A method and system for generating a personalized comfort protocol for a thermage treatment

Abstract

The application relates to the technical field of medical cosmetology equipment, and discloses a generation method and system of a Thermage nursing personalized comfort scheme. A physiological signal reflecting a patient's tolerance state is acquired in real time from a nursing instrument, and combined with historical data of the personal tolerance baseline, comfort parameters in real time and baseline dimensions are respectively generated via preset mapping rules, the two are fused into a comprehensive comfort index based on dynamic credibility evaluation, a predefined mapping table is queried according to the comprehensive comfort index, and a first round of adjustment instructions for nursing parameters are determined; a second round of evaluation based on dynamic feedback is started, second round adjustment instructions are generated and sent, and the nursing instrument applies the instructions in a locked state to complete the core adjustment of the personalized scheme. The application realizes a complete closed loop from multi-dimensional data fusion, logical coherence guarantee to knowledge self-iteration, and effectively improves the standardization, personalization and comfort management level of Thermage nursing.

Description

Technical Field

[0001] This invention relates to the field of medical aesthetic equipment technology, and discloses a method and system for generating personalized comfort plans for Thermage treatments. Background Technology

[0002] Currently, in Thermage care, the management of patient comfort and adjustment of care plans generally rely on the experience and judgment of the operating physician or simple responses based on single-dimensional data. A common practice is for physicians to manually adjust energy, cooling, and other care parameters based on subjective observations of the patient's facial expressions and verbal feedback, or on single real-time parameters such as local skin impedance fed back by the device, or through preset simple threshold rules. This method has significant limitations: First, it heavily relies on the physician's personal experience and immediate judgment, making standardization and reproducibility difficult; different physicians may provide significantly different comfort interventions for the same patient. Second, adjustments based on a single, instantaneous signal are prone to misjudgment due to brief fluctuations or interference in the signal, leading to overly frequent or incorrect parameter adjustments, disrupting the consistency and stability of care, and even affecting the final therapeutic effect. Furthermore, this method lacks a systematic integration of individual patient tolerance characteristics such as pain thresholds and psychological expectations, failing to form a truly personalized comfort control strategy that permeates the entire care process.

[0003] Furthermore, some existing automated control schemes often employ linear and isolated logic. For example, they linearly reduce energy based solely on real-time impedance increases, without considering the patient's baseline tolerance or establishing a dynamic weighting relationship between real-time response and inherent characteristics. After adjustment commands are issued, the system typically lacks a mechanism to lock in or protect the comprehensive decision-making logic upon which the adjustment is based, allowing subsequent adjustments to be easily overwritten by new transient data, resulting in a lack of a consistent strategic framework for the overall adjustment process. Simultaneously, the generated temporary parameter combinations are rarely systematically recorded, analyzed, and optimized, failing to form a self-improving knowledge base that accumulates with case studies, thus limiting the continuous improvement of nursing comfort management.

[0004] Therefore, there is an urgent need for a method to generate personalized comfort plans for Thermage care that can overcome the aforementioned shortcomings. This method should objectively and comprehensively integrate real-time physiological responses with the patient's individual tolerance baseline, generating a quantitative comfort index through structured multi-dimensional assessment and fusion calculation, and then making precise parameter adjustments accordingly. More importantly, this method needs to establish a coherent adjustment logic guarantee mechanism, such as a lock-in state, to ensure that multiple rounds of adjustments are carried out within a unified strategy framework, avoiding decision jumps. Finally, this method should also support automated backtracking analysis of historically successful plans to continuously optimize its core assessment and mapping rules, thereby forming an intelligent, standardized, and evolutionarily capable personalized comfort plan generation system. Summary of the Invention

[0005] To achieve the above objectives, this application provides the following technical solution:

[0006] According to a first aspect of the present invention, the present invention claims protection for a method for generating a personalized comfort plan for Thermage care, comprising the following steps:

[0007] Obtain the first comfort-related dataset associated with the target user, process the first comfort-related dataset based on the preset first comfort mapping rule, and generate the first dimension comfort evaluation parameters;

[0008] Obtain the pre-stored second comfort-related dataset, process the second comfort-related dataset based on the preset second comfort mapping rule, and generate the second-dimensional comfort baseline parameters;

[0009] The first dimension comfort assessment parameters and the second dimension comfort baseline parameters are fused together to generate a comprehensive comfort index.

[0010] Based on the comprehensive comfort index, a predefined comfort-parameter adjustment mapping table is queried to determine the first set of adjustment instructions for the initial basic nursing parameters;

[0011] A comfort lock command is sent to the nursing device control unit, instructing the nursing device control unit to maintain the parameter adjustment logic calculated based on the fusion of the first dimension comfort assessment parameters and the second dimension comfort baseline parameters after receiving and applying the first round of adjustment command set, until a unlock command is received;

[0012] After sending the comfort lock command, a second round of evaluation is initiated based on the dynamic feedback mechanism to generate an updated comprehensive comfort index, and a second round of adjustment command set is generated based on the updated comprehensive comfort index;

[0013] The second set of adjustment instructions is sent to the nursing device control unit. According to the comfort lock instruction, when applying the second set of adjustment instructions, the parameter adjustment logic previously determined by the first dimension comfort assessment parameter and the second dimension comfort baseline parameter is maintained.

[0014] Furthermore, after sending the second set of adjustment instructions to the nursing device control unit, the method further includes:

[0015] Receive an implementation confirmation signal from the nursing device control unit, confirming that the second round of adjustment command set has been successfully applied and a stable nursing waveform has been generated;

[0016] In response to the implementation confirmation signal, a comfort lock release command is generated and sent to the nursing device control unit. The comfort lock release command is used to instruct the nursing device control unit that if new comfort-related data is received in subsequent nursing stages, the nursing parameters should be adjusted according to the preset default parameter update strategy or the newly triggered complete multi-dimensional fusion calculation process, and the specific fusion logic established by the previous first and second rounds of evaluation should no longer be forcibly maintained.

[0017] Receive and persistently store a complete protocol data package containing the final application set of nursing parameters, the corresponding comprehensive comfort index sequence, and a nursing process summary log. The complete protocol data package is associated with the unique identifier of the target user and stored in the protocol database.

[0018] Based on the complete protocol data package, a personalized comfort protocol report document is automatically generated, which includes the evolution path of nursing parameters, the change curve of comfort index, and descriptions of key adjustment nodes.

[0019] Furthermore, the second round of evaluation based on the dynamic feedback mechanism, and the multi-round adjustment and verification process executed after sending the comfort lock command, further include the following detailed steps:

[0020] After the first set of adjustment instructions is applied to the nursing instrument control unit, a monitoring window period is initiated. The duration of the monitoring window period is dynamically set according to the type of nursing parameter being adjusted and its change range.

[0021] During the monitoring window period, the first type of physiological signal sequence of the target user is continuously collected through the first data interface at a monitoring sampling frequency higher than the initial sampling frequency to form a monitoring dataset;

[0022] The monitoring dataset is subjected to real-time fragmented analysis, and the monitoring window period is divided into multiple consecutive time segments. For each time segment, the characteristic statistical values ​​of the first type of physiological signal are calculated. The characteristic statistical values ​​include the mean, variance and the trend slope with respect to the time segment.

[0023] The characteristic statistical values ​​of each time segment are input into a preset instant response evaluation model, which is configured to output an instant tolerance stability score. The instant tolerance stability score is used to characterize the degree of fluctuation of the user's physiological response relative to the expected response within that time segment.

[0024] If the immediate tolerance stability score is lower than the preset stability threshold for at least three consecutive time segments, the physiological response induced by the first round of adjustment instructions is determined to be in an unstable state, triggering an immediate fine-tuning procedure. The immediate fine-tuning procedure includes: based on the characteristic statistical values ​​of the current time segment, deducing the fine-tuning amount of some parameters in the first round of adjustment instructions, and generating an immediate fine-tuning instruction; the immediate fine-tuning instruction is sent to the nursing device control unit within the framework of the comfort locking instruction, and the nursing device control unit performs rapid parameter correction accordingly, while recording this fine-tuning as a preparatory adjustment for the second round of assessment.

[0025] When the monitoring window period ends, or when the instantaneous tolerance stability score recovers to above the stability threshold in subsequent time segments and maintains a preset stable duration, the data sampling phase of the second round of evaluation is formally initiated.

[0026] During the data sampling phase of the second round of evaluation, the standard sampling frequency is restored, a fully updated first comfort-related dataset is collected, and the steps for generating updated first-dimensional comfort evaluation parameters and subsequent fusion calculations as described in claim 1 are executed.

[0027] After generating the second round of adjustment instruction set, it is logically compared and merged with all the preparatory adjustments recorded by the real-time fine-tuning program. If the second round of adjustment instruction set is consistent with any preparatory adjustment in the adjustment direction but different in magnitude, the one with the larger magnitude is adopted to form the final merged second round of adjustment instruction. If the directions are opposite, the second round of adjustment instruction set calculated based on the complete data sampling stage is adopted, and the preparatory adjustments with opposite directions are ignored.

[0028] The merged second-round adjustment instruction set is sent to the nursing device control unit as the final second-round adjustment instruction set.

[0029] Furthermore, the first comfort-related dataset is processed based on a preset first comfort mapping rule to generate a first-dimensional comfort evaluation parameter, specifically including the following detailed steps:

[0030] The original first type of physiological signal sequence obtained through the first data interface is subjected to signal quality verification. The signal quality verification includes detecting the signal loss rate and whether the noise amplitude exceeds the allowable range. If the signal quality verification fails, a re-acquisition request is sent to the signal acquisition module, and subsequent steps are paused until a valid signal sequence is obtained.

[0031] The original Class I physiological signal sequences that passed the verification were standardized, and the standardization process included:

[0032] The raw signal values ​​are converted into percentage changes relative to the target user's baseline values ​​at rest to eliminate absolute differences in individual baseline physiological levels.

