Trauma care training method and system based on somatosensory interaction and multi-modal perception

Abstract

The application discloses a trauma rescue training method and system based on somatosensory interaction and multi-modal perception, and particularly relates to the field of medical simulation training, and is used for solving the problems that rescue operation stability is difficult to quantify and on-site rescue and transportation risk is disconnected. First, a virtual-real space mapping matrix and a treatment benchmark library are constructed, equipment identity and pressure time sequence data are synchronously collected, pressure dispersion is extracted to generate an operation mechanics stability index; then, intervention behavior compliance is checked, the pressure dispersion is mapped to a wound fastening coefficient, and the virtual wounded blood loss rate is dynamically corrected; finally, the transportation vibration energy spectrum and the fastening coefficient are introduced for mechanics coupling checking, when the vibration intensity breaks through the maintenance threshold, the secondary bleeding logic is triggered, the clinical rescue outcome is judged, and a whole-chain closed-loop training system from accurate rescue to dynamic transportation is constructed.

Description

Technical Field

[0001] This invention relates to the field of medical simulation training technology, specifically to a trauma care training method and system based on somatosensory interaction and multimodal perception. Background Technology

[0002] In rescue scenarios involving major natural disasters or public safety emergencies, emergency responders need to provide rapid and precise treatment to large numbers of injured people under high pressure and within limited time. Trauma care not only requires rescuers to have sound medical decision-making abilities but also stable and standardized practical skills. Especially in critical areas such as hemostasis and bandaging, the force, placement, and stability of the manipulation directly affect the survival rate of the injured. Therefore, building a training system that can highly replicate real physiological responses and environmental disturbances is of great significance for improving the practical capabilities of rescue teams.

[0003] Existing trauma care training methods typically rely on static models or basic electronic feedback devices, which have substantial shortcomings in realistic simulation. First, existing methods often focus on detecting whether an action occurs, rather than conducting refined mechanical analysis of the quality of that action. For example, they cannot effectively distinguish between mechanical discrepancies such as hand tremors or unstable pressure applied by the trainee, leading to situations where actions that complete the steps but achieve poor hemostasis can still pass the assessment. Second, existing training processes are often segmented and isolated; the quality data from the on-site treatment phase is usually not transferred to the subsequent transport phase, resulting in a logical disconnect between the treatment and transport processes. In a real rescue chain, if initial bandaging and fixation are not secure, they are highly susceptible to loosening during subsequent transport, leading to secondary bleeding. Existing technology struggles to simulate this chain of injury development caused by the coupling of environmental vibration stress and operational quality. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a trauma care training method and system based on somatosensory interaction and multimodal perception, thus solving the problems mentioned above.

[0005] To achieve the above objectives, this invention employs the following technical solution: a trauma care training method based on somatosensory interaction and multimodal perception, comprising the following steps: S1. Constructing a spatial mapping matrix between a pressure sensor array on the surface of a physical mannequin and the anatomical coordinates of a virtual injured person; initializing a dynamic physiological feature dataset containing hemodynamic indicators; and configuring a treatment benchmark library containing matching device RFID coding sequences and treatment pressure thresholds according to the injury type; S2. Synchronously collecting device identification IDs obtained by a hand radio frequency reader and pressure time-series data output from the mannequin sensor nodes; extracting pressure data segments within the time period when the device identification ID is identified; extracting the pressure peak characteristics and pressure dispersion characteristics of the pressure data segments; and generating an interventional behavior dataset containing device attribute verification bits and operational mechanical stability indicators; S3. Based on spatial... The mapping matrix confirms the rescue location of the pressure time series data, verifies the matching of the device ID, pressure peak characteristics, and rescue location in the intervention behavior dataset with the treatment benchmark library. If the matching is successful, the pressure dispersion characteristics are mapped to a fastening coefficient that characterizes the firmness of wound treatment using an inverse proportional function relationship. The blood loss rate decay parameter in the dynamic physiological feature dataset is corrected according to the fastening coefficient, and the remaining blood volume value in the dataset is updated. S4. The vibration energy spectrum of the virtual transport vehicle is called to perform a mechanical coupling verification of vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the hemostasis measures are determined to be ineffective and the secondary bleeding logic is activated. The remaining blood volume value in the dynamic physiological feature dataset is deducted according to the secondary bleeding logic until the clinical endpoint is reached and the treatment outcome classification result is generated.

[0006] Furthermore, the specific process of constructing the spatial mapping matrix between the pressure sensing array on the surface of the physical simulated human and the anatomical coordinates of the virtual wounded is as follows: traverse the pressure sensing nodes on the surface of the physical simulated human, obtain the physical topological coordinates and hardware channel codes of each node, and divide the nodes into independent anatomical sensing regions based on the physical topological coordinates; load the three-dimensional mesh model of the virtual wounded, extract the set of mesh vertex indices corresponding to the anatomical structure to define the virtual anatomical region, run the region registration algorithm to establish the mapping relationship between the anatomical sensing region and the virtual anatomical region, generate a bidirectional lookup table linking the hardware channel codes and the virtual mesh vertex indices, and form a spatial mapping matrix.

[0007] Furthermore, the specific process of initializing a dynamic physiological feature dataset containing hemodynamic parameters and configuring a treatment benchmark library containing matching device RFID code sequences and treatment pressure thresholds according to the injury type is as follows: Based on the preset injury type, the corresponding pathophysiological evolution template is retrieved, and a physiological state vector containing the total blood volume benchmark value, coagulation factor activity coefficient, and vascular resistance parameters is instantiated to construct the initial dynamic physiological feature dataset; the medical device attribute database is searched, and a list of standard treatment device RFID codes matching the injury type is filtered. An effective rescue pressure range and standard rescue site identifier are set for each standard treatment device; an association index is established between the list of standard treatment device RFID codes and the effective rescue pressure range and standard rescue site identifier to construct a multidimensional treatment benchmark library.

[0008] Furthermore, the specific process of synchronously collecting the equipment identification ID obtained by the hand RF reader and the pressure time-series data output by the mannequin sensor node, and extracting the pressure data segment within the time period when the equipment identification ID is identified, is as follows: A dual-thread data monitoring mode is initiated. The main thread monitors the equipment identification ID event uploaded by the hand RF reader, while the sub-thread monitors the real-time pressure data stream output by the mannequin sensor node. When the main thread detects the equipment identification ID input, it records the current system time as the action start timestamp and activates the time capture window. During the time capture window, active node data with pressure values ​​exceeding the effective trigger threshold are filtered from the real-time pressure data stream of the sub-thread. The sequence of pressure sampling points for all active nodes from the action start timestamp to the window closing time is locked and extracted, forming a pressure data segment synchronized with the equipment identification ID time.

[0009] Furthermore, the specific process of extracting the pressure peak value and pressure dispersion features of the pressure data segment to generate an interventional behavior dataset containing device attribute verification bits and operational mechanical stability indicators is as follows: The extracted pressure data segment is traversed in the time domain, and the maximum pressure amplitude is selected as the pressure peak value feature to characterize the extreme force of the rescue operation; the statistical variance of the pressure sampling point sequence relative to the pressure mean in the pressure data segment is calculated as the pressure dispersion feature to quantify the degree of hand tremor or the stability of continuous compression during the rescue process; the hardware channel codes of active nodes in the pressure data segment are extracted, the corresponding rescue position coordinates are parsed using the spatial mapping matrix, the device identity ID is mapped to the device attribute verification bit, and the pressure peak value feature, pressure dispersion feature, and rescue position coordinates are combined to form the operational mechanical stability index, which is then encapsulated to generate the interventional behavior dataset.

