A digital twin model for orthopedic implant surgery
By obtaining and analyzing pressure and force timing data in orthopedic implant surgery, screening abnormal data, and building an optimized orthopedic implant digital twin model, the accuracy problem caused by instability in data collection in the existing technology is solved, and the surgery is improved Monitoring accuracy.
Patent Information
- Application Number
- CN202411118004.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-08-15
AI Technical Summary
The existing digital twin models in orthopedic implant surgery have unstable data acquisition due to sensor accuracy, operational technology and environmental factors, which affects the accuracy of the model and the accuracy of surgical monitoring.
By obtaining pressure timing data and force timing data at different positions on the surface of the implant contacting the bone, dividing abnormal pressure data sequence segments, performing data filtering and abnormality analysis, screening relevant force timing data, and constructing an optimized pressure timing data sequence to construct Orthopedic implantation digital twin model.
It improves the accuracy of orthopedic implant surgery monitoring and the credibility of the model, ensuring the accuracy of abnormal data detection and the accuracy of implant surgery.
Smart Images

Figure CN118634072B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a digital twin model used in orthopedic implant surgery. Background Art
[0002] Digital twinning is a method of digitizing physical entities or processes in the real world through mathematical modeling and simulation technology. By collecting various data from orthopedic implant surgery, a digital twin model of orthopedic implants is constructed to simulate and monitor orthopedic implant surgery in real time. Among them, the pressure distribution in the contact area between the implant and the bone and the force applied to the implant can be measured by sensors to ensure the stability of the implant.
[0003] Existing problems: However, due to the accuracy of sensors, operating techniques during surgery, possible interference during the operation, and factors such as the temperature and humidity of the operating room environment, the stability of data acquisition may be affected. When the collected data is abnormal, it will affect the accuracy of the orthopedic implant digital twin model, which may cause inaccurate monitoring of orthopedic implant surgery. Summary of the invention
[0004] The present invention provides a digital twin model for orthopedic implant surgery to solve the existing problems.
[0005] The digital twin model used in orthopedic implant surgery of the present invention adopts the following technical solution:
[0006] An embodiment of the present invention provides a digital twin model for orthopedic implant surgery, and the digital twin model is constructed using the following method:
[0007] In orthopedic implant surgery, pressure sensors are used to obtain the pressure time series data sequence at different positions on the contact surface between the implant and the bone; force sensors are used to obtain the force time series data sequence and force direction sequence at different positions on the implant; the vertical direction of the pressure sensor surface and the distance between the pressure sensor and the force sensor are obtained;
[0008] According to the differences of all pressure time series data sequences and the differences of corresponding moments of pressure data in the pressure time series data sequences, a number of abnormal pressure data sequence segments are divided;
[0009] According to the duration of each abnormal pressure data sequence segment and the pressure data in the abnormal pressure data sequence segment, the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment is obtained;
[0010] Perform filtering operation on each force time series data sequence to obtain an updated force time series data sequence; select the relevant force time series data sequence corresponding to each pressure time series data sequence according to the angle between the vertical direction of the pressure sensor surface and the force direction in the force direction sequence, the distance between the pressure sensor and the force sensor, and the similarity between the pressure time series data sequence and the updated force time series data sequence;
[0011] According to the similarity between each pressure time series data sequence and the corresponding related force time series data sequence and the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment, the abnormality degree of each pressure data in each abnormal pressure data sequence segment is obtained;
[0012] According to the abnormal degree of each pressure data in each abnormal pressure data sequence segment, the optimized pressure time series data sequence is obtained; according to the optimized pressure time series data sequence and the updated force time series data sequence, an orthopedic implant digital twin model is constructed to monitor orthopedic implant surgery.
[0013] Furthermore, the method of dividing a plurality of abnormal pressure data sequence segments according to the differences of all pressure time series data sequences and the differences of corresponding moments of pressure data in the pressure time series data sequences includes the following specific steps:
[0014] Perform Z-score standardization on each pressure time series data sequence to obtain a new pressure time series data sequence, and record the mean of all new pressure time series data sequences as the standard pressure time series data sequence;
[0015] Using the first-order derivative method, each pressure time series data sequence and the standard pressure time series data sequence are derived to obtain a number of local extreme values; the local extreme values include local maximum values and local minimum values;
[0016] The first The normalized value of the absolute value of the difference between the number of local extreme values of the pressure time series data sequence and the standard pressure time series data sequence is recorded as The trend anomaly of the pressure time series data sequence;
[0017] According to The trend anomaly of the pressure time series data sequence and the corresponding time of the local extreme value in the pressure time series data sequence are obtained. The trend anomaly of each local extreme value in the pressure time series data sequence;
[0018] In each pressure time series data sequence, the local extreme value whose trend abnormality is greater than the preset abnormal threshold is recorded as the abnormal local extreme value. The pressure data sequence segments between the two adjacent local extreme values of each abnormal local extreme value are counted, and the adjacent or overlapping pressure data sequence segments are merged to obtain several abnormal pressure data sequence segments.
[0019] Furthermore, according to The trend anomaly of the pressure time series data sequence and the corresponding time of the local extreme value in the pressure time series data sequence are obtained. The specific steps of determining the trend anomaly of each local extreme value in a pressure time series data sequence are as follows:
[0020] In all pressure time series data series, In the pressure time series data sequence The number of local maxima in all data at the time corresponding to the local maximum is recorded as In the pressure time series data sequence The trend consistency of the local maximum;
[0021] Will remove The pressure time series data sequence other than the first pressure time series data sequence is recorded as the reference pressure time series data sequence; in each reference pressure time series data sequence, In the pressure time series data sequence The local maximum value with the shortest time interval among the local maximum values is recorded as the reference local maximum value.
