measure the average distance between these locations, we need to calculate the distances between each pair and then find the average of these distances. For example, suppose there are three surgical locations, A, B, and C. First, calculate the distances between A and B, A and C, and B and C. Add these three distances and divide by 3. The result is the average distance between the three locations.

[0060] The preset ratio can be a value, such as 40%, set in advance based on factors such as the actual surgical requirements, the size of the target area, and the accuracy of the operation. The calculated mean distance is multiplied by this preset ratio to obtain the judgment threshold. The judgment threshold is used to determine whether the distance between the operating positions is sufficiently small.

[0061] The operation locations within the target area are compared pairwise, and the distance between them is calculated. If the distance between two operation locations is less than a judgment threshold, they are grouped into the same set. In this way, the operation locations within the target area are divided into several different sets. The distance between the operation locations in each set is relatively small, indicating that these locations are spatially close.

[0062] For each partitioned set, count the number of operating locations contained within it. This number is the point count for that set. Then, divide the number of points in each set by the total number of operating locations within the target area. The result is the point frequency corresponding to that partitioned set. The point frequency reflects the frequency of each partitioned set within the target area. A higher point frequency indicates a greater concentration of operating locations within that set.

[0063] After calculating the point frequencies of each partition set, compare these point frequencies and find the partition set with the highest point frequency. This set is the area with the most concentrated operation positions, because the operation positions appear the most frequently in this set.

[0064] For the selected partition with the highest frequency of points, calculate the center point of all its operation positions. This center point's coordinates can be obtained by averaging the coordinates of these positions, and this center point is used as the center of the circle. Next, find the operation position in the set farthest from the center of the circle, calculate the distance between this operation position and the center of the circle, and use this distance as the radius. A circular area is defined by the center and radius. This circular area is the high-frequency location area, representing the area with the highest concentration of operation positions within the target area.

[0065] Through the above steps, the area with the highest concentration of surgical locations within the target area, i.e., the high-frequency location area, can be accurately identified. This provides a more accurate reference for surgical operations, allowing doctors to more specifically select and operate within this area during surgery, thereby improving surgical precision and reducing ineffective operations in non-critical areas.

[0066] In some embodiments, the high-frequency angle region with the largest angle frequency within a target region of a single attribute may be obtained by the following steps: Divide the angle range into multiple zones according to the angle accuracy level, count the number of points where the operation direction falls into each angle range, and obtain the angle frequency corresponding to each angle range based on the ratio of the number of points to the total number of points; determine the angle range with the largest angle frequency as the high-frequency angle zone.

[0067] The angular accuracy level can be a pre-set standard that determines the degree of granularity of the angle division. For example, a higher angular accuracy level may divide the angular range into smaller angular intervals (such as 1° or 2°); a lower angular accuracy level may divide the angular range into larger angular intervals (such as 5° or 10°).

[0068] Within the target area, each surgical procedure has a corresponding operating direction, which can be expressed as an angle. For each procedure, the angle range within which the operating direction angle falls is determined, and the number of points in that angle range is accumulated. For example, if the operating direction angle is 12°, the number of points in the angle range of 10°-15° is increased by 1. By counting all operating directions, the number of points in each angle range can be obtained.

[0069] To calculate the angular frequency, divide the number of points in each angular range by the total number of points for all surgical directions within the target area. For example, if the number of points in the 10°-15° angular range is 20, and the total number of points for all surgical directions within the target area is 100, the angular frequency of this angular range is 0.2. After calculating the angular frequencies for each angular range, compare them. Find the angular range with the highest angular frequency. This angular range is the high-frequency angular range, indicating that surgical directions most frequently occur within this angular range during surgical procedures within the target area.

[0070] By dividing the angle range, counting the number of points, and calculating the angle frequency, we can accurately determine the angle range with the highest concentration of operating directions, namely the high-frequency angle zone. This provides doctors with a clear angle reference during surgery, allowing them to more accurately select the operating direction based on the information in the high-frequency angle zone, thereby improving the accuracy and success rate of the surgery.