[0033] From the standardized signal sequence, extract multiple predefined time-domain and frequency-domain features, the time-domain features including:

[0034] The rising slope of the signal during the application of the nursing pulse, the recovery slope during the pulse interval, and the peak and trough values ​​of the signal amplitude;

[0035] The frequency domain features include: power spectral density in a specific frequency band that is associated with autonomic nervous system regulatory activity;

[0036] Each extracted time-domain and frequency-domain feature is assigned a preset weight coefficient, which is predetermined based on the correlation between the feature and the subjective discomfort of Thermage treatment.

[0037] Each extracted feature is multiplied by its corresponding weight coefficient and then summed to obtain a preliminary composite physiological response index.

[0038] The preliminary composite physiological response index is input into a nonlinear transformation function to map the composite physiological response index to a scale value. This nonlinear transformation function has high sensitivity in the middle range and decreases sensitivity in the range close to the extreme value to prevent extreme jumps in the results.

[0039] The scale value processed by the nonlinear transformation function is output as the first dimension comfort evaluation parameter, which has the same dimensions and comparability as the second dimension comfort baseline parameter, facilitating subsequent fusion calculation.

[0040] Furthermore, the second type of physiological baseline data includes heart rate variability indices, baseline values ​​of skin conductivity response, and pain threshold pressure test results collected under non-thermal stimulation conditions;

[0041] The user-reported tolerance level information was collected before the nursing care using a standardized questionnaire, which included subjective feelings about previous similar energy device nursing experiences and self-expectation of pain.

[0042] Furthermore, the comfort assessment parameters of the first dimension are fused with the comfort baseline parameters of the second dimension to generate a comprehensive comfort index, which is achieved through the following formula:

[0043] CCI=α×(w1×SCA+w2×SCB)+(1-α)×Dynamic_Compensation

[0044] Wherein, CCI represents the comprehensive comfort index; SCA represents the comfort assessment parameter of the first dimension; SCB represents the comfort baseline parameter of the second dimension; w1 and w2 are weights pre-calibrated based on clinical data, and w1+w2=1;

[0045] α is a dynamic adaptation factor, whose value is automatically adjusted based on the coefficient of variation of the first dimension comfort assessment parameter SCA in the most recent three sampling periods. The larger the coefficient of variation, the smaller the α value, indicating that it is more dependent on the second dimension baseline parameter.

[0046] Dynamic_Compensation is a dynamic compensation item whose value is obtained from a predefined compensation value lookup table based on the proportion of time that care has been performed to the total expected care time and the number of historical adjustments to the current energy level.

[0047] Furthermore, the comfort-parameter adjustment mapping table is a multi-dimensional lookup table. It uses the comprehensive comfort index as the primary lookup key and the body part code of the current nursing area and the current cumulative emission energy as secondary lookup keys to output the corresponding set of parameter adjustment instructions. The set of parameter adjustment instructions not only includes adjustments to nursing parameters but also includes suggested stable observation durations after adjustments.

[0048] Furthermore, after receiving and persistently storing the complete protocol data package, it also includes a retrospective analysis and optimization cycle based on historical protocols, independent of the real-time care process, specifically including:

[0049] Periodically retrieve all complete protocol data packages related to Thermage care generated within a preset historical time period from the protocol database;

[0050] All retrieved complete solution data packets are parsed in a structured manner to extract key fields;

[0051] Construct an analysis dataset where each record corresponds to a complete nursing process, and the feature variables include statistical features derived from the aforementioned key fields;

[0052] Define optimization target variables, which are derived from the follow-up scores of nursing effect and the overall user satisfaction scores collected subsequently. After the nursing care ends, the data is collected through a safe follow-up system for a preset number of days and associated with the analysis dataset through an anonymous ID.

[0053] Using statistical correlation analysis, feature variables that are significantly positively or negatively correlated with the target variable are selected from the analysis dataset.

[0054] Based on the selected significantly relevant feature variables, the parameter weights in the first comfort mapping rule and / or the second comfort mapping rule are optimized in reverse, and / or the specific adjustment values ​​in the comfort-parameter adjustment mapping table are optimized;

[0055] For the first comfort level mapping rule, if it is found that the final user satisfaction is strongly correlated with the change of a certain physiological signal feature in the early stage of care, then the weight of that feature in the weight coefficient allocation is adjusted.

[0056] For the comfort-parameter adjustment mapping table, if statistical analysis shows that when the comprehensive comfort index is in a specific range, adopting a more aggressive or more conservative energy adjustment range than the current mapping table will result in higher subsequent satisfaction and effect scores, then the adjustment instructions corresponding to that range will be updated.

[0057] The optimized mapping rules and mapping table versions are numbered and stored in the rule base;

[0058] When generating new personalized comfort solutions for target users, the latest version of rules and mapping tables is applied by default, or operators are allowed to select a specific historical version under certain circumstances;

[0059] The retrospective analysis optimization cycle is executed automatically at preset intervals, forming a closed loop of continuous iteration and improvement of the core logic of solution generation.

[0060] Furthermore, the immediate response evaluation model is a rule-based state machine, in which the internal state switches according to the input feature statistics and multiple predefined state transition conditions;

[0061] The state transition conditions include: consistency of the trend slope sign of continuous time segments, the multiple relationship of variance to the historical window mean, and the matching degree between the combination of feature statistics and the pre-stored typical maladaptive pattern template.

[0062] According to a second aspect of the present invention, the present invention claims protection for a system for generating personalized comfort plans for Thermage treatments, comprising:

[0063] One or more processors;

[0064] A memory having stored one or more programs that, when executed by one or more processors, enable the one or more processors to implement the method for generating a personalized comfort solution for Thermage care.

[0065] This invention relates to the field of medical aesthetic equipment technology, and discloses a method and system for generating personalized comfort plans for Thermage treatments. The system acquires physiological signals reflecting the patient's tolerance status in real time from the treatment device, and combines this with the patient's personal tolerance baseline historical data. Comfort parameters for both real-time and baseline dimensions are generated through preset mapping rules. Based on dynamic reliability assessment, these two parameters are merged into a comprehensive comfort index. A predefined mapping table is then consulted to determine the first round of adjustment instructions for the treatment parameters. A second round of assessment based on dynamic feedback is initiated, generating and sending the second round of adjustment instructions. The treatment device, in a locked state, applies these instructions to complete the core adjustments of the personalized plan. This invention achieves a complete closed loop from multi-dimensional data fusion and logical coherence assurance to knowledge self-iteration, effectively improving the standardization, personalization, and comfort management level of Thermage treatments. Attached Figure Description

[0066] Figure 1 A flowchart illustrating the process of generating a personalized comfort plan for Thermage care, as claimed in an embodiment of the present invention.

[0067] Figure 2 This is a second workflow diagram of a method for generating a personalized comfort plan for Thermage care, as claimed in an embodiment of the present invention.

[0068] Figure 3 The third workflow diagram is for a method of generating a personalized comfort plan for Thermage care, as claimed in an embodiment of the present invention. Detailed Implementation

[0069] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0070] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of those features. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications in the embodiments of this application, such as up, down, left, right, front, back, etc., are only used to explain the relative positional relationships and movements between components in a specific orientation as shown in the accompanying drawings. If the specific orientation changes, the directional indications will change accordingly. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0071] References to embodiments herein mean that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0072] According to a first embodiment of the present invention, the present invention claims protection for a method for generating a personalized comfort plan for Thermage care, referring to... Figure 1 This includes the following steps:

[0073] Obtain the first comfort-related dataset associated with the target user, process the first comfort-related dataset based on the preset first comfort mapping rule, and generate the first dimension comfort evaluation parameters;

[0074] Obtain the pre-stored second comfort-related dataset, process the second comfort-related dataset based on the preset second comfort mapping rule, and generate the second-dimensional comfort baseline parameters;

[0075] The first dimension comfort assessment parameters and the second dimension comfort baseline parameters are fused together to generate a comprehensive comfort index.

[0076] Based on the comprehensive comfort index, a predefined comfort-parameter adjustment mapping table is queried to determine the first set of adjustment instructions for the initial basic nursing parameters;

[0077] A comfort lock command is sent to the nursing device control unit, instructing the nursing device control unit to maintain the parameter adjustment logic calculated based on the fusion of the first dimension comfort assessment parameters and the second dimension comfort baseline parameters after receiving and applying the first round of adjustment command set, until a unlock command is received;

[0078] After sending the comfort lock command, a second round of evaluation is initiated based on the dynamic feedback mechanism to generate an updated comprehensive comfort index, and a second round of adjustment command set is generated based on the updated comprehensive comfort index;

[0079] The second set of adjustment instructions is sent to the nursing device control unit. According to the comfort lock instruction, when applying the second set of adjustment instructions, the parameter adjustment logic previously determined by the first dimension comfort assessment parameter and the second dimension comfort baseline parameter is maintained.

[0080] In this embodiment, a first data communication link is established between the computing device and the control unit of the Thermage care device. This first data communication link is configured to transmit data in real-time or near real-time. Through this first data communication link, the computing device sends a start signal acquisition command to the control unit, specifying the first type of physiological signal to be acquired and the initial sampling frequency. Responding to this command, the control unit controls its integrated or external physiological signal sensors to apply the first or first few test pulses to the designated care area of ​​the target user according to preset initial basic care parameters, including basic energy level, standard pulse waveform, standard pulse duration, and standard cooling interval, while simultaneously acquiring the target user's physiological signals. The acquired physiological signals constitute the raw data of a first comfort-related dataset, which includes at least one of the following: a skin impedance change sequence, a local skin surface temperature change sequence in the care area, and a local muscle micro-vibration signal sequence acquired by a contact sensor. The computing device continuously reads the aforementioned raw data from the control unit through a first data interface, either as a data stream or in fixed-interval data packets. During the reading process, the computing device performs data integrity checks, examining the continuity of timestamps and the integrity of data fields for each data packet, discarding corrupt or discontinuous data packets, and requesting retransmission. The valid raw data read is temporarily cached in a designated memory area of ​​the computing device, forming the first comfort-related dataset.

