JP5539816B2 - MEDICAL INFORMATION PROVIDING DEVICE, MEDICAL INFORMATION PROVIDING METHOD, AND MEDICAL INFORMATION PROVIDING PROGRAM - Google Patents

Description

  The present invention relates to a medical information providing apparatus, a medical information providing method, and a medical information providing program for providing a diagnosis name candidate to a doctor.

  In recent medical practice, medical documents such as medical records and medical reports are handled as electronic data, making it easy to create and manage medical documents, and various measures to reduce the burden on doctors such as interpretation doctors who create medical reports. Computer systems are being introduced (see Patent Documents 1 and 2).

  Patent Literature 1 discloses a medical support system that stores a keyword in a finding sentence of a medical report in association with a diagnosis name and a certainty degree, and extracts and presents a diagnosis name candidate corresponding to the finding sentence input state. Yes. A phrase or term is extracted from a finding sentence input by handwriting using a character recognition technique, and a synonym is converted into a unique phrase. Then, the converted word and the keyword associated with the diagnosis name are collated, and the diagnosis name corresponding to the finding sentence input by handwriting is extracted and presented.

  In Patent Document 2, for each combination of a chief complaint (symptom) and a disease name (diagnostic name) to be entered in a medical record, a frequency at which a doctor diagnoses the disease name in the past is stored, and a disease name candidate according to the symptom input state A diagnosis support system for extracting and presenting the information is disclosed. The diagnosis frequency of the disease name is updated, and candidate disease names are displayed according to the frequency. Handwriting and template input are illustrated as symptom input forms, and in the case of handwriting input, the symptom is divided into words and given meaning.

JP 2007-018460 A JP 2008-027099 A

  The invention described in Patent Literature 1 uses a character recognition technique to cut out a phrase or term in a finding sentence, convert a synonym into a unique phrase, and collate the converted phrase with a keyword associated with a diagnosis name. Therefore, the validity and credibility of the diagnostic name candidates to be provided depend on the accuracy of character recognition. If the validity and credibility of a diagnostic name candidate cannot be secured, there is a risk of misdiagnosis.

  Patent Document 2 presents a candidate for a diagnosis name for the main complaint to be entered in the medical record, and does not target a medical report. Further, the symptom input template exemplified in Patent Document 2 is a simple one for selecting a sentence representing a symptom, and findings of a medical report having a relatively complicated sentence structure with many modifiers such as size and part. Cannot handle sentences.

  Therefore, the techniques described in Patent Documents 1 and 2 and a technique that combines these techniques increase the validity and credibility of diagnostic name candidates presented when creating a medical report, and increase the probability of occurrence of a medical accident such as a misdiagnosis. Cannot fully meet the demand for lowering.

  The present invention has been made in view of the above problems, and an object thereof is to improve the validity and credibility of a diagnostic name candidate provided to a doctor.

  In order to achieve the above object, the medical information providing apparatus of the present invention selects and inputs a template input area and a diagnosis name for selecting and inputting a part of a plurality of words obtained by disassembling a finding sentence constituting a medical report as an input word. A storage means for associating and storing in the first database a diagnosis name input via the input area, and searching the first database in response to a search request using the selected input state as a search key, or matching the search key, or Search / extraction means for extracting a diagnosis name associated with a similar selection input state from the first database; and display control means for displaying the diagnosis name extracted by the search / extraction means on a display. To do.

  The diagnosis name input area is provided with a GUI for selecting and inputting the certainty of the diagnosis name. The storage means stores a diagnosis name for each certainty factor. The search / extraction means extracts the certainty factor together with the diagnosis name. The display control means displays a diagnosis name for each certainty factor.

  The storage means increments the number of times of input when there is a diagnosis name corresponding to the first database every time a medical report is confirmed and saved, and if there is not, an area for registering the diagnosis name is stored in the first database. It is newly established and stores at least one of the number of times of input or the ratio of the number of times of input of the diagnosis name with respect to the total number of inputs in association with the diagnosis name. The search / extraction means also extracts the number of times of input or the ratio together with the diagnosis name. The display control means displays a diagnosis name according to the number of inputs or the ratio. For example, the number of inputs or the diagnosis name with a ratio equal to or higher than a certain threshold value is displayed. Diagnosis names may be displayed in order of the number of inputs or the ratio. Alternatively, the diagnosis name having the maximum number of inputs or the ratio is displayed separately from others.

  It is preferable to provide a second database that stores a likelihood indicating the degree of possible diagnosis name for each selected input state for each diagnosis name. The search / extraction means extracts the likelihood as well as the diagnosis name. The display control means displays the diagnosis name according to the likelihood. For example, a diagnosis name having a likelihood equal to or higher than a threshold is displayed. Diagnosis names may be displayed in descending order of likelihood. Alternatively, the diagnosis name with the maximum likelihood is displayed separately from others.

  It is preferable to further include a third database representing the relationship between deterioration and improvement of each diagnosis name. The search / extract means extracts from the third database the relationship between the deterioration of the extracted diagnosis name and the improvement of the extracted diagnosis name with respect to the diagnosis name described in the past medical report used by copying, and the display control means includes the diagnosis name. Display information on deterioration and improvement.

  The storage means and the search / extraction means are constructed in a server having a database, and the display control means is constructed in a client terminal connected to the server via a network. The server and the client terminal mutually exchange a search request to the search / extraction unit and a diagnosis name extracted by the search / extraction unit via a network.

  According to the medical information providing method of the present invention, the storage means causes the selection input state of the template input area and the diagnosis name input area to select and input a part of a plurality of words and phrases as a result of disassembling the findings constituting the medical report. The database is searched in response to a search request using the selected input state as a search key by a storing step that stores the diagnosis name input in association with the database and the search / extraction means, and matches or resembles the search key A search / extraction step for extracting a diagnosis name associated with a selected input state from a database, and a display step for displaying the diagnosis name extracted in the search / extraction step on a display by display control means. .

  The medical information providing program of the present invention is input via a selection input state of a template input area and a diagnosis name input area for selectively inputting, as input words, a part of a plurality of words obtained by disassembling a finding sentence constituting a medical report. A storage function for associating a diagnosis name with a database, and storing the database in response to a search request using the selected input state as a search key. The diagnosis name associated with the selected input state that matches or is similar to the search key is displayed. A search / extraction function extracted from a database and a display control function for displaying a diagnosis name extracted by the search / extraction function on a display are realized by a computer.