[0010] Furthermore, based on the spatial mapping matrix, the rescue location of the pressure time series data is confirmed, and the matching of the device ID, pressure peak characteristics, and rescue location in the intervention behavior dataset with the treatment benchmark library is verified. If the matching is successful, the pressure dispersion characteristics are mapped to a fastening coefficient that characterizes the firmness of wound treatment using an inverse proportional function relationship. The specific process is as follows: extract the hardware channel code of the active node in the intervention behavior dataset, retrieve the corresponding virtual anatomical region index in the spatial mapping matrix, and determine the virtual coordinates of the rescue location; construct a logic AND gate to compare the device ID with the standard device code in the treatment benchmark library, the pressure peak characteristics with the effective treatment pressure threshold, and the virtual coordinates of the rescue location with the standard rescue site identifier. A matching pass signal is output only when all three sets of comparison results are true; in response to the matching pass signal, the pressure dispersion characteristics are normalized, and the normalization result is subjected to an inverse mapping operation to generate a fastening coefficient with a monotonically increasing value.

[0011] Furthermore, the specific process of updating the remaining blood volume value in the dynamic physiological feature dataset by correcting the blood loss rate decay parameter in the dataset based on the clamping coefficient is as follows: read the current blood loss rate variable in the dynamic physiological feature dataset, substitute the clamping coefficient as a gain factor into the exponential decay algorithm, calculate the blood loss inhibition ratio at the current time step, apply the blood loss inhibition ratio to perform a downward correction on the current blood loss rate variable, and generate the real-time blood loss rate; perform an integral operation on the real-time blood loss rate within the current sampling time interval and simultaneously subtract the blood loss obtained by the integral from the remaining blood volume value in the dynamic physiological feature dataset to complete the hemodynamic state update for a single cycle.

[0012] Furthermore, the vibration energy spectrum of the virtual transport vehicle is invoked to perform a mechanical coupling verification of vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the specific process of determining that the hemostasis measure has failed and activating the secondary bleeding logic is as follows: The vibration energy spectrum of the virtual transport vehicle is analyzed, the vibration frequency and amplitude parameters are extracted, and the equivalent physical shear force acting on the wound is calculated using a friction mechanics model; the fastening coefficient is converted into a mechanical resistance limit value and defined as a critical maintenance threshold. When the equivalent physical shear force is detected to be greater than the critical maintenance threshold and the duration exceeds the preset judgment time window, a hemostasis failure interruption signal is generated; in response to the hemostasis failure interruption signal, the blood loss rate variable in the dynamic physiological feature dataset is switched from the current decay state to an uncontrolled high flow rate state, and the uncontrolled high flow rate state is locked to activate the secondary bleeding logic.

[0013] Furthermore, the specific process of subtracting the remaining blood volume value from the dynamic physiological feature dataset based on the secondary bleeding logic until the clinical endpoint is reached to generate the treatment outcome classification result is as follows: Based on the activated secondary bleeding logic, the remaining blood volume value is iteratively subtracted in subsequent time steps with the uncontrolled high flow rate state as the benchmark; the relationship between the remaining blood volume value and the preset shock threshold, organ failure threshold, and clinical termination threshold is monitored in real time; when the remaining blood volume value is less than any threshold or the training time ends, the iterative subtraction operation is terminated and the final remaining blood volume value is locked. The final remaining blood volume value is compared with the clinical prognosis grading table, and the corresponding treatment outcome classification result is output.

[0014] The trauma care training system based on somatosensory interaction and multimodal perception includes the following modules: an initial configuration module, used to construct a spatial mapping matrix between the pressure sensor array on the surface of the physical mannequin and the anatomical coordinates of the virtual patient, initialize a dynamic physiological feature dataset containing hemodynamic indicators, and configure a treatment benchmark library containing matching device RFID coding sequences and treatment pressure thresholds according to the injury type; a data acquisition module, used to simultaneously acquire device identification IDs obtained by the hand RFID reader and pressure time-series data output by the mannequin sensor nodes, extract pressure data segments within the time period when the device identification ID is identified, extract pressure peak features and pressure dispersion features of the pressure data segments, and generate an interventional behavior dataset containing device attribute verification bits and operational mechanical stability indicators; and a treatment verification module, used to confirm treatment based on the spatial mapping matrix. The rescue location in the pressure time series data is verified to match the device ID, pressure peak characteristics, and rescue location in the intervention behavior dataset with the treatment benchmark library. If the match is successful, the pressure dispersion characteristics are mapped to a fastening coefficient that characterizes the firmness of wound treatment using an inverse proportional function. The blood loss rate decay parameter in the dynamic physiological feature dataset is corrected based on the fastening coefficient, and the remaining blood volume value in the dataset is updated. The transport coupling module is used to call the vibration energy spectrum of the virtual transport vehicle and perform mechanical coupling verification between vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the hemostasis measures are deemed to have failed and the secondary bleeding logic is activated. The remaining blood volume value in the dynamic physiological feature dataset is deducted based on the secondary bleeding logic until the clinical endpoint is reached, generating a treatment outcome classification result.

[0015] The present invention has the following beneficial effects:

[0016] (1) A trauma care training method based on somatosensory interaction and multimodal perception, by constructing a spatial mapping matrix between physical and virtual entities and simultaneously collecting radio frequency identification and pressure time-series data, achieves multi-dimensional and accurate perception of rescue operations. This method can not only verify the correctness of equipment use, but also quantitatively analyze the mechanical stability during the rescue process by extracting the peak and dispersion characteristics of pressure data. This mechanism effectively overcomes the shortcomings of traditional training that only focuses on results and not process, and can accurately identify operational hazards caused by hand tremors or uneven force exertion in trainees, ensuring the objectivity and rigor of intervention behavior assessment.

[0017] (2) A trauma care training system based on somatosensory interaction and multimodal perception establishes a dynamic confrontation mechanism between environmental stress and treatment quality. It transforms the mechanical stability of the operation into a tightness coefficient that characterizes the firmness of wound treatment and performs mechanical coupling verification with the vibration energy spectrum of the virtual transport vehicle. By simulating the impact of vibration shear force on wound fixation during transportation, this method can realistically predict secondary bleeding events induced by unreliable initial treatment. It realizes a closed-loop prediction of the entire chain from on-site rescue to subsequent transport, prompting trainees to establish a holistic treatment concept of high-quality treatment to cope with subsequent transportation risks.

[0018] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0019] Figure 1 This is a flowchart of the trauma care training method based on somatosensory interaction and multimodal perception according to the present invention.

[0020] Figure 2 This is a flowchart of the trauma care training system based on somatosensory interaction and multimodal perception according to the present invention.

[0021] Figure 3 This is a timing diagram of the pressing operation pressure. Detailed Implementation

[0022] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0023] The embodiments of this application solve the problems of single operational evaluation dimensions and logical separation between treatment and transportation links in the existing training system by using a trauma care training method and system based on somatosensory interaction and multimodal perception.

[0024] The overall approach of the scheme in this application embodiment is as follows: First, spatial mapping technology is used to accurately project the interactive data of the physical simulated human surface onto the anatomical structure of the virtual patient, and physiological characteristic data is initialized; during the training process, the system simultaneously captures the identity information of the equipment and the pressure waveform during operation, and generates an intervention behavior dataset containing operational stability by analyzing the dispersion of the pressure data; then, the tightness coefficient of the wound is calculated in reverse based on the dispersion of the operation, and the patient's blood loss rate is corrected in real time using this coefficient; finally, the vibration energy spectrum of the virtual transport environment is introduced, and the vibration shear force generated by the environment is compared with the tightness coefficient of the wound for counter-verification. If the vibration intensity exceeds the limit of the treatment tightness capability, the hemostasis is determined to be ineffective and the injury deterioration logic is triggered, thereby completing the full-process dynamic rescue training including environmental interference factors.

