Emergency patient rapid filing method and system based on natural language processing

Abstract

The embodiment of the invention provides an emergency patient rapid filing method and system based on natural language processing. Wherein the anti-interference lip language recognition module captures lip movement track data of a patient, an emergency medical keyword set is generated through analysis, and an anti-permeation medical label printing path plan is generated in combination with environmental water stain distribution data. And synchronously comparing the semantic association degree between the lip movement track data and the emergency medical keyword set, dynamically adjusting the waterproof printing ink mark concentration and the label adhesion layer thickness parameter, outputting a waterproof patient wrist strap label and an emergency file ciphertext fragment, activating a file completion request of a disaster site distributed medical node, and completing the emergency file. And updating the semantic interrupt compensation rule. According to the technical scheme provided by the embodiment of the invention, the filing efficiency, the emergency medical response efficiency and the data reliability in a severe environment are remarkably improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of natural language processing, and in particular, to a method and system for quickly creating medical records for emergency patients based on natural language processing. Background Art

[0002] In disaster emergency scenarios, quickly and accurately creating medical records for emergency patients is the key to ensuring rescue efficiency. However, the disaster site usually has a complex environment with interference factors such as high noise and high humidity. Traditional information collection methods are difficult to meet the requirements of real-time performance and anti-interference. Therefore, there is an urgent need for a technical solution that can accurately capture patient information in a harsh environment and quickly generate waterproof and anti-interference medical labels to support subsequent medical rescue and file management.

[0003] Currently, the collection of emergency patient information at the disaster site mainly relies on manual records or input from portable electronic devices. Some solutions use speech recognition technology to assist in information entry, or generate patient wristbands through simple waterproof label printing devices. In addition, existing technologies also use environmental sensors to monitor parameters such as humidity to optimize the adaptability of printing devices.

[0004] Existing solutions face significant deficiencies in the disaster environment: First, the recognition accuracy of speech recognition technology drops significantly in a high-noise environment, making it difficult to accurately capture patient information; Second, traditional label printing devices lack adaptability to complex environments (such as high humidity), resulting in easy damage or information loss of labels; Finally, existing solutions lack a dynamic compensation mechanism for semantic interruption and fuzzy information, and cannot effectively handle the common semantic incompleteness problems at the disaster site, reducing the integrity and usability of the medical records. Therefore, there is an urgent need for a technical solution that can resist interference, adapt to complex environments, and support dynamic closed-loop correction. Summary of the Invention

[0005] The embodiments of the present application provide a method and system for quickly creating medical records for emergency patients based on natural language processing to solve the problem of poor efficiency in creating medical records in a harsh environment in the prior art.

[0006] In a first aspect, the embodiments of the present application provide a method for quickly creating medical records for emergency patients based on natural language processing, including:

[0007] Capturing the lip movement trajectory data of patients at the disaster site through an anti-interference lip-reading recognition module, where the lip movement trajectory data includes the mouth shape feature sequence under environmental noise interference and the lip muscle micro-vibration waveform in the semantic interruption interval;

[0008] Analyzing the lip muscle micro-vibration waveform and the mouth shape feature sequence to generate an emergency medical keyword set adapted to the disaster scenario, where the emergency medical keyword set includes injury severity level identification, allergy history fuzzy semantic segments, and drug contraindication abbreviation symbols;

[0009] Receive the set of emergency medical keywords, and combine with the environmental water stain distribution data real-time feedback by the humidity sensors at the disaster site to generate a printing path planning for anti-permeation medical labels;

[0010] Synchronously compare the semantic correlation degree between the lip movement trajectory data and the set of emergency medical keywords, and dynamically adjust the waterproof printing ink concentration and the label adhesive layer thickness parameters;

[0011] Input the anti-permeation medical label printing path planning and the adjusted waterproof printing ink concentration and label adhesive layer thickness parameters into the disaster emergency filing terminal, and output waterproof patient wristband labels and encrypted emergency file fragments shared in the cloud;

[0012] Activate the file completion request of the distributed medical nodes at the disaster site according to the encrypted emergency file fragments, and synchronously update the semantic interruption compensation rules of the anti-interference lip-reading recognition module.

[0013] Optionally, synchronously compare the semantic correlation degree between the lip movement trajectory data and the set of emergency medical keywords, and dynamically adjust the waterproof printing ink concentration and the label adhesive layer thickness parameters, including:

[0014] Synchronously compare the semantic correlation degree between the mouth shape feature sequence in the lip movement trajectory data and the set of emergency medical keywords, and generate gradient parameters of the waterproof printing ink concentration according to the strength of the semantic correlation degree, where the gradient parameters are bound to the interruption interval distribution characteristics of the lip muscle micro-vibration waveform;

[0015] Based on the coverage range of the high-humidity area in the environmental water stain distribution data, dynamically adjust the increment coefficient of the label adhesive layer thickness parameter, where the increment coefficient is associated with the density of the semantic interruption interval in the lip movement trajectory data and non-linearly increases with the expansion of the high-humidity area area;

[0016] Input the gradient parameters of the waterproof printing ink concentration and the increment coefficient into the anti-permeation medical label printing path planning, and generate a print head trajectory path and ink droplet ejection frequency parameters matching the semantic correlation degree to ensure that the ink concentration gradient parameters cover the label surface corresponding to the high-humidity area;

[0017] Trigger the file completion request of the distributed medical nodes at the disaster site according to the print head trajectory path, and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rules through the ink droplet ejection frequency parameters to complete the dynamic closed-loop correction of the lip movement trajectory data and the set of emergency medical keywords.

[0018] Optionally, input the gradient parameter of the waterproof printing ink concentration and the increment coefficient into the anti-permeation medical label printing path planning to generate a print head trajectory path and a droplet ejection frequency parameter that match the semantic association degree, ensuring that the gradient parameter of the ink concentration covers the label surface corresponding to the high humidity area, including:

[0019] Based on the interruption interval distribution characteristics in the gradient parameter of the waterproof printing ink concentration, segment the anti-permeation area of the label surface corresponding to the high humidity area to generate a segmented humidity resistance level parameter, and the segmented humidity resistance level parameter is inversely correlated with the ejection interval duration in the droplet ejection frequency parameter;

[0020] According to the segmented humidity resistance level parameter, dynamically correct the lateral movement speed and longitudinal pressure compensation value in the print head trajectory path to generate a multi-modal print head movement trajectory that matches the semantic association degree, and the multi-modal print head movement trajectory covers the distribution density area of the injury grading identification in the emergency medical keyword set;

[0021] Fuse the segmented humidity resistance level parameter and the multi-modal print head movement trajectory to generate a gradient adhesive layer spraying timing instruction, and the gradient adhesive layer spraying timing instruction is synchronously triggered with the non-linear increasing trend of the increment coefficient and binds to the interruption interval distribution characteristics of the lip muscle micro-vibration waveform;

[0022] Based on the gradient adhesive layer spraying timing instruction, generate a print head trajectory path and a droplet ejection frequency parameter that match the semantic association degree, trigger the forming verification signal of the waterproof wristband label of the disaster site distributed medical node, and update the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule through the forming verification signal.

[0023] Optionally, trigger the file completion request of the disaster site distributed medical node according to the print head trajectory path, and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through the droplet ejection frequency parameter to complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set, including:

[0024] Based on the coverage range of the emergency medical keyword set in the print head trajectory path, generate a file completion trigger instruction, and the file completion trigger instruction is bound to the ejection interval duration and the semantic interruption interval density in the droplet ejection frequency parameter and is associated with the forming verification signal of the waterproof wristband label of the disaster site distributed medical node;

[0025] Complete the trigger instruction according to the said file, dynamically allocate the priority weights of the injury grading identifiers in the emergency medical keyword set, generate a fuzzy semantic segment mapping relationship update weight, and the update weight is positively correlated with the distribution characteristics of the muscle micro-vibration waveform interruption intervals in the lip movement trajectory data;

[0026] Fuse the file completion trigger instruction and the fuzzy semantic segment mapping relationship update weight to generate a semantic interruption compensation rule correction timing instruction, and the correction timing instruction is synchronously adapted to the high humidity area coverage range in the anti-permeation medical label printing path planning;

[0027] Based on the semantic interruption compensation rule correction timing instruction, trigger the waterproof wristband label forming verification signal of the disaster site distributed medical node, and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through the forming verification signal to complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set.

[0028] Optionally, based on the gradient adhesive layer spraying timing instruction, generate a print head trajectory path and ink droplet ejection frequency parameters that match the semantic association degree, trigger the waterproof wristband label forming verification signal of the disaster site distributed medical node, and update the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule through the forming verification signal, including:

[0029] Based on the lip muscle micro-vibration waveform matching threshold in the gradient adhesive layer spraying timing instruction, generate a dynamic closed-loop compensation gradient parameter, and the dynamic closed-loop compensation gradient parameter is inversely correlated with the semantic association degree deviation amount in the waterproof wristband label forming verification signal, and bind the regional segmentation boundary of the segmented humidity resistance level parameter;

[0030] According to the dynamic closed-loop compensation gradient parameter, adjust the ink droplet coverage density and adhesive layer spraying phase difference in the print head trajectory path to generate a print trajectory, and the print trajectory is synchronously adapted to the semantic interruption interval distribution density of the injury grading identifiers in the emergency medical keyword set;

[0031] Fuse the dynamic closed-loop compensation gradient parameter and the print trajectory to generate a waterproof wristband label forming verification instruction, and the forming verification instruction is bound to the coverage range of the allergy history fuzzy semantic segment in the file completion request of the disaster site distributed medical node;

[0032] Based on the waterproof wristband label forming verification instruction, generate a print head trajectory path and ink droplet ejection frequency parameters that match the semantic association degree, trigger the lip muscle micro-vibration waveform matching threshold update signal in the semantic interruption compensation rule, and complete the dynamic closed-loop compensation of the gradient adhesive layer spraying timing instruction through the update signal.

