Program, information processing system and information processing method

Abstract

The purpose is to unify character strings that indicate the same content into predetermined terms. [Solution] A program for causing a computer to function as an acquisition unit 101 and a unification unit 102, wherein the acquisition unit 101 acquires a character string, and the unification unit 102 determines whether the character string matches a pre-registered standard medical term, and if the character string does not match the standard medical term, applies multiple character string conversion rules, the order of application of which is predetermined, to the character string in order, determines whether the converted character string obtained by applying the conversion rules matches the standard medical term, and if it matches the standard medical term, unifies the character string into the standard medical term.

Description

[Technical Field]

[0001] The present invention relates to a program, an information processing system, and an information processing method. [Background technology]

[0002] Conventionally, learning models obtained by machine learning have been used in various situations, including in the medical field. Patent Document 1 discloses a prediction and judgment model that can predict CRT non-responders with high accuracy. This prediction and judgment model is obtained by machine learning using training data of patients who are likely to become non-responders. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-183393 Summary of the Invention [Problem to be solved by the invention]

[0004] There are cases where data obtained in medical care is used, such as as training data for machine learning. In such cases, it is desirable to standardize the data format to a certain extent. However, due to differences in the electronic medical record formats used by each medical institution and differences in recording rules between doctors, there is a problem that the data generated varies even when the data has the same meaning.

[0005] The present invention has been made in consideration of these points, and aims to standardize character strings that indicate the same content into predetermined terms. [Means for solving the problem]

[0006] The program of the present invention is A program for causing a computer to function as an acquisition unit and a unification unit, The acquisition unit acquires a character string, The Ministry of Unification determining whether the character string matches a pre-registered standard medical term; If the string does not match the standard medical terminology, applying a plurality of character string conversion rules, the order of application of which is predetermined, to the character string in order; The method is characterized in that it is determined whether the converted character string obtained by applying the conversion rule matches the standard medical terminology, and if it matches the standard medical terminology, the character string is unified into the standard medical terminology.

[0007] In the program of the present invention, The Ministry of Unification When the character string is changed by applying the conversion rule, the conversion rule may be applied to the changed character string in accordance with the application order.

[0008] In the program of the present invention, the program further causes the computer to function as a learning unit, The learning unit may use learning data including the unified character string to learn a prediction model that predicts the condition of a patient.

[0009] In the program of the present invention, the program further causes the computer to function as a learning unit, The learning unit may use learning data including the unified character string to learn a prediction model that predicts a sudden change in a patient's condition.

[0010] In the program of the present invention, The plurality of conversion rules may include an expansion rule for expanding a notation that combines a plurality of disease names into a plurality of notations that correspond to the plurality of disease names.

[0011] In the program of the present invention, The collective notation of the plurality of disease names may include a numerical notation.

[0012] In the program of the present invention, The plurality of conversion rules may include a string conversion rule that converts a predetermined string into the standard medical terminology.

[0013] In the program of the present invention, The plurality of conversion rules may include a partial matching rule that deletes a predetermined character string.

[0014] In the program of the present invention, The plurality of conversion rules include: The expansion rules for expanding a notation that combines multiple disease names into multiple disease names, a partial match rule that compares a portion of the string with the standard medical term; Including, The partial matching rule may be applied before the expansion rule.

[0015] In the program of the present invention, The plurality of conversion rules include: a character string conversion rule for converting a predetermined character string into the standard medical terminology; a partial match rule that compares a portion of the string with the standard medical term; Including, The string conversion rule may be applied before the partial matching rule.

[0016] In the program of the present invention, The plurality of conversion rules include: The expansion rules for expanding a notation that combines multiple disease names into multiple disease names, a character string conversion rule for converting a predetermined character string into the standard medical terminology; Including, The string conversion rule may be applied before the expansion rule.

[0017] In the program of the present invention, The plurality of conversion rules include: The medical terminology may include a conversion rule for converting a first medical term into a second medical term that encompasses the first medical term.

