Medicine distribution charging system for anesthesiology department
By designing an anesthetic drug delivery charging system including drug image acquisition, text profile feature comparison and distribution cost retrieval module, the problem of manual accounting of delivery costs in high frequency and complex operating scenarios of the existing system is solved, and the accurate identification of drugs and distribution cost accounting is achieved, and management efficiency and accuracy are improved.
Patent Information
- Application Number
- CN202510212585.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-25
- Publication Date
- 2025-06-06
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the high frequency and complex operation scenarios, the existing anesthesia drug delivery charging system is prone to errors in manual accounting of delivery costs and is inefficient, resulting in confusion in management.
Design an anesthetic drug delivery charging system, including a drug image acquisition module, a drug text profile feature comparison module and a delivery fee acquisition module. Drug images are collected through the camera, text feature extraction method is used to extract text outline features, and drugs are identified through similarity calculations. Finally, they interact with the hospital's internal information management system through an integrated interface to retrieve the delivery price.
It realizes accurate identification of drugs and distribution cost accounting, overcomes the error problems that are prone to manual identification, and improves the efficiency and accuracy of drug management in complex scenarios.
Smart Images
Figure CN120108670A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of medical systems, and in particular to an anesthesia drug distribution and charging system. Background Art
[0002] Anesthesiology drugs are an important part of hospital drug costs, and their procurement, inventory management, and distribution are directly related to capital circulation and cost expenditure. Scientific and accurate charge management can not only clearly reflect the actual use and cost consumption of drugs, but also provide hospitals with accurate cost accounting basis, thereby helping to optimize management decisions.
[0003] At present, the drug distribution and charging system of the Department of Anesthesiology usually adopts the traditional manual accounting model. Specifically, medical staff will set up a table for drug entry and exit in advance to record in detail the name, delivery price and related key information of various drugs. During the drug entry and exit operations, medical staff classify and store the drugs according to the contents of the table, and calculate the delivery costs according to the records. However, when faced with a large number of drugs that need to be taken and placed at the same time, the limitations of the manual accounting model gradually emerge. On the one hand, human eye recognition and manual operation are prone to errors, resulting in deviations in the calculation of delivery costs; on the other hand, large-scale drug operations place higher requirements on manual efficiency, which can easily cause delays or management chaos.
[0004] In view of the above problems, the existing anesthesia drug distribution and charging system has shown obvious shortcomings when dealing with high-frequency and complex operation scenarios, and is in urgent need of improvement and optimization. Summary of the invention
[0005] The purpose of the present invention is to solve the problem of inaccurate calculation of distribution costs.
[0006] To achieve the above-mentioned purpose, the present invention provides an anesthesia medicine delivery and charging system, comprising a medicine image acquisition module, a medicine text contour feature comparison module and a delivery fee retrieval module; The drug image acquisition module uses the principle of camera optical imaging to capture images containing drugs to be stored, and establishes a storage record form. The table name corresponding to the storage record form is the drug information provided by the drug supplier, the drug information defined by the medical staff, and the drug number.
[0007] The drug text contour feature comparison module uses a text feature extraction method to extract the text contour features corresponding to the drug image in the drug image acquisition module and the drug picture in the storage record form, and calculates the similarity between different text contours. The drug image corresponding to the text contour feature with the highest similarity is determined to be the drug in the storage record form; When corresponding drug information exists in each drug category under the table name of the in-and-out record table, the existing drug information is defined as reference information. While comparing the drug images in the in-and-out record table, the text contour features in the reference information are also double-compared with the text contour features of the drug image in the drug image acquisition module. The product image with the highest similarity is the drug with the same drug image in the drug image acquisition module. The delivery cost retrieval module establishes interaction with the hospital's internal information management system through an integrated interface, senses the drug delivery request information in the hospital's internal information management system, and calculates the text contour features of the drug name in the drug delivery request information through the drug text contour feature comparison module in turn, compares it with the text contour features of the name in the inventory record table, calls out the drug information corresponding to the drug name in the inventory record table, outputs the drug number reminder signal in the drug information to medical staff through the integrated interface, and outputs the delivery price in the inventory record table to the required drug department.
