Magnetic wave knife alarm data intelligent discrimination method and system based on threshold value self-adaption

Abstract

The invention discloses a method and a system for intelligently judging alarm data of a magnetic wave knife based on threshold self-adaption, and relates to the technical field of intelligent data judgment. The magnetic wave knife alarm data intelligent discrimination method based on threshold self-adaption comprises the steps that a training instruction is sent out during preoperative pre-collection, and training double-side pressure values of sensors on the two sides of a handheld double-side pressing device are collected in real time; effectively collected training action data are analyzed, and an alarm threshold value of an alarm device is adjusted to adapt to different individual differences; real-time action data are collected in real time during the operation, and the real-time action data comprise two-side pressure values collected through a handheld double-side pressing device, and the two-side pressure values are actually measured double-side pressure values; and performing analysis according to the real-time action data acquired in real time, and outputting early warning information. According to the invention, intelligence and real-time performance of alarm data processing are realized, and the problem that in the prior art, the patient feedback effectiveness is reduced in the treatment process due to too large manual screening workload of medical staff is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of intelligent data discrimination technology, and in particular to an intelligent discrimination method and system for magnetic wave knife alarm data based on threshold adaptation. Background Technology

[0002] The magnetic resonance imaging (MRI) knife is not a real metal blade, but a cutting-edge treatment system that integrates physics, engineering, and medicine. Its core concept is to precisely remove or ablate lesions inside the body without cutting the skin or damaging normal tissue, thus achieving non-invasive treatment. Due to differences in skull density and individual tolerance among patients, there are still treatment risks during the MRI knife treatment process. For example, if the burning temperature is too high, causing unbearable pain or obvious adverse reactions in the patient, the patient must report it immediately, and a professional physician will decide whether to terminate the treatment. Therefore, the timing of the patient's report is extremely important. It is necessary to make an effective alarm by pressing the alarm device to provide timely feedback on the status. The doctor will then proceed with the surgery based on the alarm status to successfully complete the operation.

[0003] In existing technology, the alarm device used is a cylindrical device with an alarm button at the top. The alarm can only be triggered by holding the cylinder with four fingers and pressing the alarm button with the thumb. However, because the alarm button is too small, patients may have difficulty finding it in an emergency and thus fail to trigger the alarm device. In addition, the alarm is only a single button, which often leads to accidental activation during treatment, inevitably interfering with the treatment process and affecting the doctor's accurate judgment. In addition to the defects of the alarm device itself, there is also a lack of evaluation of the alarm signal. Relying solely on pressing the alarm cannot accurately identify the patient's status through the alarm signal, and it is still impossible to effectively determine whether to continue the surgery. The excessive workload of medical staff in manually verifying the alarm reduces the effectiveness of patient feedback during the treatment process. Summary of the Invention

[0004] To address the technical problem of reduced patient feedback effectiveness during treatment due to excessive manual screening by medical staff in existing technologies, this invention provides a threshold-adaptive intelligent judgment method and system for magnetic resonance imaging (MRI) knife alarm data. The technical solution is as follows:

[0005] On the one hand, a threshold-adaptive intelligent judgment method for magnetic resonance scalpel alarm data is provided. This method includes: during preoperative pre-collection, issuing training instructions and collecting training action data in real time, including bilateral pressure values ​​collected by sensors on both sides of the handheld bilateral compression device; analyzing the effectively collected training action data and adjusting the alarm threshold of the alarm device to adapt to different individual differences; during surgery, collecting real-time action data, including bilateral pressure values ​​collected by the handheld bilateral compression device, which are the actual measured bilateral pressure values; and analyzing the real-time collected action data to output early warning information.

[0006] On the other hand, a threshold-adaptive intelligent judgment system for magnetic resonance oscillator alarm data is provided. This system employs a threshold-adaptive intelligent judgment method for magnetic resonance oscillator alarm data and includes: a preoperative training module, an alarm threshold adjustment module, a motion data acquisition module, and a warning information output module. The preoperative training module issues training instructions and collects training motion data in real time during preoperative data acquisition, including bilateral pressure values ​​collected by sensors on both sides of the handheld bilateral compression device. The alarm threshold adjustment module analyzes the effectively collected training motion data and adjusts the alarm threshold of the alarm device to accommodate individual differences. The motion data acquisition module collects real-time motion data during surgery, including bilateral pressure values ​​collected by the handheld bilateral compression device, which are the measured bilateral pressure values. The warning information output module analyzes the real-time collected motion data and outputs warning information.

[0007] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. Preoperatively, the system acquires bilateral compression capacity through training instructions and adaptively adjusts the alarm threshold. During the operation, it continuously analyzes the measured bilateral pressure values ​​and outputs early warning information. Compared with the fixed threshold or manual experience-based threshold setting commonly used in existing technologies, which only make simple trigger judgments, this intelligent alarm data discrimination method can maintain the consistency and controllability of alarm triggering under different patient strength differences and left-right hand differences. This reduces the risk of not being able to trigger the alarm due to insufficient pressure or the risk of accidental triggering due to excessively low thresholds, improves the reliability and safety of intraoperative feedback, and reduces the burden of manual verification of alarm validity by medical staff.

