Stress assessment method and system for multiple spinal orthopedic surgery processes

Abstract

The invention provides a stress assessment method and system used in a repeated spinal orthopedic surgery process, and relates to the technical field of medical equipment, muscle tremor data of a patient is collected based on a piezoelectric sensor unit, and the muscle tremor data is transmitted to a cloud data center through a lower computer; and patient pain tolerance and infection rejection risk analysis is performed on the muscle tremor data based on the cloud data center. Therefore, real-time evaluation of spinal stress distribution and pain tolerance is realized, the defect that the traction force is judged according to experience traditionally is overcome, and it is ensured that the acting force of the traction device is matched with the individual tolerance of a patient; the sensitivity of the piezoelectric sensor to the temperature is utilized, the skin temperature change is analyzed through data processing, real-time early warning of infection and rejection reaction near the steel nail is achieved, and compared with a traditional regular inspection mode, the timeliness of complication intervention is remarkably improved.

Description

Technical Field

[0001] This application relates to the field of medical device technology, and more specifically, to a method and system for stress assessment during multiple spinal correction surgeries. Background Technology

[0002] Scoliosis is a common spinal deformity among adolescents, with a national prevalence of 1.2% among those aged 10-18. This prevalence has been gradually increasing in recent years, with a particularly significant increase between 2016 and 2024. In some extreme cases, the degree of spinal curvature reaches the level of a "folded person," severely impacting their quality of life and daily routines. For these patients, multiple spinal surgeries become an inevitable treatment option. In spinal correction surgeries, such as cervical spine reduction surgery, a support structure is fixed to the patient's head and lower back, with multiple traction rods placed between the supports. The cervical spine is slowly reduced by manually adjusting the displacement of the traction rods daily, a process that typically lasts 1-2 months. During this process, accurately assessing whether the force of the external reduction traction device is within the patient's tolerance range, and real-time monitoring for postoperative complications such as infection and rejection at the fixation site, are crucial for ensuring surgical safety and improving treatment outcomes.

[0003] Current technologies lack systematic stress assessment methods specifically for multiple spinal correction surgeries. Medical staff often rely on experience to judge traction force, which makes it difficult to accurately match the individual patient's tolerance. This can easily lead to insufficient force affecting the correction effect or excessive force causing severe pain to the patient. At the same time, the monitoring of postoperative infection and rejection reactions mostly relies on regular examinations, which cannot achieve real-time early warning and may delay the timing of intervention. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a method and system for stress assessment during multiple spinal correction surgeries, which can realize real-time analysis of spinal stress distribution, assessment of patient pain tolerance, and early warning of postoperative infection and rejection, thereby protecting patient safety and improving surgical progress.

[0005] In a first aspect, embodiments of this application provide a method for force assessment during multiple spinal correction surgeries, applied to a force assessment system. This system includes a piezoelectric sensor unit, a lower-level computer, and a cloud data center. The method includes the following steps: The data on muscle tremors of patients are collected using a piezoelectric sensor unit and transmitted to a cloud data center via a lower-level computer. Based on the cloud data center, we conducted separate analyses on the patient's pain tolerance and infection rejection risk using muscle tremor data.

[0006] In some embodiments, muscle tremor data and skin temperature data of the patient are collected based on a multi-channel piezoelectric sensor unit; the piezoelectric sensor unit shown is attached to the outside of the patient's spinal screw and includes a piezoelectric sensor, a drive circuit, medical non-woven tape, a pad, and a cable. The piezoelectric sensor uses a piezoelectric ceramic element, and the pad faces the skin.

[0007] In some embodiments, patient pain tolerance analysis is performed on muscle tremor data by the following steps: After denoising the muscle tremor data by mean square error comparison, low-pass filtering and high-pass filtering are performed sequentially to obtain the core signal used for frequency band extraction. The core signal is subjected to FFT transformation to obtain the frequency domain signal. The 4-6Hz component is extracted and the ratio of the power in this frequency band to the total power is calculated. The target waveform is output through end-to-end timing modeling with the ratio of the total power as a constraint condition. After sequentially performing derivative filtering, square filtering, integral filtering, and median smoothing filtering on the output target waveform, the hill climbing method is used to detect all local peaks, and the pain tremor peak and noise peak are determined by adaptive threshold. The frequency of pain tremor peaks in the 4-6 Hz band was counted, and the force on the steel nails was adjusted according to the type of spinal curvature. An early warning was triggered when the force exceeded the individual's tolerance threshold.