[0022] In the In the pressure time series data sequence, calculate the The local maximum and The sum of the time intervals of all reference local maxima of the local maximum is The inversely proportional normalized value of the trend consistency of the local maximum is consistent with the sum and the The normalized value of the product of the trend anomaly of the pressure time series data sequence is recorded as In the pressure time series data sequence The trend anomaly of the local maximum;
[0023] In each pressure time series data sequence, the trend anomaly of each local minimum value is obtained according to the method of obtaining the trend anomaly of each local maximum value.
[0024] Furthermore, the method of obtaining the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment according to the duration of each abnormal pressure data sequence segment and the pressure data in the abnormal pressure data sequence segment includes the following specific steps:
[0025] In the The first of the pressure time series data In the abnormal pressure data sequence segment, calculate the The normalized value of the pressure data is The absolute value of the difference between the pressure data in the standard pressure time series data sequence at the corresponding moment of the pressure data, the sum of the trend anomalies of all local extreme values, the The product of the duration of the abnormal pressure data sequence segment and the absolute value of the difference is recorded as The suspected abnormality of the pressure data.
[0026] Furthermore, the method of screening out the relevant force time series data sequence corresponding to each pressure time series data sequence according to the angle between the vertical direction of the pressure sensor surface and the force direction in the force direction sequence, the distance between the pressure sensor and the force sensor, and the similarity between the pressure time series data sequence and the updated force time series data sequence includes the following specific steps:
[0027] The first The normalized value of the Pearson correlation coefficient between the pressure time series data sequence and each updated force time series data sequence is recorded as The correlation coefficient between each pressure time series data sequence and each updated force time series data sequence;
[0028] First The vertical direction of the first pressure sensor surface is The minimum angle of each force direction in the force direction sequence constitutes an angle sequence;
[0029] Let the angle in the angle sequence that is greater than the preset judgment angle be 0, and the angle that is less than or equal to the preset judgment angle be 1, and a 01 sequence is obtained;
[0030] According to the number and distribution of 1 in the 01 sequence and the angle sequence, we get A pressure sensor and Directional consistency of the force sensors;
[0031] The first A pressure sensor and The inverse proportional normalized value of the distance of the force sensor, The pressure time series data sequence and the The correlation coefficient of the updated force time series data series and the A pressure sensor and The product of the directional consistency of the force sensors is recorded as The pressure time series data sequence is subject to the The degree of influence of the update force time series data series;
[0032] According to The degree to which each updated force time series data sequence is affected by each pressure time series data sequence is filtered out. The related force time series data series corresponding to the pressure time series data series.
[0033] Further, according to the number and distribution of 1 in the 01 sequence and the angle sequence, the first A pressure sensor and The directional consistency of the force sensor includes the following specific steps:
[0034] Calculate the product of the number of sequence segments consisting of adjacent 1s in the 01 sequence and the number of 1s in the 01 sequence, and record the ratio of the mean of all angles in the angle sequence to the product of the number as A pressure sensor and Directional consistency of the force sensors.
[0035] Furthermore, according to The degree to which each updated force time series data sequence is affected by each pressure time series data sequence is filtered out. The relevant force time series data sequence corresponding to the pressure time series data sequence includes the following specific steps:
[0036] In the Among the degrees to which each pressure time series data sequence is affected by all the updated force time series data sequences, the updated force time series data sequence corresponding to the greatest degree of influence is recorded as the relevant force time series data sequence.
[0037] Further, the abnormality degree of each pressure data in each abnormal pressure data sequence segment is obtained according to the similarity between each pressure time series data sequence and the corresponding related force time series data sequence and the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment, and the specific steps include the following:
[0038] In the In the pressure time series data sequence, The pressure data minus the The difference of the pressure data is recorded as The degree of trend change of each pressure data;
[0039] According to the method of obtaining the trend change degree of each pressure data in the pressure time series data sequence, the trend change degree of each force data in the related force time series data sequence is obtained;
[0040] In the In the pressure time series data sequence In the abnormal pressure data sequence segment, when the The trend change degree of the pressure data is The pressure data corresponds to the moment When the ratio of the trend change degree of the force data in the related force time series data series of the pressure time series data series is less than 0, the The abnormality degree of each pressure data is set to a preset constant;
[0041] When The trend change degree of the pressure data is The pressure data corresponds to the moment When the ratio of the trend change degree of the force data in the related force time series data series of the pressure time series data series is greater than or equal to 0, according to the The pressure data and The pressure data corresponds to the moment The trend change degree of the force data in the related force time series data series, The correlation coefficient between the pressure time series data sequence and the corresponding related force time series data sequence and the The suspected abnormality degree of the pressure data is obtained The abnormality of the pressure data.
[0042] Furthermore, according to The pressure data and The pressure data corresponds to the moment The trend change degree of the force data in the related force time series data series, The correlation coefficient between the pressure time series data sequence and the corresponding related force time series data sequence and the The suspected abnormality degree of the pressure data is obtained The specific steps of determining the abnormality of each pressure data are as follows:
[0043] The first The trend change degree of the pressure data is The pressure data corresponds to the moment The absolute value of the difference in the trend change degree of the force data in the related force time series data series, The pressure time series data sequence and the The inverse proportional value of the correlation coefficient of the pressure time series data series and the The normalized value of the product of the suspected abnormality degree of the pressure data is recorded as The abnormality of the pressure data.