[0008] 2. When setting individualized thresholds, the differences between the left and right hands are often ignored. When both sides are required to trigger simultaneously, the stronger side may be able to reach the threshold while the weaker side is not. This can lead to patients having the intention to trigger the alarm but being unable to do so stably, or the threshold being lowered too much to accommodate the weaker side, causing false triggers. By extracting and sorting the maximum pressure values ​​of both sides from the effective compression segments, and selecting the median as the representative compression force of each side, the system can characterize stable and reliable compression ability. After removing abnormally low values, a fusion strategy with weak side constraints is adopted to make the overall adjustment factor converge towards the weaker side. Finally, bilateral constraints are applied through the trigger lower limit and trigger upper limit. This ensures that the weaker side can still reach the trigger condition when both sides are pressed simultaneously, while avoiding the threshold being lowered without a bottom line, thus achieving individualized sensitivity matching.

[0009] 3. For tremor-induced jerking interference, existing technologies often use simple amplitude thresholds or fixed filters, which are often difficult to balance preventing false alarms caused by jerking and ensuring that real alarms are not disturbed. Before the operation, a natural grip command is used to determine the jerking elimination threshold. During the operation, the jerking is first masked using this threshold, and then frequency matching is further performed on the segment between the jerking threshold and the pain level threshold to distinguish between suspected jerking and suspected alarms. Compared with existing technologies, this significantly reduces false alarms induced by jerking in tremor scenarios, while providing continuous quantification of urgency and increasing trend warning information output for real alarms, which facilitates rapid triage and priority response by medical staff.

[0010] 4. Instead of judging alarms solely based on a single pressure value, it comprehensively considers multi-dimensional temporal characteristics such as excessive pressure, number of consecutive presses, and the proportion of rapid consecutive presses, achieving holographic quantification of behavioral information. Through various influences such as pressure intensity and continuous pressing trend, it outputs a continuous score value that dynamically reflects the urgency of the alarm. This score is not only used for alarm signal output but can also directly drive the multi-modal prompt intensity of sound and indicator lights, achieving dynamic early warning reminders from weak to strong trends, rather than simple fixed-level prompts. It can effectively distinguish between occasional false alarms and genuine continuous high-risk alarms, improving the accuracy of alarm judgment and the ability to prioritize interventions. It is adaptive and can automatically adjust the scoring threshold according to the patient's actual alarm behavior, reducing the limitations of manually setting levels. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of an intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation provided in an embodiment of the present invention; Figure 2 This is a flowchart of the output early warning information provided in an embodiment of the present invention; Figure 3 This is a flowchart of jitter interference determination provided in an embodiment of the present invention; Figure 4 This is a block diagram of a threshold-adaptive intelligent judgment system for magnetic wave knife alarm data provided in an embodiment of the present invention. Detailed Implementation

[0013] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0014] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0015] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0016] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0017] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0018] This invention provides a threshold-adaptive intelligent discrimination method for magnetic wave knife alarm data, such as... Figure 1 The flowchart shown is for an intelligent discrimination method of magnetic wave knife alarm data based on threshold adaptation. The processing flow of this method may include the following steps: S1: During preoperative data collection, a training instruction is issued to collect training action data in real time, including bilateral training pressure values ​​collected by sensors on both sides of the hand-held bilateral compression device.

[0019] S2: Analyze the effectively collected training action data and adjust the alarm threshold of the alarm device to adapt to different individual differences.

[0020] S3: During the operation, real-time motion data is collected, including the pressure values ​​on both sides collected by the hand-held bilateral compression device, which are the actual measured bilateral pressure values.

[0021] S4: Analyze real-time action data and output early warning information.

[0022] Furthermore, during preoperative data collection, training instructions are issued, and training movement data is collected in real time. Specifically, the training instructions include instructions categorized by pain level. After each different training instruction is issued, real-time collection of training movement data begins, and the collection stops after a preset collection time.

[0023] Record all data for all instructions within the preset collection time, discard invalid data, and re-collect instructions that were not effectively collected; the pain level instructions include: pain situations that gradually increase from low to high, and require the trainee to apply the corresponding bilateral pressure according to their personal habits, covering different pain levels.

[0024] In this embodiment, the pain level instruction refers to the system artificially dividing the pain experience into multiple levels, such as no pain, mild pain, moderate pain, and severe pain, according to pre-operative presets. The system then issues instructions to the trainee through the system interface. The trainee needs to actively simulate or recall different levels of pain according to the instructions and press the buttons on both sides of the alarm device in this state. The system collects the pressure data generated by pressing with both hands simultaneously in real time. For example, under the mild pain instruction, the trainee simulates the natural reaction intensity of mild discomfort when pressing; under the severe pain instruction, the trainee simulates the pressing intensity of a severe pain state as much as possible. Through this process, individualized movement characteristics corresponding to different pain levels can be captured, which is beneficial for the adaptive adjustment of subsequent thresholds.