[0025] Please see Figure 1 This invention provides a technical solution: a trauma care training method based on somatosensory interaction and multimodal perception, comprising the following steps: S1. Constructing a spatial mapping matrix between a pressure sensing array on the surface of a physical mannequin and the anatomical coordinates of a virtual injured person; initializing a dynamic physiological feature dataset containing hemodynamic indicators; and configuring a treatment benchmark library containing matching device RFID coding sequences and treatment pressure thresholds according to the injury type; S2. Synchronously collecting device identification IDs obtained by a hand radio frequency reader and pressure time-series data output by the mannequin sensor nodes; extracting pressure data segments within the time period when the device identification ID is identified; extracting the pressure peak characteristics and pressure dispersion characteristics of the pressure data segments; and generating an interventional behavior dataset containing device attribute verification bits and operational mechanical stability indicators; S3. Based on the spatial mapping matrix... Confirm the rescue location in the pressure time series data, verify the matching of the device ID, pressure peak characteristics, and rescue location in the intervention behavior dataset with the treatment benchmark library. If the matching is successful, use the inverse proportional function relationship to map the pressure dispersion characteristics into a fastening coefficient that characterizes the firmness of wound treatment. Correct the blood loss rate decay parameter in the dynamic physiological feature dataset according to the fastening coefficient, and update the remaining blood volume value in the dataset. S4. Call the vibration energy spectrum of the virtual transport vehicle to perform a mechanical coupling verification of vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, determine that the hemostasis measures have failed and activate the secondary bleeding logic. Deduct the remaining blood volume value in the dynamic physiological feature dataset according to the secondary bleeding logic until the clinical endpoint is reached to generate the treatment outcome classification result.

[0026] In this implementation plan, step S1 is mainly used to construct the basic environment and rule system for interaction between virtual simulation and physical entities. In this process, the spatial mapping matrix is ​​a mathematical transformation model that maps the physical contact coordinates of the simulated human body to the anatomical structure of the injured person in the virtual 3D scene. This ensures that the trainee's pressing action on the thigh of the physical model can be accurately identified by the system as treatment of the virtual injured person's thigh. The dynamic physiological feature dataset is a set of variables storing the real-time vital signs of the virtual injured person, including values ​​such as blood volume and blood pressure that change over time. The treatment benchmark library serves as a judgment standard, pre-setting the correct equipment identification code and effective physical pressing threshold that must be matched for specific injuries (such as massive arterial bleeding). This step, by initializing these mapping relationships and data standards, lays the data foundation for subsequent accurate capture of the trainee's somatosensory operations and medical logic judgment. Step S2 is mainly used to capture the trainee's specific operational behaviors in real time and extract key mechanical features. This step uses time synchronization technology to align the radio frequency signal of the hand grasping the equipment with the pressure signal acting on the simulated human body on the time axis, thereby extracting a complete segment of operational data. The pressure dispersion feature is a pressure fluctuation value calculated based on statistical principles. It is specifically used to quantitatively assess the stability and uniformity of the trainee's hand force application during rescue. The lower the dispersion, the more stable the operation. By extracting the pressure peak and pressure dispersion, the system generates an intervention behavior dataset. This dataset not only records what instruments the trainee used but also quantitatively records the mechanical quality of their operation, solving the technical problem of objectively evaluating the stability of operation in traditional training. Step S3 is mainly used to perform multidimensional verification of the effectiveness of the rescue operation and drive the physiological evolution of the injury. The system first confirms whether the trainee's rescue position is correct based on the spatial mapping matrix and makes a compliance judgment based on the equipment identity and force. On this basis, the tightness coefficient is a dimensionless physical quantity derived from an inverse proportional function relationship. It is negatively correlated with the pressure dispersion. That is, the more stable the trainee's operation (the smaller the dispersion), the higher the calculated wound treatment tightness coefficient, which means that the bandage or tourniquet is fixed more securely. This step utilizes this coefficient to dynamically correct the blood loss rate of the virtual patient, achieving a direct correlation between operational quality and physiological feedback. That is, the more precise and stable the technique, the better the hemostasis effect, thus establishing a realistic medical causal feedback mechanism during training. Step S4 is mainly used to simulate the risk of injury deterioration due to environmental interference during patient transport. The vibration energy spectrum of the virtual transport vehicle refers to the vibration frequency and energy distribution data generated by the simulated ambulance or helicopter during transport; the system converts this into physical shear force acting on the wound. Mechanical coupling verification refers to the process of comparing this external destructive force (shear force) with the wound's own resistance (tightening coefficient).If the initial handling is inadequate, resulting in a low securing coefficient and inability to withstand transport vibrations, the system will trigger a secondary bleeding logic, that is, restart the blood loss calculation program and deduct the remaining blood volume. This step, through risk simulation across the entire chain, forces trainees to pay attention to the stability of on-site handling, solving the problem of the disconnect between on-site treatment and subsequent transport in existing technologies.

[0027] Specifically, the process of constructing the spatial mapping matrix between the pressure sensing array on the surface of the physical simulated human and the anatomical coordinates of the virtual wounded is as follows: traverse the pressure sensing nodes on the surface of the physical simulated human, obtain the physical topological coordinates and hardware channel codes of each node, and divide the nodes into independent anatomical sensing regions based on the physical topological coordinates; load the three-dimensional mesh model of the virtual wounded, extract the set of mesh vertex indices corresponding to the anatomical structure to define the virtual anatomical region, run the region registration algorithm to establish the mapping relationship between the anatomical sensing region and the virtual anatomical region, generate a bidirectional lookup table linking the hardware channel codes and the virtual mesh vertex indices, and form a spatial mapping matrix.

[0028] In this implementation, the system scans every pressure-sensing node on the surface of the simulated human body using hardware scanning, reading its inherent physical topological coordinates (i.e., the actual spatial location of the sensor on the simulated human body surface, such as 5 cm from the knee on the outer side of the left thigh) and a unique hardware channel code (ID used for circuit addressing). Based on human anatomical features, the system logically divides these nodes into independent anatomical perception regions such as the head, chest, abdomen, and limbs. This step reduces the dimensionality of subsequent registration calculations and avoids incorrectly mapping hand operations to the legs. Next, a 3D mesh model of the virtual patient is loaded, consisting of thousands of triangular faces and vertices. The system extracts the set of virtual mesh vertex indices corresponding to the aforementioned physical anatomical regions, defining the virtual anatomical regions. To achieve high-precision mapping between the two, a region registration algorithm is run. This algorithm does not use simple point-to-point hard connections but instead employs a spatial transformation calculation based on weighted least squares to ensure that when the trainee presses the gaps in the physical sensors, the corresponding skin area on the virtual model also produces a smooth deformation response. The transformation model for converting physical coordinates to virtual coordinates in the spatial mapping matrix is ​​calculated using the following formula: In the formula: The calculated target vertex coordinate vectors on the virtual wounded soldier's 3D mesh model are used to drive the deformation of the virtual skin; Physical topological coordinate vector of pressure sensing node on human surface in physical simulation; The region registration rotation matrix is ​​used to correct the angular deviation in pose between the physical model and the virtual model; : Region registration translation vector, used to align the spatial geometric centers of the physical region and the virtual region; Deformation weight coefficient, used to adjust the degree of influence of physical pressure on the displacement of virtual mesh vertices. This coefficient is set according to the elastic modulus of the virtual skin, and the value range is usually between 0 and 1. The normal vector of the virtual mesh vertex is used to constrain the deformation direction to always be perpendicular to the virtual skin surface. The deformation weight coefficient is determined by measuring the ratio of the deformation of the physical silicone skin under unit pressure to the set deformation of the virtual skin model through pre-conducted material mechanics calibration experiments. This ratio is used as the basic weight and then combined with Gaussian attenuation processing based on the Euclidean distance from the mesh vertex to the center of force. Through the above calculations, the system generates a bidirectional lookup table linking the hardware channel encoding and the virtual mesh vertex index, i.e., a spatial mapping matrix. This matrix enables the system to locate a specific vertex on the virtual model within milliseconds when it receives a pressure signal from a specific hardware encoding, thereby achieving precise interaction through visual-touch fusion.

