WO2016070445A1 - Medical image interpretation method and system based on intelligent speech recognition - Google Patents

Medical image interpretation method and system based on intelligent speech recognition Download PDF

Info

Publication number
WO2016070445A1
WO2016070445A1 PCT/CN2014/090864 CN2014090864W WO2016070445A1 WO 2016070445 A1 WO2016070445 A1 WO 2016070445A1 CN 2014090864 W CN2014090864 W CN 2014090864W WO 2016070445 A1 WO2016070445 A1 WO 2016070445A1
Authority
WO
WIPO (PCT)
Prior art keywords
interpretation
medical image
speech recognition
signal
processing device
Prior art date
Application number
PCT/CN2014/090864
Other languages
French (fr)
Chinese (zh)
Inventor
张贯京
陈兴明
葛新科
Original Assignee
深圳市前海安测信息技术有限公司
深圳市易特科信息技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN201410614512.7A external-priority patent/CN104462763B/en
Priority claimed from CN201420653929.XU external-priority patent/CN204233142U/en
Application filed by 深圳市前海安测信息技术有限公司, 深圳市易特科信息技术有限公司 filed Critical 深圳市前海安测信息技术有限公司
Publication of WO2016070445A1 publication Critical patent/WO2016070445A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
    • G16Z99/00—Subject matter not provided for in other main groups of this subclass