[0025] The data collection process covering different pain levels refers to issuing multi-level pain commands sequentially, guiding the trainee to perform a complete round of compression actions. Data is collected for each level of pain command, recorded, and labeled as the corresponding level, until all set pain levels have been collected. This ensures full coverage of individual response differences and provides multi-level reference data for subsequent analysis. For example, the patient sequentially performs four simultaneous hand compressions at levels 0 (no pain), 1, 2, and 3. Each compression is collected independently, and the data is automatically stored with the corresponding pain level label after collection. If a level is invalid, only that level is re-collected, without needing to redo the entire process.

[0026] The personal habitual bilateral simultaneous pressure intensity refers to the trainee's natural pressure application while maintaining synchronization, based on their own habitual force application patterns, postures, and subjective experience of pain levels. It does not force a uniform absolute force, but records the patient's most authentic and conventional pressure application performance to ensure that the threshold setting is more in line with the individual's actual situation.

[0027] Each data collection action has a limited collection duration and automatically determines whether the collected data is continuous and complete. Data lost due to operational errors or signal abnormalities is automatically discarded and prompts for recollection, ensuring that effective samples can be obtained at each level. This not only guarantees the objectivity of individual differences but also provides scientific and comprehensive basic data for subsequent threshold adjustments.

[0028] Furthermore, within the preset collection time, invalid collection data is removed, and instructions that were not effectively collected are re-collected. The specific process is as follows: based on the preset collection time period corresponding to the training instructions, the validity of the bilateral training action data is checked. The bilateral training action data includes at least the sampled values ​​of the bilateral data channels and their timestamps.

[0029] If there is a sampling interruption and missing data within the preset collection period, the collected data is judged as invalid and discarded. For the training instructions corresponding to the discarded invalid data, the same training instructions are reissued and the collection is executed again until the bilateral training action data meets the requirement of continuous collection without missing data within the preset collection period.

[0030] In this embodiment, data validity verification refers to the system checking the completeness and continuity of the acquired bilateral training action data during the acquisition process under each training instruction. The verification is not limited to the original sampled values ​​of the pressure sensor, but also includes the timestamp information of each sampled data to ensure the temporal consistency of the data. Specific standards include, but are not limited to: within the acquisition period, the data channels of both sensors should have complete sampled data, and the data acquisition time sequence of both sides should be consistent; there should be no obvious breakpoints or omissions in the data stream, that is, there should be no data without data or abnormally increased data intervals within the preset sampling interval; the timestamp sequence should be strictly incremental, without disorder, repetition or jump. This inspection method can promptly detect data loss and data interruption problems caused by sensor detachment, patient operation errors or system failures.

[0031] For any data collection deemed invalid, the corresponding training instruction level will be automatically recorded, and the user will be prompted on the operation prompt interface that the data collection is invalid and to re-execute the current training instruction. Subsequently, the same level of training instruction will be automatically issued, and data collection will restart until both data channels obtain complete and continuous data within the specified time. This ensures that the training action data at each level is real and valid, providing a reliable foundation for subsequent personalized parameter tuning.

[0032] In one specific embodiment, if the patient experiences a data interruption due to one finger slipping during a Level 2 pain command, a pop-up window will indicate that the data acquisition has been interrupted. The patient is then prompted to press the button again and complete the action at that level once more. This process continues until the system determines that the acquired data is valid before proceeding to the next step. This ensures that subsequent threshold tuning and individualized analysis are based on real and reliable training data, thereby improving the accuracy and stability of the system's judgment.

[0033] Furthermore, the effectively collected training action data is analyzed, and the alarm threshold of the alarm device is adjusted to adapt to different individual differences. Specifically, this includes: analyzing the training action data corresponding to the pain level instructions to obtain the pain level alarm threshold adjustment factor; and scaling the pain level alarm threshold using the pain level alarm threshold adjustment factor.

[0034] The specific process for analyzing the effectively collected training action data is as follows: For each effectively collected training action data, obtain the complete waveform of the bilateral pressure values ​​changing over time and identify the duration of the press; perform effective segment detection on the waveform and extract the corresponding effective press segments on both sides.

[0035] Based on the effective press segments, the maximum pressure values ​​of all effective press segments on both sides are determined and sorted from smallest to largest. The median of the maximum pressure values ​​is selected as the representative press force on that side. The representative press force is used to reflect the repeatable and stable pressing ability. At the same time, the smallest press pressure peak value on both sides is recorded as the minimum achievable force reference. The minimum achievable force is used to represent the minimum pressure value of the effective triggering ability.

[0036] In this embodiment, "dual sides" refers to the two pressing sides respectively set on the alarm device, namely: the left button acquisition channel and the right button acquisition channel. Valid segment detection refers to analyzing the acquired complete pressure-time waveform, eliminating irrelevant or invalid segments, and retaining only the intervals that truly reflect the trainee's effective pressing behavior. Specifically, this includes: setting start and end criteria based on the pressure value change curve, eliminating short-term pulses, signal jitter, or discontinuous abnormal waveforms, and retaining only segments whose duration exceeds a preset threshold. For each valid segment, its pressure peak value and duration are calculated to form the basis for subsequent analysis. The pressing duration segment is the length of the waveform segment identified as valid on the time axis, used to determine whether the pressing behavior is stable and sufficiently sustained.