[0029] Specifically, the process of initializing a dynamic physiological feature dataset containing hemodynamic parameters and configuring a treatment benchmark library containing matching equipment RFID code sequences and treatment pressure thresholds according to the injury type is as follows: Based on the preset injury type, the corresponding pathophysiological evolution template is retrieved, and a physiological state vector containing the total blood volume benchmark value, coagulation factor activity coefficient, and vascular resistance parameters is instantiated to construct the initial dynamic physiological feature dataset; the medical device attribute database is searched, and a list of standard treatment equipment RFID codes matching the injury type is selected, and an effective rescue pressure range and standard rescue site identifier are set for each standard treatment equipment; an association index is established between the list of standard treatment equipment RFID codes and the effective rescue pressure range and standard rescue site identifier to construct a multidimensional treatment benchmark library.

[0030] In this implementation scheme, the system retrieves the corresponding pathophysiological evolution template based on preset injury types (such as penetrating femoral artery injury, tension pneumothorax, etc.). This template is not static data, but rather a set of initial values ​​for initialized differential equations. To make the training more realistic and differentiated, the system needs to instantiate physiological state vectors. Especially when determining the initial total blood volume baseline value for the injured person, a uniform fixed value cannot be used; instead, it needs to be dynamically calculated based on the set virtual injured person's vital signs (height, weight, body surface area) to ensure that the rate of shock evolution conforms to individual differences. The initial total blood volume baseline value is calculated using the following formula: In the formula: The initial total blood volume of the virtual wounded soldier, in milliliters; The set virtual height value for the wounded soldier; The set virtual weight value of the wounded soldier; The height correlation coefficient is derived from linear regression analysis based on statistical data of human physiology. The weight correlation coefficient is used to adjust the weight of the effect of weight on blood volume. The basic constant term is used to calibrate the baseline blood volume of wounded soldiers of different genders; Body surface area normalization index is used to eliminate the dimensional effects caused by individual body size differences. The initial traumatic blood loss percentage is determined by the severity level of the preset injury type (e.g., severe blast injury can be set to 0.15, meaning initial blood loss is 15%). Simultaneously, the system searches the medical device attribute database to filter out a list of standard treatment devices with RFID codes matching the injury type. When setting the effective rescue pressure range for each standard treatment device, this threshold is not fixed but is related to the patient's current systolic blood pressure. For example, when using a tourniquet, the applied pressure must be higher than the current arterial systolic blood pressure to effectively stop blood flow. The lower limit threshold of the effective rescue pressure is determined by the following formula: In the formula: The effective lower limit threshold of rescue pressure for standard treatment equipment: only when the pressure value collected by the sensor is higher than this value is it considered to be effective hemostasis; The virtual patient's current instantaneous systolic blood pressure value; The blocking coefficient depends on the width of the hemostatic instrument; the narrower the instrument, the larger the coefficient (due to uneven pressure distribution requiring greater pressure). Local tissue resistance parameters characterize the effect of the thickness and stiffness of muscles and soft tissues at the injured site on the resistance to pressure. Organizational resistance weighting factor; A safety redundancy constant is used to ensure the reliability of hemostasis during blood pressure fluctuations. Through the above steps, the system establishes an association index between the RFID code list of standard treatment equipment and dynamically calculated effective rescue pressure ranges and standard rescue site identifiers, constructing a multi-dimensional treatment benchmark library. This allows the training system to go beyond simply mechanically comparing values; it enables dynamic adjustments to assessment standards based on the patient's physiological characteristics (height and weight) and real-time status (blood pressure), greatly enhancing the scientific rigor and medical logic of the training.

[0031] Specifically, the process of synchronously collecting the device identification ID obtained by the hand-held RF reader and the pressure time-series data output by the mannequin sensor node, and extracting the pressure data segment within the time period when the device identification ID is identified, is as follows: A dual-thread data monitoring mode is initiated. The main thread monitors the device identification ID event uploaded by the hand-held RF reader, while the sub-thread monitors the real-time pressure data stream output by the mannequin sensor node. When the main thread detects the device identification ID input, it records the current system time as the action start timestamp and activates the time capture window. During the time capture window, active node data with pressure values ​​exceeding the effective trigger threshold are filtered from the real-time pressure data stream of the sub-thread. The sequence of pressure sampling points for all active nodes from the action start timestamp to the window closing time is locked and extracted, forming a pressure data segment synchronized with the device identification ID time.

[0032] In this implementation scheme, the system adopts a dual-thread architecture. The main thread focuses on capturing when and what device was used, i.e., listening to RF ID events; the sub-thread focuses on recording "what kind of mechanical action was generated," i.e., polling the pressure flow. When the main thread recognizes the device ID, it not only records a timestamp but, more importantly, activates a dynamic time capture window. The lifecycle of this window corresponds to a complete medical procedure. Within the time capture window, the system does not receive all data but needs to filter out background noise caused by non-contact or accidental contact. To this end, the system introduces valid action determination logic, recognizing a valid compression only when the sensor reading exceeds the valid trigger threshold. To adapt to sensor baseline drift in different environments, this valid trigger threshold is not a fixed value but is dynamically calculated and determined using the following formula: In the formula: The effective trigger threshold for the effective action is dynamically calculated. Only when the real-time reading of the pressure sensor is higher than this value will the generated data be marked as active node data and recorded by the system. The baseline voltage or pressure reading of the sensor under no-load conditions is usually obtained through self-test calibration during system startup. Baseline drift correction factor, used to compensate for sensor zero-point drift caused by temperature changes or aging of the simulated human material; The standard deviation of the current environmental noise is calculated from the environmental background data collected by the system within a short time window, and is used to characterize the intensity of environmental vibration or electromagnetic interference. The noise suppression factor is typically set to 3 to 5 to ensure that the threshold covers more than 99% of the random noise range. The contact sensitivity compensation constant is used to set the minimum physical contact force required to trigger a system response, preventing slight friction from clothing from being misinterpreted as operational data. Through the above calculations, the system can generate pure pressure data segments that strictly correspond to the device ID in complex electromagnetic and vibration environments, providing a high-quality data foundation for subsequent feature analysis. Figure 3 As shown in the figure, the pressure-time curve of a complete CPR operation collected by the system is displayed. The horizontal axis represents time (seconds), and the vertical axis represents pressure value (mmHg). The dashed line represents the effective trigger threshold set by the system; only data segments exceeding this threshold are considered valid operations and are truncated. The fluctuation amplitude at the peak of the curve intuitively reflects the hand tremor of the rescuer during the operation. The larger the fluctuation amplitude, the higher the calculated pressure dispersion, indicating poorer operation stability.