[0008] In some embodiments, determining the pain tremor peak and noise peak using an adaptive threshold includes the following steps: If PEAK > THRE_1, it is determined to be a painful tremor peak; where PEAK represents the local peak value found by the hill climbing method, THRE_1 represents the threshold for distinguishing painful tremor peaks from noise peaks, and THRE_1 is updated using the following formula: S_PKI = 0.125 PEAK + 0.875 S_PKI old N_PKI = 0.125 PEAK + 0.875 N_PKI old THRE_1=N_PKI+0.25(S_PKI- N_PKI) Where S_PKI represents the peak amplitude of the corresponding muscle tremor wave, S_PKI old S_PKI is the value from the last update. N_PKI represents the peak amplitude of the non-muscle tremor wave. old This refers to the N_PKI as it was last updated.

[0009] In some embodiments, adjusting the force on the steel screws in accordance with the type of spinal curvature includes the following steps: When the spine curves to the left, the right-side screws experience greater force than the left-side screws; when the spine curves to the right, the left-side screws experience greater force than the right-side screws; when the spine is folded forward for correction, the lower screws experience greater force than the upper screws; when the spine needs to be folded backward for correction, the upper screws experience greater force than the lower screws. The force applied to the steel nails was adjusted with the constraint that the number of times the pain tremor peaks occurred should not exceed the individual's tolerance threshold.

[0010] In some embodiments, the infection rejection risk analysis of muscle tremor data is performed in the following manner, including the following steps: After denoising the muscle tremor data by mean square error comparison, low-pass filtering and high-pass filtering were performed sequentially to obtain a purified signal reflecting changes in skin temperature. The purified signal is subjected to spike noise replacement, extreme value trimming, basic statistical calculation, fluctuation analysis, and data validity determination in a fixed processing window of set duration to obtain valid data reflecting changes in skin temperature. By analyzing the magnitude of changes in the effective data, it can be determined whether the patient is at risk of infection or rejection.

[0011] In some embodiments, the cloud data center receives data from the lower-level machine via TCP / IP communication and processes the data using encapsulated SDK and API functions.

[0012] Secondly, embodiments of this application provide a force assessment system for multiple spinal correction surgeries, including a piezoelectric sensor unit, a lower-level computer, and a cloud data center, for collaboratively implementing the steps of a force assessment method for multiple spinal correction surgeries as described in any one of the first aspects.

[0013] Thirdly, this application also provides an electronic device, including: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, the steps of the stress assessment method for multiple spinal correction surgeries described in any of the second aspects above are executed.

[0014] Fourthly, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of the stress assessment method for multiple spinal correction surgeries described in any one of the first aspects.

[0015] This application describes a method and system for force assessment during multiple spinal correction surgeries. It collects muscle tremor data from patients using a piezoelectric sensor unit and transmits this data to a cloud data center via a lower-level computer. The cloud data center then analyzes the muscle tremor data to assess the patient's pain tolerance and infection / rejection risk. This enables real-time assessment of spinal force distribution and pain tolerance, overcoming the shortcomings of traditional methods that rely on experience to judge traction force and ensuring that the traction device force matches the patient's individual tolerance. Furthermore, by utilizing the temperature sensitivity of the piezoelectric sensor and analyzing skin temperature changes through data processing, it provides real-time early warning of infection and rejection reactions near the steel nails. Compared to traditional periodic check-up methods, this significantly improves the timeliness of complication intervention. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart of the stress assessment method for multiple spinal correction surgeries described in the embodiments of this application is shown; Figure 2 A schematic diagram of the structure of the piezoelectric sensor unit described in an embodiment of this application is shown; Figure 3 A schematic diagram of the drive circuit described in an embodiment of this application is shown; Figure 4 This diagram illustrates the arrangement of the piezoelectric sensor units according to an embodiment of this application. Figure 5 This document illustrates a flowchart of a patient pain tolerance analysis based on muscle tremor data, as described in an embodiment of this application. Figure 6 This illustration shows a waveform diagram after median smoothing filtering of the output target waveform as described in an embodiment of this application. Figure 7 This document illustrates a flowchart of an embodiment of the present application for analyzing the risk of infection rejection from muscle tremor data. Figure 8 A schematic diagram of the structure of the electronic device described in an embodiment of this application is shown. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the accompanying drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0019] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0020] It should be noted that the term "comprising" will be used in the embodiments of this application to indicate the presence of the features declared thereafter, but does not exclude the addition of other features.

[0021] In view of the technical problems raised in the background art, this application provides a method and system for force assessment during multiple spinal correction surgeries, which can realize real-time analysis of spinal force distribution, assessment of patient pain tolerance, and early warning of postoperative infection and rejection reaction, thus protecting patient safety and improving surgical progress.

[0022] See the instruction manual appendix Figure 1 This application provides a method for stress assessment during multiple spinal correction surgeries, applied to a stress assessment system. The system includes a piezoelectric sensor unit, a lower-level computer, and a cloud data center. The method includes the following steps: S1. Collect muscle tremor data of patients based on piezoelectric sensor units and transmit it to the cloud data center through the lower-level computer; S2. Based on the cloud data center, analyze the patient's pain tolerance and infection rejection risk of muscle tremor data.