Definitions

  • the present invention relates to the field of medical imaging technology, and in particular, to a medical image interpretation method and system based on intelligent speech recognition.
  • the medical image reading system is a system used by radiologists to view, interpret images and generate interpretation reports, and plays a very important role in the field of medical imaging. After years of development, the system has been fully digitalized, which has greatly promoted the development of medical imaging. However, the existing medical image reading system still has the following shortcomings during its use:
  • the main purpose of the present invention is to simplify the operation of medical image reading and improve the efficiency and accuracy of image interpretation.
  • the present invention provides a medical image interpretation method based on intelligent speech recognition,
  • the medical image interpretation method based on intelligent speech recognition includes the following steps:
  • the generated image interpretation report is displayed.
  • the method for reading medical images based on intelligent speech recognition further comprises the steps of:
  • the step of identifying the interpretation signal and matching the interpretation result corresponding to the interpretation signal in the preset speech recognition library comprises:
  • the interpretation signal includes a preset first voice mark and a portion corresponding to the medical image file, matching the interpretation template corresponding to the part in the preset voice recognition library as a normal interpretation result;
  • the interpretation signal does not include a preset voice flag, matching the text corresponding to the interpretation signal in the preset voice recognition library as an abnormal interpretation result;
  • the interpretation signal includes the preset second voice flag
  • the text corresponding to the interpretation signal is matched in the preset medical image database as a reference interpretation result.
  • the step of displaying the generated image interpretation report comprises:
  • the method for reading medical images based on intelligent speech recognition further comprises the steps of:
  • the present invention also provides a medical image interpretation system based on intelligent speech recognition
  • the medical image interpretation system based on intelligent speech recognition includes a speech recognition device, a multi-function mouse, a pedal and a signal processing device, wherein:
  • Said a voice recognition device connected to the signal processing device for collecting and recognizing an interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and outputting the interpretation result to the signal processing device for the signal processing device Generate an image interpretation report based on the interpretation result;
  • the multi-function mouse is connected to the signal processing device, and includes at least one function button for editing the content of the image interpretation report when generating the image interpretation report;
  • the pedal is connected to the signal processing device for turning a medical image file when interpreting the medical image file;
  • the signal processing device is configured to receive the interpretation result, generate an image interpretation report according to the interpretation result, and output; and perform a corresponding operation when receiving the signal input by the multifunctional mouse and the pedal.
  • the medical image interpretation system based on intelligent speech recognition further comprises:
  • a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  • the function button of the multi-function mouse can be set to one or more of a backspace key, a blank line key, a space key, a delete key, a zoom key, a selection key, and a save key.
  • the intelligent speech recognition based medical image interpretation system further comprises:
  • a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  • the pedal includes a left pedal and a right pedal, the left pedal being used to page forward/backward when interpreting a medical image file,
  • the right pedal is used to page backward/forward when interpreting medical image files.
  • the intelligent speech recognition based medical image interpretation system further comprises:
  • a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  • the voice recognition device comprises:
  • a voice collection module for collecting an interpretation signal input by voice
  • a matching module connected to the signal processing device, for Identifying the interpretation signal, matching corresponding interpretation results in a preset speech recognition library according to the interpretation signal, and outputting the interpretation result to the signal processing device.
  • the intelligent speech recognition based medical image interpretation system further comprises:
  • a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  • the signal processing device is specifically configured to:
  • the page turning operation is performed when the page turning signal for turning the page of the medical image file is received.
  • the intelligent speech recognition based medical image interpretation system further comprises:
  • a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  • this invention After receiving the medical image file to be interpreted, in the process of medical image expert interpretation, the interpretation signal input by the voice mode is collected, the interpretation signal is recognized, and the image interpretation report is generated according to the interpretation result corresponding to the interpretation signal, and the generated image interpretation is displayed. report. No need for manual input by medical imaging experts when generating image interpretation reports, through voice recognition, It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
  • FIG. 1 is a schematic flow chart of a first embodiment of a medical image interpretation method based on intelligent speech recognition
  • FIG. 2 is a schematic diagram of the refinement process of step S20 in Figure 1;
  • FIG. 3 is a schematic flow chart of a second embodiment of a medical image interpretation method based on intelligent speech recognition
  • FIG. 4 is a schematic diagram of a system structure of a first preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention
  • FIG. 5 is a schematic diagram of a system structure of a second preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention.
  • FIG. 6 is a schematic diagram showing the system structure of a third preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention.
  • the invention provides a medical image interpretation method based on intelligent speech recognition.
  • FIG. 1 is a schematic flow chart of a first embodiment of a medical image interpretation method based on intelligent speech recognition according to the present invention.
  • the medical image interpretation method based on intelligent speech recognition includes:
  • Step S10 receiving a medical image file to be interpreted
  • Step S20 Acquiring an interpretation signal input by voice, identifying an interpretation signal, and matching an interpretation result corresponding to the interpretation signal in a preset speech recognition library;
  • Step S30 generating an image interpretation report according to the matched interpretation result
  • Step S40 displaying the generated image interpretation report.
  • the medical image data may include medical image files to be interpreted corresponding to various parts of the patient's body, and personal information of the patient, such as name, age, gender, medical history and the like.
  • the interpretation signal in the form of a voice signal can be input through a hardware device for voice input such as a microphone, the interpretation signal is collected, the interpretation signal is voice-recognized, and according to the recognized interpretation signal, The interpretation result corresponding to the interpretation signal is matched in the preset speech recognition library, and the interpretation result may include a normal interpretation result and an abnormal interpretation result. Then, based on the matched interpretation result, an image interpretation report for displaying the interpretation result is generated.
  • the received patient's name, age, gender, medical history and other information can be embedded in the front of the image interpretation report in a modular form, and the time stamp can be automatically generated at the end of the image interpretation report.
  • medical imaging experts can sign their names in the image interpretation report. In this embodiment, after the image interpretation report is generated, if the interpretation result in the image interpretation report is found to be incorrect, the erroneous content may be deleted.
  • the image interpretation report is displayed, that is, the interpretation result of the medical image expert reading the received medical image file is displayed.
  • the interpretation result is normal
  • the corresponding part is displayed as normal
  • the interpretation result is When an abnormality occurs, the specific abnormal situation is displayed according to the interpretation of the medical imaging expert.
  • the embodiment collects the interpretation signal input by the voice mode in the process of reading the medical image expert, recognizes the interpretation signal, and generates an image interpretation report according to the interpretation result corresponding to the interpretation signal, and displays the generated image.
  • Image interpretation report No need for manual input by medical imaging experts when generating image interpretation reports, through voice recognition, It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
  • FIG. 2 is a schematic diagram of the refinement process of step S20 in FIG.
  • step S20 specifically includes:
  • Step S21 collecting an interpretation signal input by a voice mode, and performing filtering and noise reduction processing on the interpretation signal;
  • the medical image expert interprets the medical image file
  • the original voice signal is first processed, such as filtering processing and noise reduction processing, eliminating the interference of the voice signal and improving the signal to noise ratio.
  • Step S22 identifying the interpretation signal, and matching the text corresponding to the interpretation signal in the preset speech recognition library as the interpretation result;
  • the processed interpretation signal is speech-recognized, and the text corresponding to the interpretation signal is matched in the preset speech recognition library as the interpretation result.
  • the preset speech recognition library collects commonly used medical terms, radiological terms, and templates for common interpretation results.
  • the template corresponds to a normal interpretation result.
  • the received interpretation signal includes a preset first voice flag and a portion corresponding to the medical image file, and the first voice flag is used as a indicator for determining whether the interpretation result is normal, that is, when When the first speech mark is included in the interpretation signal, it indicates that the interpretation result of the medical image file is normal, and the content of the first speech mark is generally selected as a word different from related medical terms and radiological terms.
  • the received interpretation signal includes 'click + lungs', indicating that the interpretation of the medical image files corresponding to the lungs is normal.
  • the template corresponding to the lungs can be matched in the speech recognition library, for example, matching 'no abnormalities in the lungs' as normal. Interpret the results.
  • the second voice flag is used as an indicator for determining whether the interpretation result is a reference interpretation result, that is, when the second voice flag is included in the interpretation signal, indicating It is necessary to match the interpretation result of the reference corresponding to the medical image file, and the content of the second voice flag can be set to a word such as 'reference', which can clearly determine the interpretation result of the instruction as the reference interpretation result.
  • the received interpretation signal is a 'reference' plus a specific image interpretation report
  • the text corresponding to the conclusion of the image interpretation report is matched in the speech recognition library, and the text is used as a reference interpretation result.
  • the received interpretation signal does not include a preset voice flag, that is, the first voice flag and the second voice flag are not included in the interpretation signal, it may be judged that the interpretation result is not a normal interpretation result.
  • the complete interpretation signal is recognized. And further matching the corresponding text in the speech recognition library according to the interpretation signal.
  • the matching text corresponding to the interpretation signal has nothing to do with the interpretation result of the medical image file, it indicates that the received interpretation signal is an invalid signal; on the contrary, it indicates that the interpretation result corresponding to the interpretation signal is an abnormal interpretation result.
  • the recognized interpretation signal is 'a plurality of soft tissue density lesions visible in both lungs, and the morphology is consistent with metastatic cancer', and the corresponding text is matched in the speech recognition library, and the text is used as an abnormal interpretation result.
  • the interpretation result corresponding to the interpretation signal is not matched in the speech recognition library, it indicates that the corresponding speech is not stored in the speech recognition library.
  • the medical imaging expert may input the corresponding through the keyboard or other means.
  • the interpretation result is added to the image interpretation report, and the input interpretation result is added to the speech recognition library, so as to match the corresponding interpretation result from the speech recognition library when the same or similar interpretation signal is received next time.
  • the method for reading medical images based on intelligent speech recognition On the basis of an embodiment, in the second embodiment, the method further includes:
  • the medical image database is matched with similar medical image data whose medical image file has similarity within a preset range, and the conclusion of the image interpretation report corresponding to the similar medical image data is input by voice as an interpretation signal.
  • a medical image database is preset, and the medical image database stores various commonly used normal or abnormal medical image data for reference.
  • the medical image After receiving the medical image file to be interpreted, the medical image may be firstly used.
  • the medical image data of the document and the medical image data in the medical image database are similarly matched.
  • the similarity is within the preset range, the medical image data in the medical image database is acquired as similar medical image data, and the preset range may be Customize settings according to the type of medical image file and related parts, for example, set to be larger than 80%.
  • the conclusion of the image interpretation report of the medical image file is obtained, and the medical imaging expert can input the voice through voice Enter the conclusion of the interpretation report as the interpretation signal.
  • the conclusions of the image interpretation report include the interpretation results of the medical image files of the patient by the medical imaging experts, and may include information such as the treatment means and the rehabilitation situation, and when generating the interpretation report of the medical image file to be interpreted, This information is displayed along with the interpretation report for reference by doctors and patients.
  • this conclusion is used as a reading signal input by the voice input method for the doctor and the patient to refer to, so that the patient can understand the treatment mode similar to his own situation, thereby providing convenience for the patient, and making the doctor based on the previous case.
  • the treatment plan for the second case provides a reference.
  • step S40 is specifically:
  • the normal interpretation result and the abnormal interpretation result are differently identified, and the reference result can be obtained at the same time. It is identified as other identifiers that can be distinguished from normal or abnormal interpretation results, such as marking different interpretation results as different fonts or different colors, so as to be displayed in a preset display manner when displayed, for example, for anomalies
  • the text content of the interpretation result is displayed in bold or highlighted.
  • the interpretation result in the image interpretation report is a normal interpretation result
  • the text content corresponding to the interpretation template may be identified as a blue font after being matched to the interpretation template indicating that the interpretation result is normal, and is normal.
  • the text content may be identified as a red font after matching the text corresponding to the abnormal interpretation result, and the text content of the red font is displayed at the time of display. Displayed in bold or highlighted to alert the patient.
  • the normal interpretation result in the image interpretation report and / Or abnormal interpretation results are marked differently, and the normal interpretation results and/or abnormal interpretation results are displayed in a preset display manner, so that the image interpretation report is displayed more clearly and is convenient for the patient to read.
  • FIG. 3 is a schematic flowchart diagram of a third embodiment of a medical image interpretation method based on intelligent speech recognition.
  • the method further includes:
  • Step S50 And receiving a marking instruction for marking the medical image file by voice input, and marking the corresponding medical image file according to the marking instruction for jumping to the corresponding medical image file after receiving the searching instruction at the time of searching.
  • the medical imaging expert can input the marking instruction by voice input, such as input 'emphasis 1 'or' mark 42 'Zhang, in this way, after receiving the marking instruction, the corresponding medical image file is marked.
  • step S50 It can be performed at any time when the medical image file is interpreted.
  • step S30 is preferably performed in the above-described first embodiment. Execution before, that is, marking the corresponding medical image file at any time before beginning to interpret the medical image file and display the generated image interpretation report.
  • marking the corresponding medical image file When receiving the marking instruction for marking the medical image file, marking the corresponding medical image file according to the marking instruction, it is convenient to accurately jump to the corresponding medical image file after receiving the searching instruction, thereby further It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
  • the invention also provides a medical image interpretation system based on intelligent speech recognition.
  • FIG. 4 is a system structure of a first preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention. Schematic.
  • the medical image interpretation system based on intelligent speech recognition comprises:
  • a voice recognition device 10 a multi-function mouse 20, a pedal 30, and a signal processing device 40, wherein:
  • the voice recognition device 10 and the signal processing device 40 Connecting, for acquiring and recognizing an interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and outputting the interpretation result to the signal processing device, where the signal processing device generates an image interpretation report according to the interpretation result;
  • the multifunctional mouse 20 and the signal processing device 40 The connection includes at least one function button for editing the content of the image interpretation report when generating the image interpretation report;
  • the pedal 30 and the signal processing device 40 Connection for turning a medical image file when interpreting a medical image file
  • the signal processing device 40 And receiving the interpretation result, generating an image interpretation report according to the interpretation result and outputting; and performing a corresponding operation when receiving the signal of the multifunctional mouse and the pedal input.
  • the pedal 30 When the doctor reads the medical image data, pass the pedal 30 When the medical image file is turned over, when a certain part is normal or abnormal, a voice input signal for evaluating a certain part is input to the voice recognition device 10, and the voice recognition device 10 is passed.
  • the voice input signal evaluated by the doctor for a certain part is collected and recognized, and is recognized as an interpretation signal, and the text corresponding to the interpretation signal is matched in the preset voice recognition library according to the interpretation signal as the interpretation result.
  • the interpretation result is output to the signal processing device 40 for the signal processing device 40.
  • the content of the image interpretation report is edited, for example, the generated image interpretation report can be deleted, modified, saved, and the like.
  • Embodiment of the present invention by the voice recognition device 10 Acquiring and recognizing the interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and outputting the interpretation result to the signal processing device 40 for the signal processing device 40 Generating an image interpretation report according to the interpretation result; editing, by the multi-function mouse 20, the content of the image interpretation report when generating the image interpretation report; Turn the medical image file into pages when interpreting medical image files.
  • the medical image interpretation system based on intelligent speech recognition provided by the invention is simple and practical in structure. It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
  • FIG. 5 is a system structure of a second preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention. Schematic.
  • the intelligent speech recognition based medical image interpretation system is based on the first preferred embodiment, preferably, The voice recognition device 10 includes:
  • the voice collection module 101 is configured to collect an interpretation signal input by voice; the voice collection module 101 Generally, it is a device capable of collecting a voice signal, such as a microphone. In a preferred embodiment, after acquiring an interpretation signal input by voice, the voice collection module 101 First, the original speech signal is processed, such as filtering processing and noise reduction processing, eliminating interference of the speech signal and improving the signal to noise ratio;
  • a matching module 102 connected to the signal processing device 40, for Identifying the interpretation signal, matching a corresponding interpretation result in a preset speech recognition library according to the interpretation signal, and outputting the interpretation result to the signal processing device 40 .
  • the preset speech recognition library collects commonly used medical terms, radiological terms, and templates for common interpretation results.
  • the template corresponds to a normal interpretation result.
  • the preset voice flag is used as an indicator for determining whether the interpretation result is normal, that is, when the speech signal is included in the interpretation signal, indicating The interpretation of the medical image file is normal, and the content of the speech mark is usually chosen to be different from the related medical terms and radiological terms. For example, when the received interpretation signal includes 'click + lungs', indicating that the interpretation of the medical image files corresponding to the lungs is normal. At this time, the template corresponding to the lungs can be matched in the speech recognition library, for example, matching 'no abnormalities in the lungs' as normal. Interpret the results.
  • the interpretation signal does not include the preset voice flag, it can be judged that the interpretation result is not a normal interpretation result.
  • the complete interpretation signal is recognized, and the corresponding text is further matched in the speech recognition library according to the interpretation signal.
  • the matching text corresponding to the interpretation signal has nothing to do with the interpretation result of the medical image file, it indicates that the received interpretation signal is an invalid signal; on the contrary, it indicates that the interpretation result corresponding to the interpretation signal is an abnormal interpretation result.
  • the recognized interpretation signal is 'a plurality of soft tissue density lesions visible in both lungs, and the morphology is consistent with metastatic cancer', and the corresponding text is matched in the speech recognition library, and the text is used as an abnormal interpretation result.
  • the multifunctional mouse 20 The function keys can be set to one or more of a backspace key, a blank key, a space key, a delete key, a zoom key, a selection key, and a save key according to actual needs.
  • the specific function of the function button of the multi-function mouse may be set in a programmatic manner, or the specific function of the function button of the multi-function mouse may be configured in a selected manner.
  • the pedal 30 includes a left pedal and a right pedal for forwarding when interpreting medical image files / Page backwards, the right pedal is used to page backward/forward when interpreting medical image files, and setting the left and right pedals makes it easier for doctors to turn forward/backward.
  • the signal processing device is specifically configured to:
  • Receiving the multifunctional mouse Perform editing operations when editing the edited signal of the content of the image interpretation report, for example, performing operations such as deleting, modifying, saving, enlarging, and reducing;
  • the page turning operation is performed, for example, the page turning forward and the page turning operation are performed.
  • the embodiment of the present invention passes the voice recognition device 10 Acquiring and recognizing the interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and quickly matching the interpretation result in the process of reading the medical image data by the doctor through the preset speech recognition library, for the signal processing device 40 And generating an image interpretation report according to the interpretation result; editing the content of the image interpretation report by using the multi-function mouse 20 when generating the image interpretation report, by using the multifunctional mouse 20
  • the function button function is set to facilitate the doctor to quickly edit the contents of the interpretation report; the pedal is used when interpreting the medical image file. Turning the medical image file, by setting the left pedal and the right pedal, it is convenient for the doctor to turn the page forward and then turn the page.
  • the medical image interpretation system based on intelligent speech recognition provided by the invention is simple and practical in structure. It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
  • FIG. 6 is a system structure of a third preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention. Schematic.
  • the intelligent speech recognition based medical image interpretation system is based on the second preferred embodiment, and further, The medical image interpretation system based on intelligent speech recognition further includes:
  • a display device 50 coupled to the signal processing device 40 for receiving the signal processing device 40
  • the output image interpretation report is displayed. Specifically, when the image interpretation report is output and displayed, the signal processing device 40 can correctly interpret the result and/or Or the abnormal interpretation result is identified, and the normal interpretation result and/or the abnormal interpretation result are displayed on the display device 50 in a preset display manner.
  • the normal interpretation result and the abnormal interpretation result are differently marked, for example, Different interpretation results are marked as different fonts or different colors so as to be displayed in a preset display manner when displayed, for example, the text content of the abnormal interpretation result is displayed in bold or highlighted.
  • the interpretation result in the image interpretation report is a normal interpretation result
  • the text content corresponding to the interpretation template may be identified as a blue font after being matched to the interpretation template indicating that the interpretation result is normal, and is normal.
  • the text content may be identified as a red font after matching the text corresponding to the abnormal interpretation result, and the text content of the red font is displayed at the time of display. Displayed in bold or highlighted to alert the patient.
  • the normal interpretation result in the image interpretation report and / Or abnormal interpretation results are marked differently, and the normal interpretation results and/or abnormal interpretation results are displayed in a preset display manner, so that the image interpretation report is displayed more clearly and is convenient for the patient to read.

Landscapes

  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Medical Treatment And Welfare Office Work (AREA)
  • Measuring And Recording Apparatus For Diagnosis (AREA)

Abstract

A medical image interpretation method and system based on intelligent speech recognition. The method comprises: receiving a medical image file to be interpreted (S10); collecting interpretation signals input by means of speech, recognizing the interpretation signals, and performing matching in a preset speech recognition library to obtain interpretation results corresponding to the interpretation signals (S20), wherein the interpretation signals are speech signals input when the medical image file to be interpreted is interpreted; generating an image interpretation report (S30); and displaying the generated image interpretation report (S40). According to the method and system, a medical image expert does not need to perform manual input to generate an image interpretation report; by means of speech recognition, the operation of medical image reading is simplified, and the efficiency and accuracy of medical image interpretation are improved.

