A real-time processing system for force feedback data of a percutaneous puncture needle

Abstract

The present application relates to the technical field of multi-dimensional analysis, and in particular to a real-time processing system for force feedback data of a percutaneous puncture needle. The system comprises a data acquisition module for acquiring force data of all dimensions of the percutaneous puncture needle; a dimension influence analysis module for determining an abnormal fluctuation coefficient of a single dimension according to the uneven change rule of the force data of the single dimension, and determining an influence coefficient of the force data of each dimension in combination with the correlation of multi-dimensional force feedback; and a feedback transmission adjustment module for weighted adjustment of all dimensions when LOF anomaly detection is performed, so as to obtain a more accurate adjusted LOF value and perform a regulation and judgment on whether transmission feedback can be performed. The present application performs single-dimensional and multi-dimensional comparative analysis on multi-dimensional force data to measure the influence credibility of each dimension in participating in an anomaly, adjusts the LOF anomaly detection, and more accurately performs transmission feedback on real-time data.

Description

Technical Field

[0001] This invention relates to the field of multidimensional analysis technology, specifically to a real-time processing system for percutaneous puncture needle force feedback data. Background Technology

[0002] With the widespread use of percutaneous needle insertion techniques, accurate force feedback data is crucial for ensuring the safety and effectiveness of the procedure. Controlling the applied force during the puncture process ensures proper and accurate insertion, and force feedback helps physicians perceive the contact between the needle and tissue in real time. Furthermore, force data provides real-time operational information, helping to adjust the puncture angle and force, preventing misoperation, and improving puncture accuracy to guarantee the safety of the procedure.

[0003] Real-time processing of percutaneous puncture needle force feedback data typically involves continuously collecting force feedback data during the procedure and inputting it into the system to form a real-time data stream. The Loophole Optimization (LOF) algorithm is then used to assess anomalies in the data points based on their local density. Force feedback data marked as abnormal by the LOF algorithm is directly filtered out by the system to prevent inaccurate feedback from affecting subsequent work.

[0004] However, fluctuations can occur when the puncture needle comes into contact with different tissues or penetrates skin of different thicknesses. In addition, due to differences in the elasticity, hardness, and density of tissues, the puncture needle may be subjected to different force feedback fluctuations during the operation. These force feedback changes represent normal physiological responses, but they can also be misjudged as outliers by the LOF algorithm and identified as abnormal data, causing the system to erroneously discard some normal data, thereby affecting the accuracy of the feedback. Summary of the Invention

[0005] To address the problem in existing technologies where normal physiological force feedback changes are misidentified as outliers by the LOF algorithm and discarded as abnormal data, the present invention aims to provide a real-time processing system for percutaneous puncture needle force feedback data. The specific technical solution adopted is as follows:

[0006] This invention provides a real-time processing system for percutaneous puncture needle force feedback data, the system comprising:

[0007] The data acquisition module is used to acquire force data of the percutaneous puncture needle in different dimensions at each acquisition moment;

[0008] The dimensional impact analysis module is used to obtain the abnormal fluctuation coefficient of the force data of each dimension based on the rate of change of the force data of each dimension in the time series before the current moment; and to determine the impact coefficient of the force data of each dimension based on the correlation of the time series changes of the force data of each dimension with other dimensions and the influence of the abnormal fluctuation coefficient.

[0009] The feedback transmission adjustment module is used to adjust the force data of each dimension based on the influence coefficient when performing LOF algorithm detection on the force data at the current moment, so as to obtain the adjusted LOF value at the current moment; and to perform transmission feedback adjustment based on the adjusted LOF value at the current moment.

[0010] Furthermore, the method for obtaining the abnormal fluctuation coefficient includes:

[0011] For any dimension of force data, the initial dimension coefficient of the force data at the current moment is obtained based on the deviation between the degree of change of the force data in the current moment and the degree of change in the time series within the preset time series before the current moment.

[0012] Based on the vibration frequency of the force data in this dimension before the current moment, and combined with the initial dimension coefficient, the instability coefficient of the force data in this dimension at the current moment is obtained.

[0013] Based on the difference in instability coefficients between the current and previous time steps of the force data in this dimension, the abnormal fluctuation coefficient of the force data in this dimension is obtained.