[0081] Subsequently, the computing device invokes its internally stored first comfort mapping rule processing engine to process the first comfort-related dataset. This processing begins with the preprocessing of the original physiological signal sequence, including applying digital filters to remove power frequency interference and high-frequency noise, and performing baseline drift correction. The preprocessed signal sequence is segmented into independent data segments corresponding to each nursing pulse and its subsequent cooling interval. For each independent data segment, the engine extracts a set of predefined quantitative features, such as: the slope of the decrease in skin impedance during pulse application and its steady-state value, the rate of increase in skin temperature after the pulse and its peak value, and the root mean square energy of the muscle micro-vibration signal and its main frequency components. The engine maintains a feature weight lookup table, which assigns a weight coefficient based on clinical research to each feature. The engine multiplies each feature value extracted from each data segment by its corresponding weight coefficient and sums them to obtain a preliminary response score for that data segment. Then, the engine averages the preliminary response scores of all data segments within the same nursing cycle, such as the first three test pulse cycles, and introduces a time decay factor to give a higher weight to the scores of the most recent cycles, ultimately calculating a weighted average score. Finally, the engine inputs the weighted average score into a standardization transformation module, which linearly or non-linearly maps the score to a value between 0 and 100, defined as the first dimension comfort assessment parameter. The higher the parameter value, the higher the level of discomfort or stress reflected in the user's real-time physiological response under the initial basic care parameters.

[0082] Simultaneously, the computing device initiates a query request for the target user's second comfort-related dataset via a second data interface, typically connected to a database or electronic health record system storing user profiles. The query request includes the target user's unique identifier. The second comfort-related dataset is not generated in real-time but pre-stored, and its content includes: the user's tolerance assessment test results before receiving Thermage treatment, such as pain threshold measured by a pressure pain meter; heart rate variability indicators obtained through resting-state electrocardiogram analysis; the user's detailed health questionnaire results, including self-assessment of pain sensitivity; subjective descriptions and ratings of previous similar energy treatments; and possible tolerance summaries from historical treatment records. After receiving this data, the computing device invokes the second comfort mapping rule processing engine. This engine first extracts keywords and performs sentiment analysis on the subjective descriptions in the questionnaire, quantifying them into a numerical level. Then, the engine maps the pain... Physiological baseline data, such as thresholds and heart rate variability indicators like SDNN, are normalized according to their respective normal ranges and converted into a baseline score of 0 to 100. Next, the engine parses historical tolerance summaries if they exist, extracts the average energy level or maximum single pulse energy that the user could tolerate in historical care, and compares it with standard population values ​​to generate a historical adaptability score. Finally, the second engine, according to a predetermined integration algorithm, assigns higher weight to physiological baseline data, uses subjective questionnaires and historical data as adjustment items, and comprehensively calculates the above quantified numerical levels, baseline scores, and historical adaptability scores to output a second-dimensional comfort baseline parameter. This parameter also falls within the range of 0 to 100, with higher values ​​indicating lower inherent tolerance baseline levels for the user.

[0083] Next, the computing device performs a fusion calculation step, handled by a dedicated fusion calculation module. This module receives first-dimensional comfort assessment parameters and second-dimensional comfort baseline parameters as input. Internally, the module first evaluates the reliability of the first-dimensional comfort assessment parameters by analyzing the variance of these parameters over the most recent consecutive sampling periods. If the variance exceeds a threshold, the real-time physiological signal is considered to have significant fluctuations or interference, and the reliability of the real-time data is reduced. Based on the calculated reliability index, the fusion module dynamically adjusts the fusion weights of the two input parameters: when the real-time data reliability is high, the weight of the first-dimensional parameter increases; conversely, it relies more on the second-dimensional baseline parameter. The specific logic for dynamic weight adjustment is implemented through an internal lookup table indexed by variance, outputting the corresponding weight allocation coefficients. After determining the dynamic weights, the module performs a weighted summation calculation to obtain a preliminary comprehensive score. Furthermore, the module considers nursing progress factors, such as whether the current stage is the initial, intermediate, or final stage of care. At different stages, the module looks up a progress compensation coefficient from another predefined table. Finally, the module combines the initial comprehensive score with the process compensation coefficient, for example by multiplying or adding them, to generate the final comprehensive comfort index.

[0084] Then, the computing device accesses a locally stored comfort-parameter adjustment map based on the calculated overall comfort index. This map is a multi-dimensional data structure whose primary key is a different range of the overall comfort index, such as 0-30, 31-60, 61-85, 86-100. For each primary key range, the map defines a corresponding set of adjustment instructions for the initial basic care parameters. These adjustment instructions are very specific, for example: reducing the energy level by 10%, shortening the pulse duration by 5%, advancing the cooling jet's start time by 100 milliseconds, and extending its duration by 20%. The computing device retrieves the corresponding first-round set of adjustment instructions based on the range the overall comfort index falls into.

[0085] After generating the first set of adjustment instructions, the computing device does not immediately send it to the nursing device control unit. Instead, it first generates a special comfort lock instruction. This instruction is a control command containing a specific identifier and a description of the lock logic. Its core instruction is: after successfully applying the subsequently received first set of adjustment instructions, the nursing device control unit must lock its internal parameter adjustment decision logic in the current state. Specifically, locking means that in the subsequent period, if the nursing device control unit's local sensors detect changes in physiological signals and trigger its built-in simplified adjustment logic, this built-in logic should not override or violate the parameter adjustment direction and approximate range framework established by the computing device based on the fusion calculation of the first and second dimensions. The computing device first sends this comfort lock instruction to the nursing device control unit through the first data communication link and waits for a confirmation acknowledgment.

[0086] Only after receiving confirmation of the lock command from the nursing device control unit does the computing device send the first set of adjustment commands to the nursing device control unit. The nursing device control unit then applies these commands to modify its nursing parameters.

[0087] Following this, the computing device initiates the second round of evaluation under the dynamic feedback mechanism. In this phase, the computing device first waits for a preset stabilization period to allow the user's body to develop a stable response to the new parameters. After the stabilization period, the computing device again requests the nursing device control unit to collect a new round of physiological signal data under the new parameters through the first data interface. This process is similar to the first round of data collection, thereby obtaining an updated first comfort-related dataset. The computing device uses the same first comfort mapping rule processing engine to process this new dataset, generating updated first-dimensional comfort assessment parameters.

[0088] Next, the computing device invokes the fusion computing module again. At this time, since the comfort lock command is still in effect, the fusion computing module is forced to use the exact same fusion logic as when generating the first round of adjustments, especially the same dynamic weight allocation rules and process compensation coefficient determination rules. Although the care process may have entered a new stage, the lock command requires ignoring the changes in compensation coefficients caused by the process changes. The module fuses the updated first-dimensional comfort assessment parameters with the previously acquired and stored second-dimensional comfort baseline parameters to generate an updated comprehensive comfort index.

[0089] Based on this updated comprehensive comfort index, the computing device queries the comfort-parameter adjustment mapping table again to obtain a second set of adjustment instructions. Finally, the computing device sends this second set of adjustment instructions to the nursing device control unit. When the nursing device control unit receives the second set of adjustment instructions, it is still in a locked state. It applies these new instructions, and during the application process, its internal logic ensures that these adjustments are made within the adjustment framework defined by the initial first and second dimension fusion logic, and will not jump to a completely different adjustment strategy due to possible instantaneous fluctuations in the new data.

[0090] Furthermore, referring to Figure 2 After sending the second set of adjustment instructions to the nursing device control unit, the method further includes:

[0091] Receive an implementation confirmation signal from the nursing device control unit, confirming that the second round of adjustment command set has been successfully applied and a stable nursing waveform has been generated;

[0092] In response to the implementation confirmation signal, a comfort lock release command is generated and sent to the nursing device control unit. The comfort lock release command is used to instruct the nursing device control unit that if new comfort-related data is received in subsequent nursing stages, the nursing parameters should be adjusted according to the preset default parameter update strategy or the newly triggered complete multi-dimensional fusion calculation process, and the specific fusion logic established by the previous first and second rounds of evaluation should no longer be forcibly maintained.

[0093] Receive and persistently store a complete protocol data package containing the final application set of nursing parameters, the corresponding comprehensive comfort index sequence, and a nursing process summary log. The complete protocol data package is associated with the unique identifier of the target user and stored in the protocol database.

[0094] Based on the complete protocol data package, a personalized comfort protocol report document is automatically generated, which includes the evolution path of nursing parameters, the change curve of comfort index, and descriptions of key adjustment nodes.

[0095] In this embodiment, the computing device enters a result confirmation waiting state. In this state, the computing device continuously monitors the data feedback channel from the nursing device control unit. After successfully applying the second set of adjustment instructions, the nursing device control unit continuously monitors the output waveform, energy stability, and contact status with the skin of the nursing head. When all these indicators remain within the error range allowed by the technical specifications for several consecutive nursing pulse cycles, the nursing device control unit determines that the nursing has entered a stable nursing stage. At this time, the nursing device control unit generates a specific implementation confirmation signal, which is a data packet containing a status code and a snapshot of the current stable parameters. After receiving this implementation confirmation signal through the first data interface, the computing device first parses the status code to confirm that it is a successful stabilization signal, rather than an error or warning signal.

[0096] Once successful implementation and stable care are confirmed, the computing device generates a comfort lock release command. This command corresponds in structure to the lock command and includes a release command identifier. The computing device sends this release command to the care device control unit. Upon receiving it, the care device control unit immediately clears the logic restriction flags set by the previous comfort lock command. This means that in subsequent care processes, if its local monitoring system detects an abnormality requiring parameter adjustment, such as a sudden increase in impedance due to poor contact, the care device control unit will restore its default, independent, and rapid adjustment capability based on a single instantaneous signal threshold. Alternatively, it will be prepared to accept different adjustment commands from the computing device, possibly based on a new round of complete multi-dimensional data fusion calculations, and will no longer be forcibly bound to the specific fusion logic framework followed in the previous locked state.