Description

基于智能语音识别的医学影像解读方法和系统  Medical image interpretation method and system based on intelligent speech recognition 基于智能语音识别的医学影像解读方法和系统  Medical image interpretation method and system based on intelligent speech recognition
技术领域 Technical field
本 发明 涉及医学影像技术领域，尤其涉及一种基于智能语音识别的医学影像解读方法和系统。 The present invention relates to the field of medical imaging technology, and in particular, to a medical image interpretation method and system based on intelligent speech recognition.
背景技术 Background technique
医学影像阅读系统是放射科医生用于查看、解读影像和生成解读报告的系统，在医学影像学领域有着非常重要的作用。经过多年的发展，该系统已全面完成数字化，极大地推动了医学影像学的发展。但是，现有的医学影像阅读系统在使用过程中还是存在以下不足： The medical image reading system is a system used by radiologists to view, interpret images and generate interpretation reports, and plays a very important role in the field of medical imaging. After years of development, the system has been fully digitalized, which has greatly promoted the development of medical imaging. However, the existing medical image reading system still has the following shortcomings during its use:
1 、在查看影像时，不能同步撰写报告，而是要切换到报告撰写的界面或程序中进行；并且，在撰写报告时，需要通过手动的方式将内容逐字逐句地输入到计算机中。对于一些较为常见的检查结果（如某部位未见异常等），仍然需要进行繁琐的输入，这样就需要消耗相当的精力和时间； 1 When viewing an image, you can't write the report synchronously, but switch to the interface or program of the report writing; and, when writing the report, you need to manually input the content word by word into the computer. For some of the more common inspection results (such as no abnormalities in a certain part), still need to make cumbersome input, which requires considerable energy and time;
2 、对影像的常见操作较为复杂，需要多次点击相关按钮或菜单才能实现，如调窗、测长度等。对于断层影像，不能精确地跳转到指定位置，需要从当前位置一张一张翻到所需的影像。 2 The common operations on images are complex and require multiple clicks of related buttons or menus, such as window adjustment and length measurement. For tomographic images, you cannot jump to the specified position accurately. You need to scroll to the desired image one by one from the current position.
3 、传统的键盘 + 鼠标的输入方式难以完美应对越来越复杂的影像阅片操作，容易发生误操作致使 阅片 效率下降的现象。 3, the traditional keyboard + mouse input method is difficult to perfectly cope with more and more complex image reading operations, prone to misuse caused by reading The phenomenon of reduced efficiency.
发明内容 Summary of the invention
本发明的主要目的在于简化医学影像阅读的操作，提高影像解读的效率和准确性。 The main purpose of the present invention is to simplify the operation of medical image reading and improve the efficiency and accuracy of image interpretation.
为实现上述目的，本发明提供一种 基于智能语音识别的医学影像解读方法， 所述 基于智能语音识别的医学影像解读方法包括以下步骤： To achieve the above object, the present invention provides a medical image interpretation method based on intelligent speech recognition, The medical image interpretation method based on intelligent speech recognition includes the following steps:
接收待解读的医学影像文件； Receiving a medical image file to be interpreted;
采集通过语音方式输入的解读信号，识别所述解读信号，在预置的语音识别库中匹配解读信号对应的解读结果；所述解读信号为对所述待解读的医学影像文件进行解读时所输入的语音信号； Acquiring an interpretation signal input by voice, identifying the interpretation signal, and matching an interpretation result corresponding to the interpretation signal in a preset speech recognition library; the interpretation signal is input when the medical image file to be interpreted is interpreted Voice signal
生成根据匹配的所述解读结果影像解读报告； Generating an image interpretation report according to the matching interpretation result;
显示生成的所述影像解读报告。 The generated image interpretation report is displayed.
优选地， 所述 基于智能语音识别的医学影像解读方法还包括步骤： Preferably, the method for reading medical images based on intelligent speech recognition further comprises the steps of:
在预置的医学影像数据库中匹配与所述医学影像文件相似度在预置范围内的相似医学影像数据，并通过语音方式输入所述相似医学影像数据对应的影像解读报告的结论，作为解读信号。 Matching similar medical image data whose similarity with the medical image file is within a preset range in a preset medical image database, and inputting the conclusion of the image interpretation report corresponding to the similar medical image data by using a voice as an interpretation signal .
优选地， 所述 识别所述解读信号，在预置的语音识别库中匹配解读信号对应的解读结果的步骤包括： Preferably, the step of identifying the interpretation signal and matching the interpretation result corresponding to the interpretation signal in the preset speech recognition library comprises:
当所述解读信号包括预置的第一语音标志和医学影像文件对应的部位时，在所述预置的语音识别库中匹配与该部位对应的解读模版，作为正常的解读结果； When the interpretation signal includes a preset first voice mark and a portion corresponding to the medical image file, matching the interpretation template corresponding to the part in the preset voice recognition library as a normal interpretation result;
当所述解读信号不包括预置的语音标志时，在所述预置的语音识别库中匹配所述解读信号对应的文字，作为异常的解读结果； When the interpretation signal does not include a preset voice flag, matching the text corresponding to the interpretation signal in the preset voice recognition library as an abnormal interpretation result;
当所述解读信号包括预置的第二语音标志时，在预设的医学影像数据库中匹配所述解读信号对应的文字，作为参考的解读结果。 When the interpretation signal includes the preset second voice flag, the text corresponding to the interpretation signal is matched in the preset medical image database as a reference interpretation result.
优选地， 所述 显示生成的所述影像解读报告的步骤包括： Preferably, the step of displaying the generated image interpretation report comprises:
对影像解读报告中的正常的解读结果和 / 或异常的解读结果和 / 或参考解读结果进行标识，以预置的显示方式显示所述正常的解读结果和 / 或异常的解读结果和 / 或参考解读结果。 Normal interpretation results and/or abnormal interpretation results in the image interpretation report and / Or by referring to the interpretation result, the normal interpretation result and/or the abnormal interpretation result and/or the reference interpretation result are displayed in a preset display manner.
优选地，所述基于智能语音识别的医学影像解读方法还包括步骤： Preferably, the method for reading medical images based on intelligent speech recognition further comprises the steps of:
接收通过语音输入的对医学影像文件进行标记的标记指令，根据该标记指令标记相应的医学影像文件，以供在查找时接收到查找指令后跳转至相应的医学影像文件。 Receiving a mark instruction for marking the medical image file by voice input, and marking the corresponding medical image file according to the mark instruction for jumping to the corresponding medical image file after receiving the search instruction at the time of searching.
此外，为实现上述目的，本发明还提供 一种 基于智能语音识别的医学影像解读系统， 所述 基于智能语音识别的医学影像解读系统包括语音识别装置、多功能鼠标、踏板以及信号处理装置，其中： In addition, in order to achieve the above object, the present invention also provides a medical image interpretation system based on intelligent speech recognition, The medical image interpretation system based on intelligent speech recognition includes a speech recognition device, a multi-function mouse, a pedal and a signal processing device, wherein:
所述 语音识别装置，与所述信号处理装置连接，用于采集并识别语音输入的解读信号，根据解读信号匹配对应的解读结果，并将所述解读结果输出至所述信号处理装置，供信号处理装置根据解读结果生成影像解读报告； Said a voice recognition device connected to the signal processing device for collecting and recognizing an interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and outputting the interpretation result to the signal processing device for the signal processing device Generate an image interpretation report based on the interpretation result;
所述多功能鼠标，与所述信号处理装置连接，包括至少一个功能按键，用于在生成影像解读报告时对所述影像解读报告的内容进行编辑； The multi-function mouse is connected to the signal processing device, and includes at least one function button for editing the content of the image interpretation report when generating the image interpretation report;
所述踏板，与所述信号处理装置连接，用于在解读医学影像文件时对医学影像文件进行翻页； The pedal is connected to the signal processing device for turning a medical image file when interpreting the medical image file;
所述信号处理装置，用于接收所述解读结果，根据解读结果生成影像解读报告并输出；以及，在接收到所述多功能鼠标和踏板输入的信号时执行相应的操作。 The signal processing device is configured to receive the interpretation result, generate an image interpretation report according to the interpretation result, and output; and perform a corresponding operation when receiving the signal input by the multifunctional mouse and the pedal.
优选地，所 述 基于智能语音识别的医学影像解读系统还包括： Preferably, the medical image interpretation system based on intelligent speech recognition further comprises:
显示装置，与所述 信号处理装置连接，用于接收 所述 信号处理装置输出的影像解读报告并显示。 And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
优选地， 所述多功能鼠标的功能按键可以设置为退格键、空行键、空格键、删除键、缩放键、选择键、保存键中的一种或多种。 Preferably, The function button of the multi-function mouse can be set to one or more of a backspace key, a blank line key, a space key, a delete key, a zoom key, a selection key, and a save key.
优选地， 所述 基于智能语音识别的医学影像解读系统还包括： Preferably, the intelligent speech recognition based medical image interpretation system further comprises:
显示装置，与所述 信号处理装置连接，用于接收 所述 信号处理装置输出的影像解读报告并显示。 And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
优选地， 所述踏板包括左踏板和右踏板，所述左踏板用于 在解读医学影像文件时向前 / 向后翻页，所述 右踏板用于 在解读医学影像文件时向后 / 向前翻页。 Preferably, the pedal includes a left pedal and a right pedal, the left pedal being used to page forward/backward when interpreting a medical image file, The right pedal is used to page backward/forward when interpreting medical image files.
优选地， 所述 基于智能语音识别的医学影像解读系统还包括： Preferably, the intelligent speech recognition based medical image interpretation system further comprises:
显示装置，与所述 信号处理装置连接，用于接收 所述 信号处理装置输出的影像解读报告并显示。 And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
优选地， 所述 语音识别装置包括： Preferably, the voice recognition device comprises:
语音采集模块，用于采集 通过语音方式输入的解读信号 ； a voice collection module for collecting an interpretation signal input by voice;
匹配模块，与所述信号处理装置连接，用于 识别所述解读信号，根据所述解读信号在预置的语音识别库中匹配对应的解读结果，并将所述解读结果输出至所述信号处理装置。 a matching module, connected to the signal processing device, for Identifying the interpretation signal, matching corresponding interpretation results in a preset speech recognition library according to the interpretation signal, and outputting the interpretation result to the signal processing device.
优选地， 所述 基于智能语音识别的医学影像解读系统还包括： Preferably, the intelligent speech recognition based medical image interpretation system further comprises:
显示装置，与所述 信号处理装置连接，用于接收 所述 信号处理装置输出的影像解读报告并显示。 And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
优选地， 所述信号处理装置具体用于： Preferably, the signal processing device is specifically configured to:
在接收到所述多功能鼠标 对影像解读报告的内容进行编辑的编辑信号时执行相应的编辑操作； Performing a corresponding editing operation when receiving an editing signal for editing the content of the image interpretation report by the multi-function mouse;
以及，在接收到所述踏板对医学影像文件进行翻页的翻页信号时执行翻页操作。 And, the page turning operation is performed when the page turning signal for turning the page of the medical image file is received.
优选地， 所述 基于智能语音识别的医学影像解读系统还包括： Preferably, the intelligent speech recognition based medical image interpretation system further comprises:
显示装置，与所述 信号处理装置连接，用于接收 所述 信号处理装置输出的影像解读报告并显示。 And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
本发明 在接收到待解读的医学影像文件后，在医学影像专家解读的过程中，采集通过语音方式输入的解读信号，识别解读信号并根据解读信号对应的解读结果生成影像解读报告，显示生成的影像解读报告。在生成影像解读报告时无需医学影像专家手动输入，通过语音识别的方式， 简化了医学影像阅读的操作，并且提高了医学影像解读的效率和准确性。 this invention After receiving the medical image file to be interpreted, in the process of medical image expert interpretation, the interpretation signal input by the voice mode is collected, the interpretation signal is recognized, and the image interpretation report is generated according to the interpretation result corresponding to the interpretation signal, and the generated image interpretation is displayed. report. No need for manual input by medical imaging experts when generating image interpretation reports, through voice recognition, It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
附图说明 DRAWINGS
图 1 为本发明 基于智能语音识别的医学影像解读方法第 一实施例的流程示意图 ； 1 is a schematic flow chart of a first embodiment of a medical image interpretation method based on intelligent speech recognition;
图 2 为 图 1 中步骤 S20 的细化流程示意图 ； Figure 2 is a schematic diagram of the refinement process of step S20 in Figure 1;
图 3 为本发明 基于智能语音识别的医学影像解读方法第 二实施例的流程示意图 ； 3 is a schematic flow chart of a second embodiment of a medical image interpretation method based on intelligent speech recognition;
图 4 为本发明 基于智能语音识别的医学影像解读系统第一优选实施例的系统结构 示意图 ； 4 is a schematic diagram of a system structure of a first preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention;