[0044] Furthermore, the step of obtaining the optimized pressure time series data sequence according to the abnormality degree of each pressure data in each abnormal pressure data sequence segment includes the following specific steps:
[0045] In each abnormal pressure data sequence segment in each pressure time series data sequence, the pressure data with an abnormality greater than a preset judgment threshold is recorded as abnormal pressure data;
[0046] All abnormal pressure data in each pressure time series data sequence are removed, and then filled using an interpolation algorithm to obtain an optimized pressure time series data sequence.
[0047] The beneficial effects of the technical solution of the present invention are:
[0048] In an embodiment of the present invention, the pressure time series data sequence at different positions of the contact surface between the implant and the bone and the force time series data sequence at different positions on the implant are obtained during orthopedic implant surgery, and the pressure time series data sequence is divided into several abnormal pressure data sequence segments. The abnormal pressure data sequence segments with abnormalities are initially divided through data change trend analysis, which reduces the scope of abnormal data detection, thereby increasing the accuracy of subsequent abnormal data detection, thereby obtaining the suspected abnormal degree of each pressure data in each abnormal pressure data sequence segment. The relevant force time series data sequence corresponding to each pressure time series data sequence is screened out, and the relevant force time series data sequence is accurately screened through the relationship between the force and the pressure direction, which is used for the subsequent comparative analysis of the force and the pressure, and further ensures the accuracy of abnormal data detection. The abnormal degree of each pressure data in each abnormal pressure data sequence segment is obtained, so as to obtain the optimized pressure time series data sequence, and construct an orthopedic implant digital twin model, thereby ensuring the credibility of the model, and then used to monitor orthopedic implant surgery. So far, the present invention has accurately detected abnormal pressure data, thereby constructing an accurate orthopedic implant digital twin model, which improves the monitoring effect of orthopedic implant surgery. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0050] Figure 1 This is a flowchart of the steps of a digital twin model used in orthopedic implant surgery according to the present invention;
[0051] Figure 2 Schematic diagram of knee implant surgery. DETAILED DESCRIPTION
[0052] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of a digital twin model for orthopedic implant surgery proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0053] Unless defined otherwise, 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 belongs.
[0054] The following is a detailed description of a specific scheme of a digital twin model for orthopedic implant surgery provided by the present invention in conjunction with the accompanying drawings.
[0055] See also Figure 1 , which shows a flowchart of the steps of a digital twin model used in orthopedic implant surgery provided by one embodiment of the present invention. The digital twin model is constructed using the following method:
[0056] Step S001: In orthopedic implant surgery, use a pressure sensor to obtain a pressure timing data sequence at different positions on the contact surface between the implant and the bone; use a force sensor to obtain a force timing data sequence and a force direction sequence at different positions on the implant; obtain the vertical direction of the pressure sensor surface and the distance between the pressure sensor and the force sensor.
[0057] The main purpose of this embodiment is to remove abnormal data and improve the accuracy of the orthopedic implant digital twin model, thereby ensuring the accuracy of orthopedic implant surgery monitoring.
[0058] In historical orthopedic implant surgeries of the knee joint, several pressure sensors placed on the surface where the implant contacts the bone and on the surrounding area are used to collect pressure timing data sequences at different locations on the surface where the implant contacts the bone and on the surrounding area during each orthopedic implant surgery of the knee joint.
[0059] Force sensors are mainly used to measure the magnitude and direction of force applied to an object. Several force sensors are placed in the support structure of the implant and the surrounding area to collect the force time series data sequence and force direction sequence of different positions in the support structure and surrounding area of the implant during each orthopedic implant surgery of the knee joint.
[0060] It should be noted that the data collection frequency in the time series data sequence is once per second, and each data corresponds to a timestamp. These data are helpful in evaluating the stability, mechanical properties and interaction of the implant with the surrounding bone tissue. Figure 2 shown.
[0061] During the orthopedic implant surgery of each knee joint, the distance between each pressure sensor and each force sensor and the vertical direction of the surface of each pressure sensor are collected after the patient's posture is fixed.
[0062] It should be noted that since pressure sensors usually determine the magnitude of pressure by measuring the pressure exerted by an object on the sensing surface, they usually do not directly provide information about the direction of the force. They are usually designed to measure pressure along a vertical direction rather than force in multiple directions. In orthopedic implant surgery, the patient's posture is usually fixed during the operation to reduce the interference of posture changes on the surgical operation. Therefore, the vertical direction of the pressure sensor surface is used to represent the force direction at different positions on the surface where the implant contacts the bone.
[0063] Step S002: according to the differences of all pressure time series data sequences and the differences of corresponding moments of pressure data in the pressure time series data sequences, a number of abnormal pressure data sequence segments are divided.
[0064] During orthopedic implant surgery, the patient's activity or load pattern is relatively stable. Although the specific location of the pressure sensor may vary slightly, it is usually placed in an area close to the bone surface to ensure that the pressure changes in the contact area between the implant and the bone can be effectively captured.
[0065] Since the human bone structure varies from person to person, the bone morphology and load distribution at the same knee joint may be different for different individuals. For example, there can be significant differences in the load distribution at the center and surrounding areas of the joint. Therefore, the pressure data measured by pressure sensors at different locations on the same bone surface may be different. However, the total force borne by the implant should be balanced with the pressure on the surface of the implant in contact with the bone to improve the stability and long-term effect of the implant.