[0014] Furthermore, the method for obtaining the initial dimension coefficients includes:

[0015] Within a preset time series range before the current moment, calculate the numerical difference of the force data in this dimension between every two adjacent sampling moments, and then use the mean of all numerical differences as the preceding mean variable for this dimension.

[0016] The difference between the force data of this dimension at the current moment and the previous sampling moment is taken as the current degree of change of this dimension;

[0017] The ratio of the current degree of change of this dimension to the mean of the preceding variables is used as the initial dimension coefficient of the force data of this dimension at the current moment.

[0018] Furthermore, the method for obtaining the instability coefficient includes:

[0019] Within a preset time series range before the current moment, the force data of the current dimension is subjected to Fourier transform to obtain the frequency domain space of that dimension; the frequency at the highest peak in the frequency domain space is normalized to obtain the high frequency anomaly of that dimension.

[0020] The product of the high-frequency anomaly of this dimension and the initial dimension coefficient is used as the instability coefficient of the force data of this dimension at the current moment.

[0021] Furthermore, obtaining the abnormal fluctuation coefficient of the force data in this dimension based on the difference in instability coefficients between the current time and previous time steps includes:

[0022] The product of the instability coefficient at the previous sampling time and the instability coefficient at the current time is used as the abnormal fluctuation coefficient.

[0023] Furthermore, the method for obtaining the influence coefficient includes:

[0024] For any one dimension of force data, the comparative usability of the force data in that dimension with the force data in each other dimension is obtained based on the degree of correlation between the time-series changes of the force data in that dimension and the force data in each other dimension.

[0025] By combining the comparability of the force data in this dimension with the force data in every other dimension, and the abnormal fluctuation coefficients of the force data in this dimension with the force data in other dimensions, the influence coefficient of the force data in this dimension is obtained.

[0026] Furthermore, the method for obtaining the comparative availability includes:

[0027] Within a predefined local range before the current moment, calculate the Pearson correlation coefficient between the force data of this dimension and the force data of each other dimension in the time series, and use the absolute value of the Pearson correlation coefficient as the correlation of each other dimension.

[0028] Calculate the correlation and sum of other dimensions for this dimension, and use this as the correlation sum of this dimension; use the ratio of the correlation of each other dimension to the correlation sum of this dimension as the usability of the force data for this dimension compared with the force data of each other dimension.

[0029] Furthermore, the step of combining the comparability of the force data in this dimension with the force data in every other dimension, and the abnormal fluctuation coefficients of the force data in this dimension with the force data in other dimensions, to obtain the influence coefficient of the force data in this dimension, includes:

[0030] Each of the other dimensions of the stress data in this dimension is taken as an analysis dimension in turn. The product of the abnormal fluctuation coefficient of the analysis dimension and the comparative availability is taken as the synchronization interference degree of the analysis dimension. The sum of the synchronization interference degrees of all other dimensions of the stress data in this dimension is taken as the relative influence degree of this dimension.

[0031] The product of the abnormal fluctuation coefficient and the relative influence of the force data in this dimension is normalized and used as the influence coefficient of the force data in this dimension.

[0032] Furthermore, the method for obtaining the adjusted LOF value includes:

[0033] Within the preset detection range before the current moment, obtain the difference distance between the force data in each dimension at every two moments;

[0034] The product of the difference distance between the force data in each dimension at every two time points and the influence coefficient is used as the dimensional distance between each two time points in each dimension; the adjustment distance between each two time points is obtained by combining the dimensional distances of all dimensions between each two time points.

[0035] Based on the adjustment distance between time points within the preset detection range before the current time, the LOF algorithm is used to obtain the adjusted LOF value at the current time.

[0036] Furthermore, the transmission feedback adjustment based on the adjusted LOF value at the current moment includes:

[0037] If the adjusted LOF value at the current moment is greater than the preset abnormal threshold, the force data at the current moment will not be transmitted or fed back.