  According to the present invention, the selection input state of the template input area and the diagnosis name are associated with each other and stored in the database, and the diagnosis name candidate corresponding to the search request is retrieved and extracted from the database and displayed on the display. The validity and credibility of the diagnostic name candidate provided to

It is a block diagram of a medical information system. It is a block diagram which shows the outline of the computer which comprises each terminal and a server. It is a figure which shows the schematic structure of a report preparation terminal and DB server, and a report edit screen. It is explanatory drawing which shows the data structure of a report. It is explanatory drawing which shows the outline of the update of a diagnostic name table, and the search / extraction process of a diagnostic name candidate. It is a figure which shows a diagnostic name table. It is a figure which shows a diagnostic name candidate display window. It is a flowchart which shows the flow of a report preparation process. It is a flowchart which shows the flow of the update process of a diagnostic name table. It is a flowchart which shows the flow of a search / extraction process of a diagnostic name candidate. It is a figure which shows a likelihood table. It is a figure which shows another example of a diagnostic name candidate display window.

  In FIG. 1, a medical information system 2 is constructed in a medical facility such as a hospital, and a report creation terminal installed in a clinical department terminal 11 installed in a clinical department 10 and a radiation examination department (hereinafter simply referred to as an examination department) 12. 13, a database (hereinafter abbreviated as DB) server 14, and a network 15 that connects these to be communicable. The network 15 is, for example, a LAN (Local Area Network) laid in the hospital.

  The clinical department terminal 11 is operated by a doctor (hereinafter referred to as a requesting doctor) of the clinical department 10 who requests examination and interpretation. The medical department terminal 11 is used when issuing an order for requesting the examination to the examination department 12 in addition to viewing and inputting the medical chart 16. The clinical department terminal 11 displays the examination image 17 and the report 18 provided from the examination department 12 and provides them for browsing by the requesting doctor.

  The report creation terminal 13 is operated by a doctor of the examination department 12 specializing in interpretation (hereinafter referred to as an interpretation doctor). The report creation terminal 13 is used when the interpretation doctor confirms the order or creates the report 18. The report creation terminal 13 supports the creation of the report 18 by displaying an image display screen and a report edit screen 44 (see FIG. 3).

  A plurality of DBs such as a medical record DB 19, an image DB 20, a report DB 21, and a diagnosis name DB 22 are constructed in the DB server 14. The medical record DB 19 stores medical record 16 data for each patient. The image DB 20 stores data of the inspection image 17 photographed by the modality 23 of the examination department 12 such as a CR device, a CT device, or an MRI device. The report DB 21 stores data of the report 18 created by the report creation terminal 13. The diagnosis name DB 22 stores the diagnosis name data extracted from the data of the report 18 in the diagnosis name table 24 (see FIG. 6).

  The DB server 14 receives the data of the inspection image 17 from the modality 23 via the network 15 and stores the received data of the inspection image 17 in the image DB 20. That is, the DB server 14 functions as a so-called PACS (Picture Archiving and Communication Systems) server and constitutes a PACS together with the modality 23.

  The data of the inspection image 17 has an image ID for identifying each inspection image 17. The data of the inspection image 17 is stored in the image DB 20 in a file format compliant with, for example, DICOM (Digital Imaging and Communication in Medicine). A DICOM tag that records incidental information including items such as a patient ID, an examination ID, an examination date, and an examination type is assigned to the file of the examination image 17. The data of the inspection image 17 stored in the image DB 20 can be searched using various items of the DICOM tag as search keys.

  The DB server 14 constitutes a medical chart system together with the medical department terminal 11 and the medical chart DB 19. The DB server 14 constitutes a report creation support system together with the report creation terminal 13, the image DB 20, and the report DB 21. The data of the report 18 has a report ID for identifying each report 18 like the data of the examination image 17 and can be searched by a search key such as a report ID, an examination ID, a patient ID, and a patient name. . In addition, in this example, although each DB19-22 was demonstrated in the example constructed | assembled in one DB server 14, you may construct | assemble each DB19-22 in a separate DB server.

  The order issued by the medical department terminal 11 has various items for recording information such as patient ID, patient name, request date, request source, examination type (CT, MRI, etc.), examination purpose, necessity of interpretation. Information such as the affiliation, name, and doctor ID of the requesting doctor such as internal medicine and brain surgery is recorded in the request source item. Information such as a therapeutic effect determination for determining a therapeutic effect on a lesion being treated and a metastasis search for examining the presence or absence of a metastatic lesion are recorded in the items for examination.

  The order issued by the clinical department terminal 11 is transmitted to an order reception terminal (not shown) installed in the examination department 12 and accepted by the examination department 12. The order receiving terminal assigns an inspection ID to the received order and manages the order data. The examination ID is transmitted from the order reception terminal to the clinical department terminal 11 together with the reception completion notification. The staff (inspection engineer) of the examination department 12 performs imaging by the modality 23 based on the order received by the order reception terminal.

  When interpretation is necessary (when an item for whether or not order interpretation is necessary), an order with an examination ID is transmitted from the order reception terminal to the report creation terminal 13. The interpretation doctor confirms the order via the report creation terminal 13 and reads the data of the examination image 17 to be interpreted from the image DB 20. Then, the interpretation results of the read inspection image 17 are collected in a report 18. When the purpose of examination is to observe the progress of a lesion such as determination of therapeutic effect, the report 18 is created by copying the previous report 18 of the patient and correcting a different part from the previous report.

  When the creation of the report 18 is completed, the interpretation doctor transmits a creation completion notice to the order issuer's department terminal 11 through the report creation terminal 13. The creation completion notification includes addresses in the DBs 20 and 21 where the inspection image 17 and the report 18 are stored. The requesting physician accesses the address included in the creation completion notification through the medical department terminal 11 and browses the examination image 17 and the report 18.

  Each of the terminals 11 and 13 and the DB server 14 is configured by installing a control program such as an operating system and an application program such as a client program or a server program based on a computer such as a personal computer, a server computer, or a workstation. The

  In FIG. 2, the computers constituting the terminals 11 and 13 and the DB server 14 have substantially the same basic configuration, and each includes a CPU 30, a memory 31, a storage device 32, a LAN port 33, and a console 34. Yes. These are interconnected via a data bus 35.

  The storage device 32 is, for example, an HDD (Hard Disk Drive). The storage device 32 stores a control program and application program (hereinafter referred to as AP) 36. In addition, the storage device 32 of the report creation terminal 13 has a selection state of each panel 51, 52, the term selection button 53 and an input state of the input box 54 (both refer to FIG. 3) (hereinafter collectively referred to as a selection input state). Accordingly, a finding sentence creation definition (not shown) and drawing data for creating a finding sentence are stored. The drawing data defines the type, arrangement, size, color, etc. of GUI (Graphical User Interface) of various operation screens displayed on the display 37.

