JP2016133821A - Information processing device, information processing method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically determine a term which shows a change state of a lesion part on a progress observation.SOLUTION: An information processing device includes: acquisition means which acquires change of a lesion part from the lesion part contained in medical information of a processing object and the lesion part contained in past medical information; and determination means which determines a term showing change of the lesion part acquired by the acquisition means.SELECTED DRAWING: Figure 2

Description

  The disclosed technology relates to an information processing apparatus, an information processing method, and a program.

  In recent years, systems such as electronic medical records that handle medical documents such as medical records and interpretation reports as electronic data are being introduced to medical sites.

  A sentence including a term such as “previous” related to follow-up observation is extracted from this electronic medical document (for example, “the right lung apex has no change from the previous time”) and displayed on the display unit A method is known (Patent Document 1). According to this technique, a doctor can confirm a matter that requires follow-up by referring to a sentence including terms related to follow-up. Then, the doctor can fill in the items necessary for the follow-up on the electronic medical record or the like without omission.

JP2011-2997A

  However, it is still a burden for the doctor to change the lesion in the follow-up observation because the doctor still needs to determine the state and input a term indicating the state of change to a system such as an electronic medical record.

  The disclosed technique has been made in view of such a situation, and an object thereof is to automatically determine a term indicating a change state of a lesioned part in follow-up observation.

  In addition, the present invention is not limited to the above-described object, and is a function and effect derived from each configuration shown in the embodiment for carrying out the present invention, which is another object of the present invention. It can be positioned as one.

  In order to solve the above-described problem, the disclosed information processing apparatus acquires a change in the lesion from the size of the lesion included in the medical information to be processed and the size of the lesion included in the past medical information. Acquisition means for determining, and a determination means for determining a term representing a change in the lesion acquired by the acquisition means.

  It is possible to automatically determine a term indicating the state of change of the lesion in follow-up observation.

FIG. 2A is a block diagram illustrating an example of a configuration of an information processing apparatus according to an embodiment. (B) is a figure showing an example of composition of a plurality of clients and a server. It is an example of data transition concerning an embodiment. It is a flowchart which shows an example of the document creation process of the information processing apparatus which concerns on embodiment. It is a flowchart which shows an example of the learning method of the information processing apparatus which concerns on embodiment. It is a figure which shows an example of the interpretation report which concerns on embodiment. It is a figure which shows an example of the analysis result of the interpretation report which concerns on embodiment. It is an example of the dictionary data which concern on embodiment. (A)-(c) is an example of the rule data which concern on embodiment. It is a figure which shows an example of a structure of the interpretation report which concerns on embodiment. It is an example of a display of the display part concerning an embodiment. It is an example of a display of the display part concerning an embodiment. It is a figure which shows an example of the medical chart which concerns on embodiment. It is a figure which shows an example of the rule data which concern on embodiment. (A), (b) is a figure which shows an example of the learning result which concerns on embodiment.

<Embodiment 1>
Hereinafter, the information processing apparatus according to the present embodiment will be described in detail with reference to the drawings. However, the components described in this embodiment are merely examples, and the scope of the technology of the present invention is not limited by the following individual embodiments.

  FIG. 1A is a block diagram illustrating an example of a configuration of a medical document creation apparatus (information processing apparatus) according to the present embodiment. The medical document creation apparatus includes a storage unit 111, an input unit 113 such as a keyboard, a mouse, and a touch panel, a data processing unit 115 that is a CPU, a display unit 116, and a network I / F unit 117.

  The storage unit 111 stores various information. For example, the storage unit 111 includes at least one of RAM, ROM, HDD, SSD, and the like. In the present embodiment, the storage unit 111 stores a word dictionary 120, a rule dictionary 121, a template 122, a learning result 123, and an analysis result 124. Note that all of these pieces of information are stored in the storage unit 111, but may not be stored in a single memory.