[0023] To clearly understand the technical solutions of the embodiments of this application, the stress assessment system can be described by way of example first.

[0024] See the instruction manual appendix Figure 2 The piezoelectric sensor unit includes a piezoelectric sensor, a driving circuit, non-woven fabric tape, a pad, and a cable. The piezoelectric sensor uses a piezoelectric ceramic element. A solid material pad is placed above the piezoelectric sensor. The piezoelectric sensor is connected to the driving circuit via a cable. The pad, piezoelectric sensor, and driving circuit are fixed in place by two layers of medical non-woven fabric tape, and the output is via the cable. (See the attached instruction manual for details.) Figure 3 The drive circuit, consisting of a distributed resistor, a field-effect transistor, and a capacitor, forms a low-pass filter used to filter out high-frequency noise. The drive circuit is placed adjacent to the piezoelectric ceramic. See the attached instruction manual. Figure 4 The piezoelectric sensor unit is typically placed on the skin surface where the spinal screws have been inserted, with the padding facing the skin. This system can support up to 16 data acquisition channels, i.e., 16 sets of piezoelectric sensor units. Each piezoelectric sensor unit corresponds to an independent number, defined as Y01-Y16. Generally, an adult spine has 26 vertebrae: seven cervical vertebrae, twelve thoracic vertebrae, five lumbar vertebrae, one sacral vertebra, and one coccygeal vertebra. From C1 to the coccyx, they are designated as numbers 1-26. With the patient facing away from the doctor, the left side is labeled 'a', and the right side is labeled 'b'. Depending on the specific surgery, after necessary postoperative disinfection, the piezoelectric sensor unit is ideally placed on the outside of the screw (away from the center of the spine), about 1-2 cm away from the screw. Simultaneously, the system links the screw location number with the piezoelectric sensor number. For example, six steel nails were placed in the thoracic vertebrae 1-3, namely: piezoelectric sensors Y01-8a, Y02-8b, Y03-9a, Y04-9b, Y05-10a, and Y06-10b.

[0025] The lower-level machine mainly consists of signal conversion and processing, processor, storage, transmission, indicator lights / display screen, and communication module. It primarily performs functions such as analog signal filtering, analog-to-digital conversion, data storage, data transmission, communication protocols, status indication, and communication. The signal conversion and processing section handles analog signals from piezoelectric ceramic sensors and digital signals from other sensors; the processor section schedules the normal operation of each part; the storage section includes a large-capacity FLASH or SD memory card to store data and parameters; the transmission section communicates with the upper-level machine via various wireless protocols. The display section uses LEDs or LCD screens of different colors to indicate the system status. The hardware kernel supports online upgrades; upgrade programs can be downloaded to the kernel via a server or mobile phone to automatically improve system performance and stability. The communication section is responsible for communicating with the cloud data center and can also communicate with mobile phones, allowing the phone to configure lower-level machine parameters or view real-time data. The transmission unit uses WIFI, Bluetooth, Ethernet cable, 2G / 3G / 4G, and other communication methods to transmit data. The WIFI module can connect to the network either as a wireless network card or via a universal twisted-pair Ethernet cable interface. The Bluetooth module can automatically adapt to Bluetooth 2.0 and Bluetooth 4.0.

[0026] The cloud data center is responsible for data communication, status communication, and interactive control with the lower-level hardware. It stores data collected by the lower-level hardware in real-time or with delay in the cloud data center and monitors the working status of the lower-level hardware in real time, issuing configuration parameters and status control commands. The server receives data from the device via TCP / IP communication and performs specific operations using the encapsulated SDK and API functions. The server can receive, process, store, and display data. To improve the server's data receiving capacity, multiple servers can be combined into a load-balanced cluster. The server can configure the device's parameters through data communication. Current relational and non-relational databases generally suffer from high data redundancy and slow write / read speeds when storing large amounts of human physiological parameters. This application provides a data storage method that stores data sequentially as a byte stream. Through a defined data structure, it can significantly improve the read and write speeds of large amounts of physiological parameters, reduce data redundancy, save storage space, and effectively reduce storage costs. The data storage file stores data sequentially as a byte stream, and the file name is in the format of "ID + sequence number". The file consists of four parts: ID1, index, description information, and data. ID1 stores the user ID and version number; the index identifies the index location of the data storage; the description information provides specific notes about the data; and the data section stores the actual data size. If the data file becomes too large, a new data storage file can be generated to store the new data. Furthermore, a time-series data compression algorithm selection mechanism is provided, allowing applications to autonomously select the most suitable compression algorithm based on the characteristics of the actual data. This significantly improves data compression efficiency while effectively reducing storage pressure and increasing write and query speeds.

