Design And Implementation Of The Oral Medical Records Recognition System Based On Keyword Extraction And Punctuation Filling Technology

When doctors diagnoses patients often writes case report by hands,which greatly reduces the efficiency of medical diagnosis.Thus it proposes ways to improve the efficiency of medical records by voice recognition.But the recognize texts by the existing voice recognition system are very unfriendly that can not make human know the keywords,and even don't raise punctuation.Therefore,we studied the keywords extracted speech recognition and text punctuation filling technology and apply it to medical dictation recognition systems.The system enables the doctor more convenient when reading medical texts after recognition,but also more efficient access to current medical records from a large number of text critical information.This greatly improves the efficiency of doctors.The main contents are as follows:1.Analyzed user needs.This paper had researched and Analyzed the current phenomenon of doctors' handwritten medical records.The results showed that the handwritten medical records reduces the efficiency of medical diagnosis,current speech recognition and text recognition system is not friendly enough.Therefore,this paper supplied the requirements of developing the system of oral medical records recognition based on the keyword extraction and punctuation.Proposed a new keyword extraction algorithm named FTW-TF-IDF.This paper focused on keyword extraction algorithm and analyzed the effect of improving the extraction step focuses on the words and points extraction algorithm.Therefore,this article applied HTK segmentation technology to the keyword extraction process segmentation step.Besides,the author improved TF-IDF algorithm,based on the increase of the feature weights of the parameters proposed.Threfore,a new keyword extraction algorithm named FTW-TF-IDF had be preposed and be proved better.Besides,the author studied punctuation filling technology,optimized for the characteristics of the project improved the punctuation filling technology,and proved the validity of punctuation optimized filling technique.3.Designed and implemented the system of oral medical records recognition based on the keyword extraction and punctuation.The system used the C / S architecture,designed and implemented a client login,identification recording,record,upload and set up modules and servers,punctuation and raise keyword extraction module.The paper analyzed the system security vulnerabilities,and used AES encryption technology to ensure system security.4.Tested and ran the system of oral medical records recognition based on the keyword extraction and punctuation.the results of the test shows that the system is stable,various functions meet the demand,can be a good help doctors improve diagnostic efficiency.