  Further, the DB server 14 is provided with, for example, a disk array in which a plurality of HDDs are connected as the DB storage device 32, in addition to the HDD storing the program. The disk array may be built in the main body of the DB server 14, or may be provided separately from the main body and connected to the main body through a cable or a network.

  The memory 31 is a work memory for the CPU 30 to execute processing. The CPU 30 loads the control program stored in the storage device 32 into the memory 31 and executes processing according to the program, thereby comprehensively controlling each part of the computer. The memory 31 temporarily stores temporary storage data representing the selected input state.

  The LAN port 33 is a network interface that performs transmission control with the network 15. The console 34 includes a display 37 and an input device 38 such as a keyboard and a mouse.

  In the medical department terminal 11, client programs such as chart software for browsing and editing the chart 16 and viewer software for browsing the examination image 17 and the report 18 are installed as the AP 36. When the client program is activated, a GUI operation screen is displayed on the display 37 of the medical department terminal 11. The operation screen includes a display screen for displaying the chart 16, the inspection image 17, and the report 18 read from the chart DB 19, the image DB 20, and the report DB 21, respectively.

  The medical department terminal 11 receives an operation instruction such as an instruction to input / edit the medical chart 16 and an instruction to input / issue an order through the input device 38. The input chart 16 and order data are stored in the chart DB 19.

  A report editing client program for supporting report creation is installed in the report creation terminal 13 as an AP 36. The report creation terminal 13 performs display processing of the inspection image 17 and editing processing of the report 18 by a client program for editing the report. In the DB server 14, a server program is installed as the AP 36, which executes processing in response to requests from the terminals 11 and 13 as clients and responds with processing results.

  In FIG. 3, when the CPU 30 of the report creation terminal 13 starts a report editing client program, as a console control unit (corresponding to display control means) 40, a DB access unit 41, an editing processing unit 42, and an order acquisition unit 43. Function. The order acquisition unit 43 acquires an order from the order reception terminal via the network 15. The acquired order is registered in an order table (not shown) provided in the storage device 32 of the report creation terminal 13, for example.

  The report creation terminal 13 has a configuration in which two displays 37 a and 37 b are connected to a terminal body provided with a CPU 30. An image display screen used for observing the inspection image 17 is output to one display 37a. A report editing screen 44 used to create the report 18 is output to the other display 37b.

  The image display screen and the report edit screen 44 constitute a GUI operation screen. The console control unit 40 reads drawing data corresponding to the operation of the input device 38 from the storage device 32, and outputs these operation screens to the respective displays 37a and 37b based on the read drawing data. The console control unit 40 receives an operation instruction input from the input device 38 through the operation screen.

  The image display screen and the report editing screen 44 are activated in conjunction with each other. When an examination ID including the examination image 17 to be interpreted is input from the report editing screen 44, the console control unit 40 acquires the data of the examination image 17 corresponding to the examination ID from the image DB 20 through the DB access unit 41. To do. The console control unit 40 activates an image display screen when outputting the acquired inspection image 17 to the display 37a.

  The image display screen is generated based on an ultrasonic image taken by an ultrasonic diagnostic apparatus, a fluoroscopic image taken by a CR image, a tomographic image taken by a CT apparatus or an MRI apparatus, or a tomographic image. Various inspection images 17 such as three-dimensional images are displayed. The image display screen can display a plurality of inspection images 17 simultaneously, such as displaying six frames of tomographic images arranged on one screen. The image display screen is provided with various operation tools constituting the GUI such as operation buttons, list boxes, and icons. Various operation instructions are input from the input device 38 through such an operation tool.

  The report editing screen 44 is provided with a basic information display area 45, a finding sentence input area 46, a diagnosis name display area 47, a term selection input area 48, a diagnosis name input area 49, and an operation button area 50. Each of these areas is various operation tools constituting the GUI. Various operation instructions are input from the input device 38 through such an operation tool. The operation instructions include an instruction to read the report 18 data from the report DB 21, an instruction to save the report 18 data in the report DB 21, an instruction to select the observation sentence input area 46 to be active (a state where input is possible), and the like. included.

  The basic information display area 45 includes basic information such as the patient name (“Shiro Tominaga”), the examination ID (“CT0803”), and the date of photographing of the examination image 17 (“10/7/30 (July 30, 2010)”). Information is displayed. The basic information is read from the order.

  In the observation sentence input area 46, an observation record such as the state of a lesion recognized by the interpretation doctor by observing the examination image 17, that is, a sentence indicating the observation (hereinafter referred to as an observation sentence) is input. In the figure, only one finding sentence input area 46 is displayed, but the finding sentence input area 46 may be added. When the finding text input area 46 is added, when there are a plurality of examination purposes (for example, treatment effect determination and metastasis search), the findings are input separately for each examination purpose. It is also possible to use a method in which a plurality of interpreting doctors input their findings, such as a resident doctor and a supervising doctor, an interpreting doctor who performs primary interpretation, and an interpreting doctor who performs secondary interpretation.

  The term selection input area 48 is an operation tool for inputting a finding sentence to the finding sentence input area 46 by a mouse click operation, and is also called a template type operation tool. The term selection input area 48 includes a large classification panel 51 for each organ such as “liver” and “bile duct”, and a small classification panel 52. The small classification panel 52 has findings items that categorize items to be entered as findings for each organ, such as “liver morphology” and “mass lesion”. Each panel 51, 52 can be switched and selected by a tab. The organ name selected as the large classification panel 51 is displayed beside the finding sentence input area 46. In FIG. 3, “liver” is selected as the large classification panel 51, and “mass lesion” is selected as the small classification panel 52.

  The small classification panel 52 is further provided with a plurality of term selection buttons 53. The term selection buttons 53 are arranged side by side for each observation item such as “number”, “boundary”, “edge”, etc., further subdividing the finding items. For example, the observation item “boundary” includes input words such as “clear”, “unclear”, and “edge” observation items “alignment” and “irregularity”. As the term of the term selection button 53, a medical term that is frequently used in each observation item is used. Further, all the terms of the term selection button 53 represent the degree and shape of the finding item and the observation item, and are terms that modify the finding item and the observation item.

  Only one term selection button 53 can be selected for each observation item. For this reason, for example, when “Border-alignment” is selected and “Border-uneven” is selected again, the selection of “Border-alignment” is automatically canceled.

  The term selection button 53 is not arranged for the observation items “part” and “size”, but an input box 54 for receiving keyboard input is provided instead. The input box 54 is also provided at the right end of the observation item “number”. The input box 54 includes words and phrases indicating the location of a mass lesion such as “right lobe”, “left lobe”, “S1” to “S8”, sizes such as “15”, “10”, “3”, A numerical value representing the location can be input by keyboard operation. The input box 54 at the right end of the observation item “number” is used when there are many lesions appearing in the examination image 17 and the site, size, and state of each lesion are exactly the same.