[0027] Data sent from the lower-level hardware module to the cloud data center can be set for a specified time period. Data within this period can be stored in memory to improve read performance. The data cache in memory is continuously updated over time. The real-time data analysis plugin can access the data in the real-time data cache module and perform relevant analysis. The results analyzed by the real-time data analysis plugin can also be written to a time-series database. Alternatively, the analysis results can be read or written to external storage via an interface. The specific implementation steps are as follows: The cloud data center SDK defines an interface for time-series data processing: IPlugin. This interface mainly includes an initialization function (Initialize), a time-series data processing function (ProcessData), and a shutdown function (Shutdown). Specific processing programs need to implement the IPlugin interface, and the implementation is placed as a dynamic link library in a designated location (hereinafter referred to as the plugin directory). When the database platform starts, it scans the dynamic link libraries in the plugin directory and loads them into memory. It checks whether these dynamic link libraries implement the IPlugin interface. If they do, it instantiates the interface and calls its initialization function (Initialize). Finally, it saves the successfully initialized plugin instance in memory. When the database platform receives new time-series data, it first saves it to a database file through the time-series data storage mechanism. Then, it sequentially sends the time-series data to the ProcessData data processing program of each plugin stored in memory, thus completing the data processing for each plugin. When the database platform stops running, it sequentially calls the Shutdown function of each plugin to notify each plugin to release resources, clear caches, and stop running.

[0028] Then, in the cloud data center, the collected data is analyzed in real time using algorithms to obtain the user's spinal stress distribution, assess the user's spinal stress tolerance, and evaluate the possibility of infection and rejection. Upon detecting any abnormalities, an alert is immediately issued to doctors or caregivers via an app or WeChat mini-program.

[0029] The following section focuses on how to analyze the collected muscle tremor data. It's important to note that tremor frequencies can be categorized as low-frequency, mid-frequency, and high-frequency. Low-frequency tremors are those less than 4 Hz, commonly seen in cerebellar tremor, red nucleus tremor, and soft palate tremor. Mid-frequency tremors range from 4 to 6 Hz, commonly seen in Parkinson's disease tremors (4-6 Hz) and essential tremor in the elderly (4-6 Hz). Dystonia and psychogenic factors can both cause mid-frequency tremors. High-frequency tremors are those above 6 Hz. Essential tremor in younger patients is high-frequency, and some patients with dystonia and psychogenic factors may also exhibit high-frequency tremors. The frequency of the tremor can roughly distinguish whether a patient has essential Parkinson's disease or essential tremor. The analysis in this application primarily focuses on muscle tremors caused by pain, with a tremor frequency of 4-6 Hz being the main focus.

[0030] See the instruction manual appendix Figure 5 The pain tolerance of patients was analyzed using muscle tremor data in the following manner, including the following steps: S201. After denoising the muscle tremor data by mean square error comparison, low-pass filtering and high-pass filtering are performed in sequence to obtain the core signal for frequency band extraction. S202. Perform FFT transformation on the core signal to obtain the frequency domain signal, extract the 4-6Hz component and calculate the ratio of the power in this frequency band to the total power, and use the ratio of the total power as a constraint condition to output the target waveform through end-to-end timing modeling. S203. After performing derivative filtering, square filtering, integral filtering and median smoothing filtering on the output target waveform in sequence, the hill climbing method is used to detect all local peaks, and the pain tremor peak and noise peak are determined by adaptive threshold. S204. Statistically count the number of occurrences of pain tremor peaks in the 4-6Hz frequency band, adjust the force on the steel nails in combination with the type of spinal curvature, and trigger an early warning when the individual's tolerance threshold is exceeded.

[0031] In one embodiment, each piezoelectric sensor unit continuously transmits data to the cloud data center at a rate of 128 data points per second. The measured value of each data point is typically in the range of 0-4096, and is an integer variable. Generally, piezoelectric data analysis and evaluation are performed on an hourly basis, and the raw muscle tremor data is output and labeled as A(t).

[0032] In step S201, data denoising is first performed to eliminate random interference and ensure the validity of the original data; then filtering is performed to purify the signal frequency and focus on the effective jitter frequency band.

[0033] Specifically, the mean squared error (MSE) of the raw muscle tremor data collected within one hour is calculated to obtain the MSE of the sensor's data change within that hour, Avg1. Starting from second 0, the MSE is calculated every 4 seconds. If the MSE of the data within these 4 seconds is greater than Avg1, this data is marked as noise data. The identified noise data portion in the raw muscle tremor data A(t) is changed to 0, resulting in a new array C(t). Then, a combined low-pass and high-pass filtering strategy is used to further remove redundant frequency components from the denoised data. First, a low-pass filter is performed, and the filtered result is denoted as L(t). The filtering algorithm is as follows: Where 'a' is the low-pass filter parameter, typically a = 0.145. Then comes the high-pass filter, the filtered result denoted as X(t), and the filtering algorithm is as follows: , where b is the high-pass filter parameter, which can generally be set to: b=0.8.