[0071] In addition, after selecting the reference area with the largest point density as the target area, the following embodiments are also included: The average evaluation difference between the reference area and the target area adjacent to the target area is calculated, and the reference area with an average evaluation difference less than the evaluation difference threshold is merged with the target area to obtain the updated target area and determine its corresponding merged attributes.

[0072] It is understandable that in order to obtain more comprehensive information, the difference between adjacent reference areas and target areas can be evaluated, the scope of the target area can be reasonably expanded, and areas with similar treatment effects and characteristics can be integrated.

[0073] Specifically, for each reference area adjacent to the target area, the difference between its evaluation coefficient and the evaluation coefficient of the target area is calculated, and then these differences are added and divided by the number of adjacent reference areas to obtain the average evaluation difference. The evaluation difference threshold is a pre-set standard value used to determine whether the adjacent reference area is sufficiently similar to the target area. When the average evaluation difference between the adjacent reference area and the target area is less than the evaluation difference threshold, it means that the difference between the reference area and the target area in terms of treatment effect is small, and they can be merged to form a new area. The merge attribute refers to the characteristics of the fusion or merger of the target area with other areas.

[0074] For the target area with merged attributes, the operation positions and operation directions that meet the outlier conditions are screened out, and the center position of the remaining operation positions is determined as the reference position, and the center angle of the remaining operation directions is determined as the reference angle. The operation reference data is obtained according to the reference position and reference angle.

[0075] Outlier conditions are set to determine whether an operation position or direction is an outlier. If an operation position or direction meets the outlier condition, it will be screened out. For example, if an operation position is significantly far from other positions in the target area, or if the angle of a certain operation direction differs from the majority of operation directions by more than a certain range, these positions can be considered outliers.

[0076] For the remaining operating positions, calculate the average coordinates of these positions to determine a central position as a reference. For example, the horizontal coordinates of all remaining operating positions can be added together and divided by the number of positions to obtain the average horizontal coordinate. Similarly, the average vertical coordinate can be calculated to determine the coordinates of the central position. This central position represents the central tendency of the remaining operating positions and is representative.

[0077] For the remaining operation directions, their angles are summarized and counted, and the average value of the angles is calculated as the reference angle. The reference position and reference angle together constitute the operation reference data, which can provide more accurate reference data.

[0078] In some embodiments, the operating positions that meet the outlier condition can be screened out by the following steps: According to the outlier screening accuracy, the target area of the merged attributes is divided into multiple screening areas, and the first quantity mean of the points in each screening area is counted; the first quantity difference between the number of points in each screening area and the first quantity mean is calculated, and the screening area whose first quantity difference is greater than the first threshold is determined to meet the outlier condition, and its corresponding operation position is screened out.

[0079] Outlier screening accuracy is a pre-set criterion that determines the level of detail in the screening area division. For example, if the outlier screening accuracy is high, the target area may be divided into a larger number of smaller screening areas to more accurately capture outliers in the data. Conversely, if the outlier screening accuracy is low, the number of screening areas will be smaller and larger. This division can be done based on spatial location, such as dividing the target area into several rectangular areas on a plane.

[0080] For each divided screening area, the number of operating points contained therein is counted. The operating points can be puncture locations. Then, the number of points in all screening areas is added up and divided by the total number of screening areas to obtain the first quantity mean of the points in each screening area. For each screening area, the first quantity mean is subtracted from its own number of points to obtain the first quantity difference of the screening area. The first threshold is a preset value used to determine whether the screening area is an outlier. When the first quantity difference of a certain screening area is greater than the first threshold, it means that the number of points in the screening area is significantly different from the average level, there is an abnormality, and the outlier condition is met. For the screening areas that meet the outlier condition, all the operating positions contained therein are screened out from the data. By screening out the outlier operating positions, the outliers and interference factors in the data are removed, so that the remaining data can better reflect the true characteristics of the target area, thereby improving the quality and reliability of the data.

