Intelligent infusion monitoring system

Abstract

The invention relates to an intelligent infusion monitoring system, which relates to the technical field of computer processing, and comprises an image acquisition and feature extraction module for acquiring a first image and a second image, the first image comprises a doctor's advice bar code on a patient wrist strap or an infusion sheet, and the second image comprises a packaging bar code of an infusion bag / bottle or a packaging box; the data preprocessing module is used for positioning a doctor's advice bar code area in the first image through a pre-trained first segmentation model, the first segmentation model is trained based on a medical special bar code data set, and training data comprises bar code labels in light reflection and wrinkle scenes; and the key data extraction module is used for positioning a package bar code area in the second image through a second segmentation model, and the second segmentation model is optimized for the curved surface package and comprises an enhanced training sample of a 45-60-degree shooting angle of the curved surface. According to the invention, information of medical advice strips and infusion liquid is automatically acquired and checked through an image recognition technology, so that the safety and accuracy of the infusion process are ensured.

Description

Technical Field

[0001] The present invention relates to the field of computer processing technology, and in particular to an intelligent infusion monitoring system. Background Art

[0002] Infusion therapy is a common treatment option. However, due to human or system factors, the doctor's instructions may not match the actual infusion, or the infusion may be expired without prompt detection. These issues can seriously impact patient treatment outcomes and safety.

[0003] Traditional methods rely primarily on visual inspection and manual verification by medical staff, but this approach is inefficient and prone to errors. With the widespread application of intelligent recognition technologies and the proliferation of smart devices, such as QR code recognition and image recognition, the application of these technologies in the medical field can effectively address the drawbacks of manual screening. By using intelligent recognition technology to quickly identify discrepancies between infusions or medications and doctor's orders, nurses can be alerted to prevent medical errors. Summary of the Invention

[0004] In response to the above technical problems, the technical solution adopted by the present invention is an intelligent infusion monitoring system, comprising:

[0005] An image acquisition and feature extraction module acquires a first image and a second image, wherein the first image includes a medical order barcode on a patient's wristband or an infusion sheet, and the second image includes a packaging barcode on an infusion bag / bottle or a packaging box;

[0006] a data preprocessing module for locating the medical order barcode region in the first image using a pre-trained first segmentation model, wherein the first segmentation model is trained based on a medical barcode dataset, wherein the training data includes barcode annotations in reflective and wrinkled scenes;

[0007] a key data extraction module that locates the package barcode region in the second image using a second segmentation model optimized for curved packaging and including enhanced training samples for curved surfaces captured at angles between 45 and 60 degrees;

[0008] The identification module identifies the medical order coding information and extracts at least the standardized drug name field and dosage specification field therefrom; identifies the liquid coding information and extracts at least the drug name, specification parameters and expiration date data therefrom.

[0009] The information comparison and identification module includes performing the following steps on the information extracted by the identification module:

[0010] S01. Normalize the drug names and convert the prescription names and packaging names into standard pharmaceutical dictionary terms;

[0011] S02, adopts the specification parameter fuzzy matching algorithm, allowing ±5% dose deviation;

[0012] S03. When the name normalization comparison is inconsistent or the specification deviation exceeds the limit, the sound and light alarm is triggered and the infusion pump is locked and started;

[0013] S04. Calculate the difference between the current time and the expiration date. If the remaining validity period is less than the estimated infusion time × 1.2 safety factor, generate an infusion speed recommendation note.

[0014] Preferably, the image acquisition and feature extraction module includes a shooting and recording device, which performs multi-angle image acquisition on the medical order barcode on the patient's wristband or infusion order and the packaging barcode on the infusion bag / bottle or packaging box.

[0015] Preferably, the positioning of the first segmentation model and the second segmentation model is obtained by training with a labeled data set including various types of medical barcodes, and the labeled data includes barcode area contour annotations.