[0037] The median of the maximum pressure values ​​is chosen as the representative pressure level to avoid interference from occasional outliers or short-term extreme force on the threshold setting. Since there may be occasional maximum values ​​in actual operation (such as misoperation, muscle tremors, etc.), the median can better reflect the most common and stable ability that an individual can achieve in multiple repetitions, thus improving the robustness of the threshold setting. Assuming that the peak values ​​of 5 effective compression segments obtained on the same side are 8N, 9N, 10N, 15N, and 20N respectively, the median is 10N. This is used as the representative pressure level, rather than directly taking the maximum value of 20N, to avoid the threshold being too high due to extreme values.

[0038] The minimum achievable force reference refers to the pressure value with the smallest peak value among all effective pressing segments. This indicator is used to evaluate the trainee's ability to effectively trigger the alarm device even when using the least force. When setting the lower limit of trigger sensitivity, it avoids the threshold being too high, which may prevent vulnerable patients from triggering the device. It also provides a basis for automatically identifying invalid force, accidental touches, and other situations, achieving adaptation to different individual differences and improving the accuracy of system judgment and the user-friendliness of human-computer interaction.

[0039] Furthermore, the pain level alarm threshold is scaled and adjusted using a pain level alarm threshold adjustment factor. The specific steps are as follows: based on the set standard reference pressure, the alarm threshold adjustment factor for both sides is calculated according to the representative pressure of both sides. At the same time, the alarm threshold adjustment factors of both sides are compared. The side with the smaller alarm threshold adjustment factor is defined as the weak side, and the side with the larger alarm threshold adjustment factor is defined as the strong side. For both sides of the button, the ratio of the representative pressure of the left side to the standard reference pressure is used as the left alarm threshold adjustment factor, and the ratio of the representative pressure of the right side to the standard reference pressure is used as the right alarm threshold adjustment factor.

[0040] After removing abnormally low values, the weak side is given a higher weight relative to the strong side, causing the pain level alarm threshold adjustment factor to converge towards the weak side, so as to ensure that the weak side can still reach the trigger condition when both sides are pressed at the same time. Based on the set alarm threshold, the alarm threshold is monotonically scaled according to the pain level alarm threshold adjustment factor. Monotonically scaling means using the pain level alarm threshold adjustment factor to proportionally adjust the set alarm threshold so that the threshold matches the patient's ability, thereby achieving individualized threshold matching between different patients.

[0041] In this embodiment, the standard reference pressure refers to a pre-set, unified benchmark value used to measure the patient's actual pressure level. This benchmark value can be set based on historical large-sample statistics or specific hardware tests. For example, it can refer to the average representative pressure value commonly seen when healthy adults complete the same training action. This standard value can serve as a comparison benchmark for all patients and can also be modified according to specific application scenarios. The alarm threshold adjustment factor is a ratio between the pressure of the patient's two fingers and the standard reference value. Specifically, the ratio of the representative pressure of the left or right side to the standard reference pressure is processed to obtain the alarm threshold adjustment factor for both sides. This coefficient reflects the actual force exertion ability of each side's fingers, facilitating normalization between different individuals. The side with the smaller alarm threshold adjustment factor is defined as the weak side, and the larger one is defined as the strong side. The weak side usually represents the side of the fingers where the patient exerts less force and is the key side that limits the overall triggering ability. Using the weak side as a constraint is to ensure that the alarm threshold setting will not cause the other side to fail to trigger effectively due to the stronger force on one side, thus maximizing the usability and safety of the patient during actual operation.

[0042] Abnormally low values ​​refer to abnormally low force values ​​caused by non-representative situations such as occasional collection errors, finger slippage, or temporary weakness. To prevent a single low value from causing the trigger threshold to be too low, statistical outlier detection is performed on all alarm threshold adjustment factors, and only typical levels are retained for fusion. This improves the representativeness and safety of the threshold. The method of fusing the two-sided alarm threshold adjustment factors is to use the weak side alarm threshold adjustment factor as the benchmark and combine it with the strong side alarm threshold adjustment factor. The fusion method is a weighted average, where the weak side has a higher weight than the strong side, and the fusion result does not exceed the smaller of the weighted average of the weak side and the strong side. The pain level alarm threshold adjustment factor obtained by weighted fusion is used as the basis for adjusting the alarm threshold, so that the parameter is closer to the weak side, but is not affected by extreme abnormally low values, thereby ensuring trigger sensitivity and reducing the risk of false alarms.

[0043] In one specific embodiment, the representative pressure applied to the left side of the patient is 8N, and to the right side is 12N. The standard reference pressure is 10N. Therefore, the alarm threshold adjustment factor for the left side is 0.8, and for the right side it is 1.2. Thus, the left side is the weak side, and the right side is the strong side. During fusion, the weak side is used as a constraint, that is, the weight of the weak value is increased when weighting the average. For example, the final coefficient = 0.9 × weak side + 0.1 × strong side, so that the final threshold is not higher than the achievable level of the weak side. In this example, 0.84 is obtained as the final adjustment factor. If the original threshold is 10N and the final adjustment factor is 0.84, then the adjusted pain level alarm threshold is 8.4N. Through the above multi-layer adjustment and constraint, different patients can obtain a pain level alarm threshold that is both safe and reliable and adaptable to individual differences based on their actual force application ability, which significantly improves the human-computer interaction and practical value of the system.