[0018] In the information processing system of the present invention, An information processing system including an acquisition unit and a unification unit, The acquisition unit acquires a character string, The Ministry of Unification determining whether the character string matches a pre-registered standard medical term; If the string does not match the standard medical terminology, applying a plurality of character string conversion rules, the order of application of which is predetermined, to the character string in order; The method is characterized in that it is determined whether the converted character string obtained by applying the conversion rule matches the standard medical terminology, and if it matches the standard medical terminology, the character string is unified into the standard medical terminology.

[0019] In the information processing method of the present invention, An information processing method performed by a computer having a control unit, a step of the control unit acquiring a character string; the control unit determines whether the character string matches a pre-registered standard medical term, and if the character string does not match the standard medical term, applies a plurality of character string conversion rules, the order of application of which is pre-determined, to the character string in order, determines whether the converted character string obtained by applying the conversion rules matches the standard medical term, and if the character string matches the standard medical term, unifies the character string into the standard medical term; The present invention is characterized in that it includes: [Effects of the Invention]

[0020] According to the program, information processing system, and information processing method of the present invention, character strings that indicate the same content can be unified into a predetermined term. [Brief explanation of the drawings]

[0021] [Figure 1] 1 is an overall configuration diagram of an information processing system. [Figure 2] 10 is a flowchart illustrating a model management process. [Figure 3] FIG. 2 is a diagram illustrating an example of a data configuration of medical data. [Figure 4] 10 is a flowchart showing detailed processing in the medical terminology standardization processing. [Figure 5] FIG. 10 is a diagram showing an example of a character string. [Figure 6] FIG. 10 is a diagram showing an example of the data structure of treatment information related to intravenous administration at Hospital A. [Figure 7] FIG. 10 is a diagram showing an example of the data structure of treatment information related to intravenous administration at Hospital B. [Figure 8] FIG. 10 is a diagram showing an example of the data structure of treatment information related to intravenous administration recorded in an interval time series format. [Figure 9] FIG. 10 is a diagram showing an example of the data structure of treatment information regarding the use of an artificial respirator in Hospital A. [Figure 10] FIG. 10 is a diagram showing an example of the data structure of treatment information related to the use of an artificial respirator in Hospital B. [Figure 11] FIG. 10 is a diagram showing an example of the data structure of treatment information related to the use of a ventilator, recorded in an interval time series format. DETAILED DESCRIPTION OF THE INVENTION

[0022] Hereinafter, the present embodiment will be described with reference to the drawings. FIG. 1 is a diagram showing the overall configuration of an information processing system 1 according to the present embodiment. The information processing system 1 according to the present embodiment acquires medical data generated by each of a plurality of medical institutions and generates a sudden change prediction model that predicts a sudden change in a patient using the medical data as learning data. Furthermore, the information processing system 1 according to the present embodiment predicts a sudden change in a patient by using the sudden change prediction model generated in this manner.

[0023] A "sudden change" is a predetermined symptom that requires new treatment. Examples of a "sudden change" include a sudden change requiring admission to an intensive care unit, a sudden change in arterial oxygen saturation that requires oxygen administration or artificial respiration, a sudden change in blood pressure or pulse rate that requires vasopressors, massive fluid infusion, or blood transfusion, the occurrence of disseminated intravascular coagulation that requires treatment, impending or reversible cardiac arrest that requires vasopressors, massive fluid infusion, blood transfusion, or aortic cross-clamping, and the initiation of cardiopulmonary resuscitation or in-hospital death as a result of any of these. A "sudden change" typically refers to a sudden change in arterial oxygen saturation that requires oxygen administration or artificial respiration, a sudden change in blood pressure or pulse rate that requires vasopressors, massive fluid infusion, or blood transfusion, the occurrence of disseminated intravascular coagulation that requires treatment, impending or reversible cardiac arrest that requires vasopressors, massive fluid infusion, blood transfusion, or aortic cross-clamping.