  Here, the word dictionary 120 is a dictionary including words to be extracted from a medical document such as an electronic medical record or an interpretation report. Specifically, the word dictionary 120 includes a word indicating a part of the human body such as “pancreas” and a word indicating a disease such as “massage”. An example of the word dictionary 120 is shown in FIG. In the word dictionary 120, word identification information (ID), word notation, word part of speech, meaning (classification) indicated by a word, and a link to a word in an inclusive relationship are associated with each other. For example, ID 1251 indicates that “pancreatic head” is a “noun” and is a term indicating a “part” of the human body. Furthermore, since “pancreatic head” is a part of “pancreas”, 1252 which is an ID of “pancreas” is shown as link information as a word included in “pancreas”. Note that the example of the dictionary shown in FIG. 8A does not include information related to the extraction of the size of a lesion or the like. For example, the notation of “mm”, “cm”, “μm” or the like indicating the size is used. It may be recorded as meaning a numerical value indicating the size. The word included in the word dictionary 120 is “mm” or the like, but the extracted word is a word including a numerical value related to “mm” or the like, for example. That is, a word such as “10 mm” is extracted.

  Next, the rule dictionary 121 is a dictionary showing the rules for the word combination method. Specifically, the part-of-speech of the word before the combination and the word to be combined are described, and further, the type of the word after the words are combined is described. For example, when the word “massage” is extracted from a medical document using the word dictionary 120, since “massiness” is a prefix from the word dictionary 120, it is combined with the word “lesion” by the rule dictionary 121. One word is generated indicating the lesion "mass lesion". The rule dictionary 121 describes the type of word after being combined with the word combined for each part of speech of the word before combining.

  The template 122 is a template used to create a sentence such as an observation sentence from words obtained using the word dictionary 120 and the rule dictionary 121. As an example of the template 122, there is a template “(word indicating a part) (word indicating a lesion) is (numerical value indicating a size) (word indicating a degree) (word indicating a comparison result with the past)”. . A sentence can be created by applying a word extracted using the dictionary 120 and the rule-dictionary 121 to this template.

  The learning result 123 is information obtained based on the analysis result for the past medical document. An example of the learning result is shown in FIG. For each region, a term indicating the change is associated with the state of change of the region of interest such as a lesion. Specifically, the type of lesion such as “mass”, the change state of the lesion such as enlargement / reduction, and the range of change are associated with each other. In addition to the above association, terms indicating changes such as lesions are associated. Specifically, a word indicating the degree of change such as “mild”, a word indicating the direction of change such as “increase”, and a word indicating the quality (change degree) of change such as “deterioration” are associated with each other. ing. In other words, the term used to describe the change in the lesion is a term that indicates whether the size of the lesion is enlarged or reduced, a term that indicates the extent of the change in the lesion, and whether the patient has recovered due to the change in the lesion. Including at least one of the following terms. Therefore, the learning result 123 indicates the change in the size of the lesion, the term indicating whether the size of the lesion is enlarged or reduced, the degree of the change in the lesion, and the change in the lesion for each lesion site. This corresponds to an example of a table in which at least one of the terms indicating whether or not the patient has recovered is associated.

  For example, in the example of FIG. 8C, when the enlargement of the size of the tumor is within 20 mm with respect to the site of the pancreas, it indicates that the region is getting worse with a slight increase. Note that “mm” is used to represent the size of a lesion or the like as a unit, but the present invention is not limited to this. In FIG. 8C, “mild” or the like is used as a word representing the degree of change, but the present invention is not limited to this, and any other word may be used as long as it represents the degree of change. Similarly, although “increase” or the like is used as a word representing the direction of change, the present invention is not limited to this, and other words may be used as long as the word represents the direction of change. Furthermore, although a word such as “deteriorated” is used as a word indicating the quality of the change, the present invention is not limited to this, and other words may be used as long as the word indicates the quality of the change.

  The analysis result 124 is, for example, a result of analysis processing performed by the data processing unit 115 on the medical document. The analysis result 124 includes, for example, a word extracted using the word dictionary 120, a word combined by the rule dictionary 121, a sentence obtained by the template 122, and the like.

  The input unit 113 receives an input from the operation of the medical document creation apparatus. For example, the input unit 113 includes a keyboard, a mouse, a touch panel, and the like. For example, the analysis result 123 is changed according to the input from the input unit 113.