[0044] like Figure 2 The diagram shows a flowchart of the output warning information provided in an embodiment of the present invention. The specific design logic is as follows: First, the pressure signals of both buttons are collected, and it is determined whether the pressure on both sides meets the standard. If not, no alarm is triggered; if so, it is compared with the alarm trigger threshold. If the standard is met, the excessive force is calculated. Then, it is determined whether there are multiple presses within a short time window. If so, a second alarm message is output with continuous sound prompts and a gradual indicator light prompt; otherwise, a first alarm message is output with a normal alarm. If the pressure does not meet the alarm trigger threshold, no alarm is triggered.

[0045] Furthermore, scaling the pain level alarm threshold using the pain level alarm threshold adjustment factor also includes outputting an alarm signal based on the pain level alarm threshold. Specifically, this involves: determining whether the corresponding pain level alarm threshold is met based on the button pressure signals from both sides; and when both sides meet the standard, calculating the excess force of the current pressing segment by comparing the net pressure peak of the pressing segment with the pain level alarm threshold.

[0046] In the case of a single press, the first alarm message is output for each valid press based on the amount of force exceeding the limit; when multiple presses occur within a preset short-time analysis window, the amount of force exceeding the limit of each press is accumulated and then a second alarm message is output.

[0047] In this embodiment, the net pressure peak value refers to the result of subtracting the resting baseline pressure from the highest pressure value within the effective compression segment. It is used to truly reflect the intensity of the patient's actual compression action. The net pressure peak value is compared with the alarm trigger baseline pressure threshold that has been individually adjusted, and the difference between the two is calculated, i.e., the force excess. The larger the force excess, the greater the patient's compression intensity and the higher the subjective urgency of pain. The first alarm information is used to indicate that the effective triggering conditions have been met in a single compression action. After completing the standard determination and calculating the force excess, the first alarm information is output. This alarm information is usually a basic level alarm. The first alarm information is based on the immediate feedback of a single action and does not consider the number of historical compressions or the continuity of the action.

[0048] The system will accumulate the excess force of all pressing actions within a short analysis window. The larger the accumulated amount, the stronger and more continuous the distress signal, and the higher the alarm level. The purpose of this accumulation process is to enhance the sensitivity to continuous distress behavior and to distinguish the risk difference between a single mild discomfort and a short-term, high-frequency distress call. The second alarm message indicates that there are continuous pressing actions with a high accumulated force within a short period of time, which is a higher level alarm message. The first alarm message is output based on this excess force to indicate a general alarm event. If multiple pressing actions occur within the preset short analysis window, the alarm level will increase with the number and intensity of continuous pressing actions, and the second alarm message will be output.

[0049] Furthermore, a second alarm message is output. The specific process is as follows: within a preset short-time analysis window, the cumulative excess amount of the pressing segment is extracted, and the continuous pressing characteristics are statistically analyzed. The continuous pressing characteristics include the number of continuous pressings and the proportion of rapid pressing. The cumulative excess amount and the continuous pressing characteristics are fused together to obtain the alarm score for this pressing event. The alarm score satisfies the condition that the alarm score increases proportionally with the force intensity, the number of continuous pressings, and the proportion of rapid pressing.

[0050] The alarm score will be used to output an early warning prompt. The early warning prompt shall include at least an audible prompt and / or an indicator light prompt. The frequency and volume of the audible prompt shall increase as the alarm score increases. The brightness and flashing frequency of the indicator light shall increase as the alarm score increases.

[0051] In this embodiment, the cumulative excess is used to quantify the extent to which the trainee's actual force exceeds the corresponding pain level alarm threshold in a single effective compression action. Its core function is to reflect the strength of the compression action so as to classify and process it in subsequent alarm judgments. The number of consecutive compressions refers to the total number of consecutive compression actions judged as effective within a short analysis window. The proportion of rapid compressions refers to the proportion of rapid compressions with time intervals less than a preset rapid threshold in all consecutive compression actions. This indicator reflects the urgency of the patient's repeated and urgent requests for help.

[0052] Fusion calculation refers to the weighted summation of three features—cumulative excess volume, number of consecutive presses, and percentage of rapid presses—after normalization, resulting in a continuous alarm score. The fusion principle requires a positive correlation between the three features and the alarm score; that is, the greater the pressure, the more consecutive presses, and the higher the percentage of rapid presses, the higher the final alarm score. Based on the final alarm score, the output warning prompts are dynamically adjusted, including audible prompts: the prompt frequency (e.g., beep interval) and prompt volume (decibels) increase with the alarm score, achieving a gradual reminder from weak to strong; and indicator light prompts: brightness (LED power) and flashing frequency (flashes / second) also increase with the score, visually highlighting the urgency and ensuring that medical staff can intuitively perceive the alarm's urgency at different times and in different situations, enabling sensitive differentiation and response to alarm events of different urgency levels. In a specific embodiment, when the alarm score is low, a low-volume, low-frequency prompt sound and a slowly flashing indicator light are used; as the score gradually increases, a high-volume, continuous ringing sound and a high-brightness, high-frequency flashing LED are used.