[0066] Therefore, the pressure time series data sequences collected by different pressure sensors on the same bone surface will show a consistent change trend, that is, the peaks and troughs of pressure may be reflected simultaneously in different pressure time series data sequences, although the specific values may be different.
[0067] Preferably, in one embodiment of the present invention, a method for obtaining the trend abnormality of each local maximum value in each pressure time series data sequence includes:
[0068] Among all the pressure time series data sequences in each orthopedic implant surgery, each pressure time series data sequence is subjected to Z-score standardization to obtain a new pressure time series data sequence, and the mean of all the new pressure time series data sequences is recorded as the standard pressure time series data sequence.
[0069] Using the first-order derivative method, each pressure time series data sequence and the standard pressure time series data sequence are derived respectively to obtain a number of local extreme values, which include local maximum values and local minimum values.
[0070] It should be noted that there is an increasing or decreasing trend between the local extreme values of adjacent quantities, and Z-score standardization is used to normalize the data values in each pressure time series data sequence to a value between 0 and 1, which is a well-known technology. The mean of multiple data sequences of equal length is a common and basic operation, that is, the data values at the corresponding positions of each data sequence are added to obtain the mean value. The first-order derivative method is a well-known technology and will not be described in detail here.
[0071] The first The normalized value of the absolute value of the difference between the number of local extreme values of the pressure time series data sequence and the standard pressure time series data sequence is recorded as The trend anomaly of a pressure time series data sequence.
[0072] What needs to be explained is: The calculation formula for the trend anomaly of a pressure time series data sequence is:
[0073]
[0074] In the formula, For the The trend anomaly of the pressure time series data sequence, For the The number of local extreme values of a pressure time series data sequence, is the number of local extreme values of the standard pressure time series data sequence, is the absolute value function, is a linear normalization function, which is used to normalize the data values between 0 and 1. Thus, the trend abnormality of each pressure time series data sequence relative to other pressure time series data sequences is obtained.
[0075] In all pressure time series data series, In the pressure time series data sequence The number of local maxima in all data at the time corresponding to the local maximum is recorded as In the pressure time series data sequence The trend consistency of the local maximum.
[0076] Will remove The pressure time series data sequence other than the first pressure time series data sequence is recorded as the reference pressure time series data sequence. In the pressure time series data sequence The local maximum with the shortest time interval among the local maximum values is recorded as the reference local maximum value.
[0077] In the In the pressure time series data sequence, calculate the The local maximum and The sum of the time intervals of all reference local maxima of the local maximum is The inverse proportional normalized value of the trend consistency of the local maximum is proportional to the sum and the The normalized value of the product of the trend anomaly of the pressure time series data sequence is recorded as In the pressure time series data sequence The trend anomaly of a local maximum.
[0078] What needs to be explained is: In the pressure time series data sequence The calculation formula for the trend anomaly of a local maximum is:
[0079]
[0080] In the formula, For the In the pressure time series data sequence The trend anomaly of the local maximum value, For the The trend anomaly of the pressure time series data sequence, For the In the pressure time series data sequence The trend consistency of the local maximum is the number of pressure time series data sequences, For the In the pressure time series data sequence The local maximum and The local maximum The time interval of the reference local maximum, For the In the pressure time series data sequence The number of reference local maxima corresponding to local maxima, It is a linear normalization function used to normalize data values between 0 and 1. for The normalized value of for Inverse proportional normalized value. The larger the time interval, the more inconsistent the trend, so the smaller the trend consistency, and the larger the time interval between the local maximum and its reference local maximum, the greater the trend abnormality.
[0081] According to the above method, the trend abnormality of each local maximum value and the trend abnormality of each local minimum value in each pressure time series data sequence can be obtained, thereby obtaining the trend abnormality of each local extreme value.
[0082] Preferably, in one embodiment of the present invention, the method for acquiring abnormal pressure data sequence segments includes:
[0083] In this embodiment, the preset abnormal threshold is 0.7, which is used as an example for description.
[0084] In each pressure time series data sequence, the local extreme value whose trend abnormality is greater than the preset abnormal threshold is recorded as the abnormal local extreme value. The pressure data sequence segments between the two adjacent local extreme values of each abnormal local extreme value are counted, and the adjacent or overlapping pressure data sequence segments are merged to obtain several abnormal pressure data sequence segments.
[0085] It should be noted that: when the abnormal local extreme value is the first or last local extreme value, the pressure data sequence segment between the first pressure data and the second local extreme value is taken, and the pressure data sequence segment between the second to last local extreme value and the last pressure data is taken, and each pressure data sequence segment is a complete fluctuation cycle. An isolated pressure data sequence segment that is not adjacent or overlapping is also an abnormal pressure data sequence segment.
[0086] Step S003: obtaining the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment according to the duration of each abnormal pressure data sequence segment and the pressure data in the abnormal pressure data sequence segment.
[0087] In all the pressure time series data sequences in each orthopedic implant surgery, the longer the duration of the abnormal pressure data sequence segment, and the greater the abnormality of the trend of the local maximum value, that is, the longer the impact time of the abnormal data in the pressure time series data sequence, and the more abnormal the data change trend. The greater the difference between the collected pressure data and the pressure data in the standard pressure time series data sequence, the more abnormal the pressure data.