  For example, the data processing unit 115 can execute various programs of the information processing apparatus, display a UI, and the like by executing a program stored in the storage unit 111. For example, the data processing unit 115 functions as the acquisition unit 11, the determination unit 12, the creation unit 13, the finding creation unit 13, and the display control unit 15 by executing a program stored in the storage unit 111. The data processing unit 115 and the storage unit 111 included in the apparatus may be one or plural. That is, when at least one processing device (CPU) and at least one storage device (memory) are connected and at least one processing device executes a program stored in at least one storage device. The information processing unit 115 functions as each means described above.

  The acquisition unit 11 acquires information such as a word indicating a lesion and the size of a lesion from an electronic medical document such as an electronic medical record or an interpretation report. Specifically, a word included in the word dictionary 120 is searched from an electronic medical document, and information such as a word indicating a lesion and a size of a lesion is extracted. The acquisition unit 11 may acquire an electronic medical document from a memory or the like (not shown). An example of a medical document is shown in FIG. FIG. 5A shows an example of an interpretation report at the first medical examination. FIG. 5B is an example of an interpretation report at the time of the second medical examination. In the example of FIG. 5, the second and subsequent interpretation reports are written centering on the portion that has changed from the previous report. FIG. 9 shows an example of another interpretation report. As shown in FIG. 9, the structure of the interpretation report is largely divided into two parts: a “column for writing findings” (900) and a “column for writing conclusions (diagnosis results)” (901). Furthermore, the findings start with “existence findings” (whether there is a lesion) (902), and if there are lesions, “qualitative findings” (squeeze the diagnosis based on the nature of the lesion) (903) and “spread findings” (lesions) (Staging of cancer, degree of inflammation spread, etc.) (904). If there are a plurality of sites with lesions, a plurality of descriptions are made (905). In this embodiment, the definition is as described above, but the present invention is not limited to this division example.

  Furthermore, the acquisition unit 11 acquires information such as the size of a lesion from the medical document that is the current processing target, and the past medical document (preferably the previous medical document) from the medical document that is the current processing target. ) To obtain information such as the size of the lesion. The acquisition unit 11 acquires a change in the lesion based on the information on the size of the lesion. Specifically, the acquisition unit 11 obtains a difference between the size of the lesion extracted from the medical document that is currently processed and the size of the lesion extracted from the past medical document. Get the change. That is, the acquisition unit 11 corresponds to an example of an acquisition unit that acquires a change in the lesion from the lesion included in the medical information to be processed and the lesion included in the past medical information. Specifically, the acquisition unit 11 corresponding to an example of the acquisition unit acquires a change in a lesion from the size of the lesion included in the medical information to be processed and the size of the lesion included in the past medical information. . For example, when a 20 mm lesion is extracted from a medical document that is currently being processed and a 10 mm lesion is extracted from a past medical document, the acquisition unit 11 displays information indicating a lesion change that is an enlargement of the 10 mm lesion. get.

  An example of information acquired from the medical document by the acquiring unit 11 is shown in FIG. More specifically, FIG. 6 is an example in which the acquisition means 11 extracts “interpretation description elements” from the interpretation report of FIG. Here, the “interpretation description element” is a classification of important words appearing in the interpretation report such as “lesion” and “part”. However, it is not limited to this classification. Here, in the interpretation report shown in FIG. 5 (a), “a mass lesion of 20 mm in size is recognized in the head of the pancreas” is described as the text as “existing findings”. “Pancreatic head” (part), “20 mm large” (size), “mass lesion” (lesion), and “recognize” (end of sentence expression: assertion, affirmation) are extracted as “interpretation description elements”. Similarly, “interpretation description elements” are extracted from “qualitative findings” and “expansion findings”.

  FIG. 7 is an example in which the acquisition unit 11 extracts “interpretation description elements” from the interpretation report of FIG. “Minor” (degree), “increase” (comparison), and “exacerbation” (comparison) are extracted as “comparison expressions”. Furthermore, expressions such as “recovery” (recovery level) or “deterioration” (recovery level) indicating whether the medical condition is getting worse or recovering appear.