[0053] Furthermore, the training instructions also include a natural grip instruction, specifically: the natural grip instruction requires the trainee to keep both hands naturally clenched without any pressing instructions; acquire resting segment data generated by natural hand contact, continuously collect resting segment data to calculate the resting baseline pressure, and combine the pressure signal to obtain a pressure jitter curve; perform statistical analysis on the amplitude fluctuation of the pressure jitter curve, use preset robust statistical indicators to determine the upper limit of the net pressure jitter amplitude, set a safety margin, obtain the jitter elimination threshold, and determine the jitter interference in the actual pressure signal based on the jitter elimination threshold.

[0054] In this embodiment, the natural grip instruction refers to the system requiring the trainee to hold the alarm device naturally with both hands without actively pressing it, in order to obtain hand pressure data in a resting state. The purpose of this instruction is to collect the contact pressure between the trainee's hands and the sensor in the most natural state without active force, which will be used as the basis for subsequent analysis such as pressure signal baseline and noise jitter. In a specific embodiment, the system prompts the trainee to relax their hands and hold the alarm device naturally. At this time, the trainee holds the device in the usual posture without actively applying force or pressing, and continuously collects pressure data for 5 seconds.

[0055] Resting segment data refers to the raw pressure signals collected when the hands are in natural contact with the device in a natural grip state. The resting baseline pressure is the median of these resting segment data within the collection period, which is used as a baseline for subsequent action discrimination. The pressure jitter curve refers to the small fluctuation curve of the pressure signal relative to the resting baseline in the trainee's resting grip state. These small fluctuations may come from natural muscle tension, small hand tremors, or unconscious movements, and are key parameters for measuring the trainee's own physiological noise and device sensitivity.

[0056] Robust statistical indicators refer to statistical quantities that are highly resistant to the influence of outliers, such as the median. By statistically analyzing the amplitude fluctuations of the pressure jitter curve using these indicators, the upper bound of the net pressure jitter amplitude is determined, which is the maximum normal fluctuation range that the pressure signal may reach under resting conditions. The safety margin refers to an additional margin added by the system on top of the calculated upper bound of the jitter amplitude to ensure the safety and robustness of subsequent discrimination. The jitter elimination threshold, which is the robust statistical upper bound combined with the safety margin, is used to discriminate the effective actions of all subsequent pressure signals, thereby effectively eliminating false alarms caused by the patient's natural grip and improving the discrimination accuracy and safety of the alarm system.

[0057] like Figure 3 The diagram shows a flowchart for jitter interference determination provided in an embodiment of the present invention. The specific design logic is as follows: Pressure signals from both sides of the buttons are collected, and then it is determined whether the signal is less than the jitter elimination threshold. If yes, it is determined to be jitter interference, and no alarm information is output; if no, it continues to determine whether the signal is less than the pain alarm threshold. If the signal is less than the pain alarm threshold, the jitter frequency of the current segment is extracted and the jitter characteristics of the resting segment are obtained. Then, it is determined whether the frequency matches the resting characteristics. If they match, it is determined to be jitter interference, and no alarm information is output; if they do not match, it is determined to be a suspected alarm, and a suspected alarm signal is output. If the signal is not less than the pain alarm threshold, the pain level determination process is initiated.

[0058] Furthermore, based on the jitter elimination threshold, jitter interference in the actual pressure signal is determined. Specifically, the button pressure signals on both sides are acquired and compared with the jitter elimination threshold. If the button pressure signal is less than the jitter elimination threshold, it is judged as jitter interference, no alarm information is output, and jitter shielding is performed.

[0059] When the key pressure signal is greater than the jitter elimination threshold and less than the pain level alarm threshold, the jitter frequency range is calculated based on the resting segment and frequency discrimination is performed on the segment: if the jitter frequency range of the segment matches the resting jitter characteristics, it is determined to be jitter interference and no alarm information is output; if the jitter frequency range of the segment does not match the resting jitter characteristics, it is considered a suspected alarm and a suspected alarm signal is output.

[0060] In this embodiment, the pressure signals of the buttons on both sides are compared with a predetermined jitter elimination threshold. This threshold is obtained by analyzing the pressure fluctuations in the trainee's resting and natural grip state. The aim is to shield false alarms caused by physiological micro-vibrations, equipment noise, etc. to the greatest extent. When the amplitude of the button pressure signal on either side collected in real time is lower than the jitter elimination threshold, it is determined to be jitter interference. There is no need to enter the subsequent action discrimination process, and no alarm information is output, thus achieving effective jitter shielding.