[0097] After sending the release command, the computing device begins the data archiving process for this personalized comfort plan. It collects all key data nodes from its internal cache throughout the entire plan generation process, including: a summary of the first comfort-related dataset initially read, the calculated sequence of first-dimensional comfort assessment parameters, key values ​​of the second comfort-related dataset, second-dimensional comfort baseline parameters, the comprehensive comfort index generated by each fusion calculation and its corresponding timestamp, detailed content of the first and second rounds of adjustment command sets generated and sent, implementation confirmation signals from the nursing device control unit, and a summary of the nursing process extracted from the nursing device control unit log, such as total nursing time, total number of pulses, and energy output range. The computing device encapsulates this data according to a predefined XML or JSON format to form a structured complete plan data package. During the encapsulation process, the computing device marks the data package with a unique identifier for the target user, such as an encrypted medical record number. Subsequently, the computing device initiates a database storage transaction through its data persistence module to write the complete plan data package into a dedicated personalized nursing plan database. This database is indexed for future retrieval by user identifier or nursing date.

[0098] After the plan is stored, the computing device starts the report generation module. This module accesses the newly stored complete plan data package and extracts key information for visualization. The report generation module first creates a new document object, and then generates the following sections according to a standard template: In the nursing parameter evolution section, it plots a curve with time on the horizontal axis and the main nursing parameters energy and pulse width on the vertical axis, clearly marking the location and adjustment amount of the first and second rounds of adjustments; in the comfort index change section, it plots the curve of the comprehensive comfort index changing over time, marking the index peaks, troughs, and threshold lines that trigger adjustments, and explaining the key adjustment nodes. The report section describes in text form the direct cause of each adjustment, such as the comprehensive index exceeding a certain range, the integration of real-time data and baseline data, and the expected adjustment goals. The report generation module integrates graphics and text into a unified document file, which can be in PDF, DOCX, or other formats. Finally, the computing device stores the generated personalized comfort plan report document in a designated directory and associates its file path information with the corresponding plan record in the database. The computing device can also send a notification to the nurse's workstation via a message interface, informing them that the personalized comfort plan for the target user has been generated and archived for review.

[0099] Furthermore, the second round of evaluation based on the dynamic feedback mechanism, and the multi-round adjustment and verification process executed after sending the comfort lock command, further include the following detailed steps:

[0100] After the first set of adjustment instructions is applied to the nursing instrument control unit, a monitoring window period is initiated. The duration of the monitoring window period is dynamically set according to the type of nursing parameter being adjusted and its change range.

[0101] During the monitoring window period, the first type of physiological signal sequence of the target user is continuously collected through the first data interface at a monitoring sampling frequency higher than the initial sampling frequency to form a monitoring dataset;

[0102] The monitoring dataset is subjected to real-time fragmented analysis, and the monitoring window period is divided into multiple consecutive time segments. For each time segment, the characteristic statistical values ​​of the first type of physiological signal are calculated. The characteristic statistical values ​​include the mean, variance and the trend slope with respect to the time segment.

[0103] The characteristic statistical values ​​of each time segment are input into a preset instant response evaluation model, which is configured to output an instant tolerance stability score. The instant tolerance stability score is used to characterize the degree of fluctuation of the user's physiological response relative to the expected response within that time segment.

[0104] If the immediate tolerance stability score is lower than the preset stability threshold for at least three consecutive time segments, the physiological response induced by the first round of adjustment instructions is determined to be in an unstable state, triggering an immediate fine-tuning procedure. The immediate fine-tuning procedure includes: based on the characteristic statistical values ​​of the current time segment, deducing the fine-tuning amount of some parameters in the first round of adjustment instructions, and generating an immediate fine-tuning instruction; the immediate fine-tuning instruction is sent to the nursing device control unit within the framework of the comfort locking instruction, and the nursing device control unit performs rapid parameter correction accordingly, while recording this fine-tuning as a preparatory adjustment for the second round of assessment.

[0105] When the monitoring window period ends, or when the instantaneous tolerance stability score recovers to above the stability threshold in subsequent time segments and maintains a preset stable duration, the data sampling phase of the second round of evaluation is formally initiated.

[0106] During the data sampling phase of the second round of evaluation, the standard sampling frequency is restored, a fully updated first comfort-related dataset is collected, and the steps for generating updated first-dimensional comfort evaluation parameters and subsequent fusion calculations as described in claim 1 are executed.

[0107] After generating the second round of adjustment instruction set, it is logically compared and merged with all the preparatory adjustments recorded by the real-time fine-tuning program. If the second round of adjustment instruction set is consistent with any preparatory adjustment in the adjustment direction but different in magnitude, the one with the larger magnitude is adopted to form the final merged second round of adjustment instruction. If the directions are opposite, the second round of adjustment instruction set calculated based on the complete data sampling stage is adopted, and the preparatory adjustments with opposite directions are ignored.

[0108] The merged second-round adjustment instruction set is sent to the nursing device control unit as the final second-round adjustment instruction set.

[0109] In this embodiment, after the nursing device control unit confirms receipt of the first set of adjustment instructions and begins application, the computing device does not immediately enter the standardized second round of data acquisition. Instead, it first initiates a carefully designed monitoring window. The length of this window is not fixed but is dynamically calculated and determined by the computing device based on the adjustment magnitude of the first set of adjustment instructions. For example, if the adjustment mainly involves a significant reduction in energy level, the monitoring window may be set shorter to quickly assess the user's immediate relief response. If the adjustment involves fine-tuning of complex parameters such as pulse waveforms, the window may be set longer to observe trends over a longer period. The computing device sends instructions to the nursing device control unit, requesting that the sampling frequency of its physiological signal sensor be increased to a monitoring sampling frequency, which is much higher than the initial sampling frequency, within the monitoring window to obtain data with higher temporal resolution.

[0110] During the monitoring window, the computing device continuously receives high-frequency physiological signal data streams through the first data interface, forming a monitoring dataset. A real-time analysis subprocess runs within the computing device. This subprocess first divides the continuous monitoring data stream into a series of equal-length time segments, each potentially only a few seconds long. For each recently passed time segment, the subprocess immediately performs rapid analysis on the physiological signal data within it. This analysis includes calculating several core statistics: the signal's mean within the segment reflects the overall level; the variance reflects the degree of fluctuation; and the slope of the trend line obtained through linear fitting reflects whether the signal is rising, falling, or stable.

[0111] The statistical combination calculated for each time segment is immediately fed into a dedicated logic unit called the immediate response assessment model. This model is not a complex machine learning model, but rather a decision tree or state machine built upon a large number of clinical experience rules. It executes a series of if-then rules based on the input statistics. For example, rule one: if the trend slope is consistently positive and large, and the variance is also increasing, then this may mean that discomfort is accumulating and intensifying, and the model outputs a low immediate tolerance stability score; rule two: if the trend slope is negative or close to zero, and the variance is small, then this means the response is stabilizing or alleviating, and the model outputs a higher score. The model outputs a simple rating such as high, medium, low, or a numerical score.

[0112] The computing device continuously tracks these real-time scores. It sets a stability threshold and a continuous counting mechanism. If the model outputs a score below the stability threshold for three consecutive or other preset number of time segments, the computing device determines that despite the application of the first round of adjustments, the user's physiological response has not stabilized and may even be developing in an unfavorable direction. At this point, the system triggers an instant fine-tuning program, interrupting passive monitoring and switching to active intervention. The instant fine-tuning program first analyzes the statistical characteristics of the latest time segment and compares them with the characteristics before the first round of adjustments. It is based on a set of simplified heuristic rules. For example, if the skin temperature rise rate is still too fast, it will suggest an additional small reduction in energy or an enhancement in cooling on top of the energy reduction in the first round, quickly generating an instant fine-tuning instruction. This instruction is preliminary and tentative. Before sending this instant fine-tuning instruction, the computing device will attach a special mark to indicate that this is a preliminary adjustment suggestion generated based on high-frequency monitoring data in a locked state. After receiving this instruction, the nursing device control unit, still in a locked state, will execute this fine-tuning, but at the same time record this adjustment as a temporary preliminary adjustment that may be overridden by subsequent formal instructions, and feed this execution record back to the computing device.

[0113] When the monitoring window period ends normally, reaching the preset dynamic duration, or after the immediate fine-tuning procedure is triggered, and after several time segments, the immediate tolerance stability score recovers and remains above the stability threshold for a preset stable duration, the computing device considers the user's state to be sufficiently stable and can enter the data sampling phase of the formal second round of evaluation. At this time, the computing device sends a command to switch the sampling frequency from the monitoring sampling frequency back to the standard sampling frequency.

[0114] During the data sampling phase of the second round of evaluation, the computing device collects sufficient physiological signal data, which may include immediate fine-tuning under the new parameters, according to similar but potentially shorter time requirements as the initial evaluation, forming an updated first comfort-related dataset. Subsequent steps, such as generating updated evaluation parameters and performing fusion calculations following locking logic, are consistent with the process described in claim 1, ultimately generating a second-round set of adjustment instructions.

[0115] Before sending the formal second-round adjustment instruction set, the computing device performs a crucial instruction coordination and merging step. First, it retrieves all immediate fine-tuning instructions issued during the monitoring window and their execution effect records from the nursing instrument control unit or local logs. Then, it compares the formal second-round adjustment instruction set with these preliminary adjustment records item by item. The comparison logic is as follows: for the same nursing parameter, such as energy level, it compares the suggested adjustment direction (increase / decrease) and adjustment magnitude of the formal and preliminary adjustment instructions. If the directions are consistent, the suggestion with the larger adjustment magnitude is adopted, forming a merged instruction to ensure sufficient adjustment strength. If the directions are opposite, the system prioritizes the formal instruction generated based on more comprehensive and formal second-round assessment data sampling, ignoring the opposing preliminary adjustment, as the preliminary adjustment may be based on transient interference signals. Through this coordination and merging process, the most reasonable merged second-round adjustment instruction set is finally determined. The computing device then sends this final instruction set to the nursing instrument control unit, completing the formal second-round adjustment.