[0088] Preferably, in one embodiment of the present invention, a method for obtaining the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment includes:
[0089] In the In the pressure time series data sequence In the abnormal pressure data sequence segment, calculate the The normalized value of the pressure data is The absolute value of the difference between the pressure data in the standard pressure time series data sequence at the corresponding moment of the pressure data, the sum of the trend anomalies of all local extreme values, the The product of the duration of the abnormal pressure data sequence segment and the absolute value of the difference is recorded as The suspected abnormality of the pressure data.
[0090] What needs to be explained is: In the pressure time series data sequence The first The calculation formula for the suspected abnormality degree of pressure data is:
[0091]
[0092] In the formula, For the In the pressure time series data sequence The first The suspected abnormality of the pressure data, For the In the pressure time series data sequence The duration of the abnormal pressure data sequence segment, For the In the pressure time series data sequence The first The trend anomaly of the local extreme value, For the In the pressure time series data sequence The number of local extreme values in the abnormal pressure data sequence segment, For the In the pressure time series data sequence The first Pressure data, is the minimum and maximum normalization function, which is a well-known technique for normalizing data values between 0 and 1. For the In the pressure time series data sequence The first Each pressure data corresponds to the pressure data in the standard pressure time series data sequence at that moment. is an absolute value function. Therefore, the longer the duration of the abnormal pressure data sequence segment, the greater the abnormality of the trend of the local extreme value within it, and the greater the difference between the pressure data and the standard pressure data, the greater the suspected abnormality of the pressure data.
[0093] Step S004: Perform filtering operation on each force time series data sequence to obtain an updated force time series data sequence; based on the angle between the vertical direction of the pressure sensor surface and the force direction in the force direction sequence, the distance between the pressure sensor and the force sensor, and the similarity between the pressure time series data sequence and the updated force time series data sequence, select the relevant force time series data sequence corresponding to each pressure time series data sequence.
[0094] Since the pressure data in all pressure time series data sequences in each orthopedic implant surgery are data under the same surgical operation and in the same surgical environment, for pressure sensors that are close to each other, there is a possibility that the abnormal data in the same time period may appear similar in the corresponding pressure time series data sequence, resulting in a smaller degree of suspected abnormality in the above calculation.
[0095] Since implants in orthopedic implant surgery are subject to the forces of the surrounding bones and muscles, as well as the operating force during the operation, that is, the force applied by the surgeon, the force sensors at different positions on the implant measure different magnitudes and directions of the forces applied to the implant. That is, different force-bearing positions on the implant have different effects on the pressure at different positions on the surface where the implant contacts the bone. Therefore, it is necessary to first screen out the force sensor that has the greatest effect on each pressure sensor for comparative analysis.
[0096] In orthopedic implant surgery, the pressure data of the contact area between the implant and the bone and the change trend of the force applied to the implant are usually related. Preferably, in one embodiment of the present invention, the method for obtaining the relevant force time series data sequence corresponding to each pressure time series data sequence includes:
[0097] An exponentially weighted moving average algorithm is used to filter each strength time series data sequence to obtain an updated strength time series data sequence.
[0098] It should be noted that the exponentially weighted moving average algorithm is a well-known technology and will not be described in detail here. The filtering operation can remove the influence of noise and ensure the credibility of the strength data.
[0099] In each orthopedic implant surgery, calculate the The normalized value of the Pearson correlation coefficient between the pressure time series data sequence and each updated force time series data sequence is taken as the The correlation coefficient between a pressure time series data series and each updated force time series data series.
[0100] It should be noted that the Pearson correlation coefficient is a well-known technology. The larger the Pearson correlation coefficient, the more similar the change trend of the data sequence. The normalized value of the Pearson correlation coefficient in this embodiment is obtained using the minimum and maximum normalization function. Since the force applied to the implant usually reflects the force transmission and load distribution between the overall bone and the implant, the greater the distance, the force will be attenuated, making the force transmission efficiency smaller, the smaller the impact on the pressure, and the more inconsistent the force of the force sensor is with the force direction of the pressure sensor, the smaller the impact on the pressure.
[0101] First A pressure sensor and Take the force sensor as an example. The vertical direction of the first pressure sensor surface is The minimum angle between each force direction in a force direction sequence constitutes an angle sequence.
[0102] In this embodiment, the preset judgment angle is 5 degrees, which is used as an example for description.
[0103] In the angle sequence, the angle that is greater than the preset judgment angle is set to 0, and the angle that is less than or equal to the preset judgment angle is set to 1, thereby obtaining a 01 sequence.
[0104] Calculate the product of the number of sequence segments consisting of adjacent 1s in the 01 sequence and the number of 1s in the 01 sequence, and record the ratio of the mean of all angles in the angle sequence to the product as A pressure sensor and Directional consistency of the force sensors.
[0105] What needs to be explained is: A pressure sensor and The calculation formula for the directional consistency of the force sensors is:
[0106]
[0107] In the formula, For the A pressure sensor and The directional consistency of the force sensors For the A pressure sensor and The mean value of all angles in the angle sequence corresponding to the force sensor, For the A pressure sensor and The number of 1s in the 01 sequence corresponding to each force sensor, For the A pressure sensor and The number of sequence segments formed by adjacent 1s in the 01 sequence corresponding to each force sensor. When the angle in the angle sequence is smaller, that is, the force of the force sensor is more consistent with the force direction of the pressure sensor, when there are more 1 values in the 01 sequence and they are more continuous, it means that the force of the force sensor has a greater and continuous impact on the pressure sensor. Therefore, the smaller the angle, the more consistent the direction, and the more 1s in the 01 sequence, and the fewer sequence segments formed by adjacent 1s, it means that the force of the force sensor and the force direction of the pressure sensor are similar for a long time and are continuous in time sequence.