[0024] The information processing system 1 generates a sudden change prediction model and uses it to predict sudden changes. However, as another example, the information processing system 1 may predict a specific condition of a patient, not limited to sudden changes. Specifically, the information processing system 1 may generate a prediction model for predicting a patient's condition and use the prediction model to predict the condition of the target patient. Examples of conditions include a drop in blood pressure, a drop in oxygen saturation, etc.

[0025] The information processing system 1 includes a model management device 10, a user terminal 20, and multiple hospital servers 30. The model management device 10 generates a sudden change prediction model and performs sudden change prediction using the sudden change prediction model. The user terminal 20 is an information processing device used by a user of the model management device 10. The two hospital servers 30 are provided in different hospitals (Hospital A and Hospital B). Note that in this embodiment, for the sake of convenience of explanation, only two hospital servers 30 are shown, but the information processing system 1 may include three or more hospital servers 30. In this case, the three or more hospital servers 30 are each provided in a different hospital.

[0026] The model management device 10, the user terminal 20, and the plurality of hospital servers 30 are connected via a network N.

[0027] The model management device 10 is configured by, for example, a computer, and mainly includes a control unit 100, a storage unit 110, and a communication unit 120.

[0028] The control unit 100 includes a processor such as a CPU (Central Processing Unit) and controls the operation of the model management device 10. The communication unit 120 includes a communication interface that communicates with external devices wirelessly or via a wired connection. The control unit 100 transmits and receives data to and from the user terminal 20 and the hospital server 30 via the communication unit 120.

[0029] The storage unit 110 includes, for example, a hard disk drive (HDD), a random access memory (RAM), a read only memory (ROM), and a solid state drive (SSD). Furthermore, the storage unit 110 is not limited to being built into the model management device 10, but may be a storage medium (for example, a USB memory) that can be detachably attached to the model management device 10. The storage unit 110 stores various data, in addition to programs executed by the control unit 100 and sudden change prediction models.

[0030] The storage unit 110 of this embodiment includes a training data DB 111 and a standard medical terminology DB 112. The training data DB 111 stores training data corresponding to medical data acquired from the hospital server 30. The standard medical terminology DB 112 stores medical terms that have been preset as standard medical terms. Medical terms are terms used in medical settings, such as disease names, procedure names, and device names. For example, Japanese disease names corresponding to ICD-10 are stored as standard medical terms (standard disease names). Note that the standard medical terminology DB 112 is not limited to Japanese disease names corresponding to ICD-10, as long as pre-registered medical terms are registered as standard medical terms.

[0031] The user terminal 20 is configured by a computer or the like, and mainly includes a communication unit 200, a control unit 210, a display unit 220, and an operation unit 230. The communication unit 200 is similar to the communication unit 120 of the model management device 10, and includes a communication interface that communicates with an external device wirelessly or via a wired connection.

[0032] The control unit 210 is similar to the control unit 100 of the model management device 10, includes a processor, and controls the operation of the user terminal 20. The display unit 220 is, for example, a monitor or the like, and displays various screens. The display unit 220 displays, for example, medical data, the results of sudden change prediction, etc. The operation unit 230 is, for example, a keyboard, and can give various commands to the control unit 210.

[0033] The hospital server 30 is configured by a computer etc. The configuration of the hospital server 30 is similar to the configuration of the model management device 10, and mainly includes a control unit, a storage unit, and a communication unit.

[0034] Next, a description will be given of the configuration of the control unit 100 of the model management device 10. The control unit 100 functions as an acquisition unit 101, a unification unit 102, a learning unit 103, and a prediction unit 104 by executing a program stored in the storage unit 110. Note that, hereinafter, the processes described as being executed by the acquisition unit 101, the unification unit 102, the learning unit 103, and the prediction unit 104 are processes performed by the control unit 100 executing the program.