[0081] In some embodiments, the operation directions that meet the outlier condition can be screened out by the following steps: According to the outlier screening accuracy, multiple screening angle intervals are divided, and the second quantity mean corresponding to all screening angle intervals is counted; the second quantity difference between the number of points corresponding to each of the screening angle intervals and the second quantity mean is calculated, and it is determined that the screening angle interval whose second quantity difference is greater than the second threshold meets the outlier condition, and the corresponding operation direction is screened out.

[0082] Outlier screening accuracy is a preset precision that determines the level of detail used in the angle interval divisions. For example, if the outlier screening accuracy is high, the angle range may be divided into more and narrower screening angle intervals, such as 5° intervals. If the outlier screening accuracy is low, the number of intervals may be smaller, and the intervals may be larger, such as 10° intervals. This divides the entire angle range into multiple different screening angle intervals, allowing for more detailed analysis of the angle data of the operating direction.

[0083] For each divided screening angle interval, the number of points of the operation direction that falls into the interval is counted. Then, the number of points of all screening angle intervals is added up, and then divided by the total number of screening angle intervals to obtain the second quantity mean corresponding to all screening angle intervals. For each screening angle interval, the second quantity mean is subtracted from its own number of points to obtain the second quantity difference of the screening angle interval. The second threshold is a preset value used to determine whether the screening angle interval is an outlier. When the second quantity difference of a certain screening angle interval is greater than the second threshold, it means that the number of points in the interval is significantly different from the average level, there is an abnormality, and the outlier condition is met. For the screening angle interval that meets the outlier condition, all the operation directions contained therein are screened out from the data. By screening out the outlier operation directions, the outliers and interference factors in the angle data are removed, so that the remaining operation direction data can better reflect the actual operation angle distribution, thereby improving the quality and reliability of the data.

[0084] It's worth noting that different methods are used to determine operational reference data based on the different attributes of the target area because target areas with different attributes have different data characteristics and distributions. For a target area with a single attribute, the data comes from a specific region, its distribution is relatively concentrated and uniform, and the data volume is usually small. For example, puncture data collected from a single small lesion area may only show some characteristics unique to that region due to the limited scope, without being affected by other areas. Due to the limited data volume, it is difficult to accurately calculate specific and highly representative values, so a range is often determined to summarize the data characteristics. For target areas with merged attributes, it involves the merging of multiple regions, and the data sources are extensive and complex, and the data characteristics of different regions may vary. The distribution of the merged data may be more dispersed, including information from multiple different regions. Because the merging increases the data volume and enriches the information, it can present more obvious patterns and trends, thus meeting the conditions for calculating specific values. For example, after merging several adjacent areas with slightly different disease conditions or anatomical structures, a comprehensive analysis of the puncture position and direction data within the merged area and the elimination of outliers can be performed to calculate a representative center position as a specific reference position, and the center angle or the most frequently occurring specific angle can be calculated as a reference angle. These specific values can more accurately reflect the overall characteristics of the merged area.

[0085] The technical solution provided by the present invention also includes: The present invention collects statistics on the historical areas of all virtual medical aids in the current patient's data history and overlays these historical areas onto the current virtual medical aid. This historical area can be considered the surgical area where acupuncture was performed on this particular patient during historical surgical procedures. The operations performed on the surgical area can be interventional procedures such as ablation and hemorrhage. Therefore, the present invention retrieves the historical area and overlays it onto the current virtual medical aid.

[0086] Determine the current area of the virtual medical aid, compare the historical area with the current area to obtain overlapping areas and non-overlapping areas, and extract the time tag of the historical area corresponding to the overlapping area. The present invention will determine the current area of the operation reference number according to the above steps. The current area can be regarded as the area where the operation can be performed this time, so the present invention will compare the historical area with the current area to obtain overlapping areas and non-overlapping areas. The overlapping area can be regarded as a certain overlap between the current and previous areas, and the non-overlapping area can be regarded as no overlap between the current and previous areas. Since the body needs to recover after each acupuncture treatment, the longer the time, the better the recovery. Therefore, the present invention will extract the time tag of the historical area corresponding to the overlapping area.