[0108] The first A pressure sensor and The inverse proportional normalized value of the distance of the force sensor, The pressure time series data sequence and the The correlation coefficient of the updated force time series data series and the A pressure sensor and The product of the directional consistency of the force sensors is recorded as The pressure time series data sequence is subject to the The degree of influence of an update force time series data series.
[0109] What needs to be explained is: The pressure time series data sequence is subject to the The calculation formula for the influence of an update force time series data sequence is:
[0110]
[0111] In the formula, For the The pressure time series data sequence is subject to the The influence degree of the update force time series data series, For the A pressure sensor and The directional consistency of the force sensors For the A pressure sensor and The distance between the force sensors, For the The pressure time series data sequence and the The correlation coefficient of the update force time series data series. is a linear normalization function, which is used to normalize data values between 0 and 1. for Therefore, the greater the direction consistency, the closer the distance, and the more similar the pressure time series data sequence is to the updated force time series data sequence, indicating that The pressure time series data sequence is subject to the The greater the influence of the update force time series data series.
[0112] In the Among the degrees to which each pressure time series data sequence is affected by all the updated force time series data sequences, the updated force time series data sequence corresponding to the greatest degree of influence is recorded as the relevant force time series data sequence.
[0113] It should be noted that if there are multiple maximum impact levels, any one of them will be selected for analysis.
[0114] Step S005: Obtain the abnormality degree of each pressure data in each abnormal pressure data sequence segment according to the similarity between each pressure time series data sequence and the corresponding related force time series data sequence and the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment.
[0115] Preferably, in one embodiment of the present invention, the method for obtaining the abnormality degree of each pressure data in each abnormal pressure data sequence segment includes:
[0116] In the In the pressure time series data sequence, The pressure data minus the The difference of the pressure data is recorded as The degree of trend change of pressure data.
[0117] It should be noted that: in this embodiment, the trend change degree of the last pressure data is set to be the trend change degree of the second to last pressure data, and this is used as an example for description.
[0118] According to the method of obtaining the trend change degree of each pressure data in the pressure time series data sequence, the trend change degree of each force data in the related force time series data sequence is obtained.
[0119] In the In the pressure time series data sequence In the abnormal pressure data sequence segment, when the The trend change degree of the pressure data is The pressure data corresponds to the moment When the ratio of the trend change degree of the force data in the related force time series data series of the pressure time series data series is less than 0, the The abnormality degree of each pressure data is set as a preset constant.
[0120] When The trend change degree of the pressure data is The pressure data corresponds to the moment When the ratio of the trend change degree of the force data in the related force time series data series of the pressure time series data series is greater than or equal to 0, the first The trend change degree of the pressure data is The pressure data corresponds to the moment The absolute value of the difference in the trend change degree of the force data in the related force time series data series, The pressure time series data sequence and the The inverse proportional value of the correlation coefficient of the pressure time series data series and the The normalized value of the product of the suspected abnormality degree of the pressure data is recorded as The abnormality of the pressure data.
[0121] It should be noted that: in this embodiment, the preset constant is 1, which is used as an example for description. The trend change degree of the pressure data is The pressure data corresponds to the moment When the ratio of the trend change degree of the force data in the related force time series data series of the pressure time series data series is greater than or equal to 0, the In the pressure time series data sequence The first The calculation formula for the abnormality of pressure data is:
[0122]
[0123] In the formula, For the In the pressure time series data sequence The first The abnormality of the pressure data, For the In the pressure time series data sequence The first The suspected abnormality of the pressure data, For the The pressure time series data sequence and the The correlation coefficient of the pressure time series data series, For the In the pressure time series data sequence The first The degree of trend change of pressure data, For the In the pressure time series data sequence The first The pressure data corresponds to the moment The degree of trend change of the force data in the related force time series data series of the pressure time series data series, is the absolute value function, is a linear normalization function, which is used to normalize data values between 0 and 1. for The inverse proportional value of The smaller, and The larger the value, the The more the data change trend of a pressure time series data sequence is different from the data change trend of its related force time series data sequence, the greater the possibility of being affected by the anomaly. right Adjust to get The abnormality degree of each pressure data is calculated, and when the trend changes in the opposite direction, the abnormality degree is directly set to 1.
[0124] Step S006: According to the abnormality degree of each pressure data in each abnormal pressure data sequence segment, an optimized pressure time series data sequence is obtained; according to the optimized pressure time series data sequence and the updated force time series data sequence, an orthopedic implant digital twin model is constructed for monitoring orthopedic implant surgery.
[0125] In this embodiment, the preset judgment threshold is 0.8, and this is taken as an example for description.
[0126] In each abnormal pressure data sequence segment in each pressure time series data sequence, the pressure data with an abnormality greater than a preset judgment threshold is recorded as abnormal pressure data.
[0127] In each orthopedic implant surgery, all abnormal pressure data in each pressure time series data sequence are removed, and then filled using an interpolation algorithm to obtain an optimized pressure time series data sequence. The interpolation algorithm is a well-known technology and will not be described in detail here.
[0128] In historical orthopedic implant surgeries, an orthopedic implant digital twin model is constructed based on the optimized pressure time series data sequence, updated force time series data sequence, and collected medical images (such as CT scans, MRI, etc.) in each orthopedic implant surgery. This is a well-known technology and will not be elaborated here.