[0035] The acquisition unit 101 acquires medical data from each of the multiple hospital servers 30 via the communication unit 120. The unification unit 102 standardizes the terms included in the medical data to standard medical terms registered in the standard medical terminology DB 112. The unification unit 102 also standardizes the format of each data included in the medical data. The learning unit 103 uses the medical data whose terms and formats have been unified by the unification unit 102 as training data to train a sudden change prediction model. The prediction unit 104 performs sudden change prediction by inputting the patient's medical data into the sudden change prediction model obtained through training by the learning unit 103. The processing of each functional unit will be described in detail later.

[0036] 2 is a flowchart showing the model management process by the model management device 10. In the model management process, first, the acquisition unit 101 acquires medical data from each of the multiple hospital servers 30 (step S100). The medical data includes, for example, character strings entered by user operation in an electronic medical record.

[0037] FIG. 3 is a diagram showing an example of the data structure of medical data. The medical data includes patient information, treatment information, and measurement results. The patient information includes basic patient information such as name, age, and blood type. The treatment information includes information related to treatment, such as the use of a ventilator, oral administration, administration of injections, and administration by intravenous injection. The measurement results include measurement results, i.e., patient biological information, such as electrocardiogram waveforms, arterial pressure waveforms, and arterial blood oxygen saturation waveforms. Furthermore, the measurement results also include information regarding whether or not there has been a sudden change in the patient's condition.

[0038] In FIG. 2, after the processing of step S100, the unifying unit 102 unifies the medical terms included in the medical data with the standard medical terms registered in the standard medical terminology DB 112 (step S102). The medical data includes medical terms used in medicine, such as disease names, drug names, procedure names, and equipment names. However, for example, the term "atlas fracture" may be written as "atlas fracture" or "C1 fracture." Furthermore, while "radius fracture" and "ulna fracture" are written as two of them, they may also be written as a single fracture, such as "radius-ulna fracture."

[0039] In order for medical data to be used as training data for a sudden change prediction model, it is preferable that medical terms with the same meaning be registered as the same term. Therefore, in this embodiment, the unifying unit 102 performs a process of unifying each medical term included in the medical data into a standard medical term.

[0040] For example, "atlas fracture" is registered as a standard medical term in the standard medical terminology DB 112. Therefore, "atlas fracture" included in the medical data is stored as "atlas fracture" as training data in the training data DB 111. On the other hand, "C1 fracture" included in the medical data is converted (unified) into the corresponding standard medical term "atlas fracture."

[0041] After the process of step S102, the unifying unit 102 unifies the format of the numerical data included in the medical data into a predetermined standard format (step S104). This process will be described in detail later. Next, the unifying unit 102 stores the medical data, whose medical terminology has been unified and whose format has been unified into the standard format by the above process, as training data in the training data DB 111 (step S106).

[0042] Next, the learning unit 103 generates a sudden change prediction model by using the learning data stored in the learning data DB 111 (step S108). Specifically, the learning unit 103 generates a sudden change prediction model by performing supervised machine learning using the learning data stored in the learning data DB 111. Next, the prediction unit 104 performs a sudden change prediction using the sudden change prediction model based on the medical data of the patient to be predicted (step S110). Thus, the model management process is completed.

[0043] Note that the execution order of the medical term unification process in step S102 and the medical data format unification process in step S104 is not limited to the embodiment. As another example, step S102 and step S104 may be performed in parallel, or the process of step S102 may be performed after the process of step S104. [[ID=​​​​​​​​​​​​​On the other hand, if the target character string does not exactly match the standard medical terminology (N in step S202), the unifying unit 102 performs line splitting processing (step S206). In the line splitting processing (step S206), the unifying unit 102 splits the target character string into multiple target character strings in accordance with line splitting rules. The line splitting rules split a target character string written across multiple lines into multiple target character strings. In accordance with the line splitting rules, for example, as shown in FIG. 5, the unifying unit 102 splits a character string displayed across two lines into two character strings, and sets each of them as a target character string.