[0061] When the amplitude of any button pressure signal exceeds the jitter elimination threshold but does not reach the pain level alarm threshold, it is necessary to further determine whether the signal is abnormal jitter or a potential alarm behavior. If the dominant frequency range of the segment is consistent with or similar to the dominant frequency of the jitter characteristics extracted from the resting segment (e.g., the frequency difference is less than the preset threshold), the signal is considered to be caused by natural physiological jitter and no alarm is needed. Jitter shielding continues, which can further improve the ability to resist accidental touches and prevent the signal from being mistaken for a valid alarm due to slight force or external disturbances. If the dominant frequency characteristics of the current pressing segment deviate significantly from the dominant frequency range of the jitter extracted from the resting segment, the signal is determined to have non-jitter characteristics. It may be that the patient actively tried to alarm but did not reach a sufficient threshold. At this time, a suspected alarm signal is output to facilitate early intervention by medical staff or further observation of the patient's condition, achieving precise differentiation between physiological jitter and active alarm actions.

[0062] This invention also provides a threshold-adaptive intelligent discrimination system for magnetic wave knife alarm data. Figure 4 This is a block diagram of a threshold-adaptive intelligent judgment system for magnetic field knife alarm data. The system is used for intelligent judgment of magnetic field knife alarm data based on a threshold-adaptive method. The system includes a preoperative training module, an alarm threshold adjustment module, a motion data acquisition module, and an early warning information output module.

[0063] The preoperative training module is used to issue training instructions and collect training action data in real time during preoperative data collection, including the training bilateral pressure values ​​collected by the sensors on both sides of the hand-held bilateral pressing device.

[0064] The alarm threshold adjustment module is used to analyze the effectively collected training action data and adjust the alarm threshold of the alarm device to adapt to different individual differences.

[0065] During the operation, the motion data acquisition module collects real-time motion data, including the pressure values ​​on both sides collected by the hand-held bilateral compression device, which are the actual measured bilateral pressure values.

[0066] The early warning information output module analyzes the real-time action data collected and outputs early warning information.

[0067] For ease of explanation, Figure 4 Only the main components of the system are shown. The system in this embodiment can be used to perform... Figure 1 The technical solutions of the method embodiments shown are similar in principle and in effect, and will not be described again here.

[0068] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable system. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0069] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0070] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0071] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0072] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0073] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, systems, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0074] In the embodiments provided by this invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems 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 various embodiments of the present invention 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 the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, 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 described in the various embodiments of this invention. 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] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligently identifying magnetic wave knife alarm data based on threshold adaptation, characterized in that, The method includes: S1: During preoperative data collection, a training instruction is issued to collect training action data in real time, including the training bilateral pressure values ​​collected by the sensors on both sides of the hand-held bilateral pressing device. S2: Analyze the effectively collected training action data and adjust the alarm threshold of the alarm device to adapt to different individual differences; S3: During the operation, real-time motion data is collected, including bilateral pressure values ​​collected by a hand-held bilateral compression device, which are the actual measured bilateral pressure values. S4: Analyze real-time action data and output early warning information.

2. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 1, characterized in that, During the preoperative data collection, a training instruction is issued, and training motion data is collected in real time, specifically as follows: The training instructions include: instructions categorized by pain level; After each different training instruction is issued, real-time collection of training motion data begins, and the collection stops after a preset collection time. Record all data from all commands within the preset collection time, discard invalid data, and re-collect commands that were not collected. The pain level instructions include: pain situations that gradually increase from low to high, and require the trainee to apply bilateral pressure at the same time according to their personal habits. The pain situations cover different pain levels.

3. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 2, characterized in that, The data collected within the preset collection time is processed by removing invalid data and re-collecting any invalid data. The specific process is as follows: Based on the preset collection time period corresponding to the training instruction, the validity of the bilateral training action data is checked. The bilateral training action data includes at least the sampled values ​​of the bilateral data channels and their timestamps. If there is a sampling interruption and data loss within the preset collection time period, the collected data will be judged as invalid and discarded. For the training instructions corresponding to the invalid collected data that were removed, the same training instructions were reissued and the data collection was performed again until the bilateral training action data met the requirement of continuous collection without any missing data within the preset collection time period.

4. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 2, characterized in that, The analysis of the effectively collected training action data and the adjustment of the alarm threshold of the alarm device to adapt to different individual differences specifically include: By analyzing the training action data corresponding to the pain level instructions, the pain level alarm threshold adjustment factor is obtained. The pain level alarm threshold is scaled and adjusted using a pain level alarm threshold adjustment factor; The specific process for analyzing the effectively collected training motion data is as follows: For each valid training action data, obtain the complete waveform of bilateral pressure values ​​changing over time and identify the duration of the press. Perform valid segment detection on the waveform and extract the corresponding valid press segments on both sides; Based on the effective pressing segments, the maximum pressure values ​​of all effective pressing segments on both sides are determined and sorted from smallest to largest. The median of the maximum pressure values ​​is selected as the representative pressing force on that side. The representative pressing force is used to reflect the repeatable and stable pressing ability. At the same time, the minimum peak pressure of the press, which is sorted on both sides, is recorded as the minimum achievable force reference. The minimum achievable force reference is used to represent the minimum pressure value for effective triggering capability.

5. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 4, characterized in that, The specific steps for scaling and adjusting the pain level alarm threshold using a pain level alarm threshold adjustment factor are as follows: Based on the set standard reference pressing force, the alarm threshold adjustment factor of both sides is calculated according to the representative pressing force of both sides. At the same time, the alarm threshold adjustment factors of both sides are compared. The side with the smaller alarm threshold adjustment factor is defined as the weak side, and the side with the larger alarm threshold adjustment factor is defined as the strong side. After removing abnormally low values, the weak side is given a higher weight relative to the strong side. The alarm threshold adjustment factors of both sides are merged into a pain level alarm threshold adjustment factor, so that the pain level alarm threshold adjustment factor converges to the weak side, so as to ensure that the weak side can still meet the triggering condition when both sides are pressed at the same time. Based on the set alarm threshold, the alarm threshold is monotonically scaled according to the pain level alarm threshold adjustment factor. The scaled alarm threshold is constrained by both the lower and upper trigger limits to obtain a pain level alarm threshold within a stably achievable trigger range, thereby enabling individualized threshold matching among different patients.

6. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 5, characterized in that, After scaling the pain level alarm threshold using the pain level alarm threshold adjustment factor, the method further includes outputting an alarm signal based on the pain level alarm threshold, specifically: Based on the pressure signals from both sides of the buttons, the corresponding pain level alarm threshold is determined. When both sides of the pressure meet the standard, the excess force of this pressure segment is calculated by comparing the net pressure peak of the pressure segment with the alarm trigger bottom line pressure threshold. In the case of a single press, the first alarm message is output for each effective press based on the amount of force exceeded. When multiple presses occur within a preset short-time analysis window, the excessive force of each press is accumulated and processed, and then a second alarm message is output.

7. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 6, characterized in that, The specific process for outputting the second alarm information is as follows: Within a preset short-time analysis window, the cumulative excess of the pressing segment is extracted, and the continuous pressing characteristics are statistically analyzed, including the number of continuous pressings and the proportion of rapid pressing. The alarm score for this pressing event is obtained by combining the cumulative excess amount with the continuous pressing characteristics. The alarm score is proportional to the increase of the cumulative excess amount, the number of continuous pressings, and the proportion of rapid pressing. The alarm score will be used to output an early warning prompt. The early warning prompt shall include at least an audible prompt and / or an indicator light prompt. The frequency and volume of the audible prompt shall increase as the alarm score increases. The brightness and flashing frequency of the indicator light shall increase as the alarm score increases.

8. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 2, characterized in that, The training instructions also include natural grip instructions, specifically: The natural grip instruction is: the trainee is required to keep both hands naturally clenched without pressing instructions; Acquire resting segment data generated by natural hand contact, continuously collect resting segment data to calculate resting baseline pressure, and combine pressure signals to obtain pressure jitter curve; Statistical analysis was performed on the amplitude fluctuation of the pressure jitter curve. A preset robust statistical index was used to determine the upper limit of the net pressure jitter amplitude and a safety margin was set to obtain the jitter elimination threshold. Based on the jitter elimination threshold, jitter interference in the actual pressure signal was judged.

9. The intelligent discrimination method for magnetic wave knife alarm data based on threshold adaptation according to claim 8, characterized in that, The determination of jitter interference in the actual pressure signal based on the jitter elimination threshold is as follows: Acquire the button pressure signals from both sides and compare them with the bounce elimination threshold; If the button pressure signal is less than the jitter elimination threshold, it is judged as jitter interference, no alarm information is output, and jitter is shielded. When the key pressure signal is greater than the jitter elimination threshold but less than the pain level alarm threshold, the jitter frequency range is calculated based on the resting segment and frequency discrimination is performed on that segment: If the jitter frequency range of the segment matches the resting jitter characteristics, it is determined to be jitter interference and no alarm information is output. If the jitter frequency range of the segment does not match the resting jitter characteristics, it is considered a suspected alarm and a suspected alarm signal is output.

10. A threshold-adaptive intelligent discrimination system for magnetic wave knife alarm data, wherein the threshold-adaptive intelligent discrimination system for magnetic wave knife alarm data is used to implement the threshold-adaptive intelligent discrimination method for magnetic wave knife alarm data as described in any one of claims 1-9, characterized in that, The system includes: a preoperative training module, an alarm threshold adjustment module, a motion data acquisition module, and an early warning information output module; The preoperative training module is used to issue training instructions and collect training action data in real time during preoperative pre-collection, including the training bilateral pressure values ​​collected by the sensors on both sides of the hand-held bilateral pressing device. The alarm threshold adjustment module is used to analyze the effectively collected training action data and adjust the alarm threshold of the alarm device to adapt to different individual differences. The motion data acquisition module is used to collect real-time motion data during the operation, including the pressure values ​​on both sides collected by the hand-held bilateral compression device, which are the actual measured bilateral pressure values. The early warning information output module is used to analyze real-time action data collected in real time and output early warning information.

Images