[0034] In step S202, pain-related muscle tremor signals, specifically the 4-6 Hz tremor frequency, are accurately extracted through frequency domain transformation and AI modeling. This is because pain-induced muscle tremors are mostly concentrated in this frequency band. First, the core signal X(t) obtained in step S201 is converted into a frequency domain signal X(f), transforming the time-amplitude relationship into a frequency-power relationship, facilitating direct location of the 4-6 Hz frequency band. After extracting the component indices with frequencies between 4 Hz and 6 Hz, the power of the selected frequency band is obtained by taking the square of the modulus of the complex spectrum and summing the results. Then calculate the total power. ,pass The energy proportion of this frequency band is quantified to preliminarily determine the intensity of the effective jitter. To address potential distortions in the original waveform, a combined model of 1D convolutional layers (extracting local time-domain features, such as jitter peak intervals) and LSTM layers (capturing long-term time-series dependencies, such as jitter frequency stability) is used to directly predict the target waveform after filtering from X(t) to 4-6Hz.

[0035] The model training uses the Adam optimizer (learning rate 1e-4) and batch size 32-128, with the loss function being a weighted sum of time-domain MSE (ensuring waveform shape matching) and frequency-domain MSE (ensuring frequency band energy matching).

[0036] in, For frequency domain loss, The waveform data predicted by the model at time i. Let N be the actual target waveform data of the model at time i, and N be the number of data points in the sample. For frequency domain loss, To predict waveforms After FFT transformation, the amplitude at frequency f, For the target waveform After FFT transformation, the amplitude at frequency f, This represents the frequency domain loss weighting coefficient.

[0037] In step S203, the target waveform output by AI is further optimized through four steps of derivative, square, integral, and median smoothing, which enhances effective tremor features (such as peak clarity) and suppresses residual noise.

[0038] Specifically, derivative filtering is performed first. The derivative filter removes the DC component of the input and achieves linear gain at high frequencies (enhancing the slope). It can be viewed as a high-pass filter. The transfer function is:

[0039] Output With input The relationship is:

[0040] in, For unit delay operators, The input data for derivative filtering, The output data of derivative filtering, The input data is from the previous moment, and the output value reflects the rate of change of the signal between the current moment and the previous moment. The greater the rate of change, the steeper the waveform slope.

[0041] Continue with square filtering. The purpose of square filtering is to make the sample values ​​positive, which further enhances the slope and highlights the wave.

[0042] Output of the square filter and input The relationship is .

[0043] Perform an integral filter (the value of N is important) to smooth the output. The output of the integral filter. and input The relationship is If the value of N is too large, it will overwhelm the wave group. If N is too small, many waves will be generated in the wave group. A value of 30 for N is suitable. The integral filter has a delay of approximately 21 samples.

[0044] Continue with median smoothing filtering, N-point median smoothing filtering. In one embodiment, see the appendix to the specification. Figure 6 The bottom waveform represents the original data, the middle waveform represents the data after square filtering, and the top waveform represents the data after integral filtering and median smoothing.

[0045] Finally, by using the hill climbing method to find peaks and adaptive threshold classification, pain-related muscle tremor peaks were accurately identified from the optimized waveform.

[0046] First, the waveform data is traversed to find all local peaks PEAK (i.e., the turning points of "uphill followed by downhill"), covering both pain tremor peaks and noise peaks. Then, a dynamically updated threshold THRE_1 is used to distinguish between muscle tremor peaks S_PKI and noise peaks N_PKI. Specifically, if PEAK > THRE_1, it is determined to be a pain tremor peak; where PEAK represents the local peak found by the hill-climbing method, THRE_1 represents the threshold for distinguishing between pain tremor peaks and noise peaks, and THRE_1 is updated using the following formula: S_PKI = 0.125 PEAK + 0.875 S_PKI old N_PKI = 0.125 PEAK + 0.875 N_PKI old THRE_1=N_PKI+0.25(S_PKI- N_PKI) Where S_PKI represents the peak amplitude of the corresponding muscle tremor wave, S_PKI old S_PKI is the value from the last update. N_PKI represents the peak amplitude of the non-muscle tremor wave. old This refers to the N_PKI as it was last updated.

[0047] In step S204, the statistical results of the muscle tremor peak S_PKI in the 4-6Hz frequency band are combined with the spinal curvature type to determine whether the force on the steel nail exceeds the patient's tolerance threshold. Excessive force on the steel nail can cause pain, which leads to muscle tremors being collected by the sensor. Therefore, the more times the 4-6Hz frequency band S_PKI occurs, the more pronounced the pain near the corresponding steel nail and the greater the force. By statistically analyzing the cumulative number of S_PKI occurrences over a certain period, the force intensity can be quantified.