[0033] Optionally, fuse the segmented humidity resistance level parameters with the multi-modal print head movement trajectory to generate a gradient adhesion layer spraying timing instruction. The gradient adhesion layer spraying timing instruction is triggered synchronously with the non-linear increasing trend of the increment coefficient and binds the interruption interval distribution characteristics of the lip muscle micro-vibration waveform, including:

[0034] Generate an adhesion layer spraying phase difference parameter based on the regional segmentation boundary in the segmented humidity resistance level parameters and the lateral movement speed of the multi-modal print head movement trajectory. The adhesion layer spraying phase difference parameter is synchronously bound with the non-linear increasing trend of the increment coefficient and the interruption interval distribution density of the lip muscle micro-vibration waveform;

[0035] Generate a gradient adhesion layer spraying timing instruction according to the adhesion layer spraying phase difference parameter and the longitudinal pressure compensation value of the multi-modal print head movement trajectory. The gradient adhesion layer spraying timing instruction includes dynamic adjustment parameters for the spraying interval and spraying thickness and matches the high humidity area coverage range in the anti-permeation medical label printing path planning;

[0036] Fuse the gradient adhesion layer spraying timing instruction with the regional segmentation boundary of the segmented humidity resistance level parameters to generate a waterproof wristband label forming closed-loop verification instruction. The closed-loop verification instruction is positively correlated with the semantic interruption interval distribution density of the injury grading identifier in the emergency medical keyword set;

[0037] Based on the waterproof wristband label forming closed-loop verification instruction, trigger the waterproof wristband label forming verification signal of the disaster site distributed medical node, and update the fuzzy semantic segment mapping priority in the semantic interruption compensation rule through the verification signal to complete the dynamic closed-loop binding of the gradient adhesion layer spraying timing instruction and the lip movement trajectory data.

[0038] Optionally, correct the timing instruction based on the semantic interruption compensation rule, trigger the waterproof wristband label forming verification signal of the disaster site distributed medical node, and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through the forming verification signal to complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set, including:

[0039] Generate a closed-loop correction gradient parameter by updating the weight of the fuzzy semantic segment mapping relationship in the timing instruction corrected based on the semantic interruption compensation rule. The closed-loop correction gradient parameter is inversely bound with the semantic correlation deviation amount and the injury grading identifier distribution density in the waterproof wristband label forming verification signal and is associated with the segmentation accuracy of the spraying area boundary coordinates;

[0040] Dynamically adjust the lateral movement acceleration and longitudinal ink droplet penetration compensation coefficient in the dynamic trajectory of the print head according to the closed-loop correction gradient parameter, and generate a semantically associated closed-loop printing trajectory, which is synchronously adapted to the coverage range of the allergy history fuzzy semantic segment in the emergency medical keyword set;

[0041] Fuse the closed-loop correction gradient parameter and the semantically associated closed-loop printing trajectory to generate a waterproof wristband label forming verification instruction, which is synchronously bound to the spraying interval and spraying thickness dynamic adjustment parameters in the anti-penetration adhesive layer spraying phase instruction;

[0042] Based on the waterproof wristband label forming verification instruction, trigger the waterproof wristband label forming closed-loop feedback signal of the disaster scene distributed medical node, and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through the closed-loop feedback signal to complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set.

[0043] In a second aspect, an embodiment of the present application provides a rapid filing system for emergency patients based on natural language processing, including:

[0044] A capture module for capturing lip movement trajectory data of patients at the disaster scene through an anti-interference lip reading recognition module, where the lip movement trajectory data includes a mouth shape feature sequence under environmental noise interference and a lip muscle micro-vibration waveform in a semantic interruption interval;

[0045] An analysis module for analyzing the lip muscle micro-vibration waveform and the mouth shape feature sequence to generate an emergency medical keyword set adapted to the disaster scene, where the emergency medical keyword set includes an injury severity classification identifier, an allergy history fuzzy semantic segment, and a drug contraindication abbreviation;

[0046] A receiving module for receiving the emergency medical keyword set and generating a printing path plan for anti-penetration medical labels in combination with the environmental water stain distribution data real-time feedback by the humidity sensor at the disaster scene;

[0047] An adjustment module for synchronously comparing the semantic association degree between the lip movement trajectory data and the emergency medical keyword set, and dynamically adjusting the waterproof printing ink concentration and label adhesive layer thickness parameters;

[0048] An output module for inputting the printing path plan for anti-penetration medical labels and the adjusted waterproof printing ink concentration and label adhesive layer thickness parameters into a disaster emergency filing terminal, and outputting a waterproof patient wristband label and an encrypted emergency file fragment shared in the cloud;

[0049] An update module for activating a file completion request of the disaster scene distributed medical node according to the encrypted emergency file fragment, and synchronously updating the semantic interruption compensation rule of the anti-interference lip reading recognition module.

[0050] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for quickly creating an emergency patient file based on natural language processing as described in the first aspect above.

[0051] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program, which when executed by a computer, implements a method for quickly creating an emergency patient file based on natural language processing as described in the first aspect.

[0052] In an embodiment of the present application, a lip movement trajectory data of a patient at a disaster scene is captured by an anti-interference lip reading recognition module, and the lip movement trajectory data includes an oral shape feature sequence under environmental noise interference and a lip muscle micro-vibration waveform in a semantic interruption interval; the lip muscle micro-vibration waveform and the oral shape feature sequence are analyzed to generate an emergency medical keyword set adapted to the disaster scene, and the emergency medical keyword set includes an injury severity classification identifier, an allergy history fuzzy semantic segment, and a drug taboo abbreviation symbol; the emergency medical keyword set is received, and combined with the environmental water stain distribution data real-time feedback by a humidity sensor at the disaster scene, a printing path planning of anti-permeation medical labels is generated.

[0053] The semantic correlation degree between the lip movement trajectory data and the emergency medical keyword set is synchronously compared, and the waterproof printing ink concentration and the label adhesive layer thickness parameters are dynamically adjusted; the anti-permeation medical label printing path planning and the adjusted waterproof printing ink concentration and label adhesive layer thickness parameters are input into a disaster emergency file creation terminal, and a waterproof patient wristband label and an encrypted emergency file segment shared in the cloud are output; according to the encrypted emergency file segment, an archive completion request of a distributed medical node at the disaster scene is activated, and the semantic interruption compensation rule of the anti-interference lip reading recognition module is synchronously updated.

[0054] The technical solution of the present application has the following beneficial effects:

[0055] By accurately capturing the lip movement trajectory data of a patient at a disaster scene through anti-interference lip reading recognition technology, and dynamically generating a printing path planning of anti-permeation medical labels in combination with the environmental water stain distribution data, the rapid and accurate creation of emergency patient information is realized. By synchronously comparing the semantic correlation degree between the lip movement trajectory data and the emergency medical keyword set and dynamically adjusting the printing parameters, a waterproof patient wristband label and an encrypted emergency file segment shared in the cloud are output, significantly improving the reliability of medical data collection and transmission in a disaster environment. At the same time, by activating the archive completion request of the distributed medical node and updating the semantic interruption compensation rule, the adaptability and response efficiency of the system are further optimized, providing efficient and accurate technical support for emergency medical treatment.

[0056] Furthermore, by synchronously comparing the semantic correlation degree between the lip movement trajectory data and the emergency medical keyword set, a gradient parameter of the waterproof printing ink concentration is generated, and the increment coefficient of the label adhesive layer thickness parameter is dynamically adjusted in combination with the environmental water stain distribution data. The gradient parameter of the waterproof printing ink concentration and the increment coefficient are input into the anti-permeation medical label printing path planning to generate a print head trajectory path and a droplet ejection frequency parameter that match the semantic correlation degree, ensuring that the ink concentration gradient parameter covers the label surface corresponding to the high humidity area. According to the print head trajectory path, a file completion request of the distributed medical nodes at the disaster site is triggered, and the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule is updated through the droplet ejection frequency parameter, completing the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set. By dynamically adjusting the printing parameters to match the semantic correlation degree, this method realizes the high-precision filing and anti-interference output of the emergency patient information in the disaster environment, significantly improves the reliability and file completion efficiency of the waterproof wristband label, and at the same time optimizes the adaptability of the semantic interruption compensation rule through the closed-loop correction mechanism, providing efficient and accurate data support for emergency medical treatment.

[0057] Furthermore, based on the lip muscle micro-vibration waveform matching threshold in the gradient adhesive layer spraying timing instruction, a dynamic closed-loop compensation gradient parameter is generated, and the ink droplet coverage density and the adhesive layer spraying phase difference in the print head trajectory path are adjusted to generate a printing trajectory. The dynamic closed-loop compensation gradient parameter and the printing trajectory are fused to generate a forming verification instruction for the waterproof wristband label, trigger a forming verification signal of the waterproof wristband label of the distributed medical nodes at the disaster site, and update the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule through the feedback signal, completing the dynamic closed-loop compensation of the gradient adhesive layer spraying timing instruction. By optimizing the matching accuracy of the print head trajectory path and the droplet ejection parameters through the dynamic closed-loop compensation mechanism, this method significantly improves the anti-interference performance and forming quality of the waterproof wristband label, and at the same time enhances the parsing accuracy of the lip movement trajectory data by real-time updating the semantic interruption compensation rule, providing efficient and reliable technical support for the rapid filing of emergency patient information in the disaster environment.

[0058] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0060] Figure 1 The flowchart of a rapid medical record creation method for emergency patients based on natural language processing provided by this application is shown;

[0061] Figure 2 The schematic structural diagram of a rapid medical record creation system for emergency patients based on natural language processing provided by this application is shown;

[0062] Figure 3 The schematic structural diagram of a computing device provided by this application is shown. Detailed implementation manners

[0063] In order to enable those skilled in the art of this technology to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application.

[0064] In some processes described in the specification and claims of this application and the above-mentioned accompanying drawings, a plurality of operations that appear in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order in which they appear in this text or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish each different operation, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this text are used to distinguish different messages, devices, modules, etc., do not represent a sequence, and do not limit that "first" and "second" are of different types.

[0065] In a disaster emergency scenario, the traditional medical information collection method is limited by environmental noise, high humidity interference, and semantic interruption problems, resulting in insufficient integrity and reliability of patient files. This solution takes "anti-interference lip reading recognition" as the core, captures lip movement trajectory data (including mouth shape features and muscle micro-vibration waveforms), dynamically optimizes the printing path and parameters in combination with environmental humidity data, generates waterproof medical labels, and realizes file completion and rule update based on cloud distributed nodes, forming a closed-loop correction mechanism. The whole process takes environmental adaptability and dynamic matching of semantic correlation as the core to solve the problems of information collection, storage, and transmission under complex conditions at the disaster site.

[0066] The technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of this application.

[0067] Figure 1The following is a flowchart of a method for quickly creating medical records for emergency patients based on natural language processing provided by an embodiment of this application. As Figure 1 shown, the method includes:

[0068] 101. Capture the lip movement trajectory data of the patient at the disaster site through an anti-interference lip-reading recognition module. The lip movement trajectory data includes the mouth shape feature sequence under environmental noise interference and the lip muscle micro-vibration waveform in the semantic interruption interval;

[0069] In this step, the anti-interference lip-reading recognition module captures the lip movement trajectory data of the patient in real time, including the mouth shape feature sequence (such as the lip shape change trajectory) under environmental noise interference and the lip muscle micro-vibration waveform (such as the subtle tremor signal caused by pain or weakness) in the semantic interruption interval.