[0016] Preferably, the recognition module includes a calibration unit, which uses fast Fourier transform to extract the spatial frequency characteristics of the barcode and eliminates harmonic interference caused by deformation of the packaging surface through frequency domain filtering.

[0017] Preferably, the calibration unit performs the following steps on the calibration of the QR code:

[0018] S100, converting the detected barcode area into a grayscale image;

[0019] S101, performing a two-dimensional FFT transform on the image to obtain a spectrum graph;

[0020] S102. Design a bandpass filter to retain the fundamental frequency component corresponding to the barcode module width. The filter passband is [1 / (3W), 1 / (0.8W)], where W is the estimated width of a single barcode module.

[0021] S103 , performing inverse FFT transformation on the filtered spectrum to obtain a corrected image.

[0022] Preferably, the data preprocessing module includes an error reporting unit, which, when the coded information cannot be identified: extracts the visual features of the liquid packaging, performs similarity matching through the drug feature database, obtains candidate drug information, and then feeds back to the nurse for confirmation to determine the final drug information.

[0023] Preferably, forming the infusion speed suggestion in S04 includes the following processing steps:

[0024] S200, calculate the remaining effective time T based on the drug validity period data and the current time r ;

[0025] S201 obtains the current infusion time T set by the infusion pump i , if T r <K×T i , then recalculate the maximum allowable infusion rate V:

[0026] V=(m×Tr) / (K×T i );

[0027] Where m is the total volume of the drug, and K is a safety factor of 1.1-1.3;

[0028] S203: Generate an early warning signal containing V and send it to the infusion pump control terminal.

[0029] The present invention has at least the following beneficial effects:

[0030] 1. Through the two-way cross-verification of the doctor's order barcode and the liquid packaging code, the automatic verification of the drug name and specification is realized, reducing the manual verification error

[0031] 2. Real-time analysis of the expiration date data in the liquid code, combined with the infusion time to calculate the remaining validity safety threshold, effectively preventing the use of expired drugs.

[0032] 3. Adopting multi-angle image acquisition combined with surface optimization segmentation model to solve the reflection and wrinkle problems of liquid packaging and improve the recognition rate;

[0033] 4. When the barcode is damaged, the visual feature matching process is initiated, and combined with the drug feature database to achieve a recall rate of more than 90% for similar drugs. By introducing the nurse confirmation link, a human-machine collaborative verification closed loop is formed. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0035] Figure 1 This is a module diagram of an intelligent infusion monitoring system provided in Example 1 of the present invention;

[0036] Figure 2 This is a flowchart of the information comparison and identification module provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0037] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0038] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way are interchangeable where appropriate so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0039] Example 1

[0040] This embodiment provides an intelligent infusion monitoring system. Figure 1-2 Shown: Includes:

[0041] An image acquisition and feature extraction module acquires a first image and a second image, wherein the first image includes the medical order barcode on the patient's wristband or infusion sheet, and the second image includes the packaging barcode of the infusion bag / bottle or packaging box;

[0042] The image acquisition and feature extraction module includes a shooting and input device, which performs multi-angle image acquisition on the medical order barcode on the patient's wristband or infusion order and the packaging barcode on the infusion bag / bottle or packaging box.

[0043] Specifically, the recording device consists of two parts: a barcode recognition module and a high-definition camera acquisition module. These two components work together. If the barcode recognition module fails to recognize a scanned QR code, the HD camera acquisition module is activated to capture the QR code. By repairing the QR code image, a clear QR code is provided that is sufficiently recognizable by the barcode recognition module, ensuring its usability. Furthermore, multi-angle image acquisition improves the accuracy and robustness of barcode recognition, especially in complex scenarios where the barcode contains reflections, wrinkles, or curved surface deformations, ensuring complete acquisition and accurate recognition of the barcode information.