  The determining means 12 determines a term representing the change in the lesion. That is, the determination unit 12 corresponds to an example of a determination unit that determines a term representing a change in a lesion acquired by the acquisition unit. For example, the determination unit 12 determines a term representing a change in a lesion based on a learning model obtained from a past medical document. Specifically, the determining unit 12 refers to the learning result 123 with respect to the change in the lesion acquired by the acquisition unit 11, and determines a term representing the change in the lesioned part. For example, when it is extracted from a medical document that the mass of the pancreas is enlarged by 60 mm by the acquisition unit 11, the determination unit 12 refers to the learning result 123 and uses the terms “severe”, “increase”, and “deterioration”. Is a term representing a change in a lesion. That is, the determination unit 123 determines a term representing the change in the lesion part based on the lesion change acquired by the acquisition unit and the table. Note that information such as the pancreas and the tumor is acquired by the acquisition unit 11 in association with the size when acquiring the size of the lesion.

  The creating unit 13 creates a learning result 123 in a table format, for example, based on terms representing the size of a lesion and changes in the lesion included in a plurality of medical documents. That is, the creation unit 13 corresponds to an example of a creation unit that creates a table based on terms representing the size of a lesioned part and changes in the lesioned part included in a plurality of past medical information. More specifically, the creation unit 13 acquires a change in the size of the lesion included in the past medical document, acquires a term indicating the change in the lesion used at that time, and determines the size of the lesion. It is possible to create the learning result 123 by associating it with the change of. For example, the creation unit 13 may associate a term indicating a change in a lesion with a change in the size of a lesion, or indicate a past lesion size, a current lesion size, and a lesion change. It is good also as matching a term.

  The finding creation means 14 creates a sentence such as a finding based on the term representing the change in the lesion determined by the decision means 12. Specifically, the finding creating unit 14 creates a finding sentence by applying a term representing a change in a lesion determined by the determining unit 12 to a template 122 that is an example of a template for creating a sentence. That is, the finding creation means 14 corresponds to an example of a finding creating means for creating a finding using a template for creating the term and sentence determined by the determining means. As an example, let us consider a case where the size of a pancreatic tumor is 10 mm to 20 mm and is enlarged by 10 mm. In this case, the determination unit 12 acquires the words “mild”, “increased”, and “deteriorated” from the learning result 123. Then, the finding creating means 14 creates a finding sentence “a pancreatic tumor is slightly increased by 20 mm” using the template 203 of FIG. 2, for example.

  The display control unit 15 performs control to display various information on the display unit 116. For example, the display control unit 15 causes the display unit 116 to display the finding sentence created by the finding creating unit 14. In other words, the display control unit 15 corresponds to an example of a display control unit that displays the findings created by the finding creation unit on the display unit. Further, the display control means 15 may display a medical document to be compared on the display unit 116. An example of a UI displayed on the display unit 116 by the display control means 15 is shown in FIG. FIG. 10A displays the previous interpretation report and the interpretation report created this time. In addition, the display control means 15 displays the interpretation description elements (lesion size) corresponding to these interpretation reports on the display unit 116 in a comparable manner. Furthermore, the interpretation description element related to the comparison expression also displays the relevance. The correspondence between the previous inspection image and the current inspection image is also displayed. FIG. 10B shows an example in which the cursor is moved to a position different from that in FIG. As in FIG. 10A, the correspondence between the interpretation description elements (lesion size) and the comparison expression, and the correspondence between the images are displayed. FIG. 11 is a diagram for explaining processing in a case where an unsuitable user determines the comparison expression displayed on the display unit 116. In response to an input from the user input unit 113, the display control unit 15 can display the candidates for the comparison expressions and allow the user to select them. Here, the correspondence and the relationship are displayed by arrows and arcs, but other display forms such as display colors and highlighting may be used. In the example shown in FIG. 11, the user can select a term indicating the degree of change in the lesion, but the present invention is not limited to this. For example, the user may be able to select another term such as a term indicating whether or not the patient has recovered due to a change in lesion. Note that the display of the correspondence between reports may be switched between display and non-display by a user operation.

  The display unit 116 is an LCD display, for example, and can display various information. The display type is not limited to LCD.

