Artificial intelligence assisted pharmacist dispensing system

Abstract

The invention relates to an artificial intelligence assisted pharmacist dispensing system. According to the technical scheme, the artificial intelligence assisted pharmacist dispensing system is in butt joint with a doctor's advice information platform to obtain prescription information. The artificial intelligence auxiliary pharmacist medicine dispensing system automatically checks the medicine information obtained in the medicine placing area with the prescription information, and when the artificial intelligence auxiliary pharmacist medicine dispensing system finds that the medicine information is inconsistent with the prescription information, prompt information is sent out in time. And the pharmacist carries out error correction according to the prompt information and then carries out medicine issuing. The beneficial effects are as follows: the artificial intelligence assisted pharmacist dispensing system adopts a deep learning algorithm to develop a target detection model of medicine packaging box information, the detection speed is high, and the accuracy is high; the prompt information can timely and accurately tell a pharmacist where the medicine to be dispensed does not accord with the prescription information, and the result is objective and reliable; external equipment of the artificial intelligence assisted pharmacist medicine dispensing system is a high-definition camera which is easy to install, small in occupied space and easy to deploy in a pharmacy medicine dispensing window.

Description

Technical Field

[0001] This invention relates to a medication dispensing device to assist hospital pharmacists, and more particularly to an artificial intelligence-assisted medication dispensing system. Background Technology

[0002] Medication dispensing errors typically refer to discrepancies between the medications dispensed by the pharmacy and those prescribed by the physician. Currently, medication dispensing errors in hospital pharmacies remain unavoidable. In hospital pharmacy practice, medications with similar packaging and medications with the same name but different manufacturers and packaging are frequently encountered. While these medications are visually identifiable, human vision is prone to fatigue, especially during peak prescription periods, inevitably leading to dispensing errors. Reducing the incidence of medication dispensing errors is a crucial aspect of improving the quality of medication dispensing in hospital pharmacies. It directly impacts the safety and effectiveness of medication use for patients and is a vital measure for enhancing the quality of hospital pharmacy services. Summary of the Invention

[0003] This invention addresses the problem of drug dispensing errors caused by visual fatigue during the dispensing process by pharmacists. It designs a complete artificial intelligence-assisted pharmacist dispensing system. The core of the system is to train a pharmacy drug packaging box recognition model using a deep learning target recognition algorithm. It has the advantages of high speed, high accuracy, and easy deployment at the dispensing window.

[0004] The technical solution is as follows: The AI-assisted pharmacist dispensing system interfaces with the hospital's medical order information platform to obtain medical order prescription information. Then, the AI-assisted pharmacist dispensing system automatically compares the information of the medications to be dispensed, previously obtained from the medication placement area, with the medical order prescription information. When the AI-assisted pharmacist dispensing system detects a discrepancy between the medication information and the medical order prescription information, it will promptly issue a prompt. The pharmacist corrects the error based on the prompt and then proceeds with the medication dispensing.

[0005] The aforementioned hospital prescription information platform refers to the information platform for prescriptions issued by various clinical departments of the hospital. The connection to this platform is completed with the assistance of the hospital's information department.

[0006] The above-mentioned prescription information includes the name, manufacturer, dosage form, and quantity of the medicine prescribed by the doctor.

[0007] The aforementioned information about the drugs to be dispensed refers to the information on the drug packaging boxes in the previous drug placement area. This information is obtained by the AI-assisted pharmacist dispensing system through the verification of the drug packaging boxes in the previous drug placement area. The information obtained includes the drug's name, manufacturer, dosage form, quantity, etc.

[0008] The verification of the drug information to be dispensed against the doctor's prescription information is automatically completed by the AI-assisted pharmacist dispensing system.

[0009] The above prompts include both voice and text prompts. The voice prompts can be customized as needed, such as "Beep...". Text prompts are based on the type of error, such as "...Quantity error, please check."

[0010] The aforementioned error correction refers to the pharmacist re-checking and dispensing medications based on prompts from the AI-assisted pharmacist dispensing system.