[0044] A data preprocessing module locates the medical order barcode region in the first image using a pre-trained first segmentation model, where the first segmentation model is trained based on a medical barcode dataset, the training data including barcode annotations in reflective and wrinkled scenes;

[0045] The above-mentioned data preprocessing module includes an error reporting unit, which, when the coded information cannot be identified: extracts the visual features of the liquid packaging, performs similarity matching through the drug feature database, obtains candidate drug information, and then feeds back to the nurse for confirmation to determine the final drug information.

[0046] Specifically, the error reporting unit provides an alternative solution when the barcode cannot be recognized. Through visual feature matching and similarity calculation, it can obtain possible drug information, provide a reference for medical staff, and reduce the risk of errors caused by the unrecognizable barcode.

[0047] The key data extraction module locates the package barcode region in the second image using a second segmentation model optimized for curved packaging and including enhanced training samples for curved surfaces captured at angles between 45 and 60 degrees.

[0048] The identification module identifies the medical order coding information and extracts at least the standardized drug name field and dosage specification field therefrom; identifies the liquid coding information and extracts at least the drug name, specification parameters and expiration date data therefrom.

[0049] The above-mentioned recognition module includes a calibration unit, which uses fast Fourier transform to extract the spatial frequency characteristics of the barcode and eliminates harmonic interference caused by deformation of the packaging surface through frequency domain filtering.

[0050] It can achieve accurate correction of QR codes, eliminate image distortion and blur caused by shooting angle, light and other factors, thereby improving the accuracy and reliability of QR code recognition.

[0051] The calibration unit's calibration process for the QR code includes:

[0052] S100, converting the detected barcode area into a grayscale image;

[0053] S101, performing a two-dimensional FFT transform on the image to obtain a spectrum graph;

[0054] S102. Design a bandpass filter to retain the fundamental frequency component corresponding to the barcode module width. The filter passband is [1 / (3W), 1 / (0.8W)], where W is the estimated width of a single barcode module.

[0055] S103 , performing inverse FFT transformation on the filtered spectrum to obtain a corrected image.

[0056] Specifically, the calibration unit eliminates harmonic interference caused by curved packaging surfaces, improving the accuracy and stability of barcode recognition. This is particularly useful for barcode recognition on curved packaging, ensuring accurate extraction of barcode information.

[0057] The information comparison and identification module includes the following steps to process the information extracted by the identification module:

[0058] S01. Normalize the drug names and convert the prescription names and packaging names into standard pharmaceutical dictionary terms;

[0059] S02, adopts the specification parameter fuzzy matching algorithm, allowing ±5% dose deviation;

[0060] S03. When the name normalization comparison is inconsistent or the specification deviation exceeds the limit, the sound and light alarm is triggered and the infusion pump is locked and started;

[0061] S04. Calculate the difference between the current time and the expiration date. If the remaining validity period is less than the estimated infusion time × 1.2 safety factor, generate an infusion speed recommendation note.

[0062] Preferably, the positioning of the first segmentation model and the second segmentation model is obtained by training with a labeled data set including various types of medical barcodes, and the labeled data includes barcode area contour annotations.

[0063] Preferably, forming the infusion speed suggestion Kou Tiao Bao in S04 includes the following processing steps:

[0064] S200, calculate the remaining effective time T based on the drug validity period data and the current time r ;

[0065] S201 obtains the current infusion time T set by the infusion pump i , if T r <K×T i , then recalculate the maximum allowable infusion rate V:

[0066] V=(m×Tr) / (K×T i );

[0067] Where m is the total volume of the drug, and K is a safety factor of 1.1-1.3;

[0068] S203: Generate an early warning signal containing V and send it to the infusion pump control terminal.

[0069] Specifically, by calculating the remaining effective time and infusion duration, the system automatically generates an infusion speed recommendation to ensure that the infusion is completed within the drug's effective period. Furthermore, by generating an early warning signal and sending it to the infusion pump control terminal, the system enables real-time monitoring and adjustment of the infusion process, improving the safety and effectiveness of infusion therapy.