图 5 为本发明 基于智能语音识别的医学影像解读系统第二优选实施例的系统结构 示意图 ； 5 is a schematic diagram of a system structure of a second preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention;
图 6 为本发明 基于智能语音识别的医学影像解读系统第三优选实施例的系统结构 示意图 。 6 is a schematic diagram showing the system structure of a third preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention.
本发明目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。 The implementation, functional features, and advantages of the present invention will be further described in conjunction with the embodiments.
具体实施方式 detailed description
应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。 It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
本发明提供一种 基于智能语音识别的医学影像解读方法。 The invention provides a medical image interpretation method based on intelligent speech recognition.
参照图 1 ， 图 1 为本发明 基于智能语音识别的医学影像解读方法第 一实施例的流程示意图。 1 is a schematic flow chart of a first embodiment of a medical image interpretation method based on intelligent speech recognition according to the present invention.
在一实施例中， 基于智能语音识别的医学影像解读方法包括： In an embodiment, the medical image interpretation method based on intelligent speech recognition includes:
步骤 S10 ，接收待解读的医学影像文件； Step S10, receiving a medical image file to be interpreted;
步骤 S20 ，采集通过语音方式输入的解读信号，识别解读信号，在预置的语音识别库中匹配解读信号对应的解读结果； Step S20 Acquiring an interpretation signal input by voice, identifying an interpretation signal, and matching an interpretation result corresponding to the interpretation signal in a preset speech recognition library;
步骤 S30 ，根据匹配的解读结果生成影像解读报告； Step S30, generating an image interpretation report according to the matched interpretation result;
步骤 S40 ，显示生成的影像解读报告。 Step S40, displaying the generated image interpretation report.
接收待解读的医学影像资料，该医学影像资料可包括患者身体各个部位对应的待解读的医学影像文件，以及患者的个人信息，如姓名、年龄、性别、病史等信息。 Receiving medical image data to be interpreted, the medical image data may include medical image files to be interpreted corresponding to various parts of the patient's body, and personal information of the patient, such as name, age, gender, medical history and the like.
在医学影像专家解读医学影像文件时，可以通过麦克风等用于语音输入的硬件设备输入语音信号形式的解读信号，采集该解读信号，对该解读信号进行语音识别，并根据识别出的解读信号，在预置的语音识别库中匹配与解读信号对应的解读结果，解读结果可包括正常的解读结果和异常的解读结果。然后，根据匹配出的解读结果，生成用于显示解读结果的影像解读报告。在生成影像解读报告时，可以将所接收到的患者的姓名、年龄、性别、病史等信息以模块化的形式嵌入到影像解读报告的前部，并且可在影像解读报告的末尾自动生成时间戳；同时，医学影像专家可以在影像解读报告中签署自己的姓名。本实施例中，在生成影像解读报告后，如发现影像解读报告中的解读结果有误，则可对错误的内容进行删除。 When the medical image expert interprets the medical image file, the interpretation signal in the form of a voice signal can be input through a hardware device for voice input such as a microphone, the interpretation signal is collected, the interpretation signal is voice-recognized, and according to the recognized interpretation signal, The interpretation result corresponding to the interpretation signal is matched in the preset speech recognition library, and the interpretation result may include a normal interpretation result and an abnormal interpretation result. Then, based on the matched interpretation result, an image interpretation report for displaying the interpretation result is generated. When generating the image interpretation report, the received patient's name, age, gender, medical history and other information can be embedded in the front of the image interpretation report in a modular form, and the time stamp can be automatically generated at the end of the image interpretation report. At the same time, medical imaging experts can sign their names in the image interpretation report. In this embodiment, after the image interpretation report is generated, if the interpretation result in the image interpretation report is found to be incorrect, the erroneous content may be deleted.
在生成影像解读报告后，显示该影像解读报告，即显示医学影像专家对接收到的医学影像文件进行解读的解读结果，当解读结果为正常时，则显示对应的部位为正常，当解读结果为异常时，则根据医学影像专家的解读显示具体的异常情况。 After the image interpretation report is generated, the image interpretation report is displayed, that is, the interpretation result of the medical image expert reading the received medical image file is displayed. When the interpretation result is normal, the corresponding part is displayed as normal, and the interpretation result is When an abnormality occurs, the specific abnormal situation is displayed according to the interpretation of the medical imaging expert.
本实施例在接收到待解读的医学影像文件后，在医学影像专家解读的过程中，采集通过语音方式输入的解读信号，识别解读信号并根据解读信号对应的解读结果生成影像解读报告，显示生成的影像解读报告。在生成影像解读报告时无需医学影像专家手动输入，通过语音识别的方式， 简化了医学影像阅读的操作，并且提高了医学影像解读的效率和准确性。 After receiving the medical image file to be interpreted, the embodiment collects the interpretation signal input by the voice mode in the process of reading the medical image expert, recognizes the interpretation signal, and generates an image interpretation report according to the interpretation result corresponding to the interpretation signal, and displays the generated image. Image interpretation report. No need for manual input by medical imaging experts when generating image interpretation reports, through voice recognition, It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
参照图 2 ， 图 2 为 图 1 中步骤 S20 的细化流程示意图。 Referring to FIG. 2, FIG. 2 is a schematic diagram of the refinement process of step S20 in FIG.
在上述实施例中，步骤 S20 具体包括： In the above embodiment, step S20 specifically includes:
步骤 S21 ，采集通过语音方式输入的解读信号，对解读信号进行滤波和降噪处理； Step S21: collecting an interpretation signal input by a voice mode, and performing filtering and noise reduction processing on the interpretation signal;
在医学影像专家解读医学影像文件时，采集到通过语音方式输入的解读信号后，首先对该原始的语音信号进行处理，如进行滤波处理和降噪处理，排除语音信号的干扰并提高信噪比。 When the medical image expert interprets the medical image file, after acquiring the interpretation signal input by the voice mode, the original voice signal is first processed, such as filtering processing and noise reduction processing, eliminating the interference of the voice signal and improving the signal to noise ratio. .
步骤 S22 ，识别解读信号，在预置的语音识别库中匹配与解读信号对应的文字，作为解读结果； Step S22, identifying the interpretation signal, and matching the text corresponding to the interpretation signal in the preset speech recognition library as the interpretation result;
对处理后的解读信号进行语音识别，并在预置的语音识别库中匹配与解读信号对应的文字，作为解读结果。本实施例中，预置的语音识别库中收集了常用的医学术语、放射学术语以及常见的解读结果的模版，通常情况下，模版所对应的为正常的解读结果。 The processed interpretation signal is speech-recognized, and the text corresponding to the interpretation signal is matched in the preset speech recognition library as the interpretation result. In this embodiment, the preset speech recognition library collects commonly used medical terms, radiological terms, and templates for common interpretation results. Generally, the template corresponds to a normal interpretation result.
具体地，在本实施例中，如接收到的解读信号中包括预置的第一语音标志和医学影像文件对应的部位，该第一语音标志作为判断解读结果是否为正常的指示标志，即当解读信号中包括该第一语音标志时，表明医学影像文件的解读结果为正常，该第一语音标志的内容通常选择为异于相关的医学术语和放射学术语的词语。例如，当接收到的解读信号包括'点击 + 肺部'，则表明肺部所对应的医学影像文件的解读结果为正常，此时，可在语音识别库中匹配与肺部对应的模板，例如匹配'肺部未见异常'作为正常的解读结果。 Specifically, in the embodiment, the received interpretation signal includes a preset first voice flag and a portion corresponding to the medical image file, and the first voice flag is used as a indicator for determining whether the interpretation result is normal, that is, when When the first speech mark is included in the interpretation signal, it indicates that the interpretation result of the medical image file is normal, and the content of the first speech mark is generally selected as a word different from related medical terms and radiological terms. For example, when the received interpretation signal includes 'click + lungs', indicating that the interpretation of the medical image files corresponding to the lungs is normal. At this time, the template corresponding to the lungs can be matched in the speech recognition library, for example, matching 'no abnormalities in the lungs' as normal. Interpret the results.
如接收到的解读信号中包括预置的第二语音标志，该第二语音标志作为判断解读结果是否为参考的解读结果的指示标志，即当解读信号中包括该第二语音标志时，表示此时需匹配与医学影像文件对应的参考的解读结果，该第二语音标志的内容可设置为'参考'等可以明显确定其指示的解读结果为参考的解读结果的词语。例如，当接收到的解读信号为'参考'加具体的影像解读报告的结论，则在语音识别库中匹配与影像解读报告的结论对应的文字，将该文字作为参考的解读结果。 If the received interpretation signal includes a preset second voice flag, the second voice flag is used as an indicator for determining whether the interpretation result is a reference interpretation result, that is, when the second voice flag is included in the interpretation signal, indicating It is necessary to match the interpretation result of the reference corresponding to the medical image file, and the content of the second voice flag can be set to a word such as 'reference', which can clearly determine the interpretation result of the instruction as the reference interpretation result. For example, when the received interpretation signal is a 'reference' plus a specific image interpretation report, the text corresponding to the conclusion of the image interpretation report is matched in the speech recognition library, and the text is used as a reference interpretation result.
如接收到的解读信号不包括预置的语音标志，即解读信号中不包括第一语音标志和第二语音标志，则可判断解读结果不是正常的解读结果，此时，识别完整的解读信号，并进一步根据解读信号在语音识别库中匹配与其对应的文字。此时，如匹配到的解读信号对应的文字与对医学影像文件的解读结果无关，则表明接收到的解读信号是无效的信号；相反，则表明解读信号对应的解读结果为异常的解读结果。例如，识别出的解读信号为'两肺可见数个软组织密度灶，形态与转移性癌相符'，则在语音识别库中匹配相应的文字，将该文字作为异常的解读结果。 If the received interpretation signal does not include a preset voice flag, that is, the first voice flag and the second voice flag are not included in the interpretation signal, it may be judged that the interpretation result is not a normal interpretation result. At this time, the complete interpretation signal is recognized. And further matching the corresponding text in the speech recognition library according to the interpretation signal. At this time, if the matching text corresponding to the interpretation signal has nothing to do with the interpretation result of the medical image file, it indicates that the received interpretation signal is an invalid signal; on the contrary, it indicates that the interpretation result corresponding to the interpretation signal is an abnormal interpretation result. For example, the recognized interpretation signal is 'a plurality of soft tissue density lesions visible in both lungs, and the morphology is consistent with metastatic cancer', and the corresponding text is matched in the speech recognition library, and the text is used as an abnormal interpretation result.
本实施例中，如在语音识别库中未匹配出与解读信号对应的解读结果，则表明该语音识别库中未存储有相应的文字，此时，医学影像专家可通过键盘或其他方式输入相应的解读结果至影像解读报告中，并将输入的解读结果添加至语音识别库中，以备在下次接收到相同或相似的解读信号时，从语音识别库中匹配对应的解读结果。 In this embodiment, if the interpretation result corresponding to the interpretation signal is not matched in the speech recognition library, it indicates that the corresponding speech is not stored in the speech recognition library. At this time, the medical imaging expert may input the corresponding through the keyboard or other means. The interpretation result is added to the image interpretation report, and the input interpretation result is added to the speech recognition library, so as to match the corresponding interpretation result from the speech recognition library when the same or similar interpretation signal is received next time.
在上述 本发明 基于智能语音识别的医学影像解读方法第 一实施例的基础上，第二实施例中，该方法还包括： In the above invention, the method for reading medical images based on intelligent speech recognition On the basis of an embodiment, in the second embodiment, the method further includes:
在预置的医学影像数据库中匹配与医学影像文件相似度在预置范围内的相似医学影像数据，并通过语音方式输入相似医学影像数据对应的影像解读报告的结论，作为解读信号。 The medical image database is matched with similar medical image data whose medical image file has similarity within a preset range, and the conclusion of the image interpretation report corresponding to the similar medical image data is input by voice as an interpretation signal.