[0116] Furthermore, the first comfort-related dataset is processed based on a preset first comfort mapping rule to generate a first-dimensional comfort evaluation parameter, specifically including the following detailed steps:

[0117] The original first type of physiological signal sequence obtained through the first data interface is subjected to signal quality verification. The signal quality verification includes detecting the signal loss rate and whether the noise amplitude exceeds the allowable range. If the signal quality verification fails, a re-acquisition request is sent to the signal acquisition module, and subsequent steps are paused until a valid signal sequence is obtained.

[0118] Reference Figure 3 The original first-type physiological signal sequences that passed the verification were standardized, and the standardization process included:

[0119] The raw signal values ​​are converted into percentage changes relative to the target user's baseline values ​​at rest to eliminate absolute differences in individual baseline physiological levels.

[0120] Extract multiple predefined time-domain and frequency-domain features from the standardized signal sequence;

[0121] The time-domain features include:

[0122] The rising slope of the signal during the application of the nursing pulse, the recovery slope during the pulse interval, and the peak and trough values ​​of the signal amplitude;

[0123] The frequency domain features include: power spectral density in a specific frequency band that is associated with autonomic nervous system regulatory activity;

[0124] Each extracted time-domain and frequency-domain feature is assigned a preset weight coefficient, which is predetermined based on the correlation between the feature and the subjective discomfort of Thermage treatment.

[0125] Each extracted feature is multiplied by its corresponding weight coefficient and then summed to obtain a preliminary composite physiological response index.

[0126] The preliminary composite physiological response index is input into a nonlinear transformation function to map the composite physiological response index to a scale value. This nonlinear transformation function has high sensitivity in the middle range and decreases sensitivity in the range close to the extreme value to prevent extreme jumps in the results.

[0127] The scale value processed by the nonlinear transformation function is output as the first dimension comfort evaluation parameter, which has the same dimensions and comparability as the second dimension comfort baseline parameter, facilitating subsequent fusion calculation.

[0128] In this embodiment, the processing begins with a rigorous data access check. The original first-type physiological signal sequence received by the computing device from the first data interface is first sent to a signal quality verification module. This module performs multiple checks: First, it checks the continuity of the data packet sequence number or timestamp; if jumps or missing values ​​exceed a preset number, the signal loss rate is deemed excessive. Second, it checks the range of the signal amplitude, for example, whether the skin impedance value is within the possible physical range of the human body, such as between 1 kiloohms and 10 megaohms; if it exceeds the range, it is considered invalid. Third, it calculates the noise level of the signal within a short time window, for example, the root mean square value of the signal after high-pass filtering; if the noise level exceeds the allowable threshold preset based on the sensor specifications, the noise amplitude is deemed excessive. If any of the above checks fails, the signal quality verification module generates an error code and sends a re-acquisition request command to the nursing device control unit through the first data interface, indicating which signal channel has a problem. At the same time, the entire first comfort mapping rule processing flow is paused, waiting for new valid data to arrive. Only when all checks pass is the original signal sequence marked as valid and proceeds to the next stage.

[0129] Next is the standardization preprocessing of the signal. For each type of physiological signal, such as skin impedance and temperature, the computing device reads the resting baseline value corresponding to the target user from the user profile or a resting period record before the start of the treatment. This baseline value may be a fixed value or the average value of the signal over a resting period. The core of the standardization process is to convert the original signal value into a rate of change relative to the individual baseline.

[0130] The specific operation is as follows: For the original signal value of each sampling point, subtract the individual's resting baseline value of that signal, then divide the difference by the resting baseline value, and multiply by 100% to obtain a percentage rate of change. For example, if the original skin impedance is 200 kΩ and the individual's resting baseline is 250 kΩ, then the standardized value is (200-250) / 250×100%=-20%. This processing eliminates the influence of absolute values ​​caused by inherent differences in skin dryness, stratum corneum thickness, etc. among different users, making the signal changes between different users comparable.

[0131] The standardized signal sequence then enters the feature extraction stage. The feature extraction engine segments the signal according to a predefined time window synchronized with the nursing pulse. It identifies the start and end points of each nursing pulse, typically indicated by a time-stamped signal synchronously sent by the nursing device control unit. For a pulse cycle, which includes the pulse duration and the subsequent cool-down / interval period, the engine extracts the following time-domain features:

[0132] 1) Rising / falling slope: During the pulse duration, calculate the linear fitting slope of the segment where the signal changes most drastically, such as the falling slope of the impedance;

[0133] 2) Recovery slope: During the cooling period after the pulse ends, calculate the linear fitting slope of the signal during the initial stage of the recovery process to the baseline;

[0134] 3) Peak and trough values: Find the maximum and minimum values ​​of the signal within the period and calculate their changes relative to the start point of the period;

[0135] 4) Area under the curve: The area enclosed by the deviation of the signal curve from the resting baseline is calculated, reflecting the dose of the stimulus. At the same time, for signals suitable for frequency domain analysis, such as muscle micro-vibrations, the engine will also perform short-time Fourier transform to extract frequency domain features, such as the signal power spectral density integral value in the low-frequency band related to sympathetic nerve activity in the 0.05 to 0.15 Hz frequency band.

[0136] After feature extraction is complete, the weight allocation module begins its work. This module embeds a feature-weight association table, which is pre-determined based on the analysis of a large amount of clinical trial data. The table assigns a fixed weight coefficient to each extracted feature, such as impedance drop slope, temperature recovery slope, and low-frequency power. The sum of the weight coefficients is usually normalized to 1. The allocation principle is: the physiological feature with a higher correlation to the patient's subjective pain score has a larger weight coefficient; for example, clinical trials may find that the skin impedance drop slope has the highest correlation with the transient pain score, so it is given the largest weight; while temperature peaks, although intuitive, may have a slightly lower correlation due to the lag in heat diffusion, and thus a correspondingly lower weight.

[0137] Then, the computational engine performs a weighted calculation. For the current pulse cycle, the actual value of each extracted feature, such as the impedance drop slope of -15% / second, is multiplied by its corresponding weighting coefficient, such as 0.4, to obtain the weighted contribution value of that feature. The weighted contribution values ​​of all features are summed to obtain the preliminary composite physiological response index for this pulse cycle.

[0138] This initial index is then fed into a non-linear transformation function, typically a piecewise function or an S-curve function. The design aims to map the initial index onto a 0-100 scale that better reflects clinical perception, being more sensitive to changes in the middle range and less sensitive to changes near extremes of comfort or discomfort, thus avoiding drastic fluctuations in the parameter. For example, when the initial index is between 30 and 70, the slope of the transformation function is steeper, meaning that small changes in the index will cause significant changes in the comfort assessment parameter. When the initial index is below 10 or above 90, the slope of the transformation function becomes gentler, suppressing changes in the assessment parameter within a smaller range even with large fluctuations in physiological responses. After this non-linear transformation, a value between 0 and 100 is finally output, which is formally defined as the first dimension of the comfort assessment parameter. The entire process is performed once for each complete nursing pulse cycle or every few cycles of aggregated data, resulting in a parameter sequence that is continuously updated as nursing care progresses.

[0139] Furthermore, the second type of physiological baseline data includes heart rate variability indices, baseline values ​​of skin conductivity response, and pain threshold pressure test results collected under non-thermal stimulation conditions;

[0140] The user-reported tolerance level information was collected before the nursing care using a standardized questionnaire, which included subjective feelings about previous similar energy device nursing experiences and self-expectation of pain.

[0141] In this embodiment, the collection environment and time point for the second type of physiological baseline data are strictly regulated. It must be collected in a quiet, temperature-appropriate room on the day of the planned Thermage treatment or within a week prior to the treatment when the target user is in a calm, relaxed, and non-care-related state.

[0142] Specific content includes: Heart rate variability indicators: Calculated from five consecutive minutes of electrocardiogram recordings, time-domain indicators such as the standard deviation of all normal sinus intervals (SDNN) and frequency-domain indicators such as the low-frequency power to high-frequency power ratio (LF / HF) reflect the balance and stress level of the autonomic nervous system. Baseline skin conductivity response: Measurements are taken continuously for two minutes on the user's palm or fingers using a standard skin conductivity meter under no external stimulation. The lowest stable average value is taken as the baseline, reflecting basic sympathetic tone. Pain threshold pressure test results: Using a pressure pain sensor with a standard pressure probe, uniformly increasing pressure is applied to a non-care area such as the forearm. The pressure value at which the user first reports a clear stinging sensation (not pressure) quantifies the user's starting point for pain sensitivity to mechanical pressure.

[0143] The user-reported tolerance level information was collected using a standardized questionnaire that had undergone reliability and validity testing. This questionnaire was completed independently by the user under guidance or in consultation with a professional before the start of treatment.

[0144] The scale includes multiple dimensions: First, a review of past treatment experiences: users are asked to describe their specific feelings when receiving any form of energy-based skin treatment in the past, such as laser hair removal or IPL skin rejuvenation, and to rate their average discomfort at that time using the Visual Analogue Scale (VAS) from 0 to 10. Second, pain expectation and anxiety assessment: this includes questions such as "How nervous or anxious are you about the upcoming Thermage treatment?", with rating options provided. Third, daily pain sensitivity: users are asked about their pain response to common stimuli such as minor bumps and insect bites in daily life, and their self-perception compared to their peers. Fourth, it may include concise psychological trait screening questions, such as the degree of attention paid to discomfort. The answers to all these questions will be converted into quantifiable scores by the subsequent second comfort mapping rule processing engine according to preset scoring rules, serving as the core input for constructing the user's subjective tolerance profile.