[0070] In summary, this embodiment 1 has the ability to realize automatic verification of drug names and specifications through two-way cross-verification of medical order barcodes and liquid packaging codes, reducing manual verification errors. It can also parse the expiration date data in the liquid code in real time, and calculate the remaining validity safety threshold in combination with the infusion time, effectively preventing the use of expired drugs. Secondly, multi-angle image acquisition is combined with a surface optimization segmentation model to solve the problems of reflection and wrinkles on liquid packaging, thereby improving the recognition rate; and when the barcode is damaged, the visual feature matching process is started, combined with the drug feature database to achieve a recall rate of more than 90% for similar drugs, and by introducing a nurse confirmation link, a human-machine collaborative verification closed loop is formed.

[0071] Example 2

[0072] An embodiment of the present invention provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by a processor to implement the steps:

[0073] Acquire a first image and a second image, wherein the first image includes a medical order barcode on a patient's wristband or an infusion sheet, and the second image includes a packaging barcode on an infusion bag / bottle or a packaging box;

[0074] Locating the medical order barcode region in the first image using a pre-trained first segmentation model, where the first segmentation model is trained based on a medical barcode dataset, the training data including barcode annotations in reflective and wrinkled scenes;

[0075] Locating the package barcode region in the second image using a second segmentation model optimized for curved packaging, including enhanced training samples for curved surfaces captured at angles between 45 and 60 degrees.

[0076] Identify the medical order coding information and extract at least the standardized drug name field and dosage specification field therefrom; identify the liquid coding information and extract at least the drug name, specification parameters and expiration date data therefrom.

[0077] It includes performing the following steps on the information extracted by the recognition module:

[0078] S01. Normalize the drug names and convert the prescription names and packaging names into standard pharmaceutical dictionary terms;

[0079] S02, adopts the specification parameter fuzzy matching algorithm, allowing ±5% dose deviation;

[0080] S03. When the name normalization comparison is inconsistent or the specification deviation exceeds the limit, the sound and light alarm is triggered and the infusion pump is locked and started;

[0081] S04. Calculate the difference between the current time and the expiration date. If the remaining validity period is less than the estimated infusion time × 1.2 safety factor, generate an infusion speed recommendation note.

[0082] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware using a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0083] Those skilled in the art will clearly understand that for the sake of convenience and brevity in description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0084] Example 3

[0085] An embodiment of the present invention provides an electronic device, including a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement the following steps:

[0086] Acquire a first image and a second image, wherein the first image includes a medical order barcode on a patient's wristband or an infusion sheet, and the second image includes a packaging barcode on an infusion bag / bottle or a packaging box;

[0087] Locating the medical order barcode region in the first image using a pre-trained first segmentation model, where the first segmentation model is trained based on a medical barcode dataset, the training data including barcode annotations in reflective and wrinkled scenes;

[0088] Locating the package barcode region in the second image using a second segmentation model optimized for curved packaging, including enhanced training samples for curved surfaces captured at angles between 45 and 60 degrees.

[0089] Identify the medical order coding information and extract at least the standardized drug name field and dosage specification field therefrom; identify the liquid coding information and extract at least the drug name, specification parameters and expiration date data therefrom.

[0090] It includes performing the following steps on the information extracted by the recognition module:

[0091] S01. Normalize the drug names and convert the prescription names and packaging names into standard pharmaceutical dictionary terms;

[0092] S02, adopts the specification parameter fuzzy matching algorithm, allowing ±5% dose deviation;

[0093] S03. When the name normalization comparison is inconsistent or the specification deviation exceeds the limit, the sound and light alarm is triggered and the infusion pump is locked and started;

[0094] S04. Calculate the difference between the current time and the expiration date. If the remaining validity period is less than the estimated infusion time × 1.2 safety factor, generate an infusion speed recommendation note.