[0087] The operation reference data is generated based on the overlapping area, non-overlapping area and time label. The present invention generates operation reference data based on the overlapping area, non-overlapping area and time label. If the overlapping area is empty, the current area of the previously determined operation reference number can be directly used as the final needle insertion area.

[0088] If the overlapping area is not empty, it is necessary to calculate the operational risk coefficient of each overlapping area and extract the time label and the number of time labels of each overlapping area.

[0089] The present invention first counts all time tags in each overlapping region and obtains the first time closest to the current time. The overlapping region may have been previously needled multiple times, so there may be multiple time tags, each corresponding to a different time. The difference between the first time at the time tag and the current time is calculated to obtain the first time period. This first time period is weighted based on the time period weight to obtain the first sub-coefficient. The number of time tags is weighted based on the number of times weight to obtain the second sub-coefficient.

[0090] If the first sub-coefficient is larger, it means that the overlapping area is closer to the current needle insertion time, and the relative risk will be higher. The present invention will use the preset value as the numerator and the first sub-coefficient to calculate the third sub-coefficient, and then add the third sub-coefficient to the second sub-coefficient to obtain the total risk coefficient. If the total risk coefficient is greater than the preset value, the present invention will add a first reminder message to the corresponding operation position, and based on the secondary selection of the screening area, obtain a screening area that was not previously selected after the secondary selection. This screening area is obtained from the above-mentioned target area, and a second reminder message is added to the target area. The staff can select the screening area based on the first and second reminder messages according to the actual situation.

[0091] See also Figure 3 , is a schematic structural diagram of a neurosurgery data processing system provided by an embodiment of the present invention, the neurosurgery data processing system comprising: The data module is used to build a medical imaging database, enter patient data, and perform labeling and classification; An evaluation module, configured to screen similar feature data according to the classification results, and determine evaluation data of the similar feature data according to a comparison result between the initial data and the treatment data of the similar feature data; The reference module is used to obtain spatial dimension variables of similar feature data corresponding to the current patient data, map the spatial dimension variables and evaluation data to the virtual medical assistive device, and generate operation reference data.

[0092] Figure 3 The apparatus of the embodiment shown can be used to perform Figure 2 The implementation principles and technical effects of the steps in the method embodiment shown are similar and will not be repeated here.

[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A neurosurgery data processing method, characterized in that: include: Build a medical imaging database, enter patient data, and label and classify it; Screening similar feature data according to the classification results, and determining evaluation data of the similar feature data according to the comparison results of the initial data and the treatment data of the similar feature data; Obtain spatial dimension variables of similar feature data corresponding to the current patient data, map the spatial dimension variables and evaluation data to the virtual medical assistive device, and generate operation reference data.

2. The method according to claim 1, characterized in that Screening similar feature data according to the classification results, and determining evaluation data of the similar feature data according to the comparison results of the initial data and the treatment data of the similar feature data, including: Determining that patient data corresponding to the same data group are similar feature data; The clearance dose is obtained by subtracting the residual dose from the initial dose, and the evaluation data of the corresponding patient data is obtained according to the ratio of the clearance dose to the initial dose, wherein the initial data includes the initial dose and the treatment data includes the residual dose.