[0011] The AI-assisted pharmacist dispensing system uses deep learning algorithms to develop a target detection model for drug packaging information, resulting in faster detection speed and higher accuracy. The notification information can promptly and accurately inform pharmacists of any discrepancies between the medication to be dispensed and the prescription information, and the results are objective and reliable. The AI-assisted pharmacist dispensing system uses a high-definition camera as an external device, which occupies little space at the pharmacy dispensing window. The AI-assisted pharmacist dispensing system is easy to install and deploy at pharmacy dispensing windows. Attached Figure Description

[0012] Appendix Figure 1 This is an overall schematic diagram of the present invention; Appendix Figure 2 This is a schematic diagram of the workstation of the present invention; Appendix Figure 3 This is a structural diagram of the high-definition camera device of the present invention; In the above diagram: 1. Workbench; 2. Workstation; 3. Speaker; 4. Drug packaging box information acquisition device; 5. Drug placement area before verification; 6. Drug placement area after verification; 7. Hospital medical order information platform server. Workstation host 2.1, keyboard 2.2, mouse 2.3, monitor 2.4, prescription information display area 2.5, AI-assisted pharmacist dispensing system detection information display area 2.6; 4.1 HD camera, 4.2 HD camera bracket. Detailed Implementation

[0013] Combined with appendix Figure 1-3 The present invention will be further described below: This invention mainly consists of a workbench 1, a workstation 2, a speaker 3, a drug packaging box information acquisition device 4, a pre-verification drug placement area 5, a post-verification drug placement area 6, and a hospital medical order information platform server 7. The workstation host 2.1 is equipped with an AI-assisted pharmacist dispensing system and interfaces with the hospital medical order information platform server 7. It acquires information about the drug packaging boxes to be dispensed via a high-definition camera 4.1 and verifies this information against the medical order prescription information, displaying the results on a monitor 2.4. When the AI-assisted pharmacist dispensing system detects a discrepancy between the drug information and the medical order prescription information, it promptly emits a "beep..." prompt through the speaker 3 and displays the inconsistency in the AI-assisted pharmacist dispensing system detection information display area 2.6. The pharmacist corrects the error based on the prompt and then dispenses the drug.

[0014] The workstation 2 consists of a host 2.1, a keyboard 2.2, a mouse 2.3, a monitor 2.4, a prescription information display area 2.5, and an AI-assisted pharmacist dispensing system detection information display area 2.6. The host is equipped with AI-assisted pharmacist dispensing system software, which uses a pharmacy drug packaging box recognition model trained with a deep learning algorithm.

[0015] The aforementioned drug packaging box information acquisition device 4 consists of a high-definition camera 4.1 and a high-definition camera bracket 4.2, which can capture and verify the information of the drug packaging boxes previously placed in the drug placement area 5.

[0016] The AI-assisted pharmacist dispensing system uses deep learning algorithms to develop a target detection model for drug packaging information, resulting in faster detection speed and higher accuracy. The notification information can promptly and accurately inform pharmacists of any discrepancies between the medication to be dispensed and the prescription information, and the results are objective and reliable. The AI-assisted pharmacist dispensing system uses a high-definition camera as an external device, which occupies little space at the pharmacy dispensing window. The AI-assisted pharmacist dispensing system is easy to install and deploy at pharmacy dispensing windows.

Claims

1. An artificial intelligence assisted pharmacist dispensing system, characterized by: It mainly consists of a workbench (1), a workstation (2), a speaker (3), a drug packaging box information acquisition device (4), a drug placement area before verification (5), a drug placement area after verification (6), and a hospital medical order information platform server (7). The workstation host (2.1) is equipped with an artificial intelligence-assisted pharmacist dispensing system that interfaces with the hospital medical order information platform server (7). It acquires the drug packaging box information to be dispensed through a high-definition camera (4.1), and verifies the drug packaging box information with the medical order prescription information, which is then displayed on the monitor (2.4).

2. A workstation (2) according to claim 1, characterized in that: The display (2.4) has a prescription information display area (2.5) and an AI-assisted pharmacist dispensing system detection information display area (2.6). When the AI-assisted pharmacist dispensing system finds that the information of the medicine to be dispensed is inconsistent with the prescription information, it will promptly issue a "beep..." prompt sound through the speaker (3) and display the inconsistent content in the AI-assisted pharmacist dispensing system detection information display area (2.6). The pharmacist corrects the error according to the prompt information and then dispenses the medicine.

3. The AI-assisted pharmacist dispensing system according to claim 1, its core algorithm is the improved YOLOv5 algorithm in deep learning algorithms. The improved model training parameters are as follows: input image size, 640×640; training iterations, 300; batch size, 4; learning rate, 0.01; data augmentation (image rotation, 45°; vertical flip, yes; horizontal flip, yes; image nesting, yes; image copying, yes).

4. The drug packaging box information acquisition device (4) according to claim 1, characterized in that: Composed of a high-definition camera 4.1 and a high-definition camera bracket 4.2, it is installed on the workbench and can capture and verify the information of the medicine packaging boxes placed in the previous medicine placement area 5.

Images