  The network I / F unit 117 is an interface for communicating with information processing apparatuses other than the information processing apparatus.

  In the above description, the information processing apparatus has been described as one apparatus, but the present invention is not limited to this. For example, the information processing apparatus described above can be regarded as a system as shown in FIG. FIG. 1B is a diagram illustrating an example of the configuration of the information processing system. FIG. 1B includes N clients 151 and 152 and a server 161. The clients 151 and 152 include at least the input unit 113 and the display unit 161 in FIG. Further, the clients 151 and 152 can receive information transmitted from the server 161 in response to an input from the input unit 113 and display the information on the display unit 161. The server 161 includes at least the storage unit 111 and the data processing unit 115 in FIG. Note that the storage unit 111 may be provided in the server 161 or may be provided outside the server 161. The clients 151 and 152 and the server 161 may include the input unit 113 and the display unit 161, respectively. The clients 151 and 152 and the server 161 are connected via a network so that they can communicate with each other. In order to communicate with each other, the clients 151 and 152 and the server 161 each include a network I / F unit 117. The network may be wired or wireless. The server 161 may be a cloud computer. Also, the server 161 is not limited to one, and a plurality of servers 161 may be used.

  An example of the operation of the information processing apparatus according to the present embodiment configured as described above will be described with reference to the flowchart shown in FIG. This flowchart is realized by the data processing unit 115 executing the control program.

  First, in step 300, the acquisition unit 11 acquires the previous interpretation report 200 corresponding to the interpretation report to be created this time. The interpretation report to be compared is an interpretation report of the same patient. An example of the previous interpretation report 200 is shown in FIG.

  Next, in step 301, the acquisition unit 11 performs a morphological analysis process, an extraction description element, a relationship between each interpretation description element, and the like on the previous interpretation report 200. Specifically, the acquisition unit 11 analyzes the previous interpretation report 200 using the word dictionary 120 and the rule dictionary 121. The acquisition unit 11 extracts “mass lesion”, “pancreatic head”, “recognize”, “pancreatic cancer”, “invasion”, and “present” as interpretation description elements as indicated by 201 in FIG.

  Further, the acquisition unit 11 uses the date of the examination data as indicated by 202 in FIG. 2 and the size of the lesion from the previous examination image “20 mm large” (204 in FIG. 2) and the size of the lesion from the current examination image “32 mm large”. (205 in FIG. 2) is acquired. Note that the size of the lesion in the previous examination may be acquired from the interpretation report 200. For example, when the size of a lesion cannot be acquired from the interpretation report 200, the size of the lesion may be acquired from examination data including the size of the lesion. In addition, the size of the lesion part in this examination may be acquired by the acquisition unit 11 from the data input by the user from the input unit 113, or acquired from the interpretation report and / or the inspection data by the acquisition unit 11. Also good. Information obtained by the above analysis is stored in the storage unit 111 as an analysis result 124.

  In step 302, the difference between the information obtained from the previous interpretation report 200 in step 301 and the information related to the current interpretation report is acquired. More specifically, the acquisition unit 11 acquires the difference between the previous lesion size and the current lesion size. In an example of this embodiment, the difference in the size of the lesion is 12 mm (32 mm-20 mm). This difference is the difference between the feature amounts of the previous inspection image and the target inspection image.

  In step 303, the determination means 12 determines a comparison expression. A comparison expression is determined from the difference acquired in step 302 and the relationship between each interpretation description element obtained in step 301. As described above, the determination unit 12 determines a comparative expression (a term representing a change in a lesion) using the learning result 123. In the example in FIG. 2, since the mass in the head of the pancreas is enlarged by 12 mm, the determining means 12 determines the terms “mild”, “increased”, and “deteriorated” as comparative expressions from FIG.

  In the above example, the comparative expression is determined based on the appearance probability obtained from a model (learning result) obtained by learning a past case in advance by machine learning, but the present invention is not limited to this. For example, it is possible to search a case where the same part has the same change from the stored cases, and to determine a term used in the searched case as a comparative expression. The former method of learning a model from a past case can be processed at high speed, and the latter search method can follow even if the appearance tendency changes. Similarly, a template group created in advance and stored in the storage unit 111 (122 in FIG. 1) used for sentence generation also estimates an appropriate template from the template appearance probability. The interpretation description element is also estimated from the interpretation description element appearance probability. Reference numeral 203 in FIG. 2 is an example of a selected template or interpretation description element. Note that a predetermined template may always be used regardless of the appearance probability.