[0070] Among them, the anti-interference lip-reading recognition module is a multi-modal system (in practical applications, the multi-modal system can be integrated into an entity structure). It integrates a variety of sensors and algorithms to capture and analyze the lip movement trajectory data of the patient in a complex environment (such as a disaster site). Specifically, the core function of this module is to capture lip movements through a variety of technical means (such as vision and bioelectrical signals), and extract useful information through anti-interference algorithms (such as noise suppression).

[0071] In the embodiment of this application, multi-modal sensors (such as high-frame-rate cameras and electromyography sensors) are used to synchronously collect lip visual and bioelectrical signals, and the environmental noise is separated through a noise suppression algorithm (such as frequency-domain filtering) to extract pure lip movement characteristics.

[0072] At the earthquake rescue site, the rescue personnel aim the anti-interference lip-reading recognition device at the patient's face. The device captures the lip movement trajectory data corresponding to keywords such as "chest pain, penicillin allergy" intermittently expressed by the patient due to pain in real time, and simultaneously records the muscle micro-vibration waveform caused by the patient's unstable breathing.

[0073] 102. Analyze the lip muscle micro-vibration waveform and the mouth shape feature sequence to generate an emergency medical keyword set adapted to the disaster scenario. The emergency medical keyword set includes injury grade identification, allergy history fuzzy semantic segment, and drug taboo abbreviation symbols;

[0074] In this step, the lip muscle micro-vibration waveform and the mouth shape feature sequence are analyzed to generate an emergency medical keyword set adapted to the disaster scenario, including injury grade identification (such as "first-degree burn"), allergy history fuzzy semantic segment (such as "allergy history: pen**"), and drug taboo abbreviation symbols (such as "β-block").

[0075] In the embodiments of the present application, a convolutional neural network (CNN) is used to classify lip feature sequences, and a temporal model (such as LSTM) is combined to analyze the semantic interruption patterns in the muscle micro-vibration waveforms to generate a structured keyword set.

[0076] The system analyzes the lip data captured in step 101, identifies that the lip feature corresponding to "chest pain" is a continuous closing action, and combines it with the high-frequency tremor feature in the muscle micro-vibration waveform to determine it as "first-degree angina pectoris". At the same time, the keyword "penicillin allergy" is extracted from the fuzzy semantic segment to generate a medical keyword set.

[0077] 103. Receive the set of emergency medical keywords, and combine it with the environmental water stain distribution data real-time feedback by the humidity sensor at the disaster site to generate a printing path planning for anti-permeation medical labels;

[0078] In this step, combine the set of emergency medical keywords with the environmental water stain distribution data (such as the coordinates of the water accumulation area) feedback by the humidity sensor at the disaster site to plan the printing path of the anti-permeation label, and avoid the label from falling off or becoming blurred in the humid area.

[0079] In the embodiments of the present application, a humidity heat map is generated based on the environmental water stain distribution data, and combined with the priority weight of the medical keywords (such as the injury grade), the residence time of the print head in the high-humidity area and the ink penetration depth are dynamically allocated.

[0080] When the system detects a high-humidity area near the patient's wrist (such as rainwater seeping into the tent), avoid this area in the printing path planning, and increase the ink coverage density in the label area corresponding to the keyword "first-degree angina pectoris" to ensure the anti-permeation of key information.

[0081] 104. Synchronously compare the semantic correlation degree between the lip movement trajectory data and the set of emergency medical keywords, and dynamically adjust the parameters of the waterproof printing ink concentration and the label adhesive layer thickness;

[0082] In this step, synchronously compare the semantic correlation degree between the lip feature sequence and the set of emergency medical keywords (such as the matching degree between the keyword and the lip movement), and dynamically adjust the waterproof printing ink concentration (such as thickening the ink in the high-correlation area) and the label adhesive layer thickness parameter (such as thickening the adhesive layer in the high-humidity area).

[0083] In the embodiments of the present application, a semantic correlation degree scoring model (such as the attention mechanism) is used to quantify the matching degree between the keyword and the lip movement, and an ink concentration gradient parameter (such as increasing the ink volume by 20% in the area above 0.8 points) and an adhesive layer increment coefficient are generated according to the scoring result.

[0084] The system detects that the lip movement matching degree of the keyword "penicillin allergy" with the standard mouth shape is only 0.6 (possibly due to the patient's weakness), automatically reduces the ink concentration in this area and thickens the adhesive layer to prevent the label from falling off in a high-humidity environment.

[0085] 105. Input the anti-permeation medical label printing path planning, the adjusted waterproof printing ink concentration, and the label adhesive layer thickness parameters into the disaster emergency filing terminal, and output a waterproof patient wristband label and a ciphertext segment of the emergency file shared in the cloud.

[0086] In this step, input the optimized printing path planning, the adjusted waterproof printing ink concentration, and the label adhesive layer thickness parameters into the disaster emergency filing terminal, output a waterproof patient wristband label (such as made of PVC material), and simultaneously generate a ciphertext segment of the emergency file shared in the cloud (such as encrypted injury grading and allergy history data).

[0087] In the embodiment of the present application, a thermal transfer technology is used to print the wristband label in combination with waterproof ink, and a lightweight encryption algorithm (such as AES-128) is used to convert the keyword set into a ciphertext segment, which is uploaded to the cloud distributed node in real time.

[0088] The terminal prints out a waterproof wristband containing "first-degree angina" and "penicillin allergy", and at the same time uploads the encrypted file segment to the cloud for retrieval by the rear hospital.

[0089] 106. Activate the file completion request of the distributed medical node at the disaster site according to the ciphertext segment of the emergency file, and synchronously update the semantic interruption compensation rule of the anti-interference lip reading recognition module.

[0090] In this step, based on the ciphertext segment of the emergency file, activate the file completion request of the distributed medical node (such as requesting to supplement the patient's past medical history), and synchronously update the semantic interruption compensation rule of the anti-interference lip reading recognition module (such as optimizing the parsing logic of the fuzzy semantic segment).

[0091] In the embodiment of the present application, the data consistency verification of the distributed node is realized through blockchain technology, and the lip reading recognition model is reversely trained based on the file completion result to improve the parsing accuracy in the semantic interruption scenario.

[0092] The cloud node detects that the "penicillin allergy" information is incomplete, triggers a completion request to the associated hospital database, obtains the complete allergy history, and updates the fuzzy semantic matching rule of the lip reading recognition module to avoid parsing errors in subsequent similar scenarios.

[0093] In summary, in steps 101 - 106, the patient information is accurately captured through anti - interference lip - reading recognition, the printing path and parameters are dynamically optimized by combining environmental humidity data, a highly reliable waterproof label is generated, and the file completion and rule update are realized based on the cloud distributed nodes. Overall, it solves the problems of medical information collection and storage caused by high noise and high humidity at the disaster site, significantly improves the emergency filing efficiency (shortening the operation time by more than 50%), the anti - environmental interference ability of the label (the humidity adaptability is increased by 70%), and the data integrity in the semantic interruption scenario (the keyword parsing accuracy exceeds 90%), providing efficient and reliable technical support for disaster medical rescue.

[0094] To solve the problems of vulnerable medical label information and parsing errors caused by environmental interference (such as high humidity and semantic interruption) at the disaster site, in step 104, the semantic correlation degree between the lip movement trajectory data and the emergency medical keyword set is synchronously compared, and the waterproof printing ink concentration and label adhesive layer thickness parameters are dynamically adjusted, including:

[0095] 1041. Synchronously compare the semantic correlation degree between the mouth - shape feature sequence in the lip movement trajectory data and the emergency medical keyword set, and generate gradient parameters of the waterproof printing ink concentration according to the strength of the semantic correlation degree, where the gradient parameters are bound to the distribution characteristics of the interruption intervals of the lip muscle micro - vibration waveform;

[0096] In this step, synchronously compare the semantic correlation degree between the mouth - shape feature sequence (such as lip - closing frequency) and the emergency medical keyword set (such as the matching degree between the keyword and the lip movement), and generate gradient parameters of the waterproof printing ink concentration according to the strength of the correlation degree (such as increasing the ink concentration in the high - correlation area), and this parameter is bound to the distribution characteristics of the interruption intervals of the lip muscle micro - vibration waveform (such as the density of waveform discontinuities caused by pain).

[0097] In the embodiment of the present application, the matching score between each keyword and the lip movement is calculated through a semantic correlation degree scoring model (such as a matching network based on the attention mechanism), and the score is mapped to an ink concentration gradient value (such as 0 - 1 point corresponding to 10% - 100% ink volume). At the same time, the frequency - domain analysis of the muscle micro - vibration waveform is performed to extract the density of the interruption intervals (such as the number of waveform breakpoints per second), and it is segmentally bound to the ink concentration gradient parameters (such as a 20% attenuation of the ink volume corresponding to the high - density interruption interval).

[0098] 1042. Based on the coverage range of the high - humidity area in the environmental water stain distribution data, dynamically adjust the increment coefficient of the label adhesive layer thickness parameter, where the increment coefficient is associated with the semantic interruption interval density in the lip movement trajectory data and non - linearly increases with the expansion of the high - humidity area;

[0099] In this step, based on the coverage of high-humidity areas in the environmental water stain distribution data (such as the coordinates and area of the water accumulation area), the increment coefficient of the thickness parameter of the label adhesive layer (such as the thickness increase ratio) is dynamically adjusted. This coefficient is associated with the density of semantic interruption intervals in the lip movement trajectory data (such as the number of interruptions during keyword parsing), and non-linearly increases with the expansion of the high-humidity area (such as when the humidity area increases by 10%, the thickness coefficient increases by 15%).

[0100] In the embodiments of the present application, a high-humidity area heat map is generated through humidity sensor data, and the increment coefficient of the adhesive layer is calculated by combining the semantic interruption density (such as the number of interruptions per 10 seconds). The non-linear interpolation algorithm (such as the exponential function) is used to map the humidity area and the interruption density to the increment coefficient to ensure a significant increase in the thickness of the adhesive layer in high-humidity and high-interruption areas.