[0145] Furthermore, the comfort assessment parameters of the first dimension are fused with the comfort baseline parameters of the second dimension to generate a comprehensive comfort index, which is achieved through the following formula:

[0146] CCI=α×(w1×SCA+w2×SCB)+(1-α)×Dynamic_Compensation

[0147] Wherein, CCI represents the comprehensive comfort index; SCA represents the comfort assessment parameter of the first dimension; SCB represents the comfort baseline parameter of the second dimension; w1 and w2 are weights pre-calibrated based on clinical data, and w1+w2=1;

[0148] α is a dynamic adaptation factor, whose value is automatically adjusted based on the coefficient of variation of the first dimension comfort assessment parameter SCA in the most recent three sampling periods. The larger the coefficient of variation, the smaller the α value, indicating that it is more dependent on the second dimension baseline parameter.

[0149] Dynamic_Compensation is a dynamic compensation item whose value is obtained from a predefined compensation value lookup table based on the proportion of time that care has been performed to the total expected care time and the number of historical adjustments to the current energy level.

[0150] In this embodiment, the fusion calculation process is managed by a dedicated fusion logic controller. This controller receives two core inputs: the real-time generated first-dimensional comfort assessment parameter denoted as SCA and the pre-stored second-dimensional comfort baseline parameter denoted as SCB.

[0151] The first step of the controller is to assess the instantaneous reliability of the SCA (Self-Care Ability). It maintains a small buffer containing the SCA values ​​corresponding to the most recent N consecutive nursing pulse cycles (e.g., N=5). The controller calculates the standard deviation of the coefficient of variation (CVA) for all SCA values ​​within this buffer, divided by the mean. This CVA is used as a quantification of real-time data volatility. Internally, the controller has a volatility-reliability lookup table that defines the real-time data reliability levels corresponding to different CVA intervals, such as high reliability, medium reliability, and low reliability.

[0152] The second step involves the controller dynamically determining the fusion weights based on the real-time data reliability level obtained in the first step. Internally, it uses a reliability-weight allocation table. For example, if the reliability level is high, the weights are allocated as follows: SCA weight (w1) = 0.7, SCB weight (w2) = 0.3. If the reliability level is low, the weights might change to w1 = 0.3, w2 = 0.7. Thus, when the real-time physiological signal is stable and reliable, the fusion result tends to reflect the current immediate state; when the real-time signal is noisy and fluctuates wildly, it relies more on baseline parameters representing the user's inherent characteristics to maintain decision stability.

[0153] The third step is for the controller to use dynamically determined weights w1 and w2 to calculate the weighted sum of SCA and SCB: weighted sum = w1 × SCA + w2 × SCB.

[0154] The fourth step involves the controller introducing a dynamic adaptation factor, denoted as α. The value of α is not fixed but determined by two factors: first, the reliability level of the real-time data calculated in the previous step; the higher the reliability, the closer α tends to be to 1; second, the current nursing stage (initial, middle, or late), which is a process status code obtained from the nursing instrument control unit. The controller obtains the specific α value by querying a two-dimensional table, one dimension representing the reliability level and the other the nursing stage. The role of α is to modulate the weighted sum.

[0155] Fifth, the controller calculates an independent dynamic compensation item, which primarily considers the impact of the care process. It queries two pieces of information: A) the percentage of currently emitted care pulses out of the total planned pulses; and B) the cumulative number of times the energy level has been increased or decreased since the start of care. Based on these two pieces of information, the controller accesses a process compensation lookup table. For example, when care completion exceeds 70% and there has been a history of multiple energy reductions, the lookup table may output a negative compensation value, meaning that a slightly more aggressive approach can be taken during fusion, as the user may have gradually adapted. Conversely, in the early stages of care, the compensation value may be zero or positive, indicating a need for greater conservatism. This compensation item is calculated independently of the weighted sum.

[0156] Finally, the controller performs the final combined calculation, which multiplies the weighted sum by the dynamic adaptation factor α, and then adds the dynamic compensation term.

[0157] The specific operational instruction sequence is as follows: First, calculate α × (weighted sum), then add (1-α) × Dynamic_Compensation. This expression is only for illustrative purposes; in practice, it involves adding a compensation term. The calculated result is normalized to the range of 0-100; if it exceeds this range, it is truncated. This final result is the Comprehensive Comfort Index (CCI). The entire process ensures that the index reflects both real-time responses and individual baselines, while also incorporating considerations of data quality and nursing context through credibility assessment and process compensation.

[0158] Furthermore, the comfort-parameter adjustment mapping table is a multi-dimensional lookup table. It uses the comprehensive comfort index as the primary lookup key and the body part code of the current nursing area and the current cumulative emission energy as secondary lookup keys to output the corresponding set of parameter adjustment instructions. The set of parameter adjustment instructions not only includes adjustments to nursing parameters but also includes suggested stable observation durations after adjustments.

[0159] In this embodiment, the mapping table is essentially a set of rules database, designed based on clinical expert experience and analysis of a large amount of nursing data. Its primary query dimension is the Comprehensive Comfort Index (CCI), and the CCI range of 0-100 is divided into several discontinuous intervals that may not be of equal width, such as: comfort zone CCI 0-30, observation zone CCI 31-60, fine-tuning intervention zone CCI 61-80, and strong intervention zone CCI 81-100. Each interval is a main entry in the table.

[0160] For each CCI interval entry, the mapping table does not directly store a single adjustment instruction, but rather a set of adjustment strategy options. The applicability of each strategy option is determined by a secondary lookup key, which includes at least: 1) Body part code: The treatment area is coded, such as the forehead, cheek, jawline, and neck. Different areas have different skin thicknesses, nerve distributions, and fat content, so their tolerance to the same CCI value and the required adjustments may differ. For example, for thin and sensitive neck skin, the mapping table may suggest a 15% energy reduction when the CCI enters the fine-tuning intervention zone; while for thicker cheek tissue, the same CCI interval may only suggest a 5% energy reduction; 2) Current cumulative emitted energy: This is an approximation of the total energy emitted from the start of treatment to the current moment, in joules, or its energy range (low, medium, high). As the cumulative energy increases, the skin's thermal accumulation may change its reactive characteristics, thus requiring a change in adjustment strategy. For example, when the cumulative energy is high in the later stages of treatment, even if the CCI only slightly enters the observation zone, the mapping table may suggest pausing treatment for 30 seconds and increasing cooling, while in the early stages of treatment, it may only suggest maintaining parameter observation.

[0161] When the system needs to query the mapping table, it inputs the current CCI value, body part code, and the current cumulative emission energy range. The mapping table retrieval logic first locates the range entry for the CCI, and then, under that entry, filters for the most suitable adjustment strategy option based on the body part code and energy range. The output of this option is very specific, a set of adjustment instructions, including: a) specific adjustments to the energy level, such as reducing it to 85% of the original value or increasing it by 3 energy units; b) specific adjustments to the pulse duration, such as shortening it by 10%; c) adjustments to cooling jet parameters, such as advancing the jet start time by 50 milliseconds, extending the duration of a single jet by 200 milliseconds, or adding an auxiliary jet during the pulse interval. In addition, the instruction set also mandates a suggested observation period for post-adjustment stabilization; for example, for minor adjustments, it is recommended to observe for the next 3-5 pulse cycles; for significant adjustments or strong interventions, it may be recommended to observe for 10 pulse cycles or longer, during which time further automated adjustment evaluations should be paused to allow the body to fully adapt to the new parameters. This observation period, as a meta-instruction, guides the computing device on when to start the next evaluation cycle.

[0162] Furthermore, after receiving and persistently storing the complete protocol data package, it also includes a retrospective analysis and optimization cycle based on historical protocols, independent of the real-time care process, specifically including:

[0163] Periodically retrieve all complete protocol data packages related to Thermage care generated within a preset historical time period from the protocol database;

[0164] All retrieved complete solution data packets are parsed in a structured manner to extract key fields;

[0165] Construct an analysis dataset where each record corresponds to a complete nursing process, and the feature variables include statistical features derived from the aforementioned key fields;

[0166] Define optimization target variables, which are derived from the follow-up scores of nursing effect and the overall user satisfaction scores collected subsequently. After the nursing care ends, the data is collected through a safe follow-up system for a preset number of days and associated with the analysis dataset through an anonymous ID.

[0167] Using statistical correlation analysis, feature variables that are significantly positively or negatively correlated with the target variable are selected from the analysis dataset.

[0168] Based on the selected significantly relevant feature variables, the parameter weights in the first comfort mapping rule and / or the second comfort mapping rule are optimized in reverse, and / or the specific adjustment values ​​in the comfort-parameter adjustment mapping table are optimized;

[0169] For the first comfort level mapping rule, if it is found that the final user satisfaction is strongly correlated with the change of a certain physiological signal feature in the early stage of care, then the weight of that feature in the weight coefficient allocation is adjusted.

[0170] For the comfort-parameter adjustment mapping table, if statistical analysis shows that when the comprehensive comfort index is in a specific range, adopting a more aggressive or more conservative energy adjustment range than the current mapping table will result in higher subsequent satisfaction and effect scores, then the adjustment instructions corresponding to that range will be updated.

[0171] The optimized mapping rules and mapping table versions are numbered and stored in the rule base;

[0172] When generating new personalized comfort solutions for target users, the latest version of rules and mapping tables is applied by default, or operators are allowed to select a specific historical version under certain circumstances;

[0173] The retrospective analysis optimization cycle is executed automatically at preset intervals, forming a closed loop of continuous iteration and improvement of the core logic of solution generation.

[0174] In this embodiment, the optimization cycle is automatically initiated by a system scheduled task trigger, for example, set to execute at midnight on the first Sunday of each quarter. After triggering, the system starts the retrospective analysis engine. The first step of the engine is data retrieval and aggregation. It sends a structured query request to the personalized care plan database to retrieve all complete plan data packages generated in the most recent quarter or a preset time period. The query conditions may also include optional filters such as nursing device model and operator level for more targeted analysis. All retrieved data packages are loaded into the engine's temporary analysis workspace.