[0048] In one embodiment, the piezoelectric sensor units near the steel nail are individually assessed to determine if there is a vibration signal near the nail. If a vibration signal is present, it indicates a force deviation in the adjustable rod near that point. The system or mobile app then issues a warning to medical staff, reminding them to adjust the force on the steel nail and the lever at that location.

[0049] Furthermore, algorithms can be used to analyze the user's spinal stress distribution in real time, enabling an assessment of the user's spinal stress tolerance. For example, by analyzing the number of 4-6Hz peaks appearing in the data collected by the piezoelectric ceramics near each steel nail over a certain period, and accumulating these counts, the stress tolerance of the steel nails at different locations can be determined. Generally, steel nails have no corrective effect if no stress is applied, while excessive stress can cause symptoms such as pain. Pain leads to muscle tremors, which are detected by the piezoelectric sensor unit, indicating high pressure near that unit. The pressure can be proportionally calculated based on the number of muscle tremors. The optimal stress level for each steel nail is determined based on the different degrees of spinal curvature. For spinal curvature to the left, it is preferable for the right-side screws to bear greater force, but this should not exceed the set tolerance value. It is recommended that the force be applied from the lumbar spine to the cervical spine, increasing or decreasing with the degree of curvature. The degree of increase or decrease is directly proportional to the curvature ratio, and the force level is determined by the number of muscle tremors mentioned above.

[0050] For spine curvature to the right, it's preferable to apply greater force to the left-side screws, but this should not exceed the set tolerance value. It's recommended that the force be applied from the lumbar spine to the cervical spine, increasing or decreasing with the degree of curvature. The degree of increase or decrease is directly proportional to the curvature ratio, and the force level is determined by the number of muscle tremors mentioned above.

[0051] For correction of forward flexion of the spine, it is better to have greater force on the lower part of the steel screws, but not exceeding the set tolerance value; it is recommended to apply force from the lumbar spine to the cervical spine, with the force increasing or decreasing with the degree of curvature. The degree of increase or decrease is directly proportional to the curvature ratio, and the force level depends on the number of muscle tremors mentioned above.

[0052] For correction of spinal kyphosis, it is preferable that the upper part of the steel screws bear greater force, but this should not exceed the set tolerance value. It is recommended that the force be applied from the lumbar spine to the cervical spine, increasing or decreasing with the degree of curvature. The degree of increase or decrease is directly proportional to the curvature ratio, and the force level is determined by the number of muscle tremors mentioned above.

[0053] See the instruction manual appendix Figure 7 The infection rejection risk analysis of muscle tremor data was performed using the following methods, including the following steps: P201. After denoising the muscle tremor data by means of variance comparison, low-pass filtering is performed sequentially to obtain a purified signal reflecting changes in skin temperature. P202. The purified signal is subjected to spike noise replacement, extreme value trimming, basic statistical calculation, fluctuation analysis, and data validity determination in a fixed processing window of a set duration to obtain effective data reflecting changes in skin temperature. P203. By analyzing the variation range of the effective data, determine whether the patient is at risk of infection and rejection.

[0054] The reason why the risk of infection and rejection can be determined by analyzing the muscle tremor data collected by the piezoelectric sensor is that the piezoelectric sensor is sensitive to temperature. Since the piezoelectric sensor is attached to the skin, the average value of the muscle tremor data collected by the piezoelectric sensor will decrease as the temperature rises.

[0055] Specifically, the denoising and filtering process in step P201 is similar to that in step S201 (high-pass filtering is no longer performed), and will not be elaborated here. In step P202, after obtaining the new array L(t) in step P201, further clutter removal is performed to eliminate interference clutter. The following sections are all... Each data point (2 seconds of data) is processed.

[0056] Spike noise handling: Iterate through the original data array, checking each point for spikes: if the difference between the current point and the previous or next point exceeds 1.5 times the sensor sensitivity (configurable, typically set to 20), and the difference between the two points is less than the sensitivity, then it is identified as a spike. Replace the value of the spike point with the average of the two points. Copy the processed data to a temporary array `dataArray`.

[0057] Data sorting and extreme value pruning: Sort the temporary array dataArray in ascending order. Set the values ​​of the first deltaCount elements after sorting to the value of the (deltaCount+1)th element (eliminating the smallest extreme value). Set the values ​​of the last deltaCount elements to the value of the (deltaCount+1)th element from the end (eliminating the largest extreme value).

[0058] Basic statistical calculations: Iterate through the original data, i.e., the data from 2 seconds. ): Calculate the maximum value max and the minimum value min; accumulate the sum sum for subsequent average calculation; calculate the moving average for every 8 elements as a group, divide the 256 points into 32 groups, and calculate the maximum value avgmax and the minimum value avgmin of the moving average in the 32 groups of data.

[0059] Volatility analysis, 2 seconds of data ( : Calculate the overall mean avg; calculate the mean absolute difference avgdelta (the mean absolute difference between each data point and the mean).