[0038] The present invention has the following beneficial effects:

[0039] This invention addresses all dimensions of the current force data acquired via percutaneous needle puncture. First, based on the unevenness of force data changes in a single dimension, it preliminarily determines the abnormal fluctuation coefficient for that dimension in real-time data anomaly detection. Then, combining the correlation between multi-dimensional force feedback and the temporal correlation of force data across different dimensions, along with the abnormal fluctuation coefficient, it determines the influence coefficient of each dimension's force data on anomaly detection analysis of the current real-time data, thereby improving the accuracy of anomaly identification. Finally, during LOF anomaly detection, all dimensions are weighted and adjusted to obtain a more accurate adjusted LOF value, allowing for a controllable judgment on whether transmission feedback is possible. This invention performs single-dimensional and multi-dimensional comparative analysis of multi-dimensional force data to measure the credibility of each dimension's influence on anomalies, adjusting LOF anomaly detection for more accurate transmission feedback of real-time data. Attached Figure Description

[0040] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a structural diagram of a real-time processing system for percutaneous puncture needle force feedback data provided in one embodiment of the present invention;

[0042] Figure 2 A schematic diagram of a single-dimensional force data provided in an embodiment of the present invention;

[0043] Figure 3A flowchart illustrating a method for obtaining an abnormal fluctuation coefficient according to an embodiment of the present invention;

[0044] Figure 4 This is a flowchart illustrating a method for obtaining an influence coefficient according to an embodiment of the present invention. Detailed Implementation

[0045] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a real-time processing system for percutaneous puncture needle force feedback data proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] The following description, in conjunction with the accompanying drawings, details the specific scheme of the real-time processing system for percutaneous puncture needle force feedback provided by the present invention.

[0048] Please see Figure 1 The diagram shows a structural diagram of a real-time processing system for percutaneous puncture needle force feedback data according to an embodiment of the present invention. The system includes: a data acquisition module 101, a dimensional influence analysis module 102, and a feedback transmission adjustment module 103.

[0049] The data acquisition module 101 is used to acquire the force data of the percutaneous puncture needle in different dimensions at each acquisition moment.

[0050] In this embodiment of the invention, a 100Hz acquisition frequency is used to continuously and in real-time acquire the force data of the puncture needle during the puncture process. Each force data point contains multiple dimensions, including the magnitude, direction, and velocity of the force, etc. Dimensionality is removed from the data in each dimension to avoid the influence of dimensions. It should be noted that the acquisition process and dimension removal are techniques well-known to those skilled in the art. The acquisition frequency can be adjusted by the implementer, and dimension removal can employ Z-score standardization or normalization, etc., without further limitation or elaboration. Please refer to [link to relevant documentation]. Figure 2 The diagram illustrates a single-dimensional force data according to an embodiment of the present invention, showing the temporal changes of the data when the dimension is the magnitude of the force.

[0051] The dimensional influence analysis module 102 is used to obtain the abnormal fluctuation coefficient of the force data of each dimension based on the rate of change of the force data of each dimension in the time series before the current moment; and to determine the influence coefficient of the force data of each dimension based on the correlation of the time series changes of the force data of each dimension with other dimensions and the influence of the abnormal fluctuation coefficient.

[0052] In percutaneous needle force feedback data, abnormal data caused by physiological fluctuations are usually due to differences in the hardness and elasticity of different tissue layers, such as skin, fat, muscle, and blood vessels. These fluctuations are normal reactions when the needle comes into contact with the tissue and do not indicate any abnormality.

[0053] Human tissues are not completely broken; they possess a certain degree of elasticity and transition between different layers. Therefore, when a puncture needle penetrates from one layer of tissue to another, even if fluctuations occur, the abrupt changes in force data are not significant. For example, when the puncture needle moves from the skin into the fat layer and then into the muscle layer, the force data in each dimension does not change drastically but transitions gradually.

[0054] Therefore, by combining the rate of change before the time series, the degree of anomaly probability that a single dimension can be used for true anomaly detection is analyzed. Preferably, in this embodiment of the invention, the method for obtaining the anomaly fluctuation coefficient is described in [reference needed]. Figure 3 The diagram illustrates a flowchart of a method for obtaining an abnormal fluctuation coefficient according to an embodiment of the present invention. The method includes the following steps:

[0055] S201: For any dimension of force data, based on the deviation between the degree of change of the force data in the current moment and the degree of change in the time series within the preset time series before the current moment, obtain the initial dimension coefficient of the force data in the current moment.

[0056] Therefore, if the force data in a certain dimension changes rapidly in the current real-time data, then the data is more likely to be abnormally fluctuating and can be used more effectively for subsequent detection of real abnormal data.