[0008] The lens of the camera in the drug image acquisition module is similar to the human eye lens. When light irradiates the drug, the lens collects and focuses the light. According to the principle of geometric optics, the light forms an inverted real image on the imaging sensor after passing through the lens. When the photodiode on the surface of the imaging sensor is irradiated by light, the valence band electrons in the photodiode absorb the energy of the photon and then transition to the conduction band, generating electron-hole pairs and forming current. The analog signal generated by the imaging sensor is converted from analog to digital to restore the image containing the drug to be stored.
[0009] The analog-to-digital conversion in the drug image acquisition module is implemented based on an ADC converter. The ADC converter samples the analog electrical signal at a certain sampling frequency, and converts the sampled analog signal amplitude value into a corresponding digital code according to a set quantization accuracy.
[0010] The text feature extraction method in the drug text contour feature comparison module removes noise in the drug image through Gaussian filtering technology, calculates gradient amplitude and direction, and refines edges through non-maximum suppression, retains pixels with the largest amplitude in the gradient direction as edge pixels, and finally detects connected edges with double thresholds and removes false edges to obtain complete text contour features.
[0011] The drug text contour feature comparison module removes the noise in the drug image captured by the drug image acquisition module, using a Gaussian kernel (in is the standard deviation, is the position of the pixel in the image) and convolved with the drug image to remove noise; The edge of the text is where the grayscale value of the pixel in the drug image changes sharply. The intensity can be measured by the gradient amplitude. The gradient direction indicates the direction in which the grayscale value changes fastest, that is, the direction of the text edge. The gradient of the image in the horizontal and vertical directions is calculated by two convolution kernels respectively. The horizontal convolution kernel is , the vertical convolution kernel is ; For each pixel in the drug image , its horizontal gradient By With The vertical gradient is obtained by convolution operation on the neighborhood pixels centered at Similarly; So the pixel The gradient magnitude , gradient direction ; Divide the neighborhood of a pixel into two parts according to its gradient direction (for example, for a horizontal edge, compare the upper and lower neighborhoods; for a 45° edge, compare the diagonal neighborhoods). If the gradient magnitude of a pixel is greater than the gradient magnitude of the pixels in the two neighborhoods, then the pixel is retained as an edge candidate, otherwise it is excluded. Set similarity threshold interval ,in , if the gradient amplitude of the pixel point> Then the pixel is determined to be an edge pixel; if the gradient amplitude of the pixel is less than , it is determined as a non-edge pixel; if <Gradient amplitude of pixel point< When , if the pixel is adjacent to the determined edge pixel, it will also be determined as an edge pixel, otherwise it will be excluded.
[0012] After the drug text contour feature comparison module extracts the text contour features of the drug image captured by the drug image acquisition module, the text contour features of the drug image in the storage table are extracted in the same manner; Starting from the same starting point of the text contour feature, traverse the contour pixels in a clockwise direction. When a pixel has only one adjacent contour pixel, it is determined as an endpoint. When a pixel has multiple adjacent contour pixels, it is determined as an intersection. After traversing the entire text contour in sequence, the number of endpoints and intersections of different text contour features are recorded respectively, and the ratio of the number of endpoints and the ratio of the number of intersections are calculated. Then the average value is taken as the similarity. The text contour feature with the highest similarity is the drug corresponding to the drug picture taken by the drug image acquisition module.
[0013] The calculation formula for the similarity between the text contour features in the drug text contour feature comparison module is: ; in and are the number of endpoints of the text contour of the drug image to be stored and the drug matching template, and They are the number of intersection points of the text outlines.
[0014] The integrated interface in the delivery fee retrieval module outputs a connection request through a login account in the hospital's internal information management system; Call the API function that matches the hospital's internal information management system and send the login account to the hospital's internal information management system; After receiving the login account, the hospital's internal information management system authenticates the login account according to its own verification mechanism; When the login account meets the verification mechanism of the hospital's internal information management system, the hospital's internal information management system sends a connection success response to the delivery fee retrieval module, indicating that the connection is established successfully, otherwise, it means that the connection fails to be established; When the connection is successful, the information in the hospital's internal information management system is obtained through the established connection.