[0048] In the example of FIG. 5, "radius fracture" is written on the first line, and "ulna fracture" is written on the second line. In this case, "radius fracture ulna fracture" is the target character string, but in the line splitting process (step S206), the target character string is decomposed into two, "radius fracture" and "ulna fracture". If the target character string has three or more lines, the target character string is similarly decomposed into three or more target character strings. If a change has occurred through this process, that is, if the target character string has been decomposed into two or more target character strings (Y in step S208), the unifying unit 102 proceeds to step S202, and performs the processes from step S202 onwards for each target character string.

[0049] On the other hand, if there is no change (N in step S208), the unifying unit 102 proceeds to step S210. In step S210, the unifying unit 102 performs a numerical value expansion process in accordance with the numerical value expansion rules. The numerical value expansion rules are used to rewrite a notation including multiple contents, such as a notation indicating a numerical range, into a notation for each content. For example, an atlas fracture may be written as a C1 fracture. Furthermore, multiple bones may be grouped together and written as "C1-3 fractures." In response to this, the unifying unit 102 expands a description such as "C1-3 fractures" into multiple notations corresponding to multiple disease names, such as "C1, 2, 3 fractures," in accordance with the numerical value expansion rules. Similarly, "TH2-7 fractures" is expanded into "TH2, 3, 4, 5, 6, 7 fractures."

[0050] If a change has occurred in the numerical expansion process, i.e., if numerical expansion has been performed (Y in step S212), the unifying unit 102 proceeds to step S202 and performs the processes from step S202 onwards with the character string after numerical expansion as the target character string. If no change has occurred (N in step S212), the unifying unit 102 proceeds to step S214. In step S214, the unifying unit 102 performs character string conversion processing in accordance with the character string conversion rules.

[0051] The character string conversion rules convert a predetermined character string into a standard medical term that is preliminarily associated with the character string. For example, if any of the characters "vertebral arch," "spinous process," "vertebral body," or "fracture" is detected following C7, they are converted into "7th cervical vertebral arch," "7th cervical vertebral spinous process," "7th cervical vertebral body," and "7th cervical vertebra fracture" according to the character string conversion rules. In addition, in the character string conversion process, a process is performed to remove predetermined characters in parentheses. Furthermore, an expression that modifies the disease name after the disease name, such as "2nd degree burn," is changed to modify the disease name from the front, such as "2nd degree burn." If a change has occurred in the character string conversion process (step S214), i.e., if character string conversion has been performed (Y in step S216), the unifying unit 102 proceeds to step S202 and performs the processes from step S202 onward using the converted character string as the target character string.

[0052] Furthermore, if there is no change (N in step S216), the unifying unit 102 proceeds to step S218. In step S218, the unifying unit 102 performs partial matching processing in accordance with the partial matching rule. The partial matching rule detects whether or not there is a partial match by comparing a part of the target character string with the standard medical terminology in the standard medical terminology DB 112. In the partial matching processing (step S218), the unifying unit 102 further considers the target character string to be a partial match in accordance with the partial matching rule if the target character string matches the standard medical terminology when a word at the end of the disease name, such as "disease," "injury," "symptom," or "syndrome," is added to the target character string.

[0053] If there is a partial match (Y in step S218), the unifying unit 102 determines that the matched character string has been recognized as a standard medical term (step S220) and completes the process. For example, if "depression" is simply written as "depression," it is converted to "depression" and recognized as a standard medical term. Furthermore, if "depression" is written, "depression" is converted to "depression" in the previous preprocessing process (step S200), and then converted to "depression" in step S220, and recognized as a standard medical term.

[0054] Furthermore, in the partial matching process (step S218), if the target character string includes predetermined characters such as "right," "left," "both," "multiple," "exacerbation," or "acute" at the beginning, the unifying unit 102 deletes these characters, and if the character string after deletion matches a standard medical term, the target character string is deemed to be a partial match. For example, if the target character string is "right femur fracture," "right" is deleted to become "femur fracture." Since "femur fracture" matches the standard medical term, it is recognized as a standard medical term. Furthermore, if the target character string is "bilateral rib fracture," "bilateral" is deleted to become "rib fracture." Since "rib fracture" matches the standard medical term, it is recognized as a standard medical term. Furthermore, for example, if the target character string is "pulmonary edema acute exacerbation," "acute" and "exacerbation" are deleted to become "pulmonary edema." Since "pulmonary edema" matches the standard medical term, it is recognized as a standard medical term.