[0095] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with the present profession can make some changes or modifications to equivalent embodiments of the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. An intelligent infusion monitoring system, characterized in that: include: An image acquisition and feature extraction module acquires a first image and a second image, wherein the first image includes a medical order barcode on a patient's wristband or an infusion sheet, and the second image includes a packaging barcode on an infusion bag / bottle or a packaging box; a data preprocessing module for locating the medical order barcode region in the first image using a pre-trained first segmentation model, wherein the first segmentation model is trained based on a medical barcode dataset, wherein the training data includes barcode annotations in reflective and wrinkled scenes; a key data extraction module that locates the package barcode region in the second image using a second segmentation model optimized for curved packaging and including enhanced training samples for curved surfaces captured at angles between 45 and 60 degrees; The identification module identifies the medical order coding information and extracts at least the standardized drug name field and dosage specification field therefrom; identifies the liquid coding information and extracts at least the drug name, specification parameters and expiration date data therefrom. The information comparison and identification module includes performing the following steps on the information extracted by the identification module: S01. Normalize the drug names and convert the prescription names and packaging names into standard pharmaceutical dictionary terms; S02, adopts the specification parameter fuzzy matching algorithm, allowing ±5% dose deviation; S03. When the name normalization comparison is inconsistent or the specification deviation exceeds the limit, the sound and light alarm is triggered and the infusion pump is locked and started; S04. Calculate the difference between the current time and the expiration date. If the remaining validity period is less than the estimated infusion time × 1.2 safety factor, generate an infusion speed recommendation note.

2. The intelligent infusion monitoring system according to claim 1, characterized in that: The image acquisition and feature extraction module includes a shooting and recording device, which respectively performs multi-angle image acquisition on the doctor's order barcode on the patient's wristband or infusion sheet and the packaging barcode on the infusion bag / bottle or packaging box.

3. The intelligent infusion monitoring system according to claim 1, characterized in that: The positioning of the first segmentation model and the second segmentation model is obtained by training a labeled data set including various types of medical barcodes, and the labeled data includes barcode area contour annotations.

4. The intelligent infusion monitoring system according to claim 1, characterized in that: The recognition module includes a calibration unit, which uses fast Fourier transform to extract the spatial frequency characteristics of the barcode and eliminates harmonic interference caused by deformation of the packaging surface through frequency domain filtering.

5. The intelligent infusion monitoring system according to claim 4, characterized in that: The calibration unit performs the following steps on the calibration of the QR code: S100, converting the detected barcode area into a grayscale image; S101, performing a two-dimensional FFT transform on the image to obtain a spectrum graph; S102. Design a bandpass filter to retain the fundamental frequency component corresponding to the barcode module width. The filter passband is [1 / (3W), 1 / (0.8W)], where W is the estimated width of a single barcode module. S103 , performing inverse FFT transformation on the filtered spectrum to obtain a corrected image.

6. The intelligent infusion monitoring system according to claim 1, characterized in that: The data preprocessing module includes an error reporting unit, which, when unable to identify the coded information, extracts the visual features of the liquid packaging, performs similarity matching through the drug feature database, obtains candidate drug information, and then feeds back to the nurse for confirmation to determine the final drug information.

7. The intelligent infusion monitoring system according to claim 1, characterized in that: The step S04 of forming the infusion speed suggestion comprises the following processing steps: S200, calculate the remaining effective time T based on the drug validity period data and the current time r ; S201 obtains the current infusion time T set by the infusion pump i , if T r <K×T i , then recalculate the maximum allowable infusion rate V: V=(m×Tr) / (K×T i ); Where m is the total volume of the drug, and K is a safety factor of 1.1-1.3; S203: Generate an early warning signal containing V and send it to the infusion pump control terminal.

8. A non-transitory computer-readable storage medium, wherein at least one instruction or at least one program is stored in the non-transitory computer-readable storage medium, characterized in that: The at least one instruction or the at least one program is loaded and executed by the processor to implement the method of the intelligent infusion monitoring system according to any one of claims 1 to 7.

9. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the method of the intelligent infusion monitoring system as described in any one of claims 1 to 7.

Images