[0129] This allows surgical procedures to be simulated and planned in a digital environment, taking into account individual patient differences and procedure-specific technical requirements.
[0130] Medical images of patients who currently need orthopedic implant surgery are collected and input into the orthopedic implant digital twin model to determine the optimal implant position and implant size. Augmented reality (AR) or virtual reality (VR) technology is then used to match and navigate the pre-planned implant position and the patient's anatomical structure during the actual surgery in real time. Sensor technology and real-time data feedback are used to monitor various parameters during the operation, such as implant position, bone stability, and surgical risk indicators.
[0131] It should be noted that: in this embodiment, when the denominator in the formula is 0, the denominator is set to 1 to ensure that the formula is valid, and this example is used for description.
[0132] So far, the present invention is completed.
[0133] In summary, in an embodiment of the present invention, the pressure time series data sequence of different positions of the contact surface between the implant and the bone and the force time series data sequence and the force direction sequence of different positions on the implant are obtained during orthopedic implant surgery, the pressure time series data sequence is divided into several abnormal pressure data sequence segments, the suspected abnormal degree of each pressure data in each abnormal pressure data sequence segment is obtained, the relevant force time series data sequence corresponding to each pressure time series data sequence is screened out, the abnormal degree of each pressure data in each abnormal pressure data sequence segment is obtained, thereby obtaining an optimized pressure time series data sequence, and constructing an orthopedic implant digital twin model for monitoring orthopedic implant surgery. The present invention accurately detects abnormal pressure data, thereby constructing an accurate orthopedic implant digital twin model, thereby improving the monitoring effect of orthopedic implant surgery.
[0134] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A digital twin model for orthopedic implant surgery, characterized in that: The digital twin model is constructed using the following methods: In the past orthopedic implant surgeries of the knee joint, a pressure sensor was used to obtain the pressure time series data sequence at different positions of the contact surface between the implant and the bone; a force sensor was used to obtain the force time series data sequence and the force direction sequence at different positions on the implant; the vertical direction of the pressure sensor surface and the distance between the pressure sensor and the force sensor were obtained; According to the differences of all pressure time series data sequences and the differences of corresponding moments of pressure data in the pressure time series data sequences, a number of abnormal pressure data sequence segments are divided; According to the duration of each abnormal pressure data sequence segment and the pressure data in the abnormal pressure data sequence segment, the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment is obtained; Perform filtering operation on each force time series data sequence to obtain an updated force time series data sequence; select the relevant force time series data sequence corresponding to each pressure time series data sequence according to the angle between the vertical direction of the pressure sensor surface and the force direction in the force direction sequence, the distance between the pressure sensor and the force sensor, and the similarity between the pressure time series data sequence and the updated force time series data sequence; According to the similarity between each pressure time series data sequence and the corresponding related force time series data sequence and the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment, the abnormality degree of each pressure data in each abnormal pressure data sequence segment is obtained; According to the abnormal degree of each pressure data in each abnormal pressure data sequence segment, an optimized pressure time series data sequence is obtained; according to the optimized pressure time series data sequence and the updated force time series data sequence, an orthopedic implant digital twin model is constructed to monitor orthopedic implant surgery; The method of dividing a plurality of abnormal pressure data sequence segments according to the differences of all pressure time series data sequences and the differences of corresponding moments of pressure data in the pressure time series data sequences includes the following specific steps: Perform Z-score standardization on each pressure time series data sequence to obtain a new pressure time series data sequence, and record the mean of all new pressure time series data sequences as the standard pressure time series data sequence; Using the first-order derivative method, each pressure time series data sequence and the standard pressure time series data sequence are derived to obtain a number of local extreme values; the local extreme values include local maximum values and local minimum values; The first The normalized value of the absolute value of the difference between the number of local extreme values of the pressure time series data sequence and the standard pressure time series data sequence is recorded as The trend anomaly of the pressure time series data sequence; According to The trend anomaly of the pressure time series data sequence and the corresponding time of the local extreme value in the pressure time series data sequence are obtained. The trend anomaly of each local extreme value in the pressure time series data sequence; In each pressure time series data sequence, the local extreme value whose trend abnormality is greater than the preset abnormal threshold is recorded as an abnormal local extreme value, and the pressure data sequence segments between two adjacent local extreme values of each abnormal local extreme value are counted, and the adjacent or overlapping pressure data sequence segments are merged to obtain several abnormal pressure data sequence segments; According to the The trend anomaly of the pressure time series data sequence and the corresponding time of the local extreme value in the pressure time series data sequence are obtained. The specific steps of detecting the trend anomaly of each local extreme value in a pressure time series data sequence are as follows: In all pressure time series data series, In the pressure time series data sequence The number of local maxima in all data at the time corresponding to the local maximum is recorded as In the pressure time series data sequence The trend consistency of the local maximum; Will remove The pressure time series data sequence other than the first pressure time series data sequence is recorded as the reference pressure time series data sequence; in each reference pressure time series data sequence, In the pressure time series data sequence The local maximum value with the shortest time interval among the local maximum values is recorded as the reference local maximum value. In the In the pressure time series data sequence, calculate the The local maximum and The sum of the time intervals of all reference local maxima of the local maximum is The inversely proportional normalized value of the trend consistency of the local maximum is consistent with the sum and the The normalized value of the product of the trend anomaly of the pressure time series data sequence is recorded as In the pressure time series data sequence The trend anomaly of the local maximum; In each pressure time series data sequence, the trend anomaly of each local minimum value is obtained according to