本实施例中，预设一医学影像数据库，该医学影像数据库中存储了常用的各种正常或异常的医学影像数据可供参考，在接收到待解读的医学影像文件后，可以首先对医学影像文件的医学影像数据和医学影像数据库中的医学影像数据进行相似度匹配，当相似度在预置的范围内时，获取医学影像数据库中的医学影像数据作为相似医学影像数据，预置的范围可以根据医学影像文件的类型和相关部位自定义设置，例如设置为大于 80% 。 In this embodiment, a medical image database is preset, and the medical image database stores various commonly used normal or abnormal medical image data for reference. After receiving the medical image file to be interpreted, the medical image may be firstly used. The medical image data of the document and the medical image data in the medical image database are similarly matched. When the similarity is within the preset range, the medical image data in the medical image database is acquired as similar medical image data, and the preset range may be Customize settings according to the type of medical image file and related parts, for example, set to be larger than 80%.
获取到相似医学影像数据后，根据该相似医学影像数据对应的医学影像文件与影像解读报告的结论的对应关系，获取该医学影像文件的影像解读报告的结论，医学影像专家可通过语音输入的方式，输入解读报告的结论作为解读信号。影像解读报告的结论中除了包括医学影像专家对患者的医学影像文件的解读结果，还可包括如采用的治疗手段及康复情况等信息，在生成待解读的医学影像文件的解读报告时，可将这些信息与解读报告同时进行显示，以供医生和患者参考。 After obtaining similar medical image data, according to the correspondence between the medical image file corresponding to the similar medical image data and the conclusion of the image interpretation report, the conclusion of the image interpretation report of the medical image file is obtained, and the medical imaging expert can input the voice through voice Enter the conclusion of the interpretation report as the interpretation signal. The conclusions of the image interpretation report include the interpretation results of the medical image files of the patient by the medical imaging experts, and may include information such as the treatment means and the rehabilitation situation, and when generating the interpretation report of the medical image file to be interpreted, This information is displayed along with the interpretation report for reference by doctors and patients.
在解读待解读的医学影像文件时，在医学影像数据库中匹配与该医学影像文件的医学影像数据的相似度在预置范围内的医学影像数据，获取医学影像数据库及其对应的影像解读报告的结论，将该结论通过语音输入方式作为解读信号输入，供医生和患者参考，使患者可了解到与自身情况的相似情况的治疗方式，从而为患者提供了便利，使医生依据以往的病例为本次病例的治疗方案提供参考。 When interpreting the medical image file to be interpreted, matching the medical image data with the similarity of the medical image data of the medical image file in the medical image database to obtain the medical image database and the corresponding image interpretation report In conclusion, this conclusion is used as a reading signal input by the voice input method for the doctor and the patient to refer to, so that the patient can understand the treatment mode similar to his own situation, thereby providing convenience for the patient, and making the doctor based on the previous case. The treatment plan for the second case provides a reference.
基于本发明上述第一、第二实施例，步骤 S40 具体为： Based on the foregoing first and second embodiments of the present invention, step S40 is specifically:
对影像解读报告中的正常的解读结果和 / 或异常的解读结果和 / 或参考解读结果进行标识，以预置的显示方式显示所述正常的解读结果和 / 或异常的解读结果和 / 或参考解读结果。 Normal interpretation results and/or abnormal interpretation results in the image interpretation report and / Or by referring to the interpretation result, the normal interpretation result and/or the abnormal interpretation result and/or the reference interpretation result are displayed in a preset display manner.
在预置的语音识别库中匹配到与医学影像文件的部位对应的解读结果后，根据解读结果是正常还是异常，对正常的解读结果和异常的解读结果进行不同的标识，同时可以将参考结果标识为能够与正常或异常的解读结果区分的其他标识，如将不同的解读结果标记为不同的字体或不同的颜色，以便在显示时，以预置的显示方式进行显示，例如，对异常的解读结果的文字内容以加粗形式显示或进行突出显示。本实施例中，如影像解读报告中的解读结果为正常的解读结果，则可以在匹配到代表解读结果为正常的解读模版后，对该解读模版对应的文字内容标识为蓝色字体，并正常显示；如影像解读报告中的解读结果为异常的解读结果，则可以在匹配到异常的解读结果对应的文字后，将该文字内容标识为红色字体，并在显示时将该红色字体的文字内容以加粗形式显示或进行突出显示，以提醒患者注意。 After matching the interpretation result corresponding to the part of the medical image file in the preset speech recognition library, according to whether the interpretation result is normal or abnormal, the normal interpretation result and the abnormal interpretation result are differently identified, and the reference result can be obtained at the same time. It is identified as other identifiers that can be distinguished from normal or abnormal interpretation results, such as marking different interpretation results as different fonts or different colors, so as to be displayed in a preset display manner when displayed, for example, for anomalies The text content of the interpretation result is displayed in bold or highlighted. In this embodiment, if the interpretation result in the image interpretation report is a normal interpretation result, the text content corresponding to the interpretation template may be identified as a blue font after being matched to the interpretation template indicating that the interpretation result is normal, and is normal. If the interpretation result in the image interpretation report is an abnormal interpretation result, the text content may be identified as a red font after matching the text corresponding to the abnormal interpretation result, and the text content of the red font is displayed at the time of display. Displayed in bold or highlighted to alert the patient.
在生成影像解读报告时，对影像解读报告中的正常的解读结果和 / 或异常的解读结果进行不同的标识，并以预置的显示方式显示正常的解读结果和 / 或异常的解读结果，使得影像解读报告的显示更为清晰，方便患者阅读。 When the image interpretation report is generated, the normal interpretation result in the image interpretation report and / Or abnormal interpretation results are marked differently, and the normal interpretation results and/or abnormal interpretation results are displayed in a preset display manner, so that the image interpretation report is displayed more clearly and is convenient for the patient to read.
参照图 3 ， 图 3 为本发明 基于智能语音识别的医学影像解读方法第 三实施例的流程示意图。 Referring to FIG. 3, FIG. 3 is a schematic flowchart diagram of a third embodiment of a medical image interpretation method based on intelligent speech recognition.
基于 本发明 基于智能语音识别的医学影像解读方法第 一实施例，在 第 三实施例中，该方法还包括： Based on the first embodiment of the method for reading a medical image based on the intelligent speech recognition of the present invention, in the third embodiment, the method further includes:
步骤 S50 ，接收通过语音输入的对医学影像文件进行标记的标记指令，根据该标记指令标记相应的医学影像文件，以供在查找时接收到查找指令后跳转至相应的医学影像文件。 Step S50 And receiving a marking instruction for marking the medical image file by voice input, and marking the corresponding medical image file according to the marking instruction for jumping to the corresponding medical image file after receiving the searching instruction at the time of searching.
在本实施例中，在医学影像专家解读医学影像文件的过程中，由于每个部位所对应的医学影像文件可能有多张，因此，在解读的过程中，如对某一张医学影像文件有疑问，或欲将该张医学影像文件作为重点，希望在查看完全部医学影像文件后查找到该张医学影像文件并重新进行解读，则医学影像专家可以通过语音输入的方式输入标记指令，如输入'重点 1 '或'标记第 42 '张，这样，在接收到该标记指令后，便对相应的医学影像文件进行标记。在医学影像专家查找之前所标记的医学影像文件时，同样可通过语音输入的方式输入查找指令，如输入'跳转至重点 1 '或'跳转至第 42 张'，便可查找到标记的相应医学影像文件，并显示该张医学影像文件。该步骤 S50 可以在解读医学影像文件的任何时候进行，本实施例优选在上述第一实施例中执行步骤 S30 之前执行，即在开始解读医学影像文件并显示生成的影像解读报告之前的任何时候，对相应的医学影像文件进行标记。 In this embodiment, in the process of interpreting the medical image file by the medical imaging expert, since there may be more than one medical image file corresponding to each part, in the process of interpretation, for example, for a certain medical image file Doubt, or want to focus on the medical image file, hope to find the medical image file and read it again after viewing the complete medical image file, then the medical imaging expert can input the marking instruction by voice input, such as input 'emphasis 1 'or' mark 42 'Zhang, in this way, after receiving the marking instruction, the corresponding medical image file is marked. When a medical imaging expert searches for a previously marked medical image file, the search command can also be input by voice input, such as input 'jump to focus 1 'or 'Go to the 42nd' to find the corresponding medical image file of the mark and display the medical image file. This step S50 It can be performed at any time when the medical image file is interpreted. In this embodiment, step S30 is preferably performed in the above-described first embodiment. Execution before, that is, marking the corresponding medical image file at any time before beginning to interpret the medical image file and display the generated image interpretation report.
当接收到对医学影像文件进行标记的标记指令，根据该标记指令标记相应的医学影像文件，可以方便在接收到查找指令后精准地跳转至相应的医学影像文件，从而进一步 简化了医学影像阅读的操作，并且提高了医学影像解读的效率和准确性。 When receiving the marking instruction for marking the medical image file, marking the corresponding medical image file according to the marking instruction, it is convenient to accurately jump to the corresponding medical image file after receiving the searching instruction, thereby further It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
本发明还提供一种 基于智能语音识别的医学影像解读系统 。 The invention also provides a medical image interpretation system based on intelligent speech recognition.
参照图 4 ， 图 4 为本发明 基于智能语音识别的医学影像解读系统第一优选实施例的系统结构 示意图。 Referring to FIG. 4, FIG. 4 is a system structure of a first preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention. Schematic.
在一实施例中， 基于智能语音识别的医学影像解读系统 包括： In an embodiment, the medical image interpretation system based on intelligent speech recognition comprises:
语音识别装置 10 、多功能鼠标 20 、踏板 30 以及信号处理装置 40 ，其中： a voice recognition device 10, a multi-function mouse 20, a pedal 30, and a signal processing device 40, wherein:
所述 语音识别装置 10 ，与所述信号处理装置 40 连接，用于采集并识别语音输入的解读信号，根据解读信号匹配对应的解读结果，并将所述解读结果输出至所述信号处理装置，供信号处理装置根据解读结果生成影像解读报告； The voice recognition device 10 and the signal processing device 40 Connecting, for acquiring and recognizing an interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and outputting the interpretation result to the signal processing device, where the signal processing device generates an image interpretation report according to the interpretation result;
所述多功能鼠标 20 ，与所述信号处理装置 40 连接，包括至少一个功能按键，用于在生成影像解读报告时对所述影像解读报告的内容进行编辑； The multifunctional mouse 20 and the signal processing device 40 The connection includes at least one function button for editing the content of the image interpretation report when generating the image interpretation report;
所述踏板 30 ，与所述信号处理装置 40 连接，用于在解读医学影像文件时对医学影像文件进行翻页； The pedal 30 and the signal processing device 40 Connection for turning a medical image file when interpreting a medical image file;
所述信号处理装置 40 ，用于接收所述解读结果，根据解读结果生成影像解读报告并输出；以及，在接收到所述多功能鼠标和踏板输入的信号时执行相应的操作。 The signal processing device 40 And receiving the interpretation result, generating an image interpretation report according to the interpretation result and outputting; and performing a corresponding operation when receiving the signal of the multifunctional mouse and the pedal input.
在医生进行医学影像资料阅读时，通过踏板 30 对医学影像文件进行翻页，在看到某个部位正常或异常时，向语音识别装置 10 输入对某个部位评价的语音输入信号， 通过 语音识别装置 10 采集并识别医生对某个部位评价的语音输入信号，并识别为解读信号，根据解读信号在预置的语音识别库中匹配与解读信号对应的文字，作为解读结果。 When the doctor reads the medical image data, pass the pedal 30 When the medical image file is turned over, when a certain part is normal or abnormal, a voice input signal for evaluating a certain part is input to the voice recognition device 10, and the voice recognition device 10 is passed. The voice input signal evaluated by the doctor for a certain part is collected and recognized, and is recognized as an interpretation signal, and the text corresponding to the interpretation signal is matched in the preset voice recognition library according to the interpretation signal as the interpretation result.
在匹配出对应的解读结果之后，将所述解读结果输出至所述信号处理装置 40 ，供信号处理装置 40 根据解读结果生成影像解读报告；在生成影像解读报告时可以通过多功能鼠 20 标对所述影像解读报告的内容进行编辑，例如可对生成的影像解读报告进行删除、修改、保存等。 After the corresponding interpretation result is matched, the interpretation result is output to the signal processing device 40 for the signal processing device 40. Generate an image interpretation report based on the interpretation result; you can pass the multi-function mouse when generating the image interpretation report. The content of the image interpretation report is edited, for example, the generated image interpretation report can be deleted, modified, saved, and the like.