3. The method according to claim 1, characterized in that Obtaining spatial dimension variables of similar feature data corresponding to the current patient data, mapping the spatial dimension variables and evaluation data to a virtual medical assistive device, and generating operation reference data, including: Mapping the operation position and operation direction obtained by parsing the spatial dimension variables to the virtual medical assistive device, and binding the operation position and operation direction to corresponding evaluation data; Dividing the virtual medical assistive device into multiple reference areas based on the regional accuracy level, selecting the reference area with the largest point density as the target area, and determining its corresponding single attribute; Obtain the high-frequency position area with the largest point frequency and the high-frequency angle area with the largest angle frequency in the target area, determine the mean of the evaluation data corresponding to the high-frequency position area and the high-frequency angle area as the reference coefficient respectively, and obtain the operation reference data according to the high-frequency position area and the high-frequency angle area and their corresponding reference coefficients.

4. The method according to claim 3, characterized in that The following steps are used to obtain the high-frequency location area with the highest point frequency within the target area of a single attribute, including: Calculate the average distance of each operating position in the target area, and offset the average distance according to a preset ratio to obtain a judgment threshold; The operation positions with a partition spacing less than the judgment threshold are grouped into the same partition set, and the corresponding point frequency is obtained based on the ratio of the number of points in each partition set to the total number; The partition set with the largest point frequency is selected, and the center point corresponding to the partition set is determined as the center of the circle. The distance corresponding to the operation position farthest from the center of the circle is the radius length to generate a high-frequency position area.

5. The method according to claim 3, characterized in that The high-frequency angle zone with the maximum angle frequency within the target zone of a single attribute is obtained by the following steps, including: Divide the angle range into multiple zones according to the angle accuracy level, count the number of points where the operation direction falls into each angle range, and calculate the angle frequency corresponding to each angle range based on the ratio of the number of points to the total number of points; The angle range with the largest angle frequency is determined as the high-frequency angle range.

6. The method according to claim 3, characterized in that After selecting the reference area with the largest point density as the target area, the following steps are also included: Calculate the average evaluation difference between the reference area and the target area adjacent to the target area, merge the reference area with the target area whose average evaluation difference is less than the evaluation difference threshold, obtain the updated target area, and determine its corresponding merged attributes; For the target area with merged attributes, the operation positions and operation directions that meet the outlier conditions are screened out, and the center position of the remaining operation positions is determined as the reference position, and the center angle of the remaining operation directions is determined as the reference angle. The operation reference data is obtained according to the reference position and reference angle.

7. The method according to claim 6, characterized in that Screen out the operating positions that meet the outlier conditions, including: Dividing the target area of the merged attributes into multiple screening areas according to the outlier screening accuracy, and calculating the first quantity mean of the points in each of the screening areas; A first quantity difference between the number of points in each screening area and the first quantity mean is calculated, and the screening area whose first quantity difference is greater than the first threshold is determined to meet the outlier condition, and the corresponding operation position is screened out.

8. The method according to claim 6, characterized in that Screen out the operation directions that meet the outlier conditions, including: Divide the screening angle intervals into multiple intervals according to the outlier screening accuracy, and calculate the second quantity means corresponding to all the screening angle intervals; Calculate the second quantity difference between the number of points corresponding to each screening angle interval and the second quantity mean, determine that the screening angle interval whose second quantity difference is greater than the second threshold meets the outlier condition, and screen out its corresponding operation direction.

9. The method according to claim 5, characterized in that Also includes: Counting the history areas of all virtual medical assistive devices in the current patient data history, and overwriting the history areas to the current virtual medical assistive device; Determine the current area of the virtual medical assistive device, compare the historical area with the current area to obtain overlapping areas and non-overlapping areas, and extract the time tags of the historical areas corresponding to the overlapping areas; Operation reference data is generated based on the overlapping area, the non-overlapping area, and the time tag.

10. A neurosurgery data processing system, characterized in that: include: The data module is used to build a medical imaging database, enter patient data, and perform labeling and classification; An evaluation module, configured to screen similar feature data according to the classification results, and determine evaluation data of the similar feature data according to a comparison result between the initial data and the treatment data of the similar feature data; The reference module is used to obtain spatial dimension variables of similar feature data corresponding to the current patient data, map the spatial dimension variables and evaluation data to the virtual medical assistive device, and generate operation reference data.

Images