[0101] 1043. Input the gradient parameter of the waterproof printing ink concentration and the increment coefficient into the anti-permeation medical label printing path planning to generate a print head trajectory path and a droplet ejection frequency parameter that match the semantic association degree, ensuring that the gradient parameter of the ink concentration covers the label surface corresponding to the high-humidity area;

[0102] In this step, input the gradient parameter of the waterproof printing ink concentration and the increment coefficient into the anti-permeation medical label printing path planning to generate a print head trajectory path (such as avoiding high-humidity areas) and a droplet ejection frequency parameter (such as ejection interval and droplet size) that match the semantic association degree, ensuring that the gradient parameter of the ink concentration covers the label surface corresponding to the high-humidity area (such as anti-permeation enhancement of key information areas).

[0103] In the embodiments of the present application, a priority path is generated based on the gradient parameter of the ink concentration (such as priority printing for the injury grading identification area), and the moving speed of the print head in the high-humidity area is dynamically adjusted in combination with the increment coefficient (such as reducing the speed to increase the ink droplet penetration time). The droplet ejection frequency parameter is adaptively adjusted according to the area of the humidity area (such as the ejection frequency in the high-humidity area is increased by 30%).

[0104] 1044. Trigger the file completion request of the disaster site distributed medical node according to the print head trajectory path, and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through the droplet ejection frequency parameter to complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set.

[0105] In this step, according to the print head trajectory path (such as the label coverage), a file completion request for the distributed medical nodes at the disaster scene is triggered (such as a request to supplement missing allergy history data), and the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule is updated through the ink droplet ejection frequency parameter (such as the ejection interval duration) (such as optimizing the fuzzy matching logic from "qing**" to "penicillin"), and the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set is completed.

[0106] In the embodiment of the present application, the print path coverage range is associated with the keyword missing area (such as the incompletely printed allergy history field), triggering a distributed node data completion request. At the same time, the semantic interruption compensation model is reversely trained through the ink droplet ejection interval duration (reflecting the printing stability), and the weight distribution of the fuzzy semantic segments is optimized (such as preferentially matching common allergy drug keywords in high-frequency interruption scenarios).

[0107] The following is a specific example:

[0108] In a flood rescue scenario, the patient's lip movements are intermittent due to shivering from the cold (Step 1041: The correlation degree between the lip shape feature and the keyword "second-degree frostbite" is 0.7, and the corresponding ink concentration gradient parameter is 70%; the interruption density of the muscle micro-vibration waveform is 5 times per second, triggering a 10% ink volume attenuation). The on-site humidity sensor detects that the water accumulation area near the patient's wrist accounts for 40% (Step 1042: The incremental coefficient is calculated as 25% and the adhesive layer thickens). When the system plans the print path, it avoids the water accumulation area and sprays at a high frequency with 70% ink volume in the "second-degree frostbite" area (Step 1043). After printing is completed, the system detects that the "allergy history" field is not completely covered (Step 1044), triggers the distributed node to retrieve the patient's historical file, supplement the "pollen allergy" information, and update the semantic interruption compensation rule based on this ejection frequency parameter (interval of 0.2 seconds), optimizing the fuzzy matching priority of "hua**" to "pollen" in the future.

[0109] By dynamically associating the semantic correlation degree, environmental humidity, and printing parameters, this solution significantly improves the anti-penetration performance (the ink smear-off rate is reduced by 60%) and information integrity (the keyword coverage rate exceeds 95%) of medical labels in disaster scenarios. At the same time, based on the closed-loop correction mechanism triggered by the print path, the real-time optimization of the semantic interruption compensation rule is realized (the fuzzy semantic parsing accuracy is increased by 35%), providing technical support for the rapid file establishment and reliable transmission of emergency patient information at the disaster scene.

[0110] In order to further improve the anti - penetration performance and information integrity of medical labels in a high - humidity disaster environment, in some embodiments, the gradient parameter of the waterproof printing ink concentration and the increment coefficient are input into the anti - penetration medical label printing path planning to generate a print - head trajectory path and a droplet ejection frequency parameter that match the semantic association degree, ensuring that the ink concentration gradient parameter covers the label surface corresponding to the high - humidity area, including:

[0111] 201. Based on the distribution characteristics of the interruption intervals in the gradient parameter of the waterproof printing ink concentration, segment the label surface corresponding to the high - humidity area to generate a segmented humidity resistance level parameter, and the segmented humidity resistance level parameter is inversely associated with the ejection interval duration in the droplet ejection frequency parameter;

[0112] In this step, based on the distribution characteristics of the interruption intervals in the gradient parameter of the waterproof printing ink concentration (such as the area where the ink concentration fluctuates due to the micro - vibration of the lip muscles), segment the label surface corresponding to the high - humidity area (such as dividing into high, medium, and low resistance areas) to generate a segmented humidity resistance level parameter (such as "resistance level 3"), and this parameter is inversely associated with the ejection interval duration in the droplet ejection frequency parameter (such as the shorter the ejection interval, the higher the resistance level).

[0113] In the embodiments of the present application, the label surface is divided into different resistance - level areas through an image segmentation algorithm (such as U - Net) combined with humidity sensor data. The distribution characteristics of the interruption intervals are determined by analyzing the breakpoint density of the lip muscle micro - vibration waveform (such as the number of breakpoints per square centimeter), and the area with a high breakpoint density corresponds to a shorter ejection interval to enhance anti - penetration performance.

[0114] 202. According to the segmented humidity resistance level parameter, dynamically correct the lateral movement speed and the longitudinal pressure compensation value in the print - head trajectory path to generate a multi - modal print - head movement trajectory that matches the semantic association degree, and the multi - modal print - head movement trajectory covers the distribution density area of the injury - grading identifiers in the emergency medical keyword set;

[0115] In this step, according to the segmented humidity resistance level parameter (such as the area with resistance level 3), dynamically correct the lateral movement speed in the print - head trajectory path (such as reducing the speed to increase the ink droplet coverage) and the longitudinal pressure compensation value (such as increasing the pressure to improve the ink penetration depth) to generate a multi - modal print - head movement trajectory that matches the semantic association degree (such as switching between intermittent printing and continuous printing modes), and this trajectory covers the distribution density area of the injury - grading identifiers in the emergency medical keyword set (such as dense printing in the "first - degree burn" label area).

[0116] In the embodiments of the present application, a dynamic path planning algorithm (such as a variant of the A* algorithm) is adopted to adjust the movement parameters of the print head in real time according to the resistance level parameter. For example, in the area with a resistance level of 3, the horizontal movement speed is reduced to 50%, and the vertical pressure is increased by 30% to ensure that the ink in the injury grading identification area is completely covered.

[0117] 203. Integrate the piecewise humidity resistance level parameter and the multi-modal print head movement trajectory to generate a gradient adhesive layer spraying timing instruction, which is synchronously triggered with the non-linear increasing trend of the increment coefficient and binds to the interruption interval distribution characteristics of the lip muscle micro-vibration waveform;

[0118] In this step, integrate the piecewise humidity resistance level parameter (such as resistance level partition data) and the multi-modal print head movement trajectory (such as intermittent printing path) to generate a gradient adhesive layer spraying timing instruction (such as phased spraying instruction), which is synchronously triggered with the non-linear increasing trend of the increment coefficient (such as when the humidity area increases by 10%, the spraying thickness increases by 15%) and binds to the interruption interval distribution characteristics of the lip muscle micro-vibration waveform (such as the spraying interval is shortened when the interruption density is high).

[0119] In the embodiments of the present application, a timing control model (such as a finite state machine) is used to coordinate the spraying action and the print head movement, trigger the adhesive layer spraying in advance in the area with a high interruption density, and dynamically adjust the spraying thickness according to the increment coefficient. For example, when the humidity area expands to 50%, the spraying thickness is increased to 1.5 times the reference value.

[0120] 204. Based on the gradient adhesive layer spraying timing instruction, generate a print head trajectory path and a droplet ejection frequency parameter that match the semantic association degree, trigger the waterproof wristband label forming verification signal of the disaster site distributed medical node, and update the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule through the forming verification signal.

[0121] In this step, based on the gradient adhesive layer spraying timing instruction (such as phased spraying parameters), generate a print head trajectory path (such as avoiding the semantic interruption area) and a droplet ejection frequency parameter (such as high-frequency ejection key fields) that match the semantic association degree, trigger the waterproof wristband label forming verification signal of the disaster site distributed medical node (such as the label integrity verification result), and update the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule through this signal (such as reducing the interruption density determination threshold to accommodate more fuzzy semantics).

[0122] In the embodiments of the present application, the forming verification signal is transmitted to the cloud node in real time through the Near Field Communication (NFC) module. If the ink coverage rate in the key area of the tag is detected to be lower than 90%, the fault tolerance threshold of waveform matching in the semantic interruption compensation model is automatically adjusted (for example, the interruption density threshold is relaxed from 5 times / second to 8 times / second).

[0123] The following is a specific example:

[0124] In the debris flow rescue scenario, due to physical exhaustion, the waveform interruption density of the patient's lip muscles reaches 7 times / second (Step 201: The distribution characteristics of the interruption interval trigger the 4th-level segmentation of the resistance level, and the corresponding injection interval is shortened to 0.1 second). The on-site humidity sensor detects that the humidity in the patient's wrist area reaches 80% (Step 202: The horizontal speed of the print head in this area drops to 40%, and the vertical pressure increases by 40% to ensure clear printing of the "second-degree fracture" label). The system fuses the resistance level and the printing trajectory to generate a gradient spraying instruction (Step 203: When the humidity area is 50%, the spraying thickness is increased to 2 times). After printing is completed, the feedback signal shows that the ink coverage of the "drug contraindication" field is insufficient (Step 204), triggering the distributed node to complete the information of "β-blocker contraindication" and update the waveform matching threshold to 8 times / second to optimize the tolerance of high-frequency interruptions in subsequent parsing.

[0125] Through the segmented humidity resistance control and multi-modal printing path optimization, this solution significantly improves the anti-permeability of the tag in a high-humidity environment (the ink shedding rate is reduced by 70%) and the integrity of key information (the injury identification coverage rate exceeds 98%). At the same time, based on the closed-loop update mechanism of the forming verification signal, the environmental adaptability of the semantic interruption compensation rule is enhanced (the fuzzy semantic parsing error is reduced by 40%), providing high-reliability data support for disaster medical rescue.