[0175] The second step is data parsing and feature engineering. The engine parses each data packet one by one, extracting fixed core fields, such as anonymized user IDs used only for data association, without exposing identity, age, gender grouping information, or nursing site. More importantly, it derives a large number of statistical features from the rich time-series and event data contained in the data packets. These features aim to condense the characteristics of a nursing process. For example, from the comprehensive comfort index sequence, it calculates its initial value, maximum value, minimum value, average value, standard deviation, volatility, and overall upward or downward trend through linear regression slope; from the parameter adjustment records, it calculates the total number of adjustments, the ratio of energy increases to decreases, and the maximum single adjustment magnitude; from the nursing log, it extracts whether there are adjustments triggered by user-initiated pauses, whether there are interruptions due to equipment contact alarms, and their types. After parsing and derivation, each complete protocol data packet is converted into a record in the analysis dataset, containing dozens or even hundreds of feature variables.

[0176] The third step is to optimize the target association. The optimization target is not generated in real time, but rather comes from an external follow-up system. At an agreed-upon timeframe after the treatment, such as 7 or 30 days, the system collects two key scores through secure follow-up platforms such as SMS-linked questionnaires and app push notifications: first, a follow-up score on the treatment effectiveness, assessed by the user or nurse based on objective improvements such as increased skin firmness; and second, an overall user satisfaction score, covering comfort experienced during the treatment, satisfaction with the results, and overall experience. This follow-up data, using the same anonymized ID, is imported into a retrospective analysis engine and precisely correlated with the analysis records generated in the second step. A complete analysis record now includes the feature variable X and the optimization target variable Y.

[0177] The fourth step is statistical correlation analysis. The engine uses statistical methods such as Pearson correlation analysis, Spearman rank correlation, or feature importance analysis based on tree models to evaluate the association strength between each derived feature variable and the two optimization target variables, nursing effect score and satisfaction score, in the analysis dataset. The engine sets a statistical significance level, such as p-value < 0.05, and filters out a list of feature variables that are significantly positively or negatively correlated with either optimization target variable. For example, the analysis may find that a small standard deviation fluctuation of the comprehensive comfort index during the nursing process is significantly positively correlated with a high overall user satisfaction score; or that a later time of the first adjustment is significantly negatively correlated with a high nursing effect score.

[0178] The fifth step is the reverse optimization of rules and mapping tables, which is the core knowledge update step. Based on the selected significant relevant features, the engine proposes adjustment hypotheses for the core components of the generated scheme.

[0179] For the first comfort level mapping rule: If it is found that in nursing records with high satisfaction, the value of a certain physiological feature, such as the skin resistance recovery slope, is generally within a narrow ideal range in the calculation process of the first dimension comfort assessment parameters, and the weight of this feature in the current rule may be too low, the engine will generate an adjustment suggestion: increase the weight coefficient of the skin resistance recovery slope feature in the feature weight comparison table. The specific increase can be suggested based on the correlation coefficient between this feature and satisfaction.

[0180] For the second comfort mapping rule: if the analysis shows that the user's self-reported pain expectation score has a weak predictive ability for discomfort events in the final care process (i.e., the correlation is not significant), while the pain threshold stress test results have a strong correlation, the engine will suggest reducing the weight of the subjective questionnaire score and correspondingly increasing the weight of the physiological pain threshold test results in the calculation formula of the second dimension comfort baseline parameters.

[0181] For the comfort-parameter adjustment mapping table: optimization is more direct. For example, statistical analysis might reveal that for cheek care, when the overall comfort index (CCI) is in the 61-70 range, historical data showing an 8% energy reduction strategy resulted in significantly higher subsequent effects and satisfaction scores compared to the current mapping table's default 5% energy reduction strategy for that range. Therefore, the engine would generate a clear update suggestion: change the energy adjustment instruction in the mapping table corresponding to the cheek, CCI range of 61-70, from a 5% reduction to an 8% reduction. Alternatively, a new rule branch might be added.

[0182] The sixth step is version management and deployment. The engine summarizes all these adjustment suggestions and generates a rule optimization suggestion report. This report needs to undergo a simulation verification or expert review process, which can be an automated A / B test simulation or a manual review interface. After approval, the system creates a new version of the first comfort mapping rule, the second comfort mapping rule, and the comfort-parameter adjustment mapping table, and assigns a unique version number, such as V2.1, to this version combination. The new version rules are stored in the rule base. When the system subsequently generates personalized comfort plans for new target users, it automatically calls the latest effective version in the rule base by default. The system also retains the ability to roll back to older versions for special cases or comparative studies. The entire retrospective analysis and optimization cycle ends here, and the system awaits the trigger of the next cycle. This process ensures that the plan generation method can continuously evolve with the accumulation of nursing cases, but it is completely independent of any real-time nursing control cycle.

[0183] Furthermore, the immediate response evaluation model is a rule-based state machine, in which the internal state switches according to the input feature statistics and multiple predefined state transition conditions;

[0184] The state transition conditions include: consistency of the trend slope sign of continuous time segments, the multiple relationship of variance to the historical window mean, and the matching degree between the combination of feature statistics and the pre-stored typical maladaptive pattern template.

[0185] In this embodiment, the immediate response evaluation model is a state machine based on deterministic rules, and its internal state transition logic is described in detail below:

[0186] The core of this model is a state machine that defines a finite number of internal states, such as a steady state, a slightly volatile dynamic state, a deteriorating trend state, and a recovering trend state. The combination of the characteristic statistics—mean, variance, and trend slope—input for each time segment serves as the input conditions driving the state transitions.

[0187] The model initializes in an unknown state. Upon receiving the first time segment of data, it enters a basic state according to a set of initial rules. Subsequently, each time a new time segment of data is input, the model performs a state evaluation and possible transitions.

[0188] Its rule base includes, for example:

[0189] Rule A: Entering / Maintaining a Steady State: If the variance of the current time segment is below the threshold V1 and the absolute value of the trend slope is below the threshold S1, then regardless of the previous state, the model will set the state to a steady state and output a high instantaneous tolerance stability score.

[0190] Rule B enters mild volatile dynamics: If the current variance is between the thresholds V1 and V2 (V2>V1), but the absolute value of the trend slope is still lower than S1, and the previous state was not a deteriorating trend state, then the model transitions to mild volatile dynamics and outputs a moderate score.

[0191] Rule C: Entering a deteriorating trend state: If the trend slope is positive for two consecutive time segments and the value exceeds the threshold S2 by a value larger than S1, it indicates that the signal continues to change in an unfavorable direction; or, if the current mean exceeds a danger threshold M1 and the variance increases at the same time, the model will be forced to enter a deteriorating trend state and output a very low score. Entering this state is one of the main conditions for triggering the immediate fine-tuning procedure.

[0192] Rule D transitions from a deteriorating trend state: When in a deteriorating trend state, if the trend slope of subsequent input time segment data becomes negative, indicating the start of recovery, and the variance decreases, the model can transition to a recovering trend state, with the output score gradually improving. The model must remain stable in the recovering trend state for a sufficient number of time segments—for example, three consecutive segments satisfying the steady-state rule A—before it will eventually transition back to a steady state.

[0193] Furthermore, the model internally stores statistical feature templates for a few typical discomfort patterns. For example, a severe pain response template might correspond to a combination of a sharply rising mean, extremely high variance, and a very high slope. The model will perform similarity matching, such as calculating Euclidean distance, between the statistical combination of the current time segment and these templates. If the matching degree exceeds a certain threshold, even if the specific slope or variance continuity rules mentioned above are not met, the model may directly jump to the worsening trend state. This is a rapid path judgment, and the model's output score is directly determined by its current state. Each state is bound to a preset score or level. This rule-based state machine has a transparent and predictable decision-making process with low computational overhead, making it very suitable for millisecond-level real-time judgment in the nursing process.

[0194] According to a second embodiment of the present invention, the present invention claims protection for a system for generating personalized comfort plans for Thermage care, comprising:

[0195] One or more processors;

[0196] A memory having stored one or more programs that, when executed by one or more processors, enable the one or more processors to implement the method for generating a personalized comfort solution for Thermage care.

[0197] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus 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 system, 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, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0198] Furthermore, the functional units in the various embodiments of 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. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0199] The specific embodiments of the invention have been described in detail above, but they are only examples, and this application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. A method for generating a personalized comfort plan for Thermage treatment, characterized in that, Includes the following steps: Obtain the first comfort-related dataset associated with the target user, process the first comfort-related dataset based on the preset first comfort mapping rule, and generate the first dimension comfort evaluation parameters; Obtain the pre-stored second comfort-related dataset, process the second comfort-related dataset based on the preset second comfort mapping rule, and generate the second-dimensional comfort baseline parameters; The first dimension comfort assessment parameters and the second dimension comfort baseline parameters are fused together to generate a comprehensive comfort index. Based on the comprehensive comfort index, a predefined comfort-parameter adjustment mapping table is queried to determine the first set of adjustment instructions for the initial basic nursing parameters; A comfort lock command is sent to the nursing device control unit, instructing the nursing device control unit to maintain the parameter adjustment logic calculated based on the fusion of the first dimension comfort assessment parameters and the second dimension comfort baseline parameters after receiving and applying the first round of adjustment command set, until a unlock command is received; After sending the comfort lock command, a second round of evaluation is initiated based on the dynamic feedback mechanism to generate an updated comprehensive comfort index, and a second round of adjustment command set is generated based on the updated comprehensive comfort index; The second set of adjustment instructions is sent to the nursing device control unit. According to the comfort lock instruction, when applying the second set of adjustment instructions, the parameter adjustment logic previously determined by the first dimension comfort assessment parameter and the second dimension comfort baseline parameter is maintained.