[0033] Specifically, the process of extracting the pressure peak value and pressure dispersion features of the pressure data segment to generate an interventional behavior dataset containing device attribute verification bits and operational mechanical stability indicators is as follows: The extracted pressure data segment is traversed in the time domain, and the maximum pressure amplitude is selected as the pressure peak value feature to characterize the extreme force of the rescue operation; the statistical variance of the pressure sampling point sequence relative to the pressure mean in the pressure data segment is calculated as the pressure dispersion feature to quantify the degree of hand tremor or the stability of continuous compression during the rescue process; the hardware channel codes of active nodes in the pressure data segment are extracted, the corresponding rescue position coordinates are parsed using the spatial mapping matrix, the device identity ID is mapped to the device attribute verification bit, and the pressure peak value feature, pressure dispersion feature, and rescue position coordinates are combined to form the operational mechanical stability index, which is then encapsulated to generate the interventional behavior dataset.

[0034] In this implementation scheme, the system performs a time-domain traversal of the captured pressure data segments. For the pressure peak characteristic, the system employs an extreme value search algorithm to directly filter out the maximum pressure value throughout the entire operation. This indicator directly corresponds to clinically relevant hard indicators such as whether the tourniquet is tightened and whether the compression depth is adequate. Secondly, for the pressure dispersion characteristic, this is crucial for assessing the trainee's proficiency. In trauma care, consistent and stable force is key to successful hemostasis; hand tremors or fluctuating pressure can lead to wound hemostasis failure. Therefore, the system quantifies this indicator by calculating the statistical dispersion of the pressure data. The pressure dispersion characteristic in the operational mechanical stability index is calculated using the following formula: In the formula: The calculated pressure dispersion characteristic value indicates that the smaller the value, the smoother and more stable the force application process; the larger the value, the more severe the operator's hand tremors or the extremely uneven force application. : The total number of sampling points in the extracted effective pressure data segment; The instantaneous pressure measurement value at the kth sampling time; The arithmetic mean pressure value for the entire pressure data segment; The time decay weighting factor is used to assign different weights to data in different time periods during calculation. The method for determining the weights is as follows: A Hanning window or Gaussian window function is used to generate the weights, making the weights in the middle of the data segment approach 1, while the weights at the edges smoothly decay to 0. Finally, the system calls the spatial mapping matrix generated in the previous steps and parses the rescue location coordinates (e.g., left lower limb-femoral artery compression point) based on the hardware encoding of the active nodes. The system encapsulates the equipment identification ID (representing what was used), pressure peak characteristics (representing the force applied), pressure dispersion characteristics (representing hand stability), and rescue location coordinates (representing correctness of the location) according to a predefined protocol. This process is equivalent to generating an electronic medical record with a detailed biomechanical examination report for each medical operation, i.e., an interventional behavior dataset, for subsequent algorithms to perform multi-dimensional medical logic verification.

[0035] Specifically, the rescue location of the pressure time series data is confirmed based on the spatial mapping matrix. The matching of the device ID, pressure peak feature, and rescue location in the intervention behavior dataset with the treatment benchmark library is verified. If the matching is successful, the pressure dispersion feature is mapped to a fastening coefficient that characterizes the firmness of the wound treatment using an inverse proportional function relationship. The specific process is as follows: extract the hardware channel code of the active node in the intervention behavior dataset, retrieve the corresponding virtual anatomical region index in the spatial mapping matrix, and determine the virtual coordinates of the rescue location; construct a logic AND gate to compare the device ID with the standard device code in the treatment benchmark library, the pressure peak feature with the effective treatment pressure threshold, and the virtual coordinates of the rescue location with the standard rescue site identifier. A matching pass signal is output only when all three sets of comparison results are true; in response to the matching pass signal, the pressure dispersion feature is normalized, and the normalization result is subjected to an inverse mapping operation to generate a fastening coefficient with a monotonically increasing value.

[0036] In this implementation scheme, the system performs position resolution and logic verification. Using the spatial mapping matrix generated in the preceding steps, the system converts the hardware channel encoding returned by the sensors into virtual anatomical coordinates. Subsequently, the system constructs a logic AND gate, a multi-condition synchronous judgment mechanism. The system simultaneously checks three indicators: whether the device ID is the specific device required for the current injury (e.g., for arterial rupture, the tourniquet ID must be identified), whether the pressure peak characteristic reaches the minimum pressure for effective blood flow occlusion (e.g., above systolic blood pressure), and whether the virtual coordinates of the rescue position are located at the effective hemostasis point proximal to the heart. Only when these three indicators are completely consistent does the system consider the operation valid and output a matching pass signal; otherwise, it is considered an invalid operation and does not trigger subsequent physiological responses. Secondly, after successful verification, the system enters the core mechanical evaluation stage. Simply achieving the required pressure does not guarantee stable hemostasis; even slight hand tremors or uneven force can cause the tourniquet to loosen. Therefore, the system introduces a tightening coefficient calculation to convert pressure dispersion (representing instability) into a tightening coefficient (representing stability). This transformation process is not a simple linear correspondence, but rather employs a nonlinear inverse proportional mapping to simulate the nonlinear viscoelastic response of biological soft tissue under pressure. The clamping coefficient is calculated using the following formula: In the formula: The calculated fastening coefficient is usually normalized to a range of 0 to 1. The closer the value is to 1, the more secure the wound treatment is and the stronger the resistance to interference. The real-time pressure dispersion feature value extracted in the previous steps reflects the fluctuation of the current operation; The ideal operational dispersion lower limit set by the system is the variance value under the absolute steady state at the robotic arm level. The maximum allowable dispersion of the system; exceeding this value will be considered an invalid operation even if the average pressure meets the target. Sensitivity adjustment factor: used to adjust the response rate of the fastening coefficient to changes in dispersion. The larger the value, the more sensitive the system is to jitter and the more stringent the evaluation criteria. The nonlinear mapping index is used to control the steepness of the normalization curve to conform to the different fit characteristics of different parts (such as areas with abundant muscle and areas with prominent bones).

[0037] Specifically, the process of updating the remaining blood volume value in the dynamic physiological feature dataset by correcting the blood loss rate decay parameter in the dynamic physiological feature dataset based on the clamping coefficient is as follows: read the current blood loss rate variable in the dynamic physiological feature dataset, substitute the clamping coefficient as a gain factor into the exponential decay algorithm, calculate the blood loss inhibition ratio at the current time step, apply the blood loss inhibition ratio to perform a downward correction on the current blood loss rate variable, and generate the real-time blood loss rate; perform an integral operation on the real-time blood loss rate within the current sampling time interval and simultaneously subtract the blood loss obtained by the integral from the remaining blood volume value in the dynamic physiological feature dataset to complete the hemodynamic state update for a single cycle.

[0038] In this implementation, the system reads the current blood loss rate variable. In the uninterrupted state, this rate may be in a constant spray state or decrease slowly as blood pressure drops. Once an effective clamping coefficient is generated, it acts as a negative feedback gain in the physiological system. Next, the system performs a downward correction calculation for the blood loss rate. Hemostasis is not instantaneous but a process of gradual slowing of blood flow as the vessel closes under pressure. The system substitutes the clamping coefficient into an exponential decay algorithm to calculate the acceleration of blood flow closure. The real-time blood loss rate is updated using the following formula: In the formula: Real-time blood loss rate after the current time step update, usually expressed in milliliters per second; The rate of blood loss at the previous time step; The maximum hemostatic efficiency coefficient is determined by the type of injury. For example, the hemostasis of a ruptured major artery is difficult, so the coefficient is small, which means that even if the tightness coefficient is high, the blood flow velocity will decrease more slowly. The fastening coefficient calculated in the preceding steps is used as the main suppression variable; The sampling time interval in system simulation is typically in the millisecond range. The physiological response time constant characterizes the delayed response of vascular smooth muscle to external pressure. Finally, the system performs an integral deduction of blood volume. Rate is only an instantaneous state; the cumulative blood loss has a substantial impact on vital signs. The system performs a discrete integral on the real-time blood loss rate and deducts it from the total blood volume. The remaining blood volume value is iteratively calculated using the following formula: In the formula: The updated remaining blood volume value is the core indicator for determining the level of shock. : Remaining blood volume at the previous time step; The compensatory correction factor is used to fine-tune the model to match this compensatory mechanism in the early stages of blood loss, preventing a purely linear decrease in blood volume. The system achieves closed-loop control from physical operation quality to virtual physiological response, allowing trainees to intuitively see that only the tighter and more stable the bandage (higher tightness coefficient) will the rate of blood loss on the monitor decrease faster, and the patient's life countdown will slow down.