[0015] Compared with the prior art, the present invention has the following beneficial effects: In the drug distribution and charging system of the anesthesiology department, drug images are collected by a drug image collection module, and an in-and-out record form is established. The table name of the in-and-out record form includes drug information provided by the drug supplier, drug information defined by medical staff, and drug numbers. Then, the drug text contour feature comparison module extracts the drug image and the text contour features of the drug image in the in-and-out record form by a text feature extraction method, distinguishes different drugs according to the text contour features, and then uses the distinguished drug image as a reference for separation again. After the drugs are distinguished, double authentication is performed, and then the distribution fee retrieval module calls out the distribution price and outputs it to the medical staff. By comparing features and using the accuracy of computer algorithms, the error problem that is prone to occur in manual identification is overcome, and drugs can be distinguished more accurately, especially in complex scenarios where a large number of drugs are taken, placed, and processed at the same time, each drug can still be quickly and accurately identified to ensure the accuracy of drug charges. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 It is the overall module principle diagram of the present invention.
[0017] The meaning of each number in the figure is: 100. Medicine image acquisition module; 200. Medicine text contour feature comparison module; 300. Delivery cost retrieval module. DETAILED DESCRIPTION
[0018] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0019] An anesthesiology drug delivery and charging system includes a drug image acquisition module 100, a drug text contour feature comparison module 200, and a delivery fee retrieval module 300; The drug image acquisition module 100 uses the principle of camera optical imaging to capture images containing drugs to be stored. The lens of the camera is similar to the lens of the human eye. When light irradiates the drugs (the light is a natural light source, an artificial light source, or other light sources), the lens collects and focuses the light. According to the principle of geometric optics, the light forms an inverted real image on the imaging sensor after passing through the lens. When the photodiode on the surface of the imaging sensor is irradiated by light, the valence band electrons in the photodiode absorb the energy of the photon and then transition to the conduction band, generating electron-hole pairs and forming current (the reflected light intensity of different areas on the drug packaging is different. The brighter area reflects a larger light intensity, and the number of electron-hole pairs generated by irradiating the photodiode is larger, so the current signal formed is stronger; while the darker area reflects weaker light, and the number of electron-hole pairs generated is small, so the signal is also weaker). The analog signal generated by the imaging sensor is converted from analog to digital by the ADC converter to restore the image containing the drugs to be stored, so as to subsequently register the entry and exit of the drugs. The drug image acquisition module 100 converts the analog signal into a digital signal through an ADC converter. The ADC converter samples the analog electrical signal at a certain sampling frequency (e.g., thousands of times per second or even higher, depending on the performance and settings of the camera), and converts the amplitude value of the sampled analog signal into a corresponding digital code according to the set quantization accuracy (e.g., the commonly used 8-bit, 10-bit, 12-bit, etc., indicating that the amplitude range of the analog signal is divided into different discrete levels). The calculation formula is as follows: ; in, Represents the converted digital code, is the analog signal voltage value input to the ADC converter, It is a reference voltage used to determine the quantization interval corresponding to the amplitude range of the analog signal. It sets the maximum voltage range that can be referenced during the conversion of the entire analog signal.
[0020] The drug image acquisition module 100 is also used to establish an entry and exit record form. The table name corresponding to the entry and exit record form is the drug information provided by the drug supplier (including drug pictures, drug instructions, drug names on drug pictures, production dates, etc.), medical staff-defined drug information (including drug inventory quantity, drug selling price, delivery price, etc.) and drug numbers. When drugs are entered and exited in the entry and exit record form, the relevant records can be accurately updated to achieve real-time tracking and management of dynamic changes in drug inventory; the drug charges of the anesthesia department are often closely related to the specific types, specifications, dosage forms, etc. of the drugs. Using the above-mentioned correlation comparison method, the drugs can be accurately identified, so as to accurately calculate the fees based on the pre-set price system (related to the corresponding selling price and other information in the entry and exit record form), avoid charging errors caused by drug identification errors, and not only ensure the reasonable income of the hospital, but also prevent patients from being overcharged or undercharged.