本发明实施例 通过所述语音识别装置 10 采集并识别语音输入的解读信号，根据解读信号匹配对应的解读结果，并将所述解读结果输出至所述信号处理装置 40 ，供所述信号处理装置 40 根据解读结果生成影像解读报告；通过所述多功能鼠标 20 在生成影像解读报告时对所述影像解读报告的内容进行编辑；通过所述踏板 30 在解读医学影像文件时对医学影像文件进行翻页。本发明提供的基于智能语音识别的医学影像解读系统结构实用简单， 简化了医学影像阅读的操作，并且提高了医学影像解读的效率和准确性。 Embodiment of the present invention, by the voice recognition device 10 Acquiring and recognizing the interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and outputting the interpretation result to the signal processing device 40 for the signal processing device 40 Generating an image interpretation report according to the interpretation result; editing, by the multi-function mouse 20, the content of the image interpretation report when generating the image interpretation report; Turn the medical image file into pages when interpreting medical image files. The medical image interpretation system based on intelligent speech recognition provided by the invention is simple and practical in structure. It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
参照图 5 ， 图 5 为本发明 基于智能语音识别的医学影像解读系统第二优选实施例的系统结构 示意图。 Referring to FIG. 5, FIG. 5 is a system structure of a second preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention. Schematic.
在一实施例中，所述 基于智能语音识别的医学影像解读系统在第一优选实施例的基础上， 优选地， 所述 语音识别装置 10 包括： In an embodiment, the intelligent speech recognition based medical image interpretation system is based on the first preferred embodiment, preferably, The voice recognition device 10 includes:
语音采集模块 101 ，用于采集 通过语音方式输入的解读信号 ；所述语音采集模块 101 通常为麦克风等能够采集语音信号的设备，在一优选实施例中， 采集到通过语音方式输入的解读信号后，所述 语音采集模块 101 首先对该原始的语音信号进行处理，如进行滤波处理和降噪处理，排除语音信号的干扰并提高信噪比 ； The voice collection module 101 is configured to collect an interpretation signal input by voice; the voice collection module 101 Generally, it is a device capable of collecting a voice signal, such as a microphone. In a preferred embodiment, after acquiring an interpretation signal input by voice, the voice collection module 101 First, the original speech signal is processed, such as filtering processing and noise reduction processing, eliminating interference of the speech signal and improving the signal to noise ratio;
匹配模块 102 ，与所述信号处理装置 40 连接，用于 识别所述解读信号，根据所述解读信号在预置的语音识别库中匹配对应的解读结果，并将所述解读结果输出至所述信号处理装置 40 。在一实施例中，预置的语音识别库中收集了常用的医学术语、放射学术语以及常见的解读结果的模版，通常情况下，模版所对应的为正常的解读结果。 a matching module 102, connected to the signal processing device 40, for Identifying the interpretation signal, matching a corresponding interpretation result in a preset speech recognition library according to the interpretation signal, and outputting the interpretation result to the signal processing device 40 . In one embodiment, the preset speech recognition library collects commonly used medical terms, radiological terms, and templates for common interpretation results. Typically, the template corresponds to a normal interpretation result.
具体地，如解读信号中包括预置的语音标志和医学影像文件对应的部位，该预置的语音标志作为判断解读结果是否为正常的指示标志，即当解读信号中包括该语音标志时，表明医学影像文件的解读结果为正常，该语音标志的内容通常选择为异于相关的医学术语和放射学术语的词语。例如，当接收到的解读信号包括'点击 + 肺部'，则表明肺部所对应的医学影像文件的解读结果为正常，此时，可在语音识别库中匹配与肺部对应的模板，例如匹配'肺部未见异常'作为正常的解读结果。 Specifically, if the pre-interpretation signal includes a preset voice symbol and a portion corresponding to the medical image file, the preset voice flag is used as an indicator for determining whether the interpretation result is normal, that is, when the speech signal is included in the interpretation signal, indicating The interpretation of the medical image file is normal, and the content of the speech mark is usually chosen to be different from the related medical terms and radiological terms. For example, when the received interpretation signal includes 'click + lungs', indicating that the interpretation of the medical image files corresponding to the lungs is normal. At this time, the template corresponding to the lungs can be matched in the speech recognition library, for example, matching 'no abnormalities in the lungs' as normal. Interpret the results.
如解读信号不包括预置的语音标志，则可判断解读结果不是正常的解读结果，此时，识别完整的解读信号，并进一步根据解读信号在语音识别库中匹配与其对应的文字。此时，如匹配到的解读信号对应的文字与对医学影像文件的解读结果无关，则表明接收到的解读信号是无效的信号；相反，则表明解读信号对应的解读结果为异常的解读结果。例如，识别出的解读信号为'两肺可见数个软组织密度灶，形态与转移性癌相符'，则在语音识别库中匹配相应的文字，将该文字作为异常的解读结果。 If the interpretation signal does not include the preset voice flag, it can be judged that the interpretation result is not a normal interpretation result. At this time, the complete interpretation signal is recognized, and the corresponding text is further matched in the speech recognition library according to the interpretation signal. At this time, if the matching text corresponding to the interpretation signal has nothing to do with the interpretation result of the medical image file, it indicates that the received interpretation signal is an invalid signal; on the contrary, it indicates that the interpretation result corresponding to the interpretation signal is an abnormal interpretation result. For example, the recognized interpretation signal is 'a plurality of soft tissue density lesions visible in both lungs, and the morphology is consistent with metastatic cancer', and the corresponding text is matched in the speech recognition library, and the text is used as an abnormal interpretation result.
在一优选的实施例中，所述多功能鼠标 20 的功能按键可以根据实际需求设置为退格键、空行键、空格键、删除键、缩放键、选择键、保存键中的一种或多种。具体地，可通过编程的方式设置所述多功能鼠标的功能按键的具体功能，或者通过选择的方式配置所述多功能鼠标的功能按键的具体功能。 In a preferred embodiment, the multifunctional mouse 20 The function keys can be set to one or more of a backspace key, a blank key, a space key, a delete key, a zoom key, a selection key, and a save key according to actual needs. Specifically, the specific function of the function button of the multi-function mouse may be set in a programmatic manner, or the specific function of the function button of the multi-function mouse may be configured in a selected manner.
在一优选的实施例中，所述踏板 30 包括左踏板和右踏板，所述左踏板用于 在解读医学影像文件时进行向前 / 向后翻页，所述 右踏板用于 在解读医学影像文件时进行向后 / 向前翻页，设置左右两个踏板更方便医生进行向前 / 向后翻页。 In a preferred embodiment, the pedal 30 includes a left pedal and a right pedal for forwarding when interpreting medical image files / Page backwards, the right pedal is used to page backward/forward when interpreting medical image files, and setting the left and right pedals makes it easier for doctors to turn forward/backward.
在一优选的实施例中， 所述信号处理装置具体用于： In a preferred embodiment, the signal processing device is specifically configured to:
在接收到所述多功能鼠标 对影像解读报告的内容进行编辑的编辑信号时执行相应的编辑操作，例如，进行删除、修改、保存、放大、缩小等操作； Receiving the multifunctional mouse Perform editing operations when editing the edited signal of the content of the image interpretation report, for example, performing operations such as deleting, modifying, saving, enlarging, and reducing;
以及，在接收到所述踏板对医学影像文件进行翻页的翻页信号时执行翻页操作，例如，进行前翻页、后翻页操作。 And, when the page turning signal for turning the page of the medical image file is received, the page turning operation is performed, for example, the page turning forward and the page turning operation are performed.
本发明实施例通过 语音识别装置 10 ，采集并识别语音输入的解读信号，根据解读信号匹配对应的解读结果，通过预置的语音识别库，对医生阅读医学影像资料过程中的解读结果进行快速匹配，供信号处理装置 40 根据解读结果生成影像解读报告；在生成影像解读报告时通过多功能鼠标 20 对所述影像解读报告的内容进行编辑，通过对所述多功能鼠标 20 功能按键功能的设置，能够方便医生对解读报告的内容进行快速编辑；在解读医学影像文件时通过踏板 30 对医学影像文件进行翻页，通过设置左踏板、右踏板，方便医生进行前翻页、后翻页。本发明提供的基于智能语音识别的医学影像解读系统结构实用简单， 简化了医学影像阅读的操作，并且提高了医学影像解读的效率和准确性。 The embodiment of the present invention passes the voice recognition device 10 Acquiring and recognizing the interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and quickly matching the interpretation result in the process of reading the medical image data by the doctor through the preset speech recognition library, for the signal processing device 40 And generating an image interpretation report according to the interpretation result; editing the content of the image interpretation report by using the multi-function mouse 20 when generating the image interpretation report, by using the multifunctional mouse 20 The function button function is set to facilitate the doctor to quickly edit the contents of the interpretation report; the pedal is used when interpreting the medical image file. Turning the medical image file, by setting the left pedal and the right pedal, it is convenient for the doctor to turn the page forward and then turn the page. The medical image interpretation system based on intelligent speech recognition provided by the invention is simple and practical in structure. It simplifies the operation of medical image reading and improves the efficiency and accuracy of medical image interpretation.
参照图 6 ， 图 6 为本发明 基于智能语音识别的医学影像解读系统第三优选实施例的系统结构 示意图。 6 is a system structure of a third preferred embodiment of a medical image interpretation system based on intelligent speech recognition according to the present invention. Schematic.
在一实施例中，所述 基于智能语音识别的医学影像解读系统在第二优选实施例的基础上， 进一步地， 所述 基于智能语音识别的医学影像解读系统还包括： In an embodiment, the intelligent speech recognition based medical image interpretation system is based on the second preferred embodiment, and further, The medical image interpretation system based on intelligent speech recognition further includes:
显示装置 50 ，与所述 信号处理装置 40 连接，用于接收 所述 信号处理装置 40 输出的影像解读报告并显示。具体地， 在对影像解读报告输出并显示时，所述 信号处理装置 40 可对正常的解读结果和 / 或异常的解读结果进行标识，以预置的显示方式在所述显示装置 50 上显示正常的解读结果和 / 或异常的解读结果。 a display device 50 coupled to the signal processing device 40 for receiving the signal processing device 40 The output image interpretation report is displayed. Specifically, when the image interpretation report is output and displayed, the signal processing device 40 can correctly interpret the result and/or Or the abnormal interpretation result is identified, and the normal interpretation result and/or the abnormal interpretation result are displayed on the display device 50 in a preset display manner.
具体地，在预置的语音识别库中匹配到与医学影像文件的部位对应的解读结果后，根据解读结果是正常还是异常，对正常的解读结果和异常的解读结果进行不同的标识，如将不同的解读结果标记为不同的字体或不同的颜色，以便在显示时，以预置的显示方式进行显示，例如，对异常的解读结果的文字内容以加粗形式显示或进行突出显示。本实施例中，如影像解读报告中的解读结果为正常的解读结果，则可以在匹配到代表解读结果为正常的解读模版后，对该解读模版对应的文字内容标识为蓝色字体，并正常显示；如影像解读报告中的解读结果为异常的解读结果，则可以在匹配到异常的解读结果对应的文字后，将该文字内容标识为红色字体，并在显示时将该红色字体的文字内容以加粗形式显示或进行突出显示，以提醒患者注意。 Specifically, after matching the interpretation result corresponding to the part of the medical image file in the preset speech recognition library, according to whether the interpretation result is normal or abnormal, the normal interpretation result and the abnormal interpretation result are differently marked, for example, Different interpretation results are marked as different fonts or different colors so as to be displayed in a preset display manner when displayed, for example, the text content of the abnormal interpretation result is displayed in bold or highlighted. In this embodiment, if the interpretation result in the image interpretation report is a normal interpretation result, the text content corresponding to the interpretation template may be identified as a blue font after being matched to the interpretation template indicating that the interpretation result is normal, and is normal. If the interpretation result in the image interpretation report is an abnormal interpretation result, the text content may be identified as a red font after matching the text corresponding to the abnormal interpretation result, and the text content of the red font is displayed at the time of display. Displayed in bold or highlighted to alert the patient.
在生成影像解读报告时，对影像解读报告中的正常的解读结果和 / 或异常的解读结果进行不同的标识，并以预置的显示方式显示正常的解读结果和 / 或异常的解读结果，使得影像解读报告的显示更为清晰，方便患者阅读。 When the image interpretation report is generated, the normal interpretation result in the image interpretation report and / Or abnormal interpretation results are marked differently, and the normal interpretation results and/or abnormal interpretation results are displayed in a preset display manner, so that the image interpretation report is displayed more clearly and is convenient for the patient to read.
以上仅为本发明的优选实施例，并非因此限制本发明的专利范围，凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本发明的专利保护范围内。 The above are only the preferred embodiments of the present invention, and are not intended to limit the scope of the invention, and the equivalent structure or equivalent process transformations made by the description of the present invention and the drawings are directly or indirectly applied to other related technical fields. The same is included in the scope of patent protection of the present invention.