  When the finding sentence input area 46 is activated and the pointer 55 is placed on the desired term selection button 53 and a click operation is performed with the mouse, the finding sentence based on the selected term is displayed in the finding sentence input area 46. Is done. The finding sentence input to the finding sentence input area 46 can be corrected and modified by operating the keyboard.

  The finding sentence is displayed when one or more term selection buttons 53 are selected. The display of the finding sentence is updated every time the term selection button 53 is additionally selected. The temporary storage data in the memory 31 indicating the selection input state is also updated every time the term selection button 53 is additionally selected. For example, when the term selection button 53 for the two observation items is selected and the term selection button 53 for the third observation item is selected, the selection of the term selection button 53 is performed for the same observation item. Including the case of redoing.

  In FIG. 3, as the term selection button 53, “number-single shot”, “boundary-unclear”, “marginal-irregularity”, and “absorption value-low absorption” are selected as indicated by diagonal lines. In the input box 54, “part-left lobe” and “size-33 mm × 25 mm” are input from the keyboard. In addition, as a finding sentence, the figure shows a state in which “a mass lesion of 33 × 25 mm in size, unclear border, irregular margin and low absorption is observed in the left lobe of the liver” is displayed in the finding sentence input area 46. Show.

  The diagnosis name input area 49 is provided with a pull-down menu 56 for inputting a certainty factor, an input box 57 for inputting a diagnosis name, and a diagnosis name candidate button 58 for displaying diagnosis name candidates. When an inverted triangle mark next to the pull-down menu 56 is clicked, options indicating the degree of certainty of “confirmed”, “suspect”, and “exclusion” are displayed in a pull-down manner. When “Confirmation” is relatively high in confidence, “Suspect” is less certain than “Confirmation”, and when “Confirmation” cannot be said to be “Confirmation”, “Exclusion” is low in certainty, but keep in mind Selected when necessary. In the input box 57, a diagnostic name derived as a result of interpretation is input by the interpretation doctor. The diagnosis name input in the input box 57 is reprinted in the diagnosis name display area 47 in the form of an appropriate single sentence. In this example, the certainty factor “suspect” is selected, the diagnosis name “hepatic hemangiomyolipoma” is input in the input box 57, and “suspected hepatic hemangiomyolipoma” is displayed in the diagnosis name display area 47. Shows the state.

  Various operation buttons 59 to 62 are provided in the operation button area 50. The finding addition button 59 is for adding a finding sentence input area 46. When the pointer 55 is placed on the finding addition button 59 and the mouse is clicked, the finding sentence input area 46 is additionally displayed on the report editing screen 44.

  The data of the finding sentence input to the finding sentence input area 46 (hereinafter referred to as finding data) has a finding ID for identifying each individual. The finding ID is used for associating the examination image 17 and the report 18 with the finding sentence, and includes a number assigned in the order in which the finding sentence input area 46 is added. A finding ID of “F−N” is assigned to the finding sentence input to the N-th finding sentence input area 46 (N is a natural number of 1 or more). FIG. 3 shows a case where a finding sentence is input in the first finding sentence input area 46, and the finding ID “F-1” is displayed above the finding sentence input area 46.

  The end button 60 is used to end the editing of the report 18. When the end button 60 is selected, the data of the report 18 is confirmed and saved. Editing of the report 18 that has been confirmed and saved is prohibited in order to prevent unauthorized tampering. The interruption button 61 is for temporarily interrupting the creation of the report 18. When the interrupt button 61 is selected, the data of the report 18 being created is temporarily saved. The cancel button 62 is selected when the report editing screen 44 is closed.

  The DB access unit 41 transmits a processing request to the DB server 14 and receives a processing result based on commands from the console control unit 40 and the editing processing unit 42. The processing request for the image DB 20 includes a search request for searching the inspection image 17 using information such as the inspection ID and the image ID as search keys. The processing request for the report DB 21 includes a storage request for storing the data of the generated report 18, and a search request for the data of the generated report 18 and the data of the report 18 being generated. Further, the processing request for the diagnosis name DB 22 includes a diagnosis name data storage request and a diagnosis name candidate search request corresponding to the click selection of the diagnosis name candidate button 58.

  The DB access unit 41 receives the data that is the target of the storage request from the editing processing unit 42 and transmits it to the DB server 14. In addition, the search target data is received from the DB server 14 and delivered to the console control unit 40.

  The CPU 30 of the DB server 14 functions as a storage processing unit 63 and a search processing unit 64 for data of the inspection image 17 and the report 18 by executing the server program. The storage processing unit 63 executes data storage processing in each of the DBs 19 to 22 in response to a data storage request from a client such as the report creation terminal 13 or the modality 23. In response to the data distribution request from the medical department terminal 11 and the report creation terminal 13, the search processing unit 64 searches and extracts the requested data from each of the DBs 19 to 22, and sends the extracted data to the request source. To deliver.

  The edit processing unit 42 receives the finding data input in the finding sentence input area 46 of the report editing screen 44 and the selection input states of the panels 51 and 52, the term selection button 53 and the like through the console control unit 40. The edit processing unit 42 blocks the received finding data for each finding text input area 46.

  The edit processing unit 42 adds a finding ID such as “F-1” or “F-2” to each of the blocked finding data, and records these in the data of the report 18. The edit processing unit 42 adds information such as examination ID, patient ID, patient name, and selection input state read from the order to the data of the report 18 in addition to the finding data. The data in the selected input state is temporary storage data stored in the memory 31 when the report 18 is confirmed and saved.

  In addition to the finding ID, the editing processing unit 42 adds a doctor ID for identifying the interpretation doctor who has input the finding sentence to each finding data. The doctor ID is input by the interpretation doctor at the time of user authentication when the report creation terminal 13 is activated. Finding data can be searched by finding ID or doctor ID.

  The edit processing unit 42 creates a finding sentence corresponding to the selection input state of each panel 51, 52, term selection button 53, etc., based on the finding sentence creation definition stored in the storage device 32. The definition for creating the observation sentence has the hierarchical structure of the data in the order of the observation item, observation item, and words for each observation item, with the organ item as the apex, the position in the observation sentence of each observation item section, and the observation item Specifies the order of words in a clause. The definition for creating a finding sentence is, for example, categorizing a finding sentence into a main part and a predicate for each observation item and observation item, and categorizing the main part and the predicate into a main part and a modifying part that modifies the main part. The words that make up the sentence are divided into parts.