2. The method according to claim 1, characterized in that, After sending the second set of adjustment instructions to the nursing device control unit, the method further includes: Receive an implementation confirmation signal from the nursing device control unit, confirming that the second round of adjustment command set has been successfully applied and a stable nursing waveform has been generated; In response to the implementation confirmation signal, a comfort lock release command is generated and sent to the nursing device control unit. The comfort lock release command is used to instruct the nursing device control unit that if new comfort-related data is received in subsequent nursing stages, the nursing parameters should be adjusted according to the preset default parameter update strategy or the newly triggered complete multi-dimensional fusion calculation process, and the specific fusion logic established by the previous first and second rounds of evaluation should no longer be forcibly maintained. Receive and persistently store a complete protocol data package containing the final application set of nursing parameters, the corresponding comprehensive comfort index sequence, and a nursing process summary log. The complete protocol data package is associated with the unique identifier of the target user and stored in the protocol database. Based on the complete protocol data package, a personalized comfort protocol report document is automatically generated, which includes the evolution path of nursing parameters, the change curve of comfort index, and descriptions of key adjustment nodes.

3. The method according to claim 1, characterized in that, The second round of evaluation, initiated based on a dynamic feedback mechanism, and the subsequent multi-round adjustment and verification process following the issuance of the comfort lock command, further include the following detailed steps: After the first set of adjustment instructions is applied to the nursing instrument control unit, a monitoring window period is initiated. The duration of the monitoring window period is dynamically set according to the type of nursing parameter being adjusted and its change range. During the monitoring window period, the first type of physiological signal sequence of the target user is continuously collected through the first data interface at a monitoring sampling frequency higher than the initial sampling frequency to form a monitoring dataset; The monitoring dataset is subjected to real-time fragmented analysis, and the monitoring window period is divided into multiple consecutive time segments. For each time segment, the characteristic statistical values ​​of the first type of physiological signal are calculated. The characteristic statistical values ​​include the mean, variance and the trend slope with respect to the time segment. The characteristic statistical values ​​of each time segment are input into a preset instant response evaluation model, which is configured to output an instant tolerance stability score. The instant tolerance stability score is used to characterize the degree of fluctuation of the user's physiological response relative to the expected response within that time segment. If the immediate tolerance stability score is lower than the preset stability threshold for at least three consecutive time segments, the physiological response induced by the first round of adjustment instructions is determined to be in an unstable state, triggering the immediate fine-tuning procedure. The instant fine-tuning procedure includes: based on the characteristic statistical values ​​of the current time segment, reversely deriving the fine-tuning amount of some parameters in the first round of adjustment instruction set, and generating an instant fine-tuning instruction; the instant fine-tuning instruction is sent to the nursing device control unit under the framework of the comfort lock instruction, and the nursing device control unit performs rapid parameter correction accordingly, while recording this fine-tuning as a preparatory adjustment for the second round of evaluation; When the monitoring window period ends, or when the instantaneous tolerance stability score recovers to above the stability threshold in subsequent time segments and maintains a preset stable duration, the data sampling phase of the second round of evaluation is formally initiated. During the data sampling phase of the second round of evaluation, the standard sampling frequency is restored, a fully updated first comfort-related dataset is collected, and the steps for generating updated first-dimensional comfort evaluation parameters and subsequent fusion calculations as described in claim 1 are executed. After generating the second round of adjustment instruction set, it is logically compared and merged with all the preparatory adjustments recorded by the real-time fine-tuning program. If the second round of adjustment instruction set is consistent with any preparatory adjustment in the adjustment direction but different in magnitude, the one with the larger magnitude is adopted to form the final merged second round of adjustment instruction. If the directions are opposite, the second round of adjustment instruction set calculated based on the complete data sampling stage is adopted, and the preparatory adjustments with opposite directions are ignored. The merged second-round adjustment instruction set is sent to the nursing device control unit as the final second-round adjustment instruction set.

4. The method according to claim 1, characterized in that, The first comfort-related dataset is processed based on a preset first comfort level mapping rule to generate first-dimensional comfort evaluation parameters. The specific steps include the following detailed steps: The original first type of physiological signal sequence obtained through the first data interface is subjected to signal quality verification. The signal quality verification includes detecting the signal loss rate and whether the noise amplitude exceeds the allowable range. If the signal quality verification fails, a re-acquisition request is sent to the signal acquisition module, and subsequent steps are paused until a valid signal sequence is obtained. The original Class I physiological signal sequences that passed the verification were standardized, and the standardization process included: The raw signal values ​​are converted into percentage changes relative to the target user's baseline values ​​at rest to eliminate absolute differences in individual baseline physiological levels. From the standardized signal sequence, extract multiple predefined time-domain and frequency-domain features, the time-domain features including: The rising slope of the signal during the application of the nursing pulse, the recovery slope during the pulse interval, and the peak and trough values ​​of the signal amplitude; The frequency domain features include: power spectral density in a specific frequency band that is associated with autonomic nervous system regulatory activity; Each extracted time-domain and frequency-domain feature is assigned a preset weight coefficient, which is predetermined based on the correlation between the feature and the subjective discomfort of Thermage treatment. Each extracted feature is multiplied by its corresponding weight coefficient and then summed to obtain a preliminary composite physiological response index. The preliminary composite physiological response index is input into a nonlinear transformation function to map the composite physiological response index to a scale value. This nonlinear transformation function has high sensitivity in the middle range and decreases sensitivity in the range close to the extreme value to prevent extreme jumps in the results. The scale value processed by the nonlinear transformation function is output as the first dimension comfort evaluation parameter, which has the same dimensions and comparability as the second dimension comfort baseline parameter, facilitating subsequent fusion calculation.

5. The method according to claim 1 or 4, characterized in that, The second category of physiological baseline data includes heart rate variability indices, baseline values ​​of skin conductivity response, and pain threshold pressure test results collected under non-thermal stimulation conditions; The user-reported tolerance level information was collected before the nursing care using a standardized questionnaire, which included subjective feelings about previous similar energy device nursing experiences and self-expectation of pain.

6. The method according to claim 1, characterized in that, The process of fusing the first-dimensional comfort assessment parameters with the second-dimensional comfort baseline parameters to generate a comprehensive comfort index is specifically achieved through the following formula: CCI=α×(w1×SCA+w2×SCB)+(1-α)×Dynamic_Compensation Wherein, CCI represents the comprehensive comfort index; SCA represents the comfort assessment parameter of the first dimension; SCB represents the comfort baseline parameter of the second dimension; w1 and w2 are weights pre-calibrated based on clinical data, and w1+w2=1; α is a dynamic adaptation factor, whose value is automatically adjusted based on the coefficient of variation of the first dimension comfort assessment parameter SCA in the most recent three sampling periods. The larger the coefficient of variation, the smaller the α value, indicating that it is more dependent on the second dimension baseline parameter. Dynamic_Compensation is a dynamic compensation item whose value is obtained from a predefined compensation value lookup table based on the proportion of time that care has been performed to the total expected care time and the number of historical adjustments to the current energy level.

7. The method according to claim 1, characterized in that, The comfort-parameter adjustment mapping table is a multi-dimensional lookup table. It uses the comprehensive comfort index as the primary lookup key and the body part code of the current nursing area and the current cumulative emission energy as secondary lookup keys to output the corresponding set of parameter adjustment instructions. The set of parameter adjustment instructions not only includes adjustments to nursing parameters but also includes suggested stable observation durations after adjustments.

8. The method according to claim 2, characterized in that, After receiving and persistently storing the complete protocol data package, it also includes a retrospective analysis and optimization cycle based on historical protocols, independent of the real-time care process, specifically including: Periodically retrieve all complete protocol data packages related to Thermage care generated within a preset historical time period from the protocol database; All retrieved complete solution data packets are parsed in a structured manner to extract key fields; Construct an analysis dataset where each record corresponds to a complete nursing process, and the feature variables include statistical features derived from the aforementioned key fields; Define optimization target variables, which are derived from the follow-up scores of nursing effect and the overall user satisfaction scores collected subsequently. After the nursing care ends, the data is collected through a safe follow-up system for a preset number of days and associated with the analysis dataset through an anonymous ID. Using statistical correlation analysis, feature variables that are significantly positively or negatively correlated with the target variable are selected from the analysis dataset. Based on the selected significantly relevant feature variables, the parameter weights in the first comfort mapping rule and / or the second comfort mapping rule are optimized in reverse, and / or the specific adjustment values ​​in the comfort-parameter adjustment mapping table are optimized; For the first comfort level mapping rule, if it is found that the final user satisfaction is strongly correlated with the change of a certain physiological signal feature in the early stage of care, then the weight of that feature in the weight coefficient allocation is adjusted. For the comfort-parameter adjustment mapping table, if statistical analysis shows that when the comprehensive comfort index is in a specific range, adopting a more aggressive or more conservative energy adjustment range than the current mapping table will result in higher subsequent satisfaction and effect scores, then the adjustment instructions corresponding to that range will be updated. The optimized mapping rules and mapping table versions are numbered and stored in the rule base; When generating new personalized comfort solutions for target users, the latest version of rules and mapping tables is applied by default, or operators are allowed to select a specific historical version under certain circumstances; The retrospective analysis optimization cycle is executed automatically at preset intervals, forming a closed loop of continuous iteration and improvement of the core logic of solution generation.

9. The method according to claim 3, characterized in that, The immediate response evaluation model is a rule-based state machine, in which the internal state switches according to the input feature statistics and multiple predefined state transition conditions; The state transition conditions include: consistency of the trend slope sign of continuous time segments, the multiple relationship of variance to the historical window mean, and the matching degree between the combination of feature statistics and the pre-stored typical maladaptive pattern template.

10. A system for generating personalized comfort plans for Thermage treatments, characterized in that, include: One or more processors; A memory having stored one or more programs that, when executed by one or more processors, cause the one or more processors to implement a method for generating a personalized comfort program for Thermage care according to any one of claims 1 to 9.

Images