[0057] In this embodiment of the invention, within a preset time range prior to the current moment, the numerical difference of the force data in this dimension between every two adjacent sampling moments is calculated. The average of all numerical differences is then used as the preceding average variable for this dimension. The overall change trend is reflected by the overall change difference between preceding adjacent sampling moments. The preset time range can be set to the range of the six sampling moments prior to the current moment. The specific range can be limited by the implementer according to the specific implementation scenario, and is not restricted here.

[0058] Furthermore, the difference between the force data of this dimension at the current moment and the previous sampling moment is taken as the current degree of change of this dimension, reflecting the instantaneous deviation at the current moment.

[0059] Finally, the ratio of the current degree of change of this dimension to the mean of the preceding variables is used as the initial dimension coefficient of the force data of this dimension at the current moment.

[0060] S202: Based on the vibration frequency of the force data in this dimension before the current moment, and combined with the initial dimension coefficient, obtain the instability coefficient of the force data in this dimension at the current moment.

[0061] Given the rapid changes in force feedback data from percutaneous puncture needles, the tissue undergoes elastic deformation when the needle contacts different tissue layers. This deformation of elastic materials typically manifests as low-frequency vibrations. This low-frequency characteristic reflects the elastic deformation of the tissue. Abnormal data in the force feedback data differs from the actual tissue response. Therefore, combining vibration frequency can help distinguish between physiological fluctuations and genuine abnormal changes, improving the reliability of dimensionality in current analyses of real abnormalities.

[0062] In this embodiment of the invention, within a preset time range prior to the current moment, the force data of the current dimension is subjected to a Fourier transform to obtain the frequency domain space of that dimension. The frequency at the highest peak in the frequency domain space is then normalized to obtain the high-frequency anomaly degree of that dimension. Since physiological fluctuations in tissue elastic deformation are mostly low-frequency, while anomalies such as equipment vibration are mostly high-frequency, converting the time domain data to the frequency domain through Fourier transform indicates that the higher the peak value is located in the high-frequency region, the more likely it is to be a true anomaly.

[0063] It should be noted that Fourier transform and normalization are techniques well known to those skilled in the art. The choice of normalization can be linear normalization or standard normalization, etc., which will not be elaborated or limited here.

[0064] Furthermore, the product of the high-frequency anomaly of this dimension and the initial dimension coefficient is used as the instability coefficient of the force data of this dimension at the current moment. By combining the fluctuation and frequency vibration, it reflects the degree of possibility that the force data of this dimension is abnormally unstable at the current moment.

[0065] S203: Based on the difference in instability coefficients between the current and previous time steps of the force data in this dimension, obtain the abnormal fluctuation coefficient of the force data in this dimension.

[0066] Genuine interference and abnormal data are mainly caused by factors such as external noise, equipment failure, operational errors, errors in signal processing, or the inertial effects of physical systems. This interference data is transient and discontinuous; once the source of interference disappears, the system quickly returns to normal. Therefore, these transient fluctuations differ from physiological signals and persistent abnormal data, and usually disappear within a short period of time.

[0067] Therefore, by observing the instantaneous differences in the instability coefficient, abnormal error situations can be further identified. In this embodiment of the invention, the product of the instability coefficient of the previous sampling time after negative correlation mapping and the instability coefficient of the current time is used as the abnormal fluctuation coefficient. When the instability coefficient of the previous sampling time is low and the instability coefficient of the current sampling time is significantly high, it indicates a higher degree of unstable abrupt change, and the current time is more likely to be a genuine abnormal interference; therefore, the abnormal fluctuation coefficient is larger.

[0068] It should be noted that negative correlation mapping is a technique well known to those skilled in the art, such as using negative exponential form or inverse proportional form, etc., and will not be elaborated or limited here.

[0069] This completes the analysis of potential anomalies in the single-dimensional data.

[0070] Due to the multidimensional nature of force feedback data, correlation analysis is used to determine the usability of comparisons based on the physical correlations between dimensions, such as the inherent relationship between the magnitude and direction of force. This allows for the fusion of relative correlation information across multiple dimensions to determine the final level of participation in anomaly detection. When comparing data from different dimensions of the puncture needle to determine the final dimensional coefficient for each dimension, if the historical data trends of two dimensions are correlated, it indicates that they may reflect similar physical phenomena or related operational situations. For example, during the use of the puncture needle, the magnitude and direction of force may have an inherent relationship, especially during the needle tip movement or insertion phase, making comparisons between dimensions usable.