[0039] Specifically, the vibration energy spectrum of the virtual transport vehicle is invoked to perform a mechanical coupling verification of vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the hemostasis measure is determined to have failed and the secondary bleeding logic is activated. The specific process is as follows: The vibration energy spectrum of the virtual transport vehicle is analyzed, and the vibration frequency and amplitude parameters are extracted. The equivalent physical shear force acting on the wound is calculated using a friction mechanics model. The fastening coefficient is converted into a mechanical resistance limit value and defined as a critical maintenance threshold. When the equivalent physical shear force is detected to be greater than the critical maintenance threshold and the duration exceeds the preset judgment time window, a hemostasis failure interruption signal is generated. In response to the hemostasis failure interruption signal, the blood loss rate variable in the dynamic physiological feature dataset is switched from the current decay state to an uncontrolled high flow rate state, and the uncontrolled high flow rate state is locked to activate the secondary bleeding logic.

[0040] In this implementation scheme, the system analyzes the vibration energy spectrum of a virtual transport vehicle (such as an ambulance or helicopter). The vibration energy spectrum contains the vibration frequency and amplitude envelope information of the vehicle under specific road conditions or flight attitudes. The system uses a physics engine to convert this environmental vibration data into an equivalent physical shear force acting on the wound site of the injured person. This shear force represents the destructive force exerted by the external environment attempting to tear or shake the wound dressing. The equivalent physical shear force is calculated using the following formula: In the formula: The calculated equivalent physical shear force acting on the wound site represents the external physical load that disrupts the stability of the bandage. Vibration transmission efficiency coefficient: Characterizes the attenuation or amplification ratio of vibration energy transmitted from the vehicle chassis to the wounded stretcher and finally to the wound. The equivalent mass of the local tissue at the injured site is used to calculate the inertial force. : The characteristic value of vibration amplitude at the current moment extracted from the vibration energy spectrum; The current moment's principal vibration frequency extracted from the vibration energy spectrum; The angle between the shock wave vector and the wound surface normal vector is used to extract the shear component along the tangential direction of the wound surface, which is the main cause of bandage loosening. Secondly, the system establishes a resistance standard. Based on the tightening coefficient generated in the previous steps, the system converts it into a critical maintenance threshold. The tightening coefficient represents the internal defense capability, while the critical maintenance threshold is the mechanical limit that this defense can withstand. The critical maintenance threshold is determined by the following formula: In the formula: The calculated critical maintenance threshold is the maximum external shear force that the bandage can withstand under the current bandaging condition. The coefficient of static friction between the bandage material and the skin is determined by the properties of the equipment. The effective average pressure value applied during rescue; The effective contact area between the device and the skin; The fastening coefficient generated by the preceding steps serves as a core variable that determines the threshold level. A nonlinear hardening index is used to simulate the physical characteristic that the shear resistance of a bandage increases nonlinearly as it tightens. Finally, the system performs real-time comparison calculations. When the external equivalent physical shear force is detected to be continuously greater than the internal critical maintenance threshold, and the duration exceeds a preset judgment window (e.g., continuous turbulence for 3 seconds), the system determines that the physical defense has collapsed and generates a hemostasis failure interruption signal. In response to this signal, the system immediately removes the attenuation correction to the blood loss rate in the physiological model, forcibly resetting the blood loss rate variable to an uncontrolled high-flow-rate state (typically 1.2 to 1.5 times the initial bleeding rate, simulating wound tearing), thereby activating the secondary bleeding logic. This process achieves deep coupling between the environment and the operation from a mechanistic perspective, forcing trainees to perform high-quality fastening on-site to ensure escort safety.

[0041] Specifically, the process of subtracting the remaining blood volume value from the dynamic physiological feature dataset based on the secondary hemorrhage logic until the clinical endpoint is reached to generate the treatment outcome classification result is as follows: Based on the activated secondary hemorrhage logic, the remaining blood volume value is iteratively subtracted in subsequent time steps with the uncontrolled high flow rate state as the benchmark; the relationship between the remaining blood volume value and the preset shock threshold, organ failure threshold, and clinical termination threshold is monitored in real time; when the remaining blood volume value is less than any threshold or the training time ends, the iterative subtraction operation is terminated and the final remaining blood volume value is locked; the final remaining blood volume value is compared with the clinical prognosis grading table, and the corresponding treatment outcome classification result is output.

[0042] In this implementation scheme, firstly, based on the activated secondary bleeding logic, the system enters the high-risk blood loss calculation stage. At this point, the blood loss rate is no longer suppressed by the clamping coefficient, but rather based on an uncontrolled high-flow-rate state, and the remaining blood volume value is iteratively reduced in subsequent time steps. To simulate disseminated intravascular coagulation (DIC) in a state of severe blood loss, the iterative reduction process includes a self-accelerating term. The remaining blood volume value is iteratively updated using the following formula: In the formula: : The remaining blood volume value after calculation at the current nth time step; : Remaining blood volume value at the previous time step; : Baseline blood loss rate under uncontrolled high flow conditions after reset; Blood loss acceleration factor, used to simulate the pathological phenomenon that bleeding from a wound intensifies over time due to the depletion of clotting factors; The duration of the second bleeding logic after activation; The system simulation time step interval. Simultaneously, the system monitors the relationship between the remaining blood volume value and various clinically critical thresholds in real time. These thresholds include a preset shock threshold (indicating the onset of shock), an organ failure threshold (indicating irreversible damage), and a clinical termination threshold (indicating cessation of vital functions). When the remaining blood volume value falls below any threshold or the prescribed training time ends, the system terminates the calculation and locks the final remaining blood volume value. Finally, the system consults a clinical prognostic grading table, mapping the final remaining blood volume value to a specific treatment outcome classification. For example, if the final blood volume is higher than the shock threshold, it is classified as physiological homeostasis (successful treatment); if it is between the shock and organ failure thresholds, it is classified as compensated shock (requiring further treatment); if it is lower than the clinical termination threshold, it is classified as death (treatment failure). This result is not only a score but also a comprehensive medical assessment of the trainee's overall treatment decision-making and operational quality.