  In the example of FIG. 3, the modification part of the main part “size of XX × XXX mm”, “boundary XX,” and the like, the main part of the main part “massic lesion”, and the modifier of the predicate “ Liver XX ”and predicate body part“ Accept ”are registered.

  When any one of the term selection buttons 53 is selected, the edit processing unit 42 first follows the organ item and the hierarchy of the finding items corresponding to the selection state of the panels 51 and 52 for the creation of the finding sentence creation. Then, the phrase of the observation item (hereinafter referred to as an active observation item) corresponding to the selected term selection button 53 is read from the finding sentence creation definition. The edit processing unit 42 converts the word / phrase of each part read out from the finding sentence creation definition into the main part modifier part, the body part, the predicate modifier part, and the body part order (the order (registration) order of the finding sentence creation definition). ) To create a finding sentence.

  Taking the selection state of each of the panels 51 and 52 and the term selection button 53 in FIG. 3 as an example, a description sentence creation process of the editing processing unit 42 will be described. Since “liver” and “mass lesion” are selected as the panels 51 and 52, the editing processing unit 42 follows the “liver” in the definition for creating a finding sentence and the “mass lesion” in the lower layer. Then, the main part of the main part of the active observation item and the modification part of the main part and the predicate are read from the definition for creating the observation sentence. When the various phrases read out in this way are connected according to a predetermined order, the finding sentence described in FIG. 3 and paragraph [0054] is obtained.

  In FIG. 4 showing the data structure of the report 18, the report 18 has patient information such as a patient ID and a patient name, examination information such as an examination ID, an examination date and an examination purpose, and finding information as data. The finding information includes an finding ID, finding data, a doctor ID of an interpretation doctor, a selection input state, a diagnosis name, a certainty factor, and the like. In addition to this, the image 18 stores the image ID of the inspection image (key image) referred to when the report 18 is created, information on the order, and the like.

  As shown in FIG. 5, the storage processing unit 63 extracts the selection information state of the finding information, the diagnosis name, and the certainty from the data of the report 18 transmitted together with the storage request from the DB access unit 41 of the report creation terminal 13. Based on the extracted data, the contents of the diagnosis name table 24 shown in FIG. 6 stored in the diagnosis name DB 22 are updated. The diagnosis name table 24 includes items such as a selection input state, a certainty factor, a diagnosis name, an input count, and a ratio of the input count of the diagnosis name to the total input count. These items are largely classified for each selected input state, and items classified in the selected input state are further classified smaller for each certainty factor. According to the diagnosis name table 24, it is possible to know for each certainty how many diagnosis names have been input in the past in a selected input state. The selection input state data stored in the diagnosis name table 24 contributes to the diagnosis, for example, by excluding the selection input state of the observation item “part” instead of all the selection input states on the report editing screen 44. You may limit to only things.

  Selection input of “number-single shot”, “boundary-unclear”, “edge-irregularity”, “absorption value-low absorption”, “size—major axis 30 mm or more and minor axis 20 mm or more” including the example of FIG. In the state, the diagnosis name “hepatocellular carcinoma” has been input 20 times in the past (ratio 40%), and “hepatic hemangiomyolipoma” has been input 10 times in the past (ratio 20%). The number of inputs of the diagnosis name “hepatocellular carcinoma” is 20 and the ratio is 40% means that the number of inputs for the selected input state and the certainty factor “determined” is 50 times. It has shown that it has made 50 times in the past.

  For the certainty factor “suspicious”, the diagnosis name “hepatocellular carcinoma” was input 14 times (ratio 25%), “hepatic hemangiomyolipoma” was 12 times (21%), and the certainty factor “exclusion” was “hepatocyte” "Cancer" is 7 times (10%) and "Hepatic hemangioma" is 16 times (23%). The diagnosis name table 24 is not limited to the above-mentioned selection input state, and other organs such as “border-clear”, “edge-alignment”, “size-major axis less than 30 mm and minor axis less than 20 mm”, etc. The diagnosis name, the number of times of input, the ratio, etc. for various combinations of selected input states in the term selection input area 48 of the finding items are registered. It should be noted that the numerical value such as the observation item “size” may not be divided into cases in the above-described case, but may be divided in increments of a minimum unit (1 mm in the above example).

  When a diagnosis name for a certain selected input state is not registered in the diagnosis name table 24, the storage processing unit 63 adds a line for registering the diagnosis name in the row of the selected input state and certainty factor of the diagnosis name table 24. It is newly established and the diagnosis name is registered with the number of inputs as one. When a diagnosis name for a certain selection input state has already been registered in the diagnosis name table 24, the storage processing unit 63 increments the number of times the diagnosis name is input in the selection input state and certainty factor row of the diagnosis name table 24. . The storage processing unit 63 calculates the ratio of each diagnosis name and updates the ratio item each time a new diagnosis name is registered and every time the number of input of an existing diagnosis name is incremented. In this way, every time the report 18 is confirmed and saved, the contents of the diagnosis name table 24 are updated by the storage processing unit 63.

  The diagnosis name candidate button 58 is selected when displaying a diagnosis name candidate suitable for the selection input state. When the diagnosis name candidate button 58 is clicked, the console control unit 40 transmits a search key for searching for a diagnosis name candidate to the DB access unit 41. Specifically, the search key is data in a selection input state when the diagnosis name candidate button 58 is clicked, that is, temporary storage data stored in the memory 31 at that time.

  In FIG. 5, the DB access unit 41 transmits a search request for a diagnosis name candidate using the received search key to the search processing unit 64 of the DB server 14. The search processing unit 64 searches the diagnosis name table 24 for a selection input state that matches the selection input state of the search key, and extracts data of the diagnosis name, the number of inputs, and the ratio of the searched row of the selection input state. The search processing unit 64 transmits the extracted data of each item to the DB access unit 41 as a search result.

  The DB access unit 41 passes the search result received from the search processing unit 64 to the console control unit 40. The console control unit 40 generates a diagnostic name candidate display window 70 shown in FIG. 7 based on the search result, and pops up this on the display 37b.

  In FIG. 7, diagnosis name candidates are listed in the diagnosis name candidate display window 70 for each certainty factor. The diagnosis name candidates are arranged in descending order of input frequency or in descending order of ratio. A check box 71 is provided next to each diagnosis name candidate. When the check box 71 for the certainty factor to be input and the diagnosis name candidate is selected and the OK button 72 is clicked, the certainty factor and the diagnosis name candidate selected by the check box 71 are automatically displayed in the pull-down menu 56 and the input box 57. Entered. When the selection of the diagnosis name candidate is stopped, the cancel button 73 is clicked.