[0071] Therefore, by combining the temporal correlation of data across multiple dimensions with the abnormal fluctuation coefficient, the final influence coefficient of the force data in each dimension is determined. Preferably, in this embodiment of the invention, the method for obtaining the influence coefficient is described in [reference needed]. Figure 4 The diagram illustrates a flowchart of a method for obtaining an influence coefficient according to an embodiment of the present invention, which includes the following steps:

[0072] S211: For any dimension of force data, based on the correlation between the force data of that dimension and the force data of each other dimension in terms of time series changes, obtain the comparability of the force data of that dimension with the force data of each other dimension.

[0073] By reflecting the correlation of changes in time-series data to indicate the degree of synchronization availability in other dimensions, this embodiment of the invention calculates the Pearson correlation coefficient between the force data of this dimension and the force data of each other dimension within a preset local range before the current time. The absolute value of the Pearson correlation coefficient is used as the correlation of each other dimension. The Pearson correlation coefficient reflects the degree of correlation of changes in time-series data. The larger the absolute value of the Pearson correlation coefficient, the higher the correlation between the two sets of data. The preset local range can be set to the range of the 10 sampling times before the current time, which can be adjusted by the implementer and is not limited here.

[0074] It should be noted that the Pearson correlation coefficient is a well-known technique in the field and is not restricted here.

[0075] Then, the correlation and sum of other dimensions of this dimension are calculated as the correlation and sum of this dimension. The ratio of the correlation of each other dimension to the correlation and sum of this dimension is used as the comparative availability of the force data of this dimension with the force data of each other dimension. Combined with the correlation ratio of all dimensions, it reflects the different degrees of synergistic influence of other dimensions.

[0076] S212: By combining the comparability of the force data in this dimension with the force data in every other dimension, and the abnormal fluctuation coefficients of the force data in this dimension with the force data in other dimensions, the influence coefficient of the force data in this dimension is obtained.

[0077] Physiological fluctuations typically originate from local physiological changes or the contact state between the needle and tissue, such as changes in tissue stiffness or increased local resistance. These factors usually affect the magnitude of the force primarily, with a smaller impact on direction or velocity. For example, when a needle encounters relatively hard tissue, the magnitude of the force may suddenly increase, while the change in direction or velocity is relatively gradual. In this case, other dimensions such as direction and velocity may not be affected to the same extent, and therefore there are no significant abrupt changes. Furthermore, physiological fluctuations are often localized and do not simultaneously have a significant impact on all dimensions; therefore, only the magnitude of the force will show a sudden change.

[0078] During anomaly detection, if significant changes occur across multiple dimensions of data under comparative availability, it typically indicates the presence of external interference or equipment malfunctions requiring detection. Therefore, the final impact coefficient is analyzed by combining comparative availability and anomaly fluctuation coefficients.

[0079] In this embodiment of the invention, each of the other dimensions of the force data in this dimension is sequentially used as an analysis dimension. The product of the abnormal fluctuation coefficient of the analysis dimension and the comparison availability is used as the synchronization interference degree of the analysis dimension, reflecting the changes in other dimensions under comparison availability. The sum of the synchronization interference degrees of all other dimensions of the force data in this dimension is used as the relative influence degree of this dimension. The higher the degree of change in the overall correlation influence of other dimensions, the more significant the abnormal problem.

[0080] Finally, the product of the abnormal fluctuation coefficient and the relative influence of the force data in this dimension is normalized and used as the influence coefficient of the force data in this dimension. By combining the correlation fluctuation changes of all dimensions, the reliability of the force data in this dimension being either normal or abnormal at the current moment is reflected.

[0081] Thus, by combining single-dimensional analysis and multi-dimensional influence analysis, the influence coefficient of the force data in each dimension on the credibility of current anomaly detection is obtained.

[0082] The feedback transmission adjustment module 103 is used to adjust the force data of each dimension based on the influence coefficient when performing LOF algorithm detection on the force data at the current moment, and obtain the adjusted LOF value at the current moment; and to perform transmission feedback adjustment based on the adjusted LOF value at the current moment.

[0083] The LOF (Land of Fiber) algorithm is used to determine the LOF value of the currently acquired real-time force feedback data. Generally, the coefficient of each dimension remains constant across all data points, but not all dimensions are effective for LOF detection. Since the puncture process is a continuous and dynamic one, the interaction between the puncture needle and tissue, such as from skin to fat to muscle, exhibits temporal correlation. The force feedback data at the current moment is closely related to the data from previous time periods. For example, if the puncture needle was in the fat layer 0.5 seconds ago, it may have just entered the muscle layer at the current moment. Abnormal changes in force feedback should be compared with the data from those 0.5 seconds to eliminate any abnormal influences.