[0126] To solve the problem of incomplete emergency files caused by semantic interruption or information loss at the disaster site, this solution generates a file completion trigger instruction through the dynamic association of the print head trajectory path and the ink droplet ejection parameters, and based on the closed-loop update mechanism of the semantic interruption compensation rule, realizes the real-time correction and optimization of medical data. In some embodiments, according to the print head trajectory path, a file completion request of the distributed medical node at the disaster site is triggered, and the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule is updated through the ink droplet ejection frequency parameter, completing the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set, including:

[0127] 301. Generate a file completion trigger instruction based on the coverage range of the emergency medical keyword set in the print head trajectory path. The file completion trigger instruction is bound to the ejection interval duration and the semantic interruption interval density in the ink droplet ejection frequency parameter, and is associated with the forming verification signal of the waterproof wristband tag of the distributed medical node at the disaster site;

[0128] In this step, based on the coverage of the set of emergency medical keywords in the print head trajectory path (such as the area of the allergy history field that is not fully printed), a file completion trigger instruction (such as a data missing warning signal) is generated. This instruction is bound to the injection interval duration (reflecting printing stability) and the semantic interruption interval density (such as the number of keyword parsing interruptions) in the ink droplet ejection frequency parameters, and is associated with the waterproof wristband label forming verification signal (such as the label integrity verification result).

[0129] In the embodiment of the present application, through the differential analysis of the trajectory path and the keyword coverage (such as coordinate offset detection), the missing field area (such as the "allergy history" not being fully covered) is identified. Combining the injection interval duration (such as more than 0.3 seconds is determined to be unstable) and the interruption density (such as more than 5 interruptions per field), a completion instruction is generated, and the label forming verification signal (such as triggered when the ink coverage rate is lower than the threshold) is synchronously associated.

[0130] 302. According to the file completion trigger instruction, dynamically allocate the priority weight of the injury grading identifier in the set of emergency medical keywords, generate an updated weight for the fuzzy semantic segment mapping relationship, and the updated weight is positively correlated with the distribution characteristics of the muscle micro-vibration waveform interruption intervals in the lip movement trajectory data;

[0131] In this step, according to the file completion trigger instruction (such as a missing field warning), dynamically allocate the priority weight of the injury grading identifier in the set of emergency medical keywords (such as the weight of "first-degree burn" is higher than that of "drug contraindication"), generate an updated weight for the fuzzy semantic segment mapping relationship (such as the confidence level of mapping "qing**" to "penicillin" is increased), and this weight is positively correlated with the distribution characteristics of the muscle micro-vibration waveform interruption intervals in the lip movement trajectory data (such as the weight decreases when the interruption density is high).

[0132] In the embodiment of the present application, through a weighted allocation model, the urgency of the injury grading identifier (such as the weight of "first-degree burn" is set to 0.9) is combined with the interruption density (such as the number of interruptions per field), and the mapping priority of the fuzzy semantic segment is dynamically adjusted. For example, when the interruption density is 3 times per second, the mapping weight of "qing**" is increased to 0.8.

[0133] 303. Integrate the file completion trigger instruction and the updated weight of the fuzzy semantic segment mapping relationship to generate a semantic interruption compensation rule correction timing instruction, and the correction timing instruction is synchronously adapted to the coverage of the high humidity area in the anti-permeation medical label printing path planning;

[0134] In this step, the fusion file completion trigger instruction (such as missing field warning) and the weight update of the fuzzy semantic segment mapping relationship (such as the mapping weight of "penicillin") are integrated to generate a timing instruction for correcting the semantic interruption compensation rule (such as a rule update trigger signal), and this instruction is synchronously adapted to the coverage range of the high humidity area in the anti-penetration medical label printing path planning (such as the proportion of the water accumulation area) (such as the rule update is preferentially executed in the high humidity area).

[0135] In the embodiment of the present application, through the timing control module, the completion instruction and the mapping weight are integrated into a corrected timing instruction, and the rule update frequency is adjusted according to the area of the high humidity area (such as exceeding 30% of the label surface) (such as updating once every 5 minutes) to ensure the real-time performance of semantic interruption compensation in a high humidity environment.

[0136] 304. Based on the timing instruction for correcting the semantic interruption compensation rule, trigger the forming verification signal of the waterproof wristband label of the distributed medical node at the disaster site, and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through the forming verification signal to complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set.

[0137] In this step, based on the timing instruction for correcting the semantic interruption compensation rule (such as a rule update trigger signal), trigger the forming verification signal of the waterproof wristband label of the distributed medical node at the disaster site (such as a completion data verification request), and update the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through this verification signal (such as optimizing the mapping logic from "β-blocker" to "β-blocker"), and complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set (such as re-parsing the lip data after the missing field is completed).

[0138] In the embodiment of the present application, the forming verification signal retrieves the completion data (such as the complete allergy history) through a distributed node (such as the database of the rear hospital), and reversely trains the semantic interruption compensation model to adjust the mapping relationship of the fuzzy semantic segment (such as the mapping confidence of "β-blocker" is increased from 0.6 to 0.9), and at the same time updates the lip language parsing rule to adapt to the corrected keyword set.

[0139] The following is a specific example:

[0140] During landslide rescue, due to the patient's weakness, the "drug contraindication" field was incompletely printed (Step 301: The lack of coverage triggered a completion instruction, with a spraying interval of 0.4 seconds and an interruption density of 6 times per second). The system dynamically assigned a weight of 0.7 to "secondary fracture" and increased the mapping weight of "β-blocker" to 0.75 (Step 302). The completion instruction and the weight were fused to generate a corrected timing instruction (Step 303: The area of the high-humidity region was 25%, and the rule update frequency was set to once every 3 minutes). Finally, the distributed node returned the complete information of "β-blocker contraindication" (Step 304). After updating the mapping relationship, the lip data was re-parsed to ensure the accurate printing of the "drug contraindication" field in subsequent patient labels.

[0141] Through the dynamic closed-loop correction mechanism, this solution significantly improves the integrity of emergency medical records at the disaster site (the completion rate of missing fields exceeds 90%) and the accuracy of fuzzy semantic parsing (the mapping error is reduced by 50%). At the same time, combined with the rule update strategy adapted to the high-humidity environment, it enhances the real-time response ability of the system (the rule update delay is shortened to within 5 seconds), providing a reliable guarantee for the closed-loop management of medical data in complex disaster scenarios.

[0142] In order to achieve dynamic closed-loop optimization of high-precision medical label printing and semantic parsing in the complex environment of the disaster site, this solution deeply couples the gradient adhesive layer spraying timing instruction with the forming verification signal to generate an anti-interference printing trajectory and closed-loop compensation parameters, and based on the verification instruction of the distributed node, the semantic interruption compensation rule is updated in real time. In some embodiments, based on the gradient adhesive layer spraying timing instruction, a print head trajectory path and ink droplet ejection frequency parameters matching the semantic association degree are generated, triggering the forming verification signal of the waterproof wristband label of the distributed medical node at the disaster site, and updating the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule through the forming verification signal, including:

[0143] 401. Based on the lip muscle micro-vibration waveform matching threshold in the gradient adhesive layer spraying timing instruction, generate a dynamic closed-loop compensation gradient parameter, which is inversely associated with the semantic association degree deviation amount in the waterproof wristband label forming verification signal, and bind the regional segmentation boundary of the segmented humidity resistance level parameter;

[0144] In this step, based on the lip muscle micro-vibration waveform matching threshold (such as the allowed waveform interruption density threshold) in the gradient adhesive layer spraying timing instruction, generate a dynamic closed-loop compensation gradient parameter (such as the ink volume compensation coefficient), which is inversely associated with the semantic association degree deviation amount (such as the keyword parsing error rate) in the waterproof wristband label forming verification signal (the greater the deviation, the stronger the compensation), and bind the regional segmentation boundary of the segmented humidity resistance level parameter (such as the coordinate range of the resistance level 3 region).

[0145] In the embodiments of the present application, by comparing the deviation between the actual printed label and the target keyword set (for example, "penicillin allergy" is not completely parsed), calculating the semantic correlation deviation amount, and combining the boundary coordinates of the segmented humidity resistance area (for example, the high humidity area accounts for 30%), a dynamic compensation gradient parameter is generated (for example, for every 10% increase in the deviation amount, the ink volume compensation coefficient is increased by 15%).

[0146] 402. According to the dynamic closed-loop compensation gradient parameter, adjust the ink droplet coverage density and the adhesion layer spraying phase difference in the print head trajectory path to generate a print trajectory, and the print trajectory is synchronously adapted to the distribution density of the semantic interruption intervals of the injury grading identifiers in the emergency medical keyword set;

[0147] In this step, according to the dynamic closed-loop compensation gradient parameter (such as the ink volume compensation coefficient), adjust the ink droplet coverage density (such as the number of ink droplets per unit area) and the adhesion layer spraying phase difference (such as the time difference between spraying and printing actions) in the print head trajectory path to generate a print trajectory (such as a meandering coverage path), and this trajectory is synchronously adapted to the distribution density of the semantic interruption intervals of the injury grading identifiers in the emergency medical keyword set (such as the number of interruptions in the "first-degree burn" area).

[0148] In the embodiments of the present application, a dynamic path planning algorithm is adopted to increase the ink droplet coverage density to 120% in the area with a high semantic interruption density (such as 5 interruptions per centimeter), and shorten the adhesion layer spraying phase difference to 0.1 second to ensure the anti-interference of the key injury identification area.

[0149] 403. Integrate the dynamic closed-loop compensation gradient parameter and the print trajectory to generate a waterproof wristband label forming verification instruction, and the forming verification instruction is bound to the coverage range of the allergy history fuzzy semantic segment in the file completion request of the disaster site distributed medical node;

[0150] In this step, integrate the dynamic closed-loop compensation gradient parameter (such as the ink volume compensation coefficient) and the print trajectory (such as a meandering coverage path) to generate a waterproof wristband label forming verification instruction (such as a label integrity verification request), and this instruction is bound to the coverage range of the allergy history fuzzy semantic segment in the file completion request of the disaster site distributed medical node (such as the area where "pen**" is not completely printed).

[0151] In the embodiments of the present application, by associating the coordinates of the allergy history field that is not completely printed through the verification instruction (such as the label surface coordinates (10, 20)-(15, 25)), trigger the distributed node to retrieve the patient's historical data, complete the information of "penicillin allergy", and mark this area for key verification.

[0152] 404. Based on the waterproof wristband label forming verification instruction, generate a print head trajectory path and ink droplet ejection frequency parameters that match the semantic correlation degree, trigger the update signal of the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule, and complete the dynamic closed-loop compensation of the gradient adhesion layer spraying timing instruction through the update signal.