[0060] Data validity determination, 2 seconds of data ( ): Data is considered invalid if any of the following conditions are met: the minimum value is greater than 1700 or the overall average value (avg) is greater than 1800. The moving average fluctuation (avgmax - avgmin) and the mean absolute difference are both less than the preset sensitivity, ShakeSensitivity.

[0061] Invalid data handling: Set all elements of the original array to zero; the data from these two seconds is invalid.

[0062] This process involves denoising, pruning extreme values, and analyzing data volatility to ultimately determine the validity of the data. Invalid data is cleared and a zero difference is returned, while valid data is returned as valid data. This achieves the filtering effect.

[0063] In step P203, the average value of the data after noise reduction is calculated over a certain time period (generally set to 60 seconds). Using one hour as the unit, the raw data yields 60 points. If the difference between the maximum and minimum values ​​among these 60 points exceeds the set threshold (temperature change threshold), it indicates a significant temperature change on the skin surface within that hour, which can serve as an alert for medical personnel.

[0064] As can be seen, the force assessment method proposed in this application for multiple spinal correction surgeries has two main advantages. First, it collects muscle tremor signals without load using multiple piezoelectric sensors and accurately extracts pain-related frequency signals in the 4-6Hz band using AI algorithms such as 1D-CNN+LSTM. This enables real-time assessment of spinal force distribution and pain tolerance, overcoming the shortcomings of traditional methods that rely on experience to judge traction force and ensuring that the traction device force matches the patient's individual tolerance. Second, it utilizes the temperature sensitivity of piezoelectric sensors to analyze skin temperature changes through data processing, enabling real-time early warning of infection and rejection reactions near the steel nails. Compared with the traditional periodic examination mode, this significantly improves the timeliness of complication intervention.

[0065] Furthermore, abnormal warnings can be provided via an app or mini-program, allowing medical staff to adjust the stress on the steel nails and tie rods in a timely manner. This protects the patient's body while accelerating the spinal correction and repositioning process, shortening the pain caused by the 1-2 month repositioning cycle. It is suitable for various scenarios involving multiple spinal correction surgeries and has significant clinical application value.

[0066] Based on the same inventive concept, this application also provides a force assessment system for multiple spinal correction surgeries, including a piezoelectric sensor unit, a lower-level computer, and a cloud data center, for collaboratively implementing the steps of the aforementioned force assessment method for multiple spinal correction surgeries. Since the principle of the system in this application's solution is similar to the aforementioned force assessment method for multiple spinal correction surgeries, the implementation of the system can refer to the implementation of the method; repeated details will not be elaborated further.

[0067] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0068] Based on the same concept of the present invention, as shown in the appendix to the specification. Figure 8As shown in the figure, an embodiment of this application provides the structure of an electronic device 800, which includes: at least one processor 801, at least one network interface 804 or other user interface 803, a memory 805, and at least one communication bus 802. The communication bus 802 is used to realize the connection and communication between these components. The electronic device 800 may optionally include a user interface 803, including a display (e.g., touch screen, LCD, CRT, holographic imaging, or projector, etc.), a keyboard, or a clicking device (e.g., mouse, trackball, touchpad, or touch screen, etc.).

[0069] Memory 805 may include read-only memory and random access memory, and provides instructions and data to processor 801. A portion of memory 805 may also include non-volatile random access memory (NVRAM).

[0070] In some implementations, memory 805 stores executable modules or data structures, or subsets thereof, or extended sets thereof: The 8051 operating system contains various system programs used to implement various basic business functions and handle hardware-based tasks. Application module 8052 contains various applications, such as launcher, media player, and browser, to implement various application functions.

[0071] In this embodiment of the application, the processor 801 is used to execute steps such as a force assessment method for multiple spinal correction surgeries by calling a program or instruction stored in the memory 805.

[0072] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs steps such as those in a stress assessment method for multiple spinal correction surgeries.

[0073] Specifically, the storage medium can be a general-purpose storage medium, such as a portable disk or hard disk. When the computer program on the storage medium is run, it can realize real-time analysis of spinal stress distribution, assessment of patient pain tolerance, and early warning of postoperative infection and rejection, which can both protect patient safety and improve surgical progress.

[0074] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interface, and the indirect coupling or communication connection of the apparatus or units may be electrical, mechanical, or other forms.

[0075] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0076] In addition, the functional units in the embodiments provided in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0077] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0078] Finally, it should be noted that the above embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A method for stress evaluation during multiple spinal orthopedic procedures, comprising: The method is applied to a stress evaluation system, the system comprising a piezoelectric sensor unit, a lower computer and a cloud data center, and the method comprising the following steps: Collecting muscle tremor data of a patient based on the piezoelectric sensor unit and transmitting the data to the cloud data center through the lower computer; Analyzing the muscle tremor data for patient pain tolerance and infection rejection risk based on the cloud data center.