[0021] The drug text contour feature comparison module 200 uses a text feature extraction method to extract the text contour features corresponding to the drug image in the drug image acquisition module 100 and the drug picture in the inventory record form, and calculates the similarity between different text contours. The drug image corresponding to the text contour feature with the highest similarity is determined to be the drug in the inventory record form.
[0022] When corresponding drug information exists in each type of drug under the table name of the inventory record table, the existing drug information is defined as reference information. While comparing the drug pictures in the inventory record table, the text contour features in the reference information are also double-compared with the text contour features of the drug image in the drug image acquisition module 100. The product picture with the highest similarity is the drug with the same drug image in the drug image acquisition module 100. The safety of drug use in the anesthesia department is extremely important. Once the wrong drug is used, it may cause a serious threat to the patient's life and health. Through strict double comparison, the delivery of drugs is accurate, which can avoid drug safety accidents caused by drug confusion from the source, so that anesthesiologists and other medical staff can use the correct drugs delivered to carry out anesthesia work with confidence, and ensure the safety of patients' medication during surgery or pain management. The text feature extraction method uses Gaussian filtering technology to remove noise from drug images, calculate gradient amplitude and direction, and use non-maximum suppression to refine edges. The pixels with the largest amplitude in the gradient direction are retained as edge pixels. Finally, double threshold detection connects edges and removes false edges to obtain complete text contour features. In order to remove the noise in the drug image captured by the drug image acquisition module 100, a Gaussian kernel is used. (in is the standard deviation, is the position of the pixel in the image) and convolved with the drug image to remove noise; The edge of the text is where the grayscale value of the pixel in the drug image changes sharply. The intensity can be measured by the gradient amplitude. The gradient direction indicates the direction in which the grayscale value changes fastest, that is, the direction of the text edge. The gradient of the image in the horizontal and vertical directions is calculated by two convolution kernels respectively. The horizontal convolution kernel is , the vertical convolution kernel is ; For each pixel in the drug image , its horizontal gradient By With The vertical gradient is obtained by convolution operation on the neighborhood pixels centered at Similarly; So the pixel The gradient magnitude , gradient direction ; After calculating the gradient amplitude, the edge obtained is relatively coarse, and there will be multiple pixels representing the same edge. The purpose of non-maximum suppression is to retain only the pixels with the local maximum amplitude in the gradient direction as edge candidates, thereby refining the edge; Divide the neighborhood of a pixel into two parts according to its gradient direction. For example, for a horizontal edge, compare the upper and lower neighborhoods; for a 45° edge, compare the oblique neighborhoods. If the gradient amplitude of a pixel is greater than the gradient amplitude of the pixels in the two neighborhoods, then the pixel is retained as an edge candidate, otherwise it is excluded. After non-maximum suppression, there may still be some false edge candidate points caused by noise or other factors. Set the similarity threshold interval ,in , if the gradient amplitude of the pixel point> Then the pixel is determined to be an edge pixel; if the gradient amplitude of the pixel is less than , it is determined as a non-edge pixel; if <Gradient amplitude of pixel point< When , if the pixel is adjacent to the determined edge pixel, it will also be determined as an edge pixel, otherwise it will be excluded.
[0023] After the medicine text contour feature comparison module 200 extracts the text contour features of the medicine image captured by the medicine image acquisition module 100, the text contour features of the medicine image in the storage table are extracted in the same manner; Synchronously starting from the same starting point of the text contour feature, traverse the contour pixels in a clockwise direction. When a pixel has only one adjacent contour pixel, it is determined as an endpoint. When a pixel has multiple (greater than 2) adjacent contour pixels, it is determined as an intersection. After traversing the entire text contour in sequence, the number of endpoints and intersections of different text contour features are recorded respectively, and the ratio of the number of endpoints and the ratio of the number of intersections are calculated. Then, the average value is taken as the similarity. The text contour feature with the highest similarity is the drug corresponding to the drug picture taken by the drug image acquisition module 100. The calculation formula for the similarity between the above text contour features is: ; in and are the number of endpoints of the text contour of the drug image to be stored and the drug matching template, and They are the number of intersection points of the text outlines.