  In this example, three diagnosis name candidates are presented for each certainty factor, but the number of diagnosis name candidates may be three or less, and the number of inputs and the ratio when the search processing unit 64 extracts them. May be narrowed down to only diagnosis names that are equal to or greater than a preset threshold. The number of inputs and the ratio may be displayed together with the diagnosis name candidates, or the diagnosis name candidates having a high number of inputs and the ratio may be displayed separately from others by bold letters or the like (see FIG. 12).

  Hereinafter, the operation of the above configuration will be described with reference to the flowcharts shown in FIGS. The requesting doctor issues an order using the medical department terminal 11. The report creation terminal 13 receives the order issued from the clinical department terminal 11 via the order reception terminal of the examination department 12.

  The interpretation doctor accesses the report creation terminal 13 to confirm the order and starts creating the report 18. When the report editing screen 44 is displayed on the display 37b, an image display screen is displayed on the display 37a in conjunction therewith. The radiologist interprets the observation text for each organ in the observation text input area 46 of the report editing screen 44 while observing the examination image 17 on the image display screen.

  The input of the finding sentence is performed by selecting the term selection button 53. In FIG. 8, when the term selection button 53 is selected (“Yes” in step (hereinafter abbreviated as “S”) 10), the console control unit 40 sends each of the panels 51 and 52, the term selection button 53, etc. The selection input state is notified. The selection input state is sent to the memory 31 through the console control unit 40, and the temporarily stored data is stored in the memory 31 (S11).

  Next, the search for the finding sentence creation definition based on the selection input state of each panel 51, 52, the term selection button 53, etc. is executed by the editing processing unit 42, and the phrase of each observation item is read from the finding sentence creation definition. (S12).

  Then, the various words that have been read out are linked in the order in which the finding sentence creation definition is arranged, and the finding sentence is created (S13). The created finding sentence is transferred from the editing processing section 42 to the console control section 40 and displayed in the finding sentence input area 46 under the control of the console control section 40 (S14). In this way, the interpreting physician inputs the finding sentence while selecting the term selection button 53 or inputting a phrase in the input box 54. These series of processes are repeatedly executed every time the term selection button 53 is additionally selected. The image interpretation doctor selects a certainty factor from the pull-down menu 56 and inputs a diagnosis name derived from the findings in the input box 57.

  When the interpretation doctor finishes inputting the finding sentence, the interpreting doctor selects the end button 60 (yes in S15). When the end button 60 is selected, a data storage request for the report 18 is transmitted from the DB access unit 41 to the DB server 14. When a storage request is received from the report creation terminal 13, in the DB server 14, the storage processing unit 63 executes storage processing for the data of the report 18. The data of the report 18 is stored in the report DB 21. This completes the process of creating a single report 18.

  In FIG. 9, during the data storage process of the report 18, the storage processing unit 63 extracts the data of the selection input state, the certainty factor, and the diagnosis name from the data of the report 18 (S <b> 20). If the diagnosis name is new (yes in S21), the storage processing unit 63 newly sets a row for the selected input state and certainty factor of the diagnosis name table 24, and the number of inputs is set to one and a new diagnosis name is created. It is registered (S22). On the other hand, if the name is an existing diagnosis name (no in S21), the selection input state, the certainty factor, and the number of times the diagnosis name is input are incremented (S23).

  When the creation of the report 18 is completed, a report of completion of creation is transmitted from the report creation terminal 13 to the medical department terminal 11 of the requesting doctor. The requesting doctor accesses the report DB 21 through the medical department terminal 11 and reads the report 18 based on the address of the report 18 included in the creation completion notification. A report display screen and an image display screen for displaying the examination image 17 related to the report 18 are output to the display 37 of the medical department terminal 11. The requesting doctor browses these screens and confirms the contents of the report 18.

  In FIG. 10, when the diagnosis name candidate button 58 is selected during the creation of the report 18 (Yes in S30), a search key is transmitted from the console control unit 40 to the DB access unit 41, and from the DB access unit 41 to the DB server 14 A search request for a diagnostic name candidate using a search key is transmitted to the search processing unit 64 (S31).

  Then, the search processing unit 64 searches and extracts the diagnosis name, the number of times of input, and the ratio of the selected input state that matches the selected input state of the search key from the diagnosis name table 24 (S32), and the extracted information is DB as a search result. It is transmitted to the access unit 41 (S33).

  The search result received by the DB access unit 41 is delivered to the console control unit 40. Then, the console control unit 40 displays the diagnosis name candidate display window 70 on the display 37b based on the search result (S34).

  As described above, the present invention stores and manages the selection input state of the term selection button 53 and the like and the diagnosis name in association with each other, and searches for, extracts, and displays the diagnosis name candidates according to the search request. Compared with the prior art that recognizes characters and cuts out phrases and terms, the extraction accuracy of diagnostic name candidates can be improved.

  Since the selection input state is used as a search key, the extracted diagnosis name candidates are always in the designated selection input state, and a search error is unlikely to occur. Therefore, there is no room for doubt about the validity and credibility of the diagnostic name candidate, and the interpretation doctor can use the diagnostic name candidate as a reference when creating the report 18 with peace of mind.

  In the above embodiment, the diagnosis name candidates are searched and extracted using the selection input state as a search key, and the search result is displayed. However, in addition to the selection input state, the selection state of the certainty factor may be used as a search key. Good. For example, when the certainty factor is selected from the pull-down menu 56 and the diagnosis name candidate button 58 is clicked, the certainty factor of the pull-down menu 56 is included in the search key. Alternatively, apart from the pull-down menu 56, a GUI may be provided that specifies the certainty factor of the diagnosis name to be searched, such as all the certainty factors, the certainty factors of “suspect” and “exclusion”. In this case, the search processing unit 64 extracts a diagnostic name candidate in the row of the selection input state and the certainty level specified as the search key. The diagnosis name candidate display window 70 displays the diagnosis name candidates with the specified certainty factor.

  In the above embodiment, when the diagnosis name candidate button 58 is clicked, a diagnosis name candidate is searched, extracted, and displayed. Instead of providing the diagnosis name candidate button 58, for example, a term selection button 53 and an input box 54 are provided. When all of the selected items are input or when more than half of the term selection buttons 53 and the input boxes 54 are selected and input, the console control unit 40 automatically searches the DB access unit 41 for diagnosis name candidates. May be directed. However, in this case, the diagnostic name candidates are displayed even when the interpretation doctor does not need them, which may be annoying. Therefore, it is preferable that the diagnosis name candidate button 58 is provided as in the above-described embodiment, or whether the search instruction is manually performed or automatic can be selected.