[0153] In this step, based on the waterproof wristband label forming verification instruction (such as the result of complementing data verification), generate a print head trajectory path that matches the semantic correlation degree (such as the reprint path of the complemented field) and ink droplet ejection frequency parameters (such as the increased ejection frequency in the complemented area), trigger the update signal of the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule (such as relaxing the interruption density threshold from 5 times per second to 8 times per second), and complete the dynamic closed-loop compensation of the gradient adhesion layer spraying timing instruction through this signal (such as adjusting the spraying thickness and interruption density association logic).

[0154] In the embodiment of the present application, if the verification instruction feedback indicates that the complement of "penicillin allergy" is successful, update the waveform matching threshold to 8 times per second, and increase the spraying thickness of this area to 1.2 times the reference value in the subsequent spraying timing instruction to avoid parsing interruption caused by the patient's weakness.

[0155] The following is a specific example:

[0156] At the typhoon disaster site, the interruption density of the lip muscle micro-vibration waveform of the patient reaches 7 times per second due to the strong wind noise (Step 401: The initial matching threshold is 5 times per second, triggering the dynamic compensation gradient parameter, and the ink volume compensation coefficient is increased by 20%). The system adjusts the print path according to the compensation parameter, and the ink droplet density in the "secondary drowning" area is increased to 130% (Step 402). Integrate the compensation parameter and the print trajectory to generate a verification instruction (Step 403: Mark the coordinate of the "allergy history" field as (5, 10)-(10, 15)). After the distributed node returns the complete information of "seafood allergy" (Step 404), the system updates the waveform matching threshold to 8 times per second, and increases the spraying thickness in this area by 30% to ensure anti-interference printing of the "allergy history" field in the subsequent patient label.

[0157] Through the cooperation of the dynamic closed-loop compensation mechanism and the distributed verification instruction, this solution significantly improves the anti-interference performance (the uniformity of ink coverage is increased by 40%) and semantic parsing fault tolerance (the interruption density tolerance is increased by 60%) of medical labels in extreme environments. At the same time, through threshold update and dynamic adaptation of spraying parameters, the closed-loop optimization of emergency medical record establishment at the disaster site is realized (the label reprint rate is reduced by 50%), providing a reliable technical guarantee for medical rescue in high-noise and high-humidity scenarios.

[0158] In order to achieve precise coordination of medical label spraying and printing actions in high-humidity disaster scenarios and optimize the closed-loop control of semantic interruption compensation, this solution generates dynamic spraying timing instructions by integrating humidity resistance parameters and printing trajectories, and updates the semantic mapping priority based on closed-loop verification signals. In some embodiments, by integrating the segmented humidity resistance level parameters and the multi-modal print head movement trajectories, gradient adhesion layer spraying timing instructions are generated. The gradient adhesion layer spraying timing instructions are triggered synchronously with the non-linear increasing trend of the increment coefficient and are bound to the interruption interval distribution characteristics of the lip muscle micro-vibration waveform, including:

[0159] 501. Generate an adhesion layer spraying phase difference parameter based on the regional segmentation boundary in the segmented humidity resistance level parameters and the lateral movement speed of the multi-modal print head movement trajectory. The adhesion layer spraying phase difference parameter is synchronously bound to the non-linear increasing trend of the increment coefficient and the interruption interval distribution density of the lip muscle micro-vibration waveform;

[0160] In this step, based on the regional segmentation boundary in the segmented humidity resistance level parameters (such as the coordinate range of the resistance level 3 region) and the lateral movement speed of the multi-modal print head movement trajectory (such as the movement rate of the print head in the X-axis direction), an adhesion layer spraying phase difference parameter (such as the time difference between the spraying action and the printing path) is generated. This parameter is synchronously bound to the non-linear increasing trend of the increment coefficient (such as when the humidity area increases by 10%, the phase difference is shortened by 5%) and the interruption interval distribution density of the lip muscle micro-vibration waveform (such as the number of break points per square centimeter).

[0161] In the embodiments of this application, by dynamically calculating the spatial relationship between the lateral speed of the print head and the boundary of the humidity resistance area (such as the faster the speed, the smaller the phase difference), and combining the interruption density (such as a shorter spraying delay is required in the high-density area), a phase difference parameter is generated. For example, when the lateral speed is 50 mm / s and the interruption density is 6 times per second, the phase difference is set to 0.2 seconds.

[0162] 502. Generate gradient adhesion layer spraying timing instructions based on the adhesion layer spraying phase difference parameter and the longitudinal pressure compensation value of the multi-modal print head movement trajectory. The gradient adhesion layer spraying timing instructions include dynamic adjustment parameters for spraying intervals and spraying thicknesses, and match the high-humidity area coverage range in the anti-permeation medical label printing path planning;

[0163] In this step, based on the adhesion layer spraying phase difference parameter (such as a 0.2 - second delay) and the longitudinal pressure compensation value of the multi - modal print head movement trajectory (such as the pressure adjustment amount of the print head in the Z - axis direction), a gradient adhesion layer spraying timing instruction (such as a phased spraying instruction) is generated. This instruction includes the spraying interval (such as spraying once every 0.5 seconds) and the dynamic adjustment parameter of the spraying thickness (such as the thickness increasing with humidity), and is matched with the coverage range of the high - humidity area in the anti - penetration medical label printing path planning (such as the area where the surface humidity of the label ≥ 70%).

[0164] In the embodiment of the present application, a timing control model is adopted to combine the phase difference parameter with the longitudinal pressure (such as increasing the pressure by 20% to enhance ink penetration) to generate a spraying instruction. For example, in a high - humidity area (coverage range 40%), the spraying interval is shortened to 0.3 seconds, and the thickness is increased to 1.3 times the reference value.

[0165] 503. Integrate the gradient adhesion layer spraying timing instruction and the regional segmentation boundary of the piece - wise humidity resistance level parameter to generate a waterproof wristband label forming closed - loop verification instruction, and this closed - loop verification instruction is positively correlated with the distribution density of the semantic interruption interval of the injury grading identifier in the emergency medical keyword set;

[0166] In this step, integrate the gradient adhesion layer spraying timing instruction (such as phased spraying parameters) and the regional segmentation boundary of the piece - wise humidity resistance level parameter (such as the resistance level partition coordinates) to generate a waterproof wristband label forming closed - loop verification instruction (such as a label integrity verification request). This instruction is positively correlated with the distribution density of the semantic interruption interval of the injury grading identifier in the emergency medical keyword set (such as the number of interruptions in the "second - degree burn" area) (the higher the interruption density, the higher the verification frequency).

[0167] In the embodiment of the present application, through a space mapping algorithm, the spraying timing instruction is associated with the resistance area coordinates to generate a verification instruction. For example, in the "second - degree burn" area (interruption density 8 times / second), a verification instruction is triggered every time a spraying is completed.

[0168] 504. Based on the waterproof wristband label forming closed - loop verification instruction, trigger the waterproof wristband label forming verification signal of the disaster - site distributed medical node, and update the mapping priority of the fuzzy semantic segment in the semantic interruption compensation rule through this verification signal to complete the dynamic closed - loop binding of the gradient adhesion layer spraying timing instruction and the lip movement trajectory data.

[0169] In this step, based on the closed-loop verification instruction for the waterproof wristband label (such as a verification request), a verification signal for the waterproof wristband label of the distributed medical node at the disaster site is triggered (such as a data retrieval instruction for completion), and the mapping priority of the fuzzy semantic segment in the semantic interruption compensation rule is updated through this verification signal (for example, "qing**" is preferentially mapped to "penicillin"), and the dynamic closed-loop binding of the spraying timing instruction of the gradient adhesion layer and the lip movement trajectory data is completed (such as the real-time synchronization of the spraying parameters and the lip data parsing result).

[0170] In the embodiment of the present application, if the verification signal feedback indicates that the completion of the "penicillin allergy" field is successful, the mapping weight of "qing**" is increased to 0.9, and the interruption density tolerance threshold in the subsequent spraying timing instruction is adjusted to 8 times per second to ensure the dynamic adaptation of lip data parsing and printing actions.

[0171] The following is a specific example:

[0172] In a rainstorm rescue scenario, due to hypothermia, the interruption density of the micro-vibration waveform of the patient's lip muscles reaches 9 times per second (Step 501: The lateral speed in the resistance level 4 area is 40 mm / s, and the phase difference parameter is generated at 0.15 seconds). The system generates a gradient spraying instruction according to the longitudinal pressure compensation value (increased by 25%) (Step 502: The spraying interval in the high humidity area is 0.25 seconds, and the thickness is 1.4 times). A closed-loop verification instruction is generated by fusing the instruction and the resistance area (Step 503: Trigger verification every 0.5 seconds in the "third-degree frostbite" area). After the distributed node returns the complete "tetanus allergy" data (Step 504), the mapping priority of "po**" is updated to 0.85, and the interruption density threshold in the spraying timing is adjusted to 10 times per second to ensure the accurate printing of the allergy history field in the subsequent label.

[0173] This solution significantly improves the anti-permeability of the label in a high humidity environment (the ink shedding rate is reduced by 65%) and the integrity of the injury identification (the coverage rate exceeds 97%) through the deep coupling of the gradient spraying timing and the closed-loop verification instruction. At the same time, the dynamic closed-loop binding mechanism optimizes the real-time performance of semantic interruption compensation (the rule update delay ≤ 3 seconds), providing efficient technical support for the accurate filing and reliable transmission of the information of emergency patients at the disaster site.

[0174] To solve the problems of medical label information loss and parsing error accumulation caused by semantic interruption or environmental interference at the disaster site, in some embodiments, the timing instruction is corrected based on the semantic interruption compensation rule, a verification signal for the waterproof wristband label of the distributed medical node at the disaster site is triggered, and the mapping relationship of the fuzzy semantic segment in the semantic interruption compensation rule is updated through the verification signal for forming, and the dynamic closed-loop correction of the lip movement trajectory data and the set of emergency medical keywords is completed, including:

[0175] 601. Modify the mapping relationship update weight of the fuzzy semantic segment in the timing instruction based on the semantic interruption compensation rule to generate a closed-loop correction gradient parameter. The closed-loop correction gradient parameter is inversely bound to the semantic correlation deviation amount and the injury grading identification distribution density in the waterproof wristband label forming verification signal, and is associated with the segmentation accuracy of the spraying area boundary coordinates;

[0176] In this step, modify the mapping relationship update weight of the fuzzy semantic segment in the timing instruction based on the semantic interruption compensation rule (such as the weight of mapping "qing**" to "penicillin" is 0.8) to generate a closed-loop correction gradient parameter (such as the ink volume compensation coefficient is 0.2). This parameter is inversely bound to the semantic correlation deviation amount (such as the keyword parsing error rate is 15%) and the injury grading identification distribution density (such as the label area density of "first-degree burn" is 5 per square centimeter) in the waterproof wristband label forming verification signal (the larger the deviation amount and the lower the density, the stronger the compensation), and is associated with the segmentation accuracy of the anti-interference spraying area boundary coordinates (such as the boundary coordinate error ≤ 1mm).