[0043] Please see Figure 2The trauma care training system based on somatosensory interaction and multimodal perception includes the following modules: an initial configuration module, used to construct a spatial mapping matrix between the pressure sensor array on the surface of the physical mannequin and the anatomical coordinates of the virtual patient, initialize a dynamic physiological feature dataset containing hemodynamic indicators, and configure a treatment benchmark library containing matching device RFID coding sequences and treatment pressure thresholds according to the injury type; a data acquisition module, used to simultaneously acquire device identification IDs obtained by the hand radio frequency reader and pressure time-series data output by the mannequin sensor nodes, extract pressure data segments within the time period when the device identification ID is identified, extract pressure peak features and pressure dispersion features of the pressure data segments, and generate an interventional behavior dataset containing device attribute verification bits and operational mechanical stability indicators; and a treatment verification module, used to confirm treatment based on the spatial mapping matrix. The rescue location in the pressure time series data is verified to match the device ID, pressure peak characteristics, and rescue location in the intervention behavior dataset with the treatment benchmark library. If the match is successful, the pressure dispersion characteristics are mapped to a fastening coefficient that characterizes the firmness of wound treatment using an inverse proportional function. The blood loss rate decay parameter in the dynamic physiological feature dataset is corrected based on the fastening coefficient, and the remaining blood volume value in the dataset is updated. The transport coupling module is used to call the vibration energy spectrum of the virtual transport vehicle and perform mechanical coupling verification between vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the hemostasis measures are deemed to have failed and the secondary bleeding logic is activated. The remaining blood volume value in the dynamic physiological feature dataset is deducted based on the secondary bleeding logic until the clinical endpoint is reached, generating a treatment outcome classification result.

[0044] In this implementation plan, the initial configuration module primarily undertakes the functions of digital reconstruction and rule setting of the training environment. As the foundational layer of the system, this module first establishes an interactive bridge between the physical world and the virtual scene, ensuring that every touchpoint on the physical simulator accurately maps to the specific anatomical location of the virtual injured person. Simultaneously, based on a preset disaster injury background, it instantiates a virtual physiological model with specific initial values ​​for vital signs and loads the corresponding standard treatment plan and equipment parameters for that injury. This provides an anatomically logical interactive foundation and objective automated assessment standards for subsequent training, ensuring the medical rigor and data consistency of the entire training scenario. The data acquisition module, as the system's multimodal perception front-end, is responsible for capturing and analyzing the trainee's specific operational behaviors in real time. Through multi-threaded monitoring and spatiotemporal synchronization technology, it precisely aligns the radio frequency identification signal representing "equipment type" with the pressure time-series data representing "operational intensity," thereby extracting effective operational segments. More importantly, this module not only records operational values ​​but also extracts peak and dispersion features of pressure through algorithms. It quantifies and analyzes the trainee's hand tremors and the stability of sustained pressure during rescue, transforming abstract operational feel into a dataset of interventional behaviors that can be processed by a computer, providing high-precision raw data support for subsequent quality assessment. The treatment verification module undertakes the core tasks of logical judgment and physiological calculation. It first performs multi-dimensional compliance checks, comparing whether the trainee's operation position, equipment selection, and force intensity meet the requirements of clinical emergency guidelines. Once compliance is determined, it further uses an inverse proportional function algorithm to convert the pressure dispersion extracted by the previous module into a tightness coefficient characterizing the firmness of wound dressing. Subsequently, this module uses this coefficient to dynamically adjust the blood loss model of the virtual patient, reducing the blood loss rate in real time according to the reliability of the treatment. This establishes a direct causal feedback of "high-quality operation leading to efficient hemostasis" during training, realizing a logical closed loop from physical operation to physiological response. The transport coupling module is mainly used to simulate the environmental interference risks during transport and the final treatment outcome. This module incorporates the vibration energy spectrum generated during ambulance or helicopter transport, converting it into physical shear forces that attempt to disrupt wound fixation. This force is then mechanically counteracted against the tightening coefficient generated by the treatment verification module. If the initial bandaging of the trainee is insufficiently secure (low tightening coefficient) and cannot withstand transport vibrations, the module will automatically determine hemostasis failure and activate secondary bleeding logic, simulating the deterioration of the injury. Ultimately, it determines the clinical endpoint based on changes in blood volume throughout the process, generating a treatment outcome classification, thereby completing a comprehensive evaluation of the entire "on-site first aid - en route transport" treatment effectiveness.

[0045] In summary, this application has at least the following effects:

[0046] This trauma care training method and system, based on somatosensory interaction and multimodal perception, constructs a precise spatial mapping matrix between a physical simulator and a virtual patient. Utilizing radio frequency identification and multimodal heterogeneous fusion technology based on pressure time-series data, it achieves full-dimensional quantitative perception of treatment operations, encompassing "instrument selection, force application characteristics, operation position, and mechanical stability." This effectively overcomes the technical shortcomings of traditional training methods, which cannot objectively evaluate the stability and compliance of trainees' operational techniques. Furthermore, this application innovatively establishes a dynamic adversarial mechanism between environmental stress and treatment quality. It mechanically couples and verifies the wound fixation coefficient generated during on-site treatment with the vibration energy spectrum during virtual transport. This realistically simulates the chain reaction of injuries caused by inadequate initial bandaging and fixation leading to secondary bleeding in a bumpy transport environment. This breaks down the logical separation between "on-site first aid" and "transportation" in existing technologies, forcing trainees to develop a holistic treatment concept oriented towards addressing the risks of the entire transportation chain. This significantly enhances the medical logic realism and practical guidance value of trauma care training.

[0047] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0048] This invention is described with reference to flowchart illustrations and / or block diagrams of systems, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0049] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0050] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0051] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0052] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A trauma care training method based on somatosensory interaction and multimodal perception, characterized in that, Includes the following steps: S1. Construct a spatial mapping matrix between the pressure sensing array on the surface of the physical simulated human and the anatomical coordinates of the virtual wounded, initialize a dynamic physiological feature dataset containing hemodynamic indicators, and configure a treatment benchmark library containing RFID coding sequences of matching equipment and treatment pressure thresholds according to the injury type. S2. Synchronously collect the device identification ID obtained by the hand radio frequency reader and the pressure time series data output by the simulated human sensor node, extract the pressure data segment within the time period when the device identification ID is identified, extract the pressure peak feature and pressure dispersion feature of the pressure data segment, and generate an intervention behavior dataset containing device attribute verification position and operation mechanical stability index. S3. Confirm the rescue location of the pressure time series data according to the spatial mapping matrix, and verify the matching of the equipment ID, pressure peak characteristics and rescue location in the intervention behavior dataset with the treatment benchmark library. If the matching is successful, use the inverse proportional function relationship to map the pressure dispersion characteristics into the fastening coefficient that characterizes the firmness of wound treatment. Correct the blood loss rate decay parameter in the dynamic physiological characteristic dataset according to the fastening coefficient, and update the remaining blood volume value in the dataset. S4. Call the vibration energy spectrum of the virtual transport vehicle to perform mechanical coupling verification of vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the hemostasis measures are deemed to have failed and the secondary bleeding logic is activated. The remaining blood volume value in the dynamic physiological feature dataset is deducted according to the secondary bleeding logic until the clinical endpoint is reached and the treatment outcome classification result is generated.

2. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 1, characterized in that: The specific process of constructing the spatial mapping matrix between the pressure sensing array on the surface of the physical simulated human body and the anatomical coordinates of the virtual wounded person is as follows: The pressure sensing nodes on the surface of the physical simulated human are traversed to obtain the physical topological coordinates and hardware channel codes of each node. Based on the physical topological coordinates, the nodes are divided into independent anatomical sensing areas. Load the virtual wounded soldier's 3D mesh model, extract the set of mesh vertex indices corresponding to the anatomical structure to define the virtual anatomical region, run the region registration algorithm to establish the mapping relationship between the anatomical perception region and the virtual anatomical region, generate a bidirectional lookup table linking the hardware channel code and the virtual mesh vertex index, and form a spatial mapping matrix.

3. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 1, characterized in that: The specific process of initializing a dynamic physiological feature dataset containing hemodynamic parameters and configuring a treatment benchmark library containing matching equipment RFID coding sequences and treatment pressure thresholds according to the injury type is as follows: Based on the preset injury type, the corresponding pathophysiological evolution template is retrieved, and a physiological state vector containing the total blood volume baseline value, coagulation factor activity coefficient and vascular resistance parameter is instantiated to construct a dynamic physiological feature dataset at the initial moment. Search the medical device attribute database, filter the standard treatment equipment RFID code list that matches the injury type, and set the effective rescue pressure range and standard rescue site identification for each standard treatment equipment; Establish an index linking the RFID code list of standard treatment equipment with the effective rescue pressure range and standard rescue location identification, and construct a multi-dimensional treatment benchmark database.

4. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 1, characterized in that: The specific process of simultaneously acquiring the device identification ID from the hand-held RFID reader and the pressure time-series data output from the simulated human sensor node, and extracting the pressure data segment within the time period when the device identification ID is identified, is as follows: Start the dual-thread data monitoring mode: the main thread listens for the device identification ID event uploaded by the hand RF reader, and the child thread listens for the real-time pressure data stream output by the simulated human sensor node. When the main thread detects the input of the equipment identity ID, it records the current system time as the start timestamp of the action and activates the time capture window; During the time capture window, active node data with pressure values ​​exceeding the effective trigger threshold are filtered from the real-time pressure data stream of the sub-thread. The sequence of all active node pressure sampling points from the start timestamp of the action to the closing time of the window is locked and extracted to form a pressure data segment synchronized with the equipment identity ID.

5. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 1, characterized in that: The specific process of extracting the pressure peak value and pressure dispersion features of the pressure data segment to generate an interventional behavior dataset that includes instrument attribute verification positions and operational mechanical stability indicators is as follows: The extracted pressure data segments are traversed in the time domain, and the maximum pressure amplitude is selected as the pressure peak feature to characterize the extreme force of the rescue operation. The statistical variance of the pressure sampling point sequence relative to the pressure mean in the pressure data segment is calculated as the pressure dispersion feature, which is used to quantitatively characterize the degree of hand tremor or the stability of continuous compression during the rescue process. Extract the hardware channel code of active nodes in the pressure data segment, call the spatial mapping matrix to parse the corresponding rescue position coordinates, map the equipment identity ID to the equipment attribute verification bit, combine the pressure peak feature, pressure dispersion feature and rescue position coordinates to form the operation mechanical stability index, and encapsulate to generate the intervention behavior dataset.

6. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 1, characterized in that: The rescue location is confirmed based on the spatial mapping matrix of the pressure time series data. The matching of the equipment ID, pressure peak characteristics, and rescue location in the intervention behavior dataset with the treatment benchmark library is verified. If the matching is successful, the pressure dispersion characteristics are mapped to a tightening coefficient that characterizes the firmness of the wound treatment using an inverse proportional function relationship. The specific process is as follows: Extract the hardware channel codes of active nodes from the intervention behavior dataset, retrieve the corresponding virtual anatomical region index in the spatial mapping matrix, and determine the virtual coordinates of the rescue location; Construct a logic AND gate to compare the device ID with the standard device code in the treatment benchmark library, the pressure peak characteristics with the effective treatment pressure threshold, and the virtual coordinates of the rescue location with the standard rescue site identifier. Only when all three sets of comparison results are true will a matching pass signal be output. The response matching process normalizes the pressure dispersion characteristics through the signal, performs an inverse mapping operation on the normalization result, and generates a tightening coefficient with a monotonically increasing value.

7. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 6, characterized in that: The specific process of correcting the blood loss rate decay parameter in the dynamic physiological characteristic dataset based on the clamping coefficient and updating the remaining blood volume value in the dataset is as follows: Read the current blood loss rate variable from the dynamic physiological feature dataset, substitute the tightness coefficient as a gain factor into the exponential decay algorithm, calculate the blood loss inhibition ratio at the current time step, apply the blood loss inhibition ratio to perform a downward correction on the current blood loss rate variable, and generate the real-time blood loss rate. The system performs an integral operation on the real-time blood loss rate within the current sampling time interval and simultaneously subtracts the blood loss amount obtained from the integral from the remaining blood volume value of the dynamic physiological feature dataset, thus completing the hemodynamic state update for a single cycle.

8. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 1, characterized in that: The vibration energy spectrum of the virtual transport vehicle is invoked to perform a mechanical coupling verification of vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the hemostasis measure is deemed to have failed and the secondary bleeding logic is activated. The specific process is as follows: The vibration energy spectrum of the virtual transport vehicle was analyzed, the vibration frequency and amplitude parameters were extracted, and the equivalent physical shear force acting on the wound was calculated using a tribomechanical model. The fastening coefficient is converted into a mechanical resistance limit value and defined as a critical maintenance threshold. When the equivalent physical shear force is detected to be greater than the critical maintenance threshold and the duration exceeds the preset judgment time window, a hemostasis failure interruption signal is generated. In response to the hemostasis failure interruption signal, the blood loss rate variable in the dynamic physiological feature dataset is switched from the current decay state to the uncontrolled high flow rate state, and the uncontrolled high flow rate state is locked to activate the secondary bleeding logic.

9. The trauma care training method based on somatosensory interaction and multimodal perception according to claim 8, characterized in that: The specific process of subtracting the remaining blood volume value from the dynamic physiological feature dataset based on the secondary hemorrhage logic until the clinical endpoint is reached to generate the treatment outcome classification result is as follows: Based on the activated secondary bleeding logic, the remaining blood volume is iteratively reduced in subsequent time steps, using the uncontrolled high flow rate state as a benchmark. Real-time monitoring of the relationship between remaining blood volume values ​​and preset shock thresholds, organ failure thresholds, and clinical termination thresholds; When the remaining blood volume value is less than any threshold or the training time ends, the iterative deduction operation is terminated and the final remaining blood volume value is locked. The final remaining blood volume value is then compared with the clinical prognosis grading table, and the corresponding treatment outcome classification result is output.

10. A trauma care training system based on somatosensory interaction and multimodal perception, applied to the trauma care training method based on somatosensory interaction and multimodal perception as described in any one of claims 1-9, characterized in that, Includes the following modules: The initial configuration module is used to construct a spatial mapping matrix between the pressure sensing array on the surface of the physical simulated human and the anatomical coordinates of the virtual wounded person, initialize a dynamic physiological feature dataset containing hemodynamic indicators, and configure a treatment benchmark library containing RFID coding sequences of matching equipment and treatment pressure thresholds according to the injury type. The data acquisition module is used to synchronously acquire the device identification ID obtained by the hand radio frequency reader and the pressure time series data output by the simulated human sensor node, extract the pressure data segment within the period when the device identification ID is identified, extract the pressure peak feature and pressure dispersion feature of the pressure data segment, and generate an interventional behavior dataset that includes device attribute verification bits and operation mechanical stability index. The treatment verification module is used to confirm the rescue location of the pressure time series data according to the spatial mapping matrix, and to verify the matching of the equipment ID, pressure peak characteristics and rescue location in the intervention behavior dataset with the treatment benchmark library. If the matching is successful, the pressure dispersion characteristics are mapped to the fastening coefficient that characterizes the firmness of the wound treatment using the inverse proportional function relationship. The blood loss rate decay parameter in the dynamic physiological characteristic dataset is corrected according to the fastening coefficient, and the remaining blood volume value in the dataset is updated. The transport coupling module is used to call the vibration energy spectrum of the virtual transport vehicle, perform mechanical coupling verification of vibration shear force and fastening coefficient. If the vibration shear force intensity is higher than the critical maintenance threshold determined by the fastening coefficient, the hemostasis measures are deemed to have failed and the secondary bleeding logic is activated. The remaining blood volume value in the dynamic physiological feature dataset is deducted according to the secondary bleeding logic until the clinical endpoint is reached and the treatment outcome classification result is generated.

Images