[0084] To improve detection accuracy, the influence coefficient of each dimension in the anomaly detection of the current real-time data is used to weight each dimension during LOF value analysis for data at any two time points, adjusting the local distribution distance between the data to make the LOF value more accurate. In this embodiment of the invention, the method for adjusting the LOF value includes: within a preset detection range before the current time, obtaining the difference distance between the force data at each dimension for every two time points. The difference distance reflects the numerical difference between the force data in a single dimension, characterizing the degree of deviation between two time points in this dimension. The preset detection range can be set to the range of 1 second before the current time, and the specific value can be adjusted by the implementer.

[0085] Furthermore, the product of the difference distance between the force data in each dimension at every two time points and the influence coefficient is used as the dimensional distance in each dimension at every two time points. The credibility of each dimension for local anomaly detection is adjusted by the influence coefficient, making the anomaly analysis more accurate.

[0086] Further combining the dimensional distances of all dimensions between every two time points, the adjustment distance between each two time points is obtained. In this embodiment of the invention, the sum of the dimensional distances of all dimensions between every two time points is used as the adjustment distance between each two time points, reflecting the overall degree of deviation. Finally, based on the adjusted deviation distance between time points, and based on the adjustment distance between time points within a preset detection range before the current time point, the LOF algorithm is used to obtain the adjusted LOF value for the current time point. The LOF value reflects the probability of an outlier at the current time point; a larger LOF value indicates a higher probability of it being an outlier. It should be noted that the LOF algorithm is a well-known technique familiar to those skilled in the art and will not be elaborated upon here.

[0087] Therefore, the transmission feedback is adjusted based on the adjusted LOF value at the current moment. In this embodiment of the invention, if the adjusted LOF value at the current moment is greater than the preset abnormal threshold, the force data at the current moment will not be transmitted for feedback. The preset abnormal threshold is set to 2, which can be adjusted by the implementer. Data exceeding the preset abnormal threshold is determined to be interference abnormal data. Its transmission cannot help the subsequent analysis of force feedback data and is therefore interference data. So, it is not necessary to transmit for feedback to improve the accuracy and reliability of the force feedback data.

[0088] In summary, this invention, for all dimensions of the current force data collected by percutaneous puncture needles, first preliminarily determines the abnormal fluctuation coefficient for this dimension used in real-time data anomaly detection based on the unevenness of the force data changes in a single dimension. Then, combining the correlation of multi-dimensional force feedback, it uses the temporal correlation comparison of force data in different dimensions, combined with the abnormal fluctuation coefficient, to determine the influence coefficient of each dimension's force data on the current real-time data anomaly detection analysis, thereby improving the accuracy of anomaly point identification. Finally, during LOF anomaly detection, all dimensions are weighted and adjusted to obtain a more accurate adjusted LOF value, which is used to determine whether transmission feedback is possible. This invention performs single-dimensional and multi-dimensional comparative analysis of multi-dimensional force data to measure the credibility of the influence of each dimension on anomalies, adjusts LOF anomaly detection, and more accurately transmits feedback on real-time data.