  Note that at the beginning of use of the medical information system 2, the data of the report 18 stored in the report DB 21 is small, and the number of diagnosis names registered in the diagnosis name table 24 is small. Therefore, a default diagnosis name prepared in advance may be displayed as a diagnosis name candidate until the number of cases in the diagnosis name table 24 is accumulated. Alternatively, the operation of the diagnosis name candidate button 58 may be validated when the number of cases in the diagnosis name table 24 is accumulated.

  The diagnosis name table 24 may be managed for each interpretation doctor using a doctor ID input by the interpretation doctor at the time of user authentication at the time of starting the report creation terminal 13 or the like. Diagnosis name candidates specialized for each interpretation doctor can be presented.

  In the above embodiment, the number of inputs and the ratio are registered together with the diagnosis name, and diagnosis name candidates corresponding to these are displayed. However, the present invention is not limited to this, and instead of or in addition to these, The degree may be adopted.

The likelihood is an index representing the likelihood of the diagnosis X in the selection input state A. In general, from the Bayes' theorem, the posterior probability P (X | A) of the diagnosis X is proportional to the likelihood P (A | X) and the prior probability P (X) of the diagnosis X as follows.
P (X | A) ∝P (A | X) * P (X)
The likelihood P (A | X) is the probability of being in the selective input state A when the diagnosis is X, and the number of times Nx was the diagnostic X and the selective input state A when the diagnosis was X From the number of times Nxa,
P (A | X) = Nxa / Nx
It can be expressed. Likelihood is calculated using the above formula and stored in the diagnosis name DB 22 in association with the diagnosis name and the selected input state, as in the likelihood table 75 shown in FIG.

  Among liver cancers, hepatocellular carcinoma accounts for about 90%, but hepatoblastoma is few. There are 200 patients with hepatocellular carcinoma, 5 patients with hepatoblastoma in the past examination, 20 patients diagnosed with hepatocellular carcinoma in selective input state A, and 2 hepatoblastomas , The ratio of the posterior probability of hepatocellular carcinoma and hepatoblastoma in the selected input state A is 2/20, and hepatoblastoma is only 1/10 of the probability of hepatocellular carcinoma. For example, hepatoblastoma is an extremely rare diagnosis. However, since the likelihood of the diagnosis name of hepatocellular carcinoma in the selective input state A is (20/200) = 0.1 and in the case of hepatoblastoma is 2/5 = 0.4, the likelihood in the selective input state A is Hepatoblastoma is higher than hepatocellular carcinoma.

  Although rare as a whole, the likelihood of a diagnosis name that increases the probability of diagnosis depending on the selected input state increases. Conversely, the general diagnosis name has a large number of patients, and the likelihood of the diagnosis name that is likely to be input regardless of the specific selection input state is low. If diagnosis name candidates are extracted and presented according to the likelihood having such characteristics, the diagnosis name candidates are presented more in accordance with the selection input state than when the diagnosis name candidates are presented by the number of inputs and the ratio. be able to. In addition, it is possible to make the interpretation doctor aware of rare diagnosis names that are often overlooked compared to popular diagnosis names, and also helps prevent misdiagnosis.

  The search processing unit 64 extracts the diagnosis name from the diagnosis name table 24 as in the above embodiment, and then extracts the likelihood of the extracted diagnosis name from the likelihood table 75. Then, the likelihood is transmitted to the DB access unit 41 together with the search result of the diagnosis name.

  As a method of displaying the diagnosis name candidates, the diagnosis name candidates may be arranged in descending order of likelihood, as in the case of the number of inputs and the ratio in the above embodiment. When the likelihood is adopted in addition to the number of times of input and the ratio, a GUI for designating which one is to be searched / extracted is provided, or the diagnosis name is displayed as in the diagnosis name candidate display window 80 shown in FIG. The number of times of input and the likelihood may be displayed beside the candidate, and the diagnosis name candidate having the largest number may be displayed separately from the others.

  In addition, the diagnosis name “malignant tumor” is deteriorated with respect to the diagnosis name “benign tumor”, and vice versa, and the relationship between deterioration and improvement of each diagnosis name is registered in the diagnosis name DB 22 in a table format. When the past report 18 is used as a copy, information of deterioration and improvement may be displayed together when displaying the diagnosis name candidate. In this case, the relationship between the diagnosis name extracted by the search processing unit 64 and the diagnosis name of the past report 18 is checked for the relationship between deterioration and improvement in the above table, and the information on the deterioration and improvement that has been checked is included in the search result. Transmit to the DB access unit 41. By displaying information on the deterioration and improvement along with the diagnosis name candidate, it is possible to make a comparative judgment with a past examination at the time of follow-up examination such as determination of therapeutic effect.

  Note that a GUI such as an add button may be provided in the diagnosis name input area, and a plurality of certainty factors and diagnosis names may be selected and input by clicking the add button. A plurality of check boxes in the diagnosis name candidate display window can be selected, and when a plurality of check boxes are selected, a diagnosis name input area may be automatically added. Similarly, observation items such as “part” and “size” in the term selection input area may be added so that a plurality of lesion states can be input in one term selection input area. Alternatively, the term selection input area itself may be newly added, or the selection input state may be copied to add the term selection input area.

  In the above embodiment, an example is given in which the diagnostic name candidate is provided to the radiogram interpretation doctor at the report creation terminal 13, but the diagnostic name candidate may be provided while the client doctor is browsing the report 18 at the clinical department terminal 11. In this case, a search key input box is provided on the display screen of the medical department terminal 11, and the DB server 14 searches for and extracts diagnosis name candidates corresponding to the selection input state input in the input box as in the above embodiment. The search result is transmitted to the medical department terminal 11. Alternatively, diagnostic name candidates may be automatically transmitted together with the report 18 completion notification. By providing the diagnosis name candidates for browsing by the requesting doctor, the requesting doctor can determine the basis and validity of the diagnosis name entered in the report 18. In addition, it is also an assist when the requesting doctor wants to change the diagnosis name.

  In the above embodiment, the keyboard and the mouse are described as examples of the input device. However, the selection of the term selection button 53 may be received by voice using a microphone as the input device.

  In the above-described embodiment, the examination performed by the examination department 12 is described as an example, but the examination type is not limited to these, and may be a PET (Positron Emission Tomography) examination, an endoscopic examination, or the like. In addition, although the report 18 exemplifies a report that summarizes the interpretation results of the inspection image 17, it may be a report that summarizes findings on inspection data other than the inspection image. Examples of the inspection data other than the inspection image include numerical data obtained by specimen inspection such as pathological inspection, physiological inspection, and the like, or a waveform diagram such as an electrocardiogram. Moreover, the report in which the findings regarding various data such as inspection data of different inspection types are written may be used.