Each probability is calculated as follows, but is not limited to this formula. The importance of the interpretation description element may be calculated as a weight, or the correlation between the interpretation report constituent parts and the interpretation description element may be obtained by principal component analysis. The posterior probability may be calculated from the prior probabilities of the interpretation description elements input by Bayes' theorem. Thus, there is not one equivalent calculation means. Comparison expressions, templates, and interpretation interpretation elements whose appearance probabilities thus obtained are higher than a predetermined threshold are determined.
Comparison expression probability:

Template appearance probability:

Interpretation element appearance probability:

P (C, S): Comparison expression appearance probability for a condition P (T): Template appearance probability P (I): Interpretation element appearance probability n (C, S): Comparison expression appearance frequency for a condition n (T): Template Appearance Frequency n (I): Interpretation Description Element Appearance Frequency n (I _k ): Appearance Frequency of kth Interpretation Description Element n (J _k ): Appearance Frequency of kth Comparison Expression M: Number of Interpretation Description Elements L: Number of comparison expressions S: Conditions such as region, lesion type and size range Next, in step 304, the finding creating means 14 generates a sentence using the template acquired in the above-described processing.

  Specifically, the finding creation means 14 generates a sentence by applying the comparison expression determined by the determination means 12 to the template acquired in step 303. 2 in FIG. 2 is an example of a sentence created by the finding creating means 14.

  Next, in step 305, the display control means 15 displays the sentence created in step 304 on the display unit 116 and presents it to the user. 207 in FIG. 2 is a display example in which a sentence is presented.

  Here, an example of a method of creating the learning result 123 in the present embodiment will be described along the flowchart of FIG.

  First, in step 310, the acquisition unit 11 acquires the target interpretation report and the interpretation report at the previous examination from the past interpretation report group stored in the storage device 116. The acquisition unit 11 also acquires the corresponding inspection image at the same time.

  In step 311, the acquiring unit 11 analyzes the previous interpretation report and the target interpretation report acquired in step 310, extracts the interpretation description element and the comparison expression, and associates them. Further, the acquisition unit 11 extracts a lesion area, a size, and the like from the previous examination image and the target examination image, and extracts a difference between the image feature amounts of the previous examination lesion area and the target lesion area. The acquisition unit 11 extracts an image corresponding to the description of the previous interpretation report and a lesion area size in the image from the image. Further, the acquisition unit 11 extracts the size of association between the lesion areas of the target examination image corresponding to the acquired image.

  In step 312, the acquisition unit 11 extracts the relationship between the comparison expression and the interpretation description element and the difference in the feature amount of the inspection image from the target document.

  In step 313, it is determined whether or not the acquisition unit 11 has acquired all data (medical documents). When the acquisition of the same patient data is completed, the next patient data is targeted. If all patient data has been processed, go to step 314. If all the data has not been acquired, the process returns to step 310.

  In step 314, the creation means 13 learns the relationship between the interpretation description elements of the stored interpretation report and the comparison expression, and creates a learning result 123. The creating unit 13 may learn the relationship with the feature amount difference of the inspection image extracted in step 311 at the same time, and create the learning result 123.

  In the above embodiment, FIG. 8C is given as an example of the learning result 123, but the learning result 123 is not limited to the form of FIG. 8C. For example, the creation unit 13 can use the part, its position (part), the type of lesion, the sex and age of the patient, the enlargement / reduction direction, the examination name, the examination interval from the previous time, the examination conditions, and the like as learning features. Moreover, as learning target data, it is possible to cope individually by learning by hospital name, department, and doctor name and generating a model. It is also possible to determine whether or not to adopt as learning target data by a doctor's carrier. For example, only the interpretation report of a doctor who has worked for more than 10 years is used as a learning target, so that it is more effective when presented to a doctor with a shallow career.