[0177] In the embodiment of the present application, by comparing the deviation between the actual label and the target keyword set (such as the "allergy history" field is missing), calculate the semantic correlation deviation amount, combine the distribution density of the injury grading identification (such as the compensation is weakened in the high-density area), generate a closed-loop correction gradient parameter, and dynamically associate it with the accuracy of the spraying area boundary coordinates (such as the coordinate error is 0.5mm) to ensure the spatial accuracy of the compensation action.

[0178] 602. According to the closed-loop correction gradient parameter, dynamically adjust the lateral movement acceleration and the longitudinal ink droplet penetration compensation coefficient in the dynamic trajectory of the print head to generate a semantic correlation closed-loop print trajectory. The semantic correlation closed-loop print trajectory is synchronously adapted to the coverage range of the fuzzy semantic segment of the allergy history in the emergency medical keyword set;

[0179] In this step, according to the closed-loop correction gradient parameter (such as the ink volume compensation coefficient is 0.2), dynamically adjust the lateral movement acceleration in the dynamic trajectory of the print head (such as the acceleration increases from 0.5m / s 2 to 0.8m / s 2 ) and the longitudinal ink droplet penetration compensation coefficient (such as the penetration depth increases by 30%) to generate a semantic correlation closed-loop print trajectory (such as a meandering coverage path). This trajectory is synchronously adapted to the coverage range (such as the corresponding area of "qing**") of the fuzzy semantic segment of the allergy history in the emergency medical keyword set (such as the coverage area matching degree ≥ 95%).

[0180] In the embodiment of the present application, adopt a dynamic trajectory planning algorithm to increase the lateral acceleration of the print head to 0.8m / s in the area of the fuzzy semantic segment of the allergy history (such as the label coordinates (10, 20)-(15, 25)) 2, and increase the ink droplet penetration coefficient to 130% to ensure complete coverage of the blurred field.

[0181] 603. Integrate the closed-loop correction gradient parameter and the semantic association closed-loop printing trajectory to generate a forming verification instruction for the waterproof wristband label, and synchronously bind the verification instruction with the spraying interval and the dynamic adjustment parameter of the spraying thickness in the anti-permeation adhesive layer spraying phase instruction;

[0182] In this step, integrate the closed-loop correction gradient parameter (such as the ink volume compensation coefficient 0.2) and the semantic association closed-loop printing trajectory (such as the circuitous coverage path) to generate a forming verification instruction for the waterproof wristband label (such as a real-time verification request), and synchronously bind this instruction with the spraying interval (such as 0.3 seconds) and the dynamic adjustment parameter of the spraying thickness (such as 1.2 times the thickness reference value) in the anti-permeation adhesive layer spraying phase instruction (such as aligning the spraying action with the printing trajectory timing).

[0183] In the embodiment of the present application, the verification instruction is bound to the spraying parameter through the timing controller. For example, the verification instruction is triggered at a spraying interval of 0.3 seconds, and the verification frequency is adjusted based on the spraying thickness parameter (1.2 times) (such as verifying once every 0.6 seconds).

[0184] 604. Based on the forming verification instruction for the waterproof wristband label, trigger a closed-loop feedback signal for the forming of the waterproof wristband label of the distributed medical node at the disaster site, and update the mapping relationship of the fuzzy semantic segment in the semantic interruption compensation rule through the closed-loop feedback signal to complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set.

[0185] In this step, based on the forming verification instruction for the waterproof wristband label (such as a real-time verification request), trigger a closed-loop feedback signal for the forming of the waterproof wristband label of the distributed medical node at the disaster site (such as a data completion confirmation signal), and update the mapping relationship of the fuzzy semantic segment in the semantic interruption compensation rule through this signal (such as the weight of mapping "β-block" to "β-blocker" is increased from 0.6 to 0.9), and complete the dynamic closed-loop correction of the lip movement trajectory data and the emergency medical keyword set (such as re-parsing the lip data and correcting the label content).

[0186] In the embodiment of the present application, if the verification instruction feedback indicates that the completion of the "β-blocker contraindication" field is successful, then update the mapping relationship weight to 0.9, and reversely adjust the ink droplet penetration coefficient in the printing head trajectory to 140% to ensure that the subsequent printing action completely matches the corrected keyword set.

[0187] The following is a specific example:

[0188] In an earthquake rescue scenario, due to dust interference, there is a deviation in the lip data analysis of the patient (Step 601: The semantic correlation deviation is 20%, and the closed-loop correction gradient parameter of 0.3 is generated). The system dynamically adjusts the lateral acceleration of the print head to 1.0 m / s 2 , and increases the penetration coefficient to 150% in the "allergy history" area (Step 602). The fusion parameter generates a verification instruction (Step 603: Verify once every 0.5 seconds, and the spraying thickness is 1.3 times). After the distributed node returns the complete data of "pollen allergy" (Step 604), update the mapping weight of "flower **" to 0.85, and adjust the subsequent printing trajectory to ensure 100% coverage of the allergy history field in the label.

[0189] Through the coordination of the dynamic closed-loop correction mechanism and the real-time verification instruction, this solution significantly improves the semantic integrity (the coverage rate of fuzzy fields is increased to 98%) and parsing accuracy (the mapping error is reduced by 55%) of medical labels. At the same time, the closed-loop feedback mechanism ensures the dynamic adaptation of printing parameters to lip data (the label reprint rate is reduced by 60%), providing technical support for the reliable collection and efficient management of emergency patient information in disaster environments.

[0190] Figure 2 The following is a schematic structural diagram of a rapid medical record system for emergency patients based on natural language processing provided by an embodiment of the present application. As Figure 2 shown, the device includes:

[0191] A capture module 21, configured to capture the lip movement trajectory data of the patient at the disaster site through an anti-interference lip-reading recognition module. The lip movement trajectory data includes a sequence of mouth shape features under environmental noise interference and the micro-vibration waveform of the lip muscles in the semantic interruption interval;

[0192] An analysis module 22, configured to analyze the micro-vibration waveform of the lip muscles and the sequence of mouth shape features, and generate a set of emergency medical keywords adapted to the disaster scenario. The set of emergency medical keywords includes injury grade identification, allergy history fuzzy semantic segments, and drug taboo abbreviations;

[0193] A receiving module 23, configured to receive the set of emergency medical keywords, and generate a printing path plan for anti-permeation medical labels in combination with the environmental water stain distribution data real-time feedback by the humidity sensor at the disaster site;

[0194] An adjustment module 24, configured to synchronously compare the semantic correlation between the lip movement trajectory data and the set of emergency medical keywords, and dynamically adjust the waterproof printing ink concentration and label adhesive layer thickness parameters;

[0195] An output module 25, configured to input the waterproof printing ink concentration and the label adhesive layer thickness parameters after the anti-permeation medical label printing path planning and adjustment into a disaster emergency filing terminal, and output a waterproof patient wristband label and a ciphertext segment of an emergency file shared in the cloud;

[0196] An update module 26, configured to activate a file completion request of a disaster scene distributed medical node according to the ciphertext segment of the emergency file, and synchronously update the semantic interruption compensation rule of the anti-interference lip reading recognition module.

[0197] Figure 2 The described emergency patient rapid filing device based on natural language processing can execute Figure 1 The emergency patient rapid filing method based on natural language processing described in the illustrated embodiment, and its implementation principle and technical effects will not be elaborated. For the emergency patient rapid filing device based on natural language processing in the above embodiment, the specific manners in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0198] In a possible design, Figure 2 The emergency patient rapid filing device based on natural language processing in the illustrated embodiment can be implemented as a computing device, such as Figure 3 shown, and the computing device may include a storage component 31 and a processing component 32;

[0199] The storage component 31 stores one or more computer instructions, and among them, the one or more computer instructions are called and executed by the processing component 32.

[0200] The processing component 32 is used for the Figure 1 emergency patient rapid filing method based on natural language processing in the above

[0201] embodiment. Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above method.

[0202] The storage component 31 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0203] Of course, the computing device may necessarily further include other components, such as an input / output interface, a display component, a communication component, etc.

[0204] The input / output interface provides an interface between the processing component and the peripheral interface module, and the peripheral interface module may be an output device, an input device, etc.

[0205] The communication component is configured to facilitate communication between the computing device and other devices in a wired or wireless manner, etc.

[0206] Among them, the computing device may be a physical device or an elastic computing host provided by a cloud computing platform, etc. At this time, the computing device may refer to a cloud server, and the above-mentioned processing component, storage component, etc. may be basic server resources leased or purchased from a cloud computing platform.

[0207] The embodiment of the present application also provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the above Figure 1 A method for quickly creating a medical record for emergency patients based on natural language processing shown in the embodiment.

[0208] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0209] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0210] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0211] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for quickly creating files for emergency patients based on natural language processing, characterized in that: include: The lip movement trajectory data of patients at the disaster site are captured by an anti-interference lip reading recognition module, wherein the lip movement trajectory data includes a mouth shape feature sequence under environmental noise interference and a lip muscle micro-vibration waveform in a semantic interruption interval; Analyze the lip muscle micro-vibration waveform and the lip shape feature sequence to generate an emergency medical keyword set adapted to the disaster scene, wherein the emergency medical keyword set includes an injury classification mark, an allergy history fuzzy semantic segment, and a drug contraindication abbreviation; Receiving the emergency medical keyword set, combining it with the environmental water stain distribution data fed back in real time by the humidity sensor at the disaster site, and generating a printing path plan for the anti-penetration medical label; Synchronously comparing the semantic relevance of the lip motion trajectory data with the emergency medical keyword set, and dynamically adjusting the waterproof printing ink concentration and label adhesive layer thickness parameters; Input the anti-penetration medical label printing path planning and the adjusted waterproof printing ink concentration and label adhesive layer thickness parameters into the disaster emergency filing terminal, and output the waterproof patient wristband label and the emergency file ciphertext fragment shared in the cloud; The file completion request of the distributed medical node at the disaster site is activated according to the emergency file ciphertext fragment, and the semantic interruption compensation rule of the anti-interference lip reading recognition module is synchronously updated.