2. The method for stress evaluation during multiple spinal orthopedic surgical procedures of claim 1, wherein, Wherein, Collecting muscle tremor data of a patient based on a multi-channel piezoelectric sensor unit; the piezoelectric sensor unit is attached to the outside of the patient's spinal steel nail and comprises a piezoelectric sensor, a driving circuit, a medical non-woven adhesive tape, a cushion block and a cable; the piezoelectric sensor adopts a piezoelectric ceramic sheet element; the cushion block faces the skin.

3. The method for stress evaluation during multiple spinal orthopedic surgical procedures of claim 2, wherein, The patient pain tolerance analysis of the muscle tremor data is performed by the following method, comprising the following steps: After denoising the muscle tremor data by mean square difference comparison, sequentially performing low-pass filtering and high-pass filtering to obtain a core signal for frequency band extraction; Performing FFT transformation on the core signal to obtain a frequency domain signal, extracting a 4-6Hz component and calculating the ratio of the power of the frequency band to the total power, and taking the ratio of the total power as a constraint condition, outputting a target waveform through end-to-end time series modeling; After sequentially performing derivative filtering, square filtering, integral filtering and median smoothing filtering on the output target waveform, detecting all local peaks by hill climbing method, and determining pain tremor peaks and noise peaks by adaptive threshold; Counting the number of pain tremor peaks in the 4-6Hz frequency band, adjusting the stress of the steel nail according to the type of spinal curvature, and triggering an early warning when the individual tolerance threshold is exceeded.

4. The method for stress evaluation during multiple spinal orthopedic surgical procedures of claim 3, wherein, The pain tremor peaks and noise peaks are determined by an adaptive threshold, comprising the following steps: If PEAK>THRE_1, it is determined as a pain tremor peak; wherein PEAK represents a local peak found by hill climbing method, THRE_1 represents a threshold value for distinguishing pain tremor peaks from noise peaks, and THRE_1 is updated by the following formula: S PKI = 0.125 PEAK + 0.875 S PKI old N PKI = 0.125 PEAK + 0.875 N PKI old THRE_1=N_PKI+0.25(S_PKI-N_PKI) wherein S_PKI represents the peak amplitude of the corresponding muscle tremor wave, S_PKI old is the S_PKI updated last time, and N_PKI represents the peak amplitude of the non-muscle tremor wave, N_PKI old is the N_PKI updated last time.

5. The method for stress evaluation during multiple spinal orthopedic surgical procedures of claim 3, wherein, The steel nail stress is adjusted according to the type of spinal curvature, comprising the following steps: When the spine bends to the left, the right steel nail is stressed more than the left steel nail; when the spine bends to the right, the left steel nail is stressed more than the right steel nail; when the spine is folded forward for correction, the lower steel nail is stressed more than the upper steel nail; when the spine needs to be folded backward for correction, the upper steel nail is stressed more than the lower steel nail; The stress of the steel nail is adjusted with the constraint that the number of pain tremor peaks does not exceed the individual tolerance threshold.

6. The method for stress evaluation during multiple spinal orthopedic surgical procedures of claim 1, wherein, The infection rejection risk analysis of the muscle tremor data is performed by the following method, comprising the following steps: After denoising the muscle tremor data by mean square difference comparison, sequentially performing low-pass filtering to obtain a purified signal reflecting skin temperature changes; The purified signal is sequentially subjected to spike noise replacement, extreme value pruning, basic statistical calculation, volatility analysis and data validity determination with a fixed processing window of a set time length to obtain valid data reflecting skin temperature changes; By analyzing the change amplitude of the valid data, it is determined whether the patient has infection rejection risk.

7. The method for stress evaluation during multiple spinal orthopedic surgical procedures of claim 1, wherein, The cloud data center receives the lower machine data through TCP / IP communication, and performs data processing through the encapsulated SDK and API function.

8. A stress evaluation system for use in multiple spinal orthopedic procedures, comprising: The piezoelectric sensor unit, the lower machine and the cloud data center are used to cooperatively realize the steps of the stress evaluation method for the multiple spinal orthopedic surgery process according to any one of claims 1 to 7.

9. An electronic device, comprising: The piezoelectric sensor unit, the lower machine and the cloud data center are used to cooperatively realize the steps of the stress evaluation method for the multiple spinal orthopedic surgery process according to any one of claims 1 to 7. The piezoelectric sensor unit, the lower machine and the cloud data center are used to cooperatively realize the steps of the stress evaluation method for the multiple spinal orthopedic surgery process according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The piezoelectric sensor unit, the lower machine and the cloud data center are used to cooperatively realize the steps of the stress evaluation method for the multiple spinal orthopedic surgery process according to any one of claims 1 to 7.

Images