  FIG. 14 shows another example of the learning result 123. FIG. 14A shows a learning result 123 in consideration of patient age, examination interval, and the like. For example, when the size of a tumor in the lung increases by 5 mm within 3 months, the importance varies depending on the age. This is because, for example, children's lungs are smaller than adults. Therefore, in the example shown in FIG. 14A, when there is an increase in a 5 mm mass in a 40-year-old patient, “mild” (degree), “increased” (comparison), “deteriorated” as “comparison expression” ”(Comparison) is extracted by the determination means 12. On the other hand, when there is an increase in a 5 mm mass in a patient aged 10 years, “serious” (degree), “increase” (comparison), and “deterioration” (comparison) are determined as the “comparison expression”. Extracted by That is, since it is possible to determine a comparative expression in consideration of patient information such as age as well as a change in the size of a lesion, more appropriate comparative expression can be performed. Note that the age is not limited to the example of FIG. 14A, and it is also possible to give a range to the age such as teens and 20s.

  FIG. 14A also shows a learning result 123 considering the examination period. For example, when the number of tumors in the lung increases or when the size of the tumor changes, the importance varies depending on the examination interval. This is because even if it is a change in the same lesion, if it is a change in a short period, it may be worse than a change in a long period. Therefore, in the example shown in FIG. 14A, when the number of masses in the lung increases by 1 within one year of the examination interval, “mild” (degree), “exacerbation” (comparison), “Deterioration” (comparison) is extracted by the determination means 12. On the other hand, when the number of masses in the lung increases by 1 within one month of the examination interval, “severe” (degree), “exacerbation” (comparison), and “deterioration” (comparison) are determined as comparative expressions. 12 is extracted. That is, since it is possible to determine the comparison expression in consideration of not only the change in the size of the lesion part but also examination information such as the examination interval, more appropriate comparison expression can be performed. That is, the deciding unit 12 decides a term representing a change in a lesion based on patient information or examination information related to medical information to be processed. Here, as described above, patient information and examination information are information other than changes in lesions. The inspection interval is not limited to the example of FIG. 14A, and the inspection interval may be widened such as 1 to 3 months, half a year or more. In the example shown in FIG. 14 (a), the change in the size of the tumor is 5 mm and the increase in the number is 1. However, the present invention is not limited to these descriptions. It is good also as conditions.

  Further, FIG. 14B shows another example of the learning result 123. This is a learning result 123 in consideration of inspection items. Since the threshold for determining whether a patient's medical condition is mild or serious differs depending on the examination item, it is useful to learn a comparative expression for each examination item. In the example shown in FIG. 14B, even if the test is performed on the same part “blood”, the learning result is different when the test target is γ-GDP and leukocytes. Specifically, even when the value of γ-GDP is increased by 200 and the value of white blood cells is increased by 10, the change of 10 in white blood cells is more important. Therefore, when the white blood cell value increases by 10, the determination means 12 extracts “serious” (degree), “increased” (comparison), and “deterioration” (comparison) as “comparison expressions”. On the other hand, when the numerical value of γ-GDP increases by 200, “mild” (degree) and “deteriorated” (comparison) are extracted by the determining means 12 as “comparison expressions”. That is, even if the examination site is the same, it is possible to determine the comparison expression in consideration of the examination object, and thus it is possible to perform a more appropriate comparison expression.

  Moreover, the example which used the electrocardiogram and hearing ability as a test | inspection item has shown the case where a special change is not seen. In such a case, it may be decided to use the term “follow-up observation”. That is, the determination means 12 does not always determine the comparison expression, and when there is no change in the patient's condition, it may be determined that a word other than the comparison expression is used.

  Thus, by analyzing the previous medical document and the like, estimating the comparative expression centering on the change from the previous time, and presenting the candidate of the medical document that is the current target, it is possible to save the input of the doctor. In addition, the report creation efficiency is improved without disturbing the doctor's thinking.

<Embodiment 2>
In the first embodiment, the interpretation report is used as an example of the medical document, but the processing target of the present invention is not limited to the interpretation report. It can be applied to other medical documents such as medical records and case reports by selecting the difference between the inspection item and the inspection value as the analysis target or learning feature instead of the difference between the characteristics of the inspection image.