  In the above-described embodiment, an example in which there is one report creation terminal has been described. However, a plurality of report creation terminals may be used. Moreover, although the diagnosis name DB was constructed | assembled in DB server and the medical information provision apparatus of this invention was demonstrated with the report preparation terminal and DB server, the report preparation terminal (console control part) and DB server (storage processing part) were demonstrated. , A search processing unit) function may be integrated, and a dedicated device for providing diagnosis name candidates may be provided.

  In the case of a client server type information system comprising a client (report creation terminal) and a server (DB server) as in the above embodiment, a dedicated program is used as a client program for report editing. Alternatively, a general-purpose browser corresponding to a WWW (World Wide Web) protocol such as HTTP (Hyper Text Transfer Protocol) may be used.

  When a dedicated program is used, the report edit screen is generated based on screen data defined by the dedicated program. When a general-purpose browser is used, for example, the report edit screen data is stored in the Web server, and the client accesses the Web server and stores the report edit screen data processed into the Web page format. to download. The client browser interprets the source code of the received Web page and generates a report editing screen. The Web server may be shared by the DB server or may be a server different from the DB server. When a general-purpose browser is used, the CPU of the Web server configures an editing processing unit, a console control unit, and the like in cooperation with the CPU of the client or independently.

  The data storage device in which the DB is constructed may be other than the DB server, and for example, a storage device connected via a network such as NAS (Network Attached Storage) or SAN (Storage Area Network) may be used. Thus, the physical configuration of the computer system can be changed as appropriate.

  In the above embodiment, the LAN is described as an example of the network. However, when the clinical department and the examination department are distributed at a plurality of bases, a combination of the LAN and the WAN (Wide Area Network) is used as the network. May be used.

  Note that, as shown in the above embodiment, the present invention naturally extends to the form of a program and further to a storage medium for storing the program.

2 Medical Information System 13 Report Creation Terminal 14 Database (DB) Server 22 Diagnostic Name DB
24 diagnosis name table 30 CPU
31 Memory 32 Storage Device 34 Console 36 Application Program (AP)
37 Display 38 Input Device 40 Console Control Unit 44 Report Editing Screen 48 Term Selection Input Area 49 Diagnosis Name Input Area 53 Term Selection Button 54 Input Box 56 Pull-down Menu 57 Input Box 58 Diagnosis Name Candidate Button 63 Storage Processing Unit 64 Search Processing Unit 70 , 80 Diagnosis name candidate display window 75 Likelihood table

Claims (8)

The first database that associates the selected input state of the template input area for selecting and inputting a part of a plurality of words and phrases as a result of disassembling the findings constituting the medical report and the diagnosis name input via the diagnosis name input area Storage means for storing in,
Search / extraction means for searching the first database in response to a search request using the selected input state as a search key, and extracting from the first database a diagnostic name associated with the selected input state that matches or is similar to the search key;
Display control means for displaying the diagnosis name extracted by the search / extraction means on a display ;
With a third database showing the relationship between the deterioration and improvement of each diagnosis name ,
The search / extract means extracts from the third database the relationship between the deterioration of the extracted diagnosis name and the improvement, with respect to the diagnosis name described in the past medical report used by copying,
The medical information providing apparatus, wherein the display control means displays information of deterioration and improvement along with a diagnosis name . In the diagnosis name input area, a GUI for selecting and inputting the certainty of the diagnosis name is provided.
The storage means stores a diagnosis name for each certainty factor,
The searching / extracting means extracts a certainty factor together with a diagnosis name,
The medical information providing apparatus according to claim 1, wherein the display control unit displays a diagnosis name for each certainty factor. 3. The medical information provision according to claim 2 , wherein the search / extraction unit searches / extracts a diagnosis name in response to a search request using a selection state of certainty as a search key in addition to a selection input state. apparatus. The storage means increments the number of times of input when there is a diagnosis name corresponding to the first database every time a medical report is confirmed and saved, and if there is not, an area for registering the diagnosis name is stored in the first database. Establish and store at least one of the input count or the ratio of the input count of the diagnostic name to the total input count in association with the diagnostic name,
The search / extract means also extracts the number of times of input or the ratio together with the diagnosis name,
The medical information providing apparatus according to any one of claims 1 to 3, wherein the display control means displays a diagnosis name according to the number of inputs or a ratio. A second database storing the likelihood for each diagnosis name indicating the degree to which the diagnosis name can be obtained for a selected input state,
The search / extraction means extracts the likelihood as well as the diagnosis name,
Wherein the display control unit medical information providing apparatus according to any of claims 1, characterized in that to display the diagnosis according to the likelihood 4. The storage means and the search / extraction means are built in a server having a database,
The display control means is constructed in a client terminal connected to the server via a network,
6. The server and the client terminal mutually exchange a search request to the search / extraction unit and a diagnosis name extracted by the search / extraction unit via a network. The medical information provision apparatus of crab. The storage means associates the selected input state of the template input area for selecting and inputting a part of a plurality of words and phrases as a result of disassembling the findings constituting the medical report with the diagnosis name input through the diagnosis name input area. Storing step to store in the database,
A search / extraction step of searching the database in response to a search request using the selected input state as a search key by the search / extracting means, and extracting from the database a diagnosis name associated with the selected input state that matches or is similar to the search key When,
A display step of causing the display control means to display the diagnosis name extracted in the search / extraction step,
In the search / extraction step, the relationship between the deterioration of the extracted diagnosis name and the improvement of the diagnosis name described in the past medical report that has been used for copying is expressed as a third relationship that indicates the relationship between the deterioration of each diagnosis name and the improvement of the improvement. Extract from the database,
In the display step, medical information providing method is characterized in that information of deterioration and improvement is displayed together with a diagnosis name . The selection input state of the template input area for selecting and inputting a part of a plurality of phrases obtained by disassembling the finding sentences constituting the medical report as input words and the diagnosis name input through the diagnosis name input area are associated with each other and stored in the database. Storage function to
A search / extraction function that searches a database in response to a search request using a selected input state as a search key, and extracts a diagnosis name associated with the selected input state that matches or is similar to the search key from the database,
A display control function for displaying a diagnosis name extracted by the search / extraction function on a display is realized in a computer.
The search / extract function is a third method for expressing the relationship between the deterioration and improvement of the extracted diagnosis name and the relationship between deterioration and improvement of each diagnosis name with respect to the diagnosis name described in the past medical report used by copying. Extract from the database,
The medical information providing program, wherein the display control function displays information of deterioration and improvement along with a diagnosis name .

Images