[0024] With the continuous advancement of medical technology, such as the development of emerging fields such as telemedicine and smart medicine, hospitals usually establish a complete internal information management system to connect various departments (including internal medicine, surgery, obstetrics and gynecology, anesthesiology, surgery, etc.). The anesthesia department can submit drug delivery requests through the hospital's internal information management system, detailing the name, specifications, quantity and other information of the required drugs; The distribution cost retrieval module 300 establishes interaction with the hospital's internal information management system through an integrated interface, senses the drug distribution request information (drug names, quantities, and other information required by different departments) in the hospital's internal information management system, and calculates the drug name text contour features in the drug distribution request information through the drug text contour feature comparison module 200, compares it with the name text contour features of the entry and exit record table, retrieves the drug information corresponding to the drug name in the entry and exit record table (including but not limited to drug instructions, drug name, and drug number), outputs the drug number reminder signal in the drug information to the medical staff through the integrated interface, and outputs the distribution price in the entry and exit record table to the required drug department; The delivery fee retrieval module 300 outputs a connection request through a login account in the hospital's internal information management system; Call the API function that matches the hospital's internal information management system and send the login account to the hospital's internal information management system; After receiving the login account, the hospital's internal information management system authenticates the login account according to its own verification mechanism; When the login account meets the verification mechanism of the hospital internal information management system, the hospital internal information management system sends a connection success response to the delivery fee retrieval module 300, indicating that the connection is established successfully, otherwise, it means that the connection fails to be established; When the connection is successful, the information in the hospital's internal information management system is obtained through the established connection.
[0025] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.
Claims
1. An anesthesia drug distribution charging system, characterized in that: It comprises a medicine image acquisition module (100), a medicine text contour feature comparison module (200) and a distribution fee retrieval module (300); The drug image acquisition module (100) uses the principle of camera optical imaging to capture images containing drugs to be stored, and creates a storage entry and exit record form, wherein the table names corresponding to the storage entry and exit record form are drug information provided by the drug supplier, drug information defined by medical staff, and drug numbers; The drug text contour feature comparison module (200) uses a text feature extraction method to extract text contour features corresponding to the drug image in the drug image acquisition module (100) and the drug image in the storage record table, and calculates the similarity between different text contours. The drug image corresponding to the text contour feature with the highest similarity is determined to be the drug in the storage record table; When corresponding drug information exists in each drug category under the table name of the inventory record table, the existing drug information is defined as reference information. While comparing the drug image in the inventory record table, the text contour features in the reference information are also double-compared with the text contour features of the drug image in the drug image acquisition module (100). The product image with the highest similarity is the drug with the same drug image in the drug image acquisition module (100); The delivery cost retrieval module (300) establishes interaction with the hospital's internal information management system through an integrated interface, senses the drug delivery request information in the hospital's internal information management system, and calculates the drug name text contour features in the drug delivery request information through the drug text contour feature comparison module (200), compares it with the name text contour features of the entry and exit record table, retrieves the drug information corresponding to the drug name in the entry and exit record table, outputs a drug number reminder signal in the drug information to medical staff through the integrated interface, and outputs the delivery price in the entry and exit record table to the required drug department.
2. The anesthesia drug distribution charging system according to claim 1 is characterized by: The lens of the camera in the drug image acquisition module (100) is similar to the lens of a human eye. When light is irradiated on the drug, the lens collects and focuses the light. According to the principle of geometric optics, after the light passes through the lens, an inverted real image is formed on the imaging sensor. When the photodiode on the surface of the imaging sensor is irradiated with light, the valence band electrons in the photodiode absorb the energy of the photon and then transition to the conduction band, generating electron-hole pairs and forming current. The analog signal generated by the imaging sensor is converted from analog to digital to restore the image containing the drug to be stored.