2. The method according to claim 1, characterized in that Synchronously comparing the semantic relevance of the lip motion trajectory data with the emergency medical keyword set, and dynamically adjusting the waterproof printing ink concentration and the label adhesive layer thickness parameters, including: Synchronously comparing the lip shape feature sequence in the lip motion trajectory data with the semantic association of the emergency medical keyword set, and generating a gradient parameter of the waterproof printing ink concentration according to the strength of the semantic association, wherein the gradient parameter is bound to the interruption interval distribution feature of the lip muscle micro-vibration waveform; Based on the coverage of the high humidity area in the environmental water stain distribution data, dynamically adjust the incremental coefficient of the label adhesive layer thickness parameter, wherein the incremental coefficient is associated with the density of the semantic interruption interval in the lip motion trajectory data and increases nonlinearly with the expansion of the high humidity area; Input the gradient parameter of the waterproof printing ink concentration and the incremental coefficient into the anti-penetration medical label printing path planning, generate a print head trajectory path and ink droplet ejection frequency parameter that matches the semantic association, and ensure that the ink concentration gradient parameter covers the label surface corresponding to the high humidity area; The file completion request of the distributed medical node at the disaster site is triggered according to the print head trajectory path, and the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule is updated through the ink droplet ejection frequency parameter to complete the dynamic closed-loop correction of the lip motion trajectory data and the emergency medical keyword set.

3. The method according to claim 2, characterized in that Inputting the gradient parameter of the waterproof printing ink concentration and the incremental coefficient into the anti-penetration medical label printing path planning to generate a print head trajectory path and ink droplet ejection frequency parameters matching the semantic association, including: Based on the interruption interval distribution characteristics in the gradient parameter of the waterproof printing ink concentration, the label surface corresponding to the high humidity area is segmented into anti-penetration areas to generate a sliced ​​humidity resistance level parameter, and the sliced ​​humidity resistance level parameter is inversely correlated with the ejection interval duration in the ink drop ejection frequency parameter; According to the sliced ​​humidity resistance level parameter, dynamically correct the lateral movement speed and the longitudinal pressure compensation value in the print head trajectory path, generate a multimodal print head motion trajectory matching the semantic association, and the multimodal print head motion trajectory covers the distribution density area of ​​the injury classification mark in the emergency medical keyword set; The sliced ​​humidity resistance level parameter and the multi-modal print head motion trajectory are integrated to generate a gradient adhesive layer spraying timing instruction, the gradient adhesive layer spraying timing instruction is synchronously triggered with the nonlinear increasing trend of the incremental coefficient, and is bound to the interruption interval distribution characteristics of the lip muscle micro-vibration waveform; Based on the gradient adhesive layer spraying timing instructions, a print head trajectory path and ink droplet ejection frequency parameters matching the semantic association are generated to trigger a waterproof wristband tag forming verification signal of the distributed medical node at the disaster site, and the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule is updated through the forming verification signal.

4. The method according to claim 2, characterized in that: Triggering a file completion request of the distributed medical node at the disaster site according to the print head trajectory path, and updating the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule through the ink droplet ejection frequency parameter, including: Based on the coverage of the emergency medical keyword set in the print head trajectory path, a file completion trigger instruction is generated, and the file completion trigger instruction is bound to the injection interval duration and semantic interruption interval density in the ink droplet injection frequency parameter, and is associated with the waterproof wristband tag forming verification signal of the distributed medical node at the disaster site; According to the file completion trigger instruction, the priority weight of the injury classification identifier in the emergency medical keyword set is dynamically allocated to generate an update weight of the fuzzy semantic segment mapping relationship, and the update weight is positively correlated with the muscle micro-vibration waveform interruption interval distribution characteristics in the lip movement trajectory data; The file completion trigger instruction and the fuzzy semantic segment mapping relationship are integrated to update the weight, and a semantic interruption compensation rule correction timing instruction is generated, and the correction timing instruction is synchronously adapted to the high humidity area coverage range in the anti-penetration medical label printing path planning; Based on the semantic interruption compensation rule, the timing instruction is corrected to trigger the waterproof wristband tag forming verification signal of the distributed medical node at the disaster site, and the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule is updated through the forming verification signal to complete the dynamic closed-loop correction of the lip motion trajectory data and the emergency medical keyword set.

5. The method according to claim 3, characterized in that: Based on the gradient adhesive layer spraying timing instruction, a print head trajectory path and ink droplet ejection frequency parameters matching the semantic association are generated, a waterproof wristband tag forming verification signal of the distributed medical node at the disaster site is triggered, and the lip muscle micro-vibration waveform matching threshold in the semantic interruption compensation rule is updated through the forming verification signal, including: Based on the lip muscle micro-vibration waveform matching threshold in the gradient adhesive layer spraying timing instruction, a dynamic closed-loop compensation gradient parameter is generated, and the dynamic closed-loop compensation gradient parameter is reversely correlated with the semantic correlation deviation in the waterproof wristband label molding verification signal, and is bound to the regional segmentation boundary of the sliced ​​humidity resistance level parameter; According to the dynamic closed-loop compensation gradient parameter, the ink drop coverage density and the adhesive layer spraying phase difference in the print head trajectory path are adjusted to generate a printing trajectory, and the printing trajectory is synchronously adapted to the semantic interruption interval distribution density of the injury classification mark in the emergency medical keyword set; The dynamic closed-loop compensation gradient parameter and the printing trajectory are integrated to generate a waterproof wristband label forming verification instruction, and the forming verification instruction is bound to the coverage range of the fuzzy semantic segment of the allergy history in the file completion request of the distributed medical node at the disaster site; Based on the waterproof wristband label forming verification instruction, a print head trajectory path and ink droplet ejection frequency parameters matching the semantic association are generated, triggering the lip muscle micro-vibration waveform matching threshold update signal in the semantic interruption compensation rule, and completing the dynamic closed-loop compensation of the gradient adhesive layer spraying timing instruction through the update signal.

6. The method according to claim 3, characterized in that The sliced ​​humidity resistance level parameters and the multi-modal print head motion trajectory are integrated to generate gradient adhesive layer spraying timing instructions, including: Based on the regional segmentation boundary in the slice-type humidity resistance level parameter and the lateral movement speed of the multi-modal print head motion trajectory, an adhesive layer spraying phase difference parameter is generated, and the adhesive layer spraying phase difference parameter is synchronously bound to the nonlinear increasing trend of the incremental coefficient and the interruption interval distribution density of the lip muscle micro-vibration waveform; Generate a gradient adhesive layer spraying timing instruction according to the adhesive layer spraying phase difference parameter and the longitudinal pressure compensation value of the multimodal print head motion trajectory, wherein the gradient adhesive layer spraying timing instruction includes dynamic adjustment parameters of spraying interval and spraying thickness, and matches the high humidity area coverage range in the anti-penetration medical label printing path planning; The gradient adhesive layer spraying timing instruction and the regional segmentation boundary of the sliced ​​humidity resistance level parameter are integrated to generate a closed-loop verification instruction for waterproof wristband label molding, and the closed-loop verification instruction is positively correlated with the semantic interruption interval distribution density of the injury classification mark in the emergency medical keyword set; Based on the waterproof wristband label forming closed-loop verification instruction, the waterproof wristband label forming verification signal of the distributed medical node at the disaster site is triggered, and the fuzzy semantic segment mapping priority in the semantic interruption compensation rule is updated through the verification signal to complete the dynamic closed-loop binding of the gradient adhesive layer spraying timing instruction and the lip motion trajectory data.

7. The method according to claim 4, characterized in that Based on the semantic interruption compensation rule, the timing instruction is corrected, a waterproof wristband tag forming verification signal of the distributed medical node at the disaster site is triggered, and the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule is updated through the forming verification signal, including: Based on the semantic interruption compensation rule, the fuzzy semantic segment mapping relationship in the timing instruction is corrected and the weight is updated to generate a closed-loop correction gradient parameter. The closed-loop correction gradient parameter is reversely bound to the semantic association deviation and the injury classification mark distribution density in the waterproof wristband label forming verification signal, and is associated with the segmentation accuracy of the spraying area boundary coordinates; According to the closed-loop correction gradient parameter, the lateral movement acceleration and the longitudinal ink drop penetration compensation coefficient in the dynamic trajectory of the print head are dynamically adjusted to generate a semantically associated closed-loop printing trajectory, and the semantically associated closed-loop printing trajectory is synchronously adapted to the coverage of the fuzzy semantic segment of the allergy history in the emergency medical keyword set; The closed-loop correction gradient parameter and the semantically associated closed-loop printing trajectory are integrated to generate a waterproof wristband label forming verification instruction, and the verification instruction is synchronously bound to the spraying interval and spraying thickness dynamic adjustment parameters in the anti-penetration adhesive layer spraying phase instruction; Based on the waterproof wristband label forming verification instruction, the waterproof wristband label forming closed-loop feedback signal of the distributed medical node at the disaster site is triggered, and the fuzzy semantic segment mapping relationship in the semantic interruption compensation rule is updated through the closed-loop feedback signal to complete the dynamic closed-loop correction of the lip motion trajectory data and the emergency medical keyword set.

8. A rapid file creation system for emergency patients based on natural language processing, characterized in that: include: A capture module is used to capture the lip movement trajectory data of patients at the disaster site through an anti-interference lip reading recognition module, wherein the lip movement trajectory data includes a lip shape feature sequence under environmental noise interference and a lip muscle micro-vibration waveform in a semantic interruption interval; A parsing module, used to parse the lip muscle micro-vibration waveform and the lip shape feature sequence, and generate an emergency medical keyword set adapted to the disaster scene, wherein the emergency medical keyword set includes an injury classification mark, an allergy history fuzzy semantic segment, and a drug contraindication abbreviation; A receiving module, configured to receive the emergency medical keyword set, generate an anti-penetration medical label printing path plan in combination with the environmental water stain distribution data fed back in real time by the humidity sensor at the disaster site; An adjustment module, used for synchronously comparing the semantic relevance of the lip motion trajectory data with the emergency medical keyword set, and dynamically adjusting the waterproof printing ink concentration and the label adhesive layer thickness parameters; An output module is used to input the anti-penetration medical label printing path planning and the adjusted waterproof printing ink concentration and label adhesive layer thickness parameters into the disaster emergency filing terminal, and output the waterproof patient wristband label and the emergency file ciphertext fragment shared in the cloud; The updating module is used to activate the file completion request of the distributed medical node at the disaster site according to the emergency file ciphertext fragment, and synchronously update the semantic interruption compensation rule of the anti-interference lip reading recognition module.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a method for quickly creating files for emergency patients based on natural language processing as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, a method for quickly creating files for emergency patients based on natural language processing as described in any one of claims 1 to 7 is implemented.

Images