Claims (15)

  1. 一种 基于智能语音识别的医学影像解读方法， 其特征在于，所述 基于智能语音识别的医学影像解读方法包括以下步骤：Medical image interpretation method based on intelligent speech recognition, characterized in that The medical image interpretation method based on intelligent speech recognition includes the following steps:
    接收待解读的医学影像文件；Receiving a medical image file to be interpreted;
    采集通过语音方式输入的解读信号，识别所述解读信号，在预置的语音识别库中匹配解读信号对应的解读结果；所述解读信号为对所述待解读的医学影像文件进行解读时所输入的语音信号；Acquiring an interpretation signal input by voice, identifying the interpretation signal, and matching an interpretation result corresponding to the interpretation signal in a preset speech recognition library; the interpretation signal is input when the medical image file to be interpreted is interpreted Voice signal
    根据匹配的所述解读结果生成影像解读报告；Generating an image interpretation report according to the matched interpretation result;
    显示生成的所述影像解读报告。 The generated image interpretation report is displayed.
  2. 如权利要求 1 所述的基于智能语音识别的医学影像解读方法， 其特征在于，所述 基于智能语音识别的医学影像解读方法还包括步骤：The method for reading medical image based on intelligent speech recognition according to claim 1, wherein said The medical image interpretation method based on intelligent speech recognition further includes the steps of:
    在预置的医学影像数据库中匹配与所述医学影像文件相似度在预置范围内的相似医学影像数据，并通过语音方式输入所述相似医学影像数据对应的影像解读报告的结论，作为解读信号。Matching similar medical image data whose similarity with the medical image file is within a preset range in a preset medical image database, and inputting the conclusion of the image interpretation report corresponding to the similar medical image data by using a voice as an interpretation signal .
  3. 如权利要求 2 所述的基于智能语音识别的医学影像解读方法， 其特征在于，所述 识别所述解读信号，在预置的语音识别库中匹配解读信号对应的解读结果的步骤包括：The method for reading medical image based on intelligent speech recognition according to claim 2, wherein The step of identifying the interpretation signal and matching the interpretation result corresponding to the interpretation signal in the preset speech recognition library includes:
    当所述解读信号包括预置的第一语音标志和医学影像文件对应的部位时，在所述预置的语音识别库中匹配与该部位对应的解读模版，作为正常的解读结果；When the interpretation signal includes a preset first voice mark and a portion corresponding to the medical image file, matching the interpretation template corresponding to the part in the preset voice recognition library as a normal interpretation result;
    当所述解读信号不包括预置的语音标志时，在所述预置的语音识别库中匹配所述解读信号对应的文字，作为异常的解读结果；When the interpretation signal does not include a preset voice flag, matching the text corresponding to the interpretation signal in the preset voice recognition library as an abnormal interpretation result;
    当所述解读信号包括预置的第二语音标志时，在预设的医学影像数据库中匹配所述解读信号对应的文字，作为参考的解读结果。When the interpretation signal includes the preset second voice flag, the text corresponding to the interpretation signal is matched in the preset medical image database as a reference interpretation result.
  4. 如权利要求 3 所述的基于智能语音识别的医学影像解读方法， 其特征在于，所述 显示生成的所述影像解读报告的步骤包括：The method for interpreting medical image based on intelligent speech recognition according to claim 3, wherein The steps of displaying the generated image interpretation report include:
    对影像解读报告中的正常的解读结果和 / 或异常的解读结果和 / 或参考解读结果进行标识，以预置的显示方式显示所述正常的解读结果和 / 或异常的解读结果和 / 或参考解读结果。Normal interpretation results and/or abnormal interpretation results in the image interpretation report and / Or by referring to the interpretation result, the normal interpretation result and/or the abnormal interpretation result and/or the reference interpretation result are displayed in a preset display manner.
  5. 如权利要求 1 所述的基于智能语音识别的医学影像解读方法， 其特征在于， 所述基于智能语音识别的医学影像解读方法还包括步骤：The method for reading medical image based on intelligent speech recognition according to claim 1, wherein: The medical image interpretation method based on intelligent speech recognition further comprises the steps of:
    接收通过语音输入的对医学影像文件进行标记的标记指令，根据该标记指令标记相应的医学影像文件，以供在查找时接收到查找指令后跳转至相应的医学影像文件。Receiving a mark instruction for marking the medical image file by voice input, and marking the corresponding medical image file according to the mark instruction for jumping to the corresponding medical image file after receiving the search instruction at the time of searching.
  6. 一种 基于智能语音识别的医学影像解读系统， 其特征在于，所述 基于智能语音识别的医学影像解读系统包括语音识别装置、多功能鼠标、踏板以及信号处理装置，其中：Medical image interpretation system based on intelligent speech recognition, characterized in that The medical image interpretation system based on intelligent speech recognition includes a speech recognition device, a multi-function mouse, a pedal and a signal processing device, wherein:
    所述 语音识别装置，与所述信号处理装置连接，用于采集并识别语音输入的解读信号，根据解读信号匹配对应的解读结果，并将所述解读结果输出至所述信号处理装置，供信号处理装置根据解读结果生成影像解读报告；Said a voice recognition device connected to the signal processing device for collecting and recognizing an interpretation signal of the voice input, matching the corresponding interpretation result according to the interpretation signal, and outputting the interpretation result to the signal processing device for the signal processing device Generate an image interpretation report based on the interpretation result;
    所述多功能鼠标，与所述信号处理装置连接，包括至少一个功能按键，用于在生成影像解读报告时对所述影像解读报告的内容进行编辑；The multi-function mouse is connected to the signal processing device, and includes at least one function button for editing the content of the image interpretation report when generating the image interpretation report;
    所述踏板，与所述信号处理装置连接，用于在解读医学影像文件时对医学影像文件进行翻页；The pedal is connected to the signal processing device for turning a medical image file when interpreting the medical image file;
    所述信号处理装置，用于接收所述解读结果，根据解读结果生成影像解读报告并输出；以及，在接收到所述多功能鼠标和踏板输入的信号时执行相应的操作。The signal processing device is configured to receive the interpretation result, generate an image interpretation report according to the interpretation result, and output; and perform a corresponding operation when receiving the signal input by the multifunctional mouse and the pedal.
  7. 如权利要求 6 所述的基于智能语音识别的医学影像解读系统， 其特征在于，所述 基于智能语音识别的医学影像解读系统还包括：The medical image interpretation system based on intelligent speech recognition according to claim 6, wherein said The medical image interpretation system based on intelligent speech recognition further includes:
    显示装置，与所述 信号处理装置连接，用于接收 所述 信号处理装置输出的影像解读报告并显示。And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  8. 如权利要求6所述的基于智能语音识别的医学影像解读系统，其特征在于，所述多功能鼠标的功能按键可以设置为退格键、空行键、空格键、删除键、缩放键、选择键、保存键中的一种或多种。 The medical image interpretation system based on intelligent speech recognition according to claim 6, wherein the function keys of the multi-function mouse can be set as a backspace key, a blank line key, a space key, a delete key, a zoom key, and a selection. One or more of the keys and save keys.
  9. 权利要求8所述的基于智能语音识别的医学影像解读系统，其特征在于，所述基于智能语音识别的医学影像解读系统还包括：The medical image interpretation system based on intelligent speech recognition according to claim 8, wherein the medical image interpretation system based on intelligent speech recognition further comprises:
    显示装置，与所述信号处理装置连接，用于接收所述信号处理装置输出的影像解读报告并显示。And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  10. 如权利要求6所述的基于智能语音识别的医学影像解读系统，其特征在于，所述踏板包括左踏板和右踏板，所述左踏板用于在解读医学影像文件时向前/向后翻页，所述右踏板用于在解读医学影像文件时向后/向前翻页。The intelligent speech recognition based medical image interpretation system of claim 6 wherein said pedal comprises a left pedal and a right pedal, said left pedal being used to page forward/backward when interpreting medical image files The right pedal is used to page backward/forward when interpreting medical image files.
  11. 权利要求10所述的基于智能语音识别的医学影像解读系统，其特征在于，所述基于智能语音识别的医学影像解读系统还包括：The medical image interpretation system based on intelligent speech recognition according to claim 10, wherein the medical image interpretation system based on intelligent speech recognition further comprises:
    显示装置，与所述信号处理装置连接，用于接收所述信号处理装置输出的影像解读报告并显示。And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  12. 权利要求6所述的基于智能语音识别的医学影像解读系统，其特征在于，所述语音识别装置包括：The intelligent speech recognition-based medical image interpretation system of claim 6, wherein the speech recognition device comprises:
    语音采集模块，用于采集通过语音方式输入的解读信号；a voice collection module, configured to collect an interpretation signal input by voice;
    匹配模块，与所述信号处理装置连接，用于识别所述解读信号，根据所述解读信号在预置的语音识别库中匹配对应的解读结果，并将所述解读结果输出至所述信号处理装置。a matching module, coupled to the signal processing device, for identifying the read signal, matching a corresponding interpretation result in a preset speech recognition library according to the read signal, and outputting the interpretation result to the signal processing Device.
  13. 权利要求12所述的基于智能语音识别的医学影像解读系统，其特征在于，所述基于智能语音识别的医学影像解读系统还包括：The intelligent speech recognition based medical image interpretation system of claim 12, wherein the intelligent speech recognition based medical image interpretation system further comprises:
    显示装置，与所述信号处理装置连接，用于接收所述信号处理装置输出的影像解读报告并显示。And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
  14. 如权利要求6所述的基于智能语音识别的医学影像解读系统，其特征在于，所述信号处理装置具体用于：The intelligent speech recognition based medical image interpretation system according to claim 6, wherein the signal processing device is specifically configured to:
    在接收到所述多功能鼠标对影像解读报告的内容进行编辑的编辑信号时执行相应的编辑操作；Performing a corresponding editing operation when receiving the editing signal that the multi-function mouse edits the content of the image interpretation report;
    以及，在接收到所述踏板对医学影像文件进行翻页的翻页信号时执行翻页操作。And, the page turning operation is performed when the page turning signal for turning the page of the medical image file is received.
  15. 如权利要求14所述的基于智能语音识别的医学影像解读系统，其特征在于，所述基于智能语音识别的医学影像解读系统还包括：The medical image interpretation system based on intelligent speech recognition according to claim 14, wherein the medical image interpretation system based on intelligent speech recognition further comprises:
    显示装置，与所述信号处理装置连接，用于接收所述信号处理装置输出的影像解读报告并显示。And a display device connected to the signal processing device for receiving and displaying an image interpretation report output by the signal processing device.
PCT/CN2014/090864 2014-11-04 2014-11-12 Medical image interpretation method and system based on intelligent speech recognition WO2016070445A1 (en)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201410614512.7A CN104462763B (en) 2014-11-04 2014-11-04 Medical image interpretation method and device based on intelligent speech recognition
CN201420653929.XU CN204233142U (en) 2014-11-04 2014-11-04 Intelligent medical image reading system
CN201420653929.X 2014-11-04
CN201410614512.7 2014-11-04