3. The anesthesia drug distribution charging system according to claim 1 is characterized by: The analog-to-digital conversion in the drug image acquisition module (100) is implemented based on an ADC converter. The ADC converter samples the analog electrical signal at a certain sampling frequency, and converts the amplitude value of the sampled analog signal into a corresponding digital code according to a set quantization accuracy.
4. The anesthesia drug distribution charging system according to claim 1 is characterized in that: The text feature extraction method in the drug text contour feature comparison module (200) uses Gaussian filtering technology to remove noise in the drug image, calculate the gradient amplitude and direction, and use non-maximum suppression to refine the edge, retaining the pixel with the largest amplitude in the gradient direction as the edge pixel, and finally double-threshold detection to connect the edge and remove the false edge to obtain a complete text contour feature.
5. The anesthesia drug distribution charging system according to claim 4 is characterized by: The drug text contour feature comparison module (200) removes noise in the drug image captured by the drug image acquisition module (100) by using a Gaussian kernel (in is the standard deviation, is the position of the pixel in the image) and convolved with the drug image to remove noise; The edge of the text is where the grayscale value of the pixel in the drug image changes sharply. The intensity can be measured by the gradient amplitude. The gradient direction indicates the direction in which the grayscale value changes fastest, that is, the direction of the text edge. The gradient of the image in the horizontal and vertical directions is calculated by two convolution kernels respectively. The horizontal convolution kernel is , the vertical convolution kernel is ; For each pixel in the drug image , its horizontal gradient By With The vertical gradient is obtained by convolution operation on the neighborhood pixels centered at Similarly; So the pixel The gradient magnitude , gradient direction ; Divide the neighborhood of a pixel into two parts according to its gradient direction (for example, for a horizontal edge, compare the upper and lower neighborhoods; for a 45° edge, compare the diagonal neighborhoods). If the gradient magnitude of a pixel is greater than the gradient magnitude of the pixels in the two neighborhoods, then the pixel is retained as an edge candidate, otherwise it is excluded. Set similarity threshold interval ,in , if the gradient amplitude of the pixel point> Then the pixel is determined to be an edge pixel; if the gradient amplitude of the pixel is less than , it is determined as a non-edge pixel; if <Gradient amplitude of pixel point< When , if the pixel is adjacent to the determined edge pixel, it will also be determined as an edge pixel, otherwise it will be excluded.
6. The anesthesia drug distribution charging system according to claim 5 is characterized by: After the medicine text contour feature comparison module (200) extracts the text contour features of the medicine image captured by the medicine image acquisition module (100), the text contour features of the medicine image in the storage table are extracted in the same manner; Synchronously starting from the same starting point of the text contour feature, traversing the contour pixels in a clockwise direction, when a pixel point has only one adjacent contour pixel point, it is determined as an endpoint, and when a pixel point has multiple adjacent contour pixels, it is determined as an intersection point, after traversing the entire text contour in sequence, the number of endpoints and intersection points of different text contour features are recorded respectively, and the ratio of the number of endpoints and the ratio of the number of intersection points are calculated, and then the average value is taken as the similarity. The text contour feature with the highest similarity is the drug corresponding to the drug picture taken by the drug image acquisition module (100).
7. The anesthesia drug distribution charging system according to claim 6 is characterized by: The calculation formula for the similarity between the character contour features in the drug character contour feature comparison module (200) is: ； in and are the number of endpoints of the text contour of the drug image to be stored and the drug matching template, and They are the number of intersection points of the text outlines.
8. The anesthesia drug distribution charging system according to claim 1 is characterized by: The integrated interface in the delivery fee retrieval module (300) outputs a connection request through a login account in the hospital's internal information management system; Call the API function that matches the hospital's internal information management system and send the login account to the hospital's internal information management system; After receiving the login account, the hospital's internal information management system authenticates the login account according to its own verification mechanism; When the login account meets the verification mechanism of the hospital internal information management system, the hospital internal information management system sends a connection success response to the delivery fee retrieval module (300), indicating that the connection is established successfully, otherwise, it means that the connection fails to be established; When the connection is successful, the information in the hospital's internal information management system is obtained through the established connection.