[0089] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0090] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A real-time processing system for percutaneous puncture needle force feedback data, characterized in that, The system includes: The data acquisition module is used to acquire force data of the percutaneous puncture needle in different dimensions at each acquisition moment; The dimensional impact analysis module is used to obtain the abnormal fluctuation coefficient of the force data of each dimension based on the rate of change of the force data of each dimension in the time series before the current moment; and to determine the impact coefficient of the force data of each dimension based on the correlation of the time series changes of the force data of each dimension with other dimensions and the influence of the abnormal fluctuation coefficient. The feedback transmission adjustment module is used to adjust the force data of each dimension based on the influence coefficient when performing LOF algorithm detection on the force data at the current moment, so as to obtain the adjusted LOF value at the current moment; and to perform transmission feedback adjustment based on the adjusted LOF value at the current moment. The method for obtaining the abnormal fluctuation coefficient includes: for any dimension of force data, obtaining the initial dimension coefficient of the force data at the current moment based on the deviation between the degree of change of the force data at the current moment and the degree of change in the time series within a preset time series before the current moment; obtaining the instability coefficient of the force data at the current moment based on the vibration frequency of the force data at the current moment and the initial dimension coefficient; and obtaining the abnormal fluctuation coefficient of the force data at the current moment based on the difference in the instability coefficient between the force data at the current moment and the previous moment. The method for obtaining the initial dimension coefficients includes: within a preset time series range before the current time, calculating the numerical difference of the force data of this dimension between every two adjacent sampling times, and taking the mean of all numerical differences as the preceding mean variable of this dimension; taking the numerical difference of the force data of this dimension between the current time and the previous sampling time as the current degree of change of this dimension; and taking the ratio of the current degree of change of this dimension to the preceding mean variable as the initial dimension coefficient of the force data of this dimension at the current time. The method for obtaining the instability coefficient includes: within a preset time range before the current moment, performing a Fourier transform on the force data of the current dimension to obtain the frequency domain space of that dimension; normalizing the frequency at the highest peak in the frequency domain space to obtain the high-frequency anomaly of that dimension; and using the product of the high-frequency anomaly of that dimension and the initial dimension coefficient as the instability coefficient of the force data of that dimension at the current moment. Based on the difference in instability coefficients between the current time and previous time in the force data of this dimension, the abnormal fluctuation coefficient of the force data of this dimension is obtained, including: the product of the instability coefficient of the previous sampling time after negative correlation mapping and the instability coefficient of the current time, which is used as the abnormal fluctuation coefficient.

2. The real-time processing system for percutaneous puncture needle force feedback data according to claim 1, characterized in that, Methods for obtaining the influence coefficient include: For any one dimension of force data, the comparative usability of the force data in that dimension with the force data in each other dimension is obtained based on the degree of correlation between the time-series changes of the force data in that dimension and the force data in each other dimension. By combining the comparability of the force data in this dimension with the force data in every other dimension, and the abnormal fluctuation coefficients of the force data in this dimension with the force data in other dimensions, the influence coefficient of the force data in this dimension is obtained.

3. The real-time processing system for percutaneous puncture needle force feedback data according to claim 2, characterized in that, Methods for comparing availability include: Within a predefined local range before the current moment, calculate the Pearson correlation coefficient between the force data of this dimension and the force data of each other dimension in the time series, and use the absolute value of the Pearson correlation coefficient as the correlation of each other dimension. Calculate the correlation and sum of other dimensions for this dimension, and use this as the correlation sum of this dimension; use the ratio of the correlation of each other dimension to the correlation sum of this dimension as the usability of the force data for this dimension compared with the force data of each other dimension.

4. The real-time processing system for percutaneous puncture needle force feedback data according to claim 3, characterized in that, By combining the comparability of the force data in this dimension with the force data in every other dimension, and the abnormal fluctuation coefficients of the force data in this dimension with the force data in other dimensions, the influence coefficient of the force data in this dimension is obtained, including: Each of the other dimensions of the stress data in this dimension is taken as an analysis dimension in turn. The product of the abnormal fluctuation coefficient of the analysis dimension and the comparative availability is taken as the synchronization interference degree of the analysis dimension. The sum of the synchronization interference degrees of all other dimensions of the stress data in this dimension is taken as the relative influence degree of this dimension. The product of the abnormal fluctuation coefficient and the relative influence of the force data in this dimension is normalized and used as the influence coefficient of the force data in this dimension.

5. The real-time processing system for percutaneous puncture needle force feedback data according to claim 1, characterized in that, Methods for adjusting the LOF value include: Within the preset detection range before the current moment, obtain the difference distance between the force data in each dimension at every two moments; The product of the difference distance between the force data in each dimension at every two time points and the influence coefficient is used as the dimensional distance between each two time points in each dimension; the adjustment distance between each two time points is obtained by combining the dimensional distances of all dimensions between each two time points. Based on the adjustment distance between time points within the preset detection range before the current time, the LOF algorithm is used to obtain the adjusted LOF value at the current time.

6. The real-time processing system for percutaneous puncture needle force feedback data according to claim 1, characterized in that, Based on the current adjusted LOF value, the transmission feedback is adjusted, including: If the adjusted LOF value at the current moment is greater than the preset abnormal threshold, the force data at the current moment will not be transmitted or fed back.

Images