the method of obtaining the trend anomaly of each local maximum value; The method of obtaining the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment according to the duration of each abnormal pressure data sequence segment and the pressure data in the abnormal pressure data sequence segment includes the following specific steps: In the The first of the pressure time series data In the abnormal pressure data sequence segment, calculate the The normalized value of the pressure data is The absolute value of the difference between the pressure data in the standard pressure time series data sequence at the corresponding moment of the pressure data, the sum of the trend anomalies of all local extreme values, the The product of the duration of the abnormal pressure data sequence segment and the absolute value of the difference is recorded as The suspected abnormality of each pressure data; The specific steps of obtaining the abnormality degree of each pressure data in each abnormal pressure data sequence segment according to the similarity between each pressure time series data sequence and the corresponding related force time series data sequence and the suspected abnormality degree of each pressure data in each abnormal pressure data sequence segment are as follows: In the In the pressure time series data sequence, The pressure data minus the The difference of the pressure data is recorded as The degree of trend change of each pressure data; According to the method of obtaining the trend change degree of each pressure data in the pressure time series data sequence, the trend change degree of each force data in the related force time series data sequence is obtained; In the In the pressure time series data sequence In the abnormal pressure data sequence segment, when the The trend change degree of the pressure data is The pressure data corresponds to the moment When the ratio of the trend change degree of the force data in the related force time series data series of the pressure time series data series is less than 0, the The abnormality degree of each pressure data is set to a preset constant; When The trend change degree of the pressure data is The pressure data corresponds to the moment When the ratio of the trend change degree of the force data in the related force time series data series of the pressure time series data series is greater than or equal to 0, according to the The pressure data and The pressure data corresponds to the moment The trend change degree of the force data in the related force time series data series, The correlation coefficient between the pressure time series data sequence and the corresponding related force time series data sequence and the The suspected abnormality degree of the pressure data is obtained The degree of abnormality of each pressure data; According to the The pressure data and The pressure data corresponds to the moment The trend change degree of the force data in the related force time series data series, The correlation coefficient between the pressure time series data sequence and the corresponding related force time series data sequence and the The suspected abnormality degree of the pressure data is obtained The specific steps of determining the abnormality of each pressure data are as follows: The first The trend change degree of the pressure data is The pressure data corresponds to the moment The absolute value of the difference in the trend change degree of the force data in the related force time series data series, The pressure time series data sequence and the The inverse proportional value of the correlation coefficient of the pressure time series data series and the The normalized value of the product of the suspected abnormality degree of the pressure data is recorded as The abnormality of the pressure data.
2. A digital twin model for orthopedic implant surgery according to claim 1, characterized in that: The method of screening out the relevant force time series data sequence corresponding to each pressure time series data sequence according to the angle between the vertical direction of the pressure sensor surface and the force direction in the force direction sequence, the distance between the pressure sensor and the force sensor, and the similarity between the pressure time series data sequence and the updated force time series data sequence includes the following specific steps: The first The normalized value of the Pearson correlation coefficient between the pressure time series data sequence and each updated force time series data sequence is recorded as The correlation coefficient between each pressure time series data sequence and each updated force time series data sequence; First The vertical direction of the first pressure sensor surface is The minimum angle of each force direction in the force direction sequence constitutes an angle sequence; Let the angle in the angle sequence that is greater than the preset judgment angle be 0, and the angle that is less than or equal to the preset judgment angle be 1, and a 01 sequence is obtained; According to the number and distribution of 1 in the 01 sequence and the angle sequence, we get A pressure sensor and Directional consistency of the force sensors; The first A pressure sensor and The inverse proportional normalized value of the distance of the force sensor, The pressure time series data sequence and the The correlation coefficient of the updated force time series data series and the A pressure sensor and The product of the directional consistency of the force sensors is recorded as The pressure time series data sequence is subject to the The degree of influence of the update force time series data series; According to The degree to which each updated force time series data sequence is affected by each pressure time series data sequence is filtered out. The related force time series data series corresponding to the pressure time series data series.
3. A digital twin model for orthopedic implant surgery according to claim 2, characterized in that: According to the number and distribution of 1 in the 01 sequence and the angle sequence, the A pressure sensor and The directional consistency of the force sensor includes the following specific steps: Calculate the product of the number of sequence segments consisting of adjacent 1s in the 01 sequence and the number of 1s in the 01 sequence, and record the ratio of the mean of all angles in the angle sequence to the product of the number as A pressure sensor and Directional consistency of the force sensors.
4. A digital twin model for orthopedic implant surgery according to claim 2, characterized in that: According to the The degree to which each updated force time series data sequence is affected by each pressure time series data sequence is filtered out. The relevant force time series data sequence corresponding to the pressure time series data sequence includes the following specific steps: In the Among the degrees to which each pressure time series data sequence is affected by all the updated force time series data sequences, the updated force time series data sequence corresponding to the greatest degree of influence is recorded as the relevant force time series data sequence.
5. A digital twin model for orthopedic implant surgery according to claim 1, characterized in that: The step of obtaining an optimized pressure time series data sequence according to the abnormality degree of each pressure data in each abnormal pressure data sequence segment includes the following specific steps: In each abnormal pressure data sequence segment in each pressure time series data sequence, the pressure data with an abnormality greater than a preset judgment threshold is recorded as abnormal pressure data; All abnormal pressure data in each pressure time series data sequence are removed, and then filled using an interpolation algorithm to obtain an optimized pressure time series data sequence.
