| Chronic diseases are usually lifelong diseases,and patients need long-term or even life-long medication to control the disease.Due to poor medication compliance or poor self-discipline in life,patients are prone to acute complications and hospitalization.For the continuation of medical services,the hospital hopes to track the health status of patients at home for a long time after discharge,and safe medication is the top priority.This thesis designs and implements a smart home medication supervision system.As the auxiliary of hospital information system,the system aims to better monitor the medication situation of patients and improve their medication compliance.The main contents of the thesis are as follows:1.This thesis investigates the background and the current research status at home and abroad of the research subject,and analyzes the existing problems in medication supervision system.On the basis of that,it analyzes the actual needs of different users of the system,and proposes the overall system development plan,which based on B/S architecture and We Chat mini program.2.In view of the data closure of hospital information system,the paper medication list or prescription of patients at hospital or discharge is the only way to connect the data of the two systems.This thesis uses Optical Character Recognition(OCR)technology to realize the information identification of paper medication list or prescription list,and solves the data interaction between the two systems.Firstly,medication list image is preprocessed,including graying,binarization and image correction.Then,the cell region is extracted by using the line intersection characteristic,and a text region extraction method based on morphological processing is designed for some special cells.Finally,the extracted text regions are sent to the Tesseract recognition engine to complete the recognition processing.In addition,this thesis conducts targeted font training on Tesseract engine to improve the accuracy of the engine in recognition of medication list,and proposes an error correction scheme for the recognition results of medication list.Through the experimental comparison of the results before and after error correction,the effectiveness of the scheme is verified,and the accuracy of the recognition of the medication list can be improved.3.According to the demand analysis,this thesis has completed the design of the system function module and database.The realization method of the medication list identification module is elaborated,and the important functional modules of the medication management platform and We Chat mini program are discussed specifically.In order to test and verify the effectiveness and reliability of the home medication supervision system,test cases are designed to carry out functional and non-functional tests on the system.The test results show that the system meets the demand and achieves the expected goal. |
PDF Full Text Request