Publications (1)

Publication Number Publication Date
WO2016070445A1 true WO2016070445A1 (en) 2016-05-12

Family

ID=55908451

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/090864 WO2016070445A1 (en) 2014-11-04 2014-11-12 Medical image interpretation method and system based on intelligent speech recognition

Country Status (1)

Country Link
WO (1) WO2016070445A1 (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1296705A (en) * 1999-03-05 2001-05-23 皇家菲利浦电子有限公司 Ultrasonic diagnostic imaging system with digital video image marking
CN1615489A (en) * 2001-11-21 2005-05-11 韦克福里斯特大学健康科学院 Image reporting method and system
CN201041737Y (en) * 2007-03-29 2008-03-26 李大明 Electronic music book machine
CN101297351A (en) * 2005-10-27 2008-10-29 皇家飞利浦电子股份有限公司 Method and system for processing dictated information
US20090106047A1 (en) * 2007-10-19 2009-04-23 Susanne Bay Integrated solution for diagnostic reading and reporting
EP2169577A1 (en) * 2008-09-25 2010-03-31 Algotec Systems Ltd. Method and system for medical imaging reporting
CN101770302A (en) * 2010-02-25 2010-07-07 吴谦平 Multifunctional mouse
CN103460212A (en) * 2011-03-25 2013-12-18 皇家飞利浦有限公司 Generating a report based on image data
US20130339051A1 (en) * 2012-06-18 2013-12-19 George M. Dobrean System and method for generating textual report content
WO2014016726A2 (en) * 2012-07-24 2014-01-30 Koninklijke Philips N.V. System and method for generating a report based on input from a radiologist

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1296705A (en) * 1999-03-05 2001-05-23 皇家菲利浦电子有限公司 Ultrasonic diagnostic imaging system with digital video image marking
CN1615489A (en) * 2001-11-21 2005-05-11 韦克福里斯特大学健康科学院 Image reporting method and system
CN101297351A (en) * 2005-10-27 2008-10-29 皇家飞利浦电子股份有限公司 Method and system for processing dictated information
CN201041737Y (en) * 2007-03-29 2008-03-26 李大明 Electronic music book machine
US20090106047A1 (en) * 2007-10-19 2009-04-23 Susanne Bay Integrated solution for diagnostic reading and reporting
EP2169577A1 (en) * 2008-09-25 2010-03-31 Algotec Systems Ltd. Method and system for medical imaging reporting
CN101770302A (en) * 2010-02-25 2010-07-07 吴谦平 Multifunctional mouse
CN103460212A (en) * 2011-03-25 2013-12-18 皇家飞利浦有限公司 Generating a report based on image data
US20130339051A1 (en) * 2012-06-18 2013-12-19 George M. Dobrean System and method for generating textual report content
WO2014016726A2 (en) * 2012-07-24 2014-01-30 Koninklijke Philips N.V. System and method for generating a report based on input from a radiologist

Similar Documents

Publication Publication Date Title
WO2016129940A1 (en) Device and method for inputting note information into image of photographed object
WO2019165691A1 (en) Method, apparatus and device for automatically generating test case, and readable storage medium
WO2019080406A1 (en) Television voice interaction method, voice interaction control device and storage medium
WO2016195208A1 (en) Ultrasound apparatus and method of displaying ultrasound images
WO2014010998A1 (en) Method for transmitting and receiving data between memo layer and application and electronic device using the same
WO2014011000A1 (en) Method and apparatus for controlling application by handwriting image recognition
WO2016017987A1 (en) Method and device for providing image
WO2020214006A1 (en) Apparatus and method for processing prompt information
WO2014046513A1 (en) Ultrasound apparatus and information providing method of the ultrasound apparatus
WO2017148112A1 (en) Fingerprint entry method, and terminal
WO2013122355A1 (en) Tablet having user interface
WO2019169814A1 (en) Method, apparatus and device for automatically generating chinese annotation, and storage medium
WO2015160207A1 (en) System and method for detecting region of interest
WO2013125863A1 (en) Method and device for generating captured image for display windows
WO2015126044A1 (en) Method for processing image and electronic apparatus therefor
WO2020197257A1 (en) Translating method using visually represented elements, and device therefor
WO2016108407A1 (en) Annotation providing method and device
WO2015012629A1 (en) Method of processing input and electronic device thereof
WO2017054488A1 (en) Television play control method, server and television play control system
WO2018188196A1 (en) Data version control method, data version controller, device and computer-readable storage medium
WO2012028079A1 (en) Method and device for importing backup data of mobile terminal
WO2022080659A1 (en) Electronic device and control method therefor
WO2018188342A1 (en) Method, apparatus and device for generating script file, and computer-readable storage medium
WO2019062112A1 (en) Method and device for controlling air conditioner, air conditioner, and computer readable storage medium
WO2015178716A1 (en) Search method and device

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14905656

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14905656

Country of ref document: EP

Kind code of ref document: A1