  In the second embodiment, an example of a medical record will be described as a medical document.

  Since the configuration and the processing flow are the same as those in the first embodiment, description thereof is omitted.

  FIG. 12 is an example of a medical chart according to the second embodiment. FIG. 13 is an example of a rule in the second embodiment. A comparative expression is estimated from conditions such as differences in test values such as blood tests and symptoms, prescriptions, and treatments. As in the first embodiment, it is also possible to estimate using machine learning. Moreover, it is not limited to generating the second time from the first time. It is also possible to generate a medical document for the second and subsequent times by learning from the first time and the second and subsequent times.

  In this way, by analyzing the previous medical document, estimating the comparative expression centered on changes from the previous time, and presenting the candidate for the medical document that is the subject of this time, it is possible to save the input of the doctor It becomes possible. In addition, doctors can concentrate on medical care. That is, report creation efficiency is improved because the doctor's thinking is not hindered.

<Examples of other embodiments>
The purpose of the disclosed technology is to supply a recording medium recording a program code for realizing the functions of the above-described embodiments to a system or apparatus, and the computer of the system or apparatus reads and executes the control program code stored in the recording medium Is also achieved. In this case, the control program code itself read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the control program code constitutes the present invention.

  As a storage medium for supplying the control program code, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD-R, a magnetic tape, a nonvolatile memory card, a ROM, a DVD, or the like is used. Can do.

  Moreover, the case where the functions of the above-described embodiments are realized by executing the control program code read by the computer is also included. Further, the case where the OS running on the computer performs actual processing based on the instruction of the control program code and the functions of the above-described embodiments are realized by the processing is included.

  Furthermore, after the control program code read from the storage medium is written to the memory provided on the board inserted into the computer, the board performs actual processing based on the instruction of the program code, This includes cases where functions are realized. The same applies to the units connected to the computer.

DESCRIPTION OF SYMBOLS 11 Acquisition means 12 Determination means 13 Creation means 14 Finding creation means 15 Display control means

Claims (10)

Acquisition means for acquiring a change in the lesion from a lesion included in the medical information to be processed and the lesion included in past medical information;
Determining means for determining a term representing a change in the lesion acquired by the acquiring means;
An information processing apparatus comprising:   The said acquisition means acquires the change of the said lesioned part from the magnitude | size of the lesioned part contained in the said medical information of the said process target, and the magnitude | size of the said lesioned part contained in the said past medical information. The information processing apparatus described.   The term representing the change in the lesion is a term indicating whether the size of the lesion is enlarged or reduced, a term indicating the degree of the change in the lesion, and a change in the lesion by which the patient recovers. The information processing apparatus according to claim 2, wherein the information processing apparatus includes at least one of terms indicating whether or not the information is present. For each lesion site, a change in the size of the lesion, a term indicating whether the size of the lesion is enlarged or reduced, a term indicating the degree of change in the lesion, and a change in the lesion A table associated with at least one of the terms indicating whether or not the patient has recovered by
The information processing apparatus according to claim 3, wherein the determination unit determines a term representing the change in the lesioned part based on the change in the lesioned part acquired by the acquisition unit and the table.   The information processing apparatus according to claim 4, further comprising a creation unit that creates the table based on a term representing a size of a lesion part and a change in the lesion part included in a plurality of past medical information.   6. The information processing apparatus according to claim 1, further comprising a finding creation unit that creates a finding using a template for creating the term and sentence determined by the determination unit. .   The information processing apparatus according to claim 6, further comprising a display control unit that causes a display unit to display a finding created by the finding creating unit.   The determining means determines a term representing a change in the lesion based on patient information or examination information related to medical information to be processed, and the patient information and examination information are other than the lesion change. The information processing apparatus according to claim 1, wherein the information processing apparatus is information. An acquisition step of acquiring a change in the lesion from the size of the lesion included in the medical information to be processed and the size of the lesion included in the past medical information;
A determination step of determining a term representing a change in the lesion acquired in the acquisition step;
An information processing method comprising:   A program for causing a computer to execute the information processing method according to claim 9.

Images