[0055] On the other hand, there are cases where a character string containing the character "multiple" (multiple myeloma), such as "multiple myeloma," is registered as a standard medical term. For such character strings, the process of checking whether there is an exact match (step S202) immediately after the preprocessing process (step S200) recognizes them as matching the standard medical term. Therefore, it is possible to prevent the necessary "multiple" (multiple) from being deleted by the partial match process.

[0056] Furthermore, the partial matching process is performed after the character string conversion process. For example, in the character string conversion process, "burn second degree" is converted to "second-degree burn" (postmodification into prefix). If the partial matching process were performed before this process, "burn" would match the standard medical term, and the "burn" would be determined to be a partial match, resulting in the information about "second degree" being missing. However, by performing the character string conversion process before the partial matching process, this information loss can be prevented.

[0057] Furthermore, in partial matching, a word is replaced with a word of a broader concept, and then a partial match is determined again. For example, "ascending colon" is part of "colon," which is part of "large intestine." That is, "colon" encompasses "ascending colon," and "large intestine" encompasses "colon." In contrast, if the target string is "ascending colon cancer," but "ascending colon cancer" is not registered as a standard medical term, it is collapsed to "colon cancer." In such a case, partial matching is performed on the target string "ascending colon cancer." Since there is no match, "ascending colon" is converted to the broader concept of "colon," which encompasses "ascending colon," and partial matching is performed again on "colon cancer." If there is no match in this case, partial matching is performed again on "colon cancer." A match is then determined.

[0058] For example, if "ascending colon" were to be uniformly converted to "large intestine" in the character string conversion process, all standard medical terms containing "ascending colon" would also be converted to "large intestine," which is undesirable. Therefore, in this embodiment, conversion to encompassing medical terms is performed sequentially in the partial matching process. This allows accurate conversion to standard medical terms.

[0059] In the partial match process, if no partial match is found (N in step S218), the unifying unit 102 proceeds to step S222. In step S222, the unifying unit 102 performs expansion processing. In the expansion processing (step S222), the unifying unit 102 divides the collectively written disease names into individual disease names.

[0060] As a result, for example, the description "radius-ulna fracture" is expanded to "radius fracture" and "ulna fracture." Furthermore, the description "distal radius shaft fracture" is expanded to "distal radius fracture" and "radius shaft fracture." Furthermore, the description "C1,2 fracture" is expanded to "C1 fracture" and "C2 fracture." Note that "C1 fracture" is converted to "atlas fracture" in the character string conversion process (step S214) that is subsequently executed as the processes of steps S202 to S228 are repeated, as described below. Similarly, "C2 fracture" is converted to "axis fracture" in the character string conversion process (step S214).

[0061] If a change occurs in the expansion process (step S222), i.e., if expansion has been performed (Y in step S224), the unifying unit 102 proceeds to step S202 and performs the processes from step S202 onwards using the expanded character string as the target character string. In this way, the character string conversion process is applied after the expansion process. This makes it possible to prevent, for example, "C1,2 fracture" from being converted to only "atlas fracture" or only "axis fracture".

[0062] The expansion process is performed after the partial matching process. For example, suppose the target character string is "first-degree back burn." This character string is registered as a standard medical term. On the other hand, its decomposition into "back burn" and "first-degree burn" are also registered as standard medical terms. For this reason, if the expansion process is performed before the partial matching process, "first-degree back burn" will be decomposed into "back burn" and "first-degree burn," contrary to the author's intention. In contrast, in this embodiment, the expansion process is performed after the partial matching process, so it is possible to prevent such expansion that goes against the author's intention.

[0063] Furthermore, the expansion process is performed after the character string conversion process. For example, suppose the target character string is "second-degree burns (TBSA 48%)." In this case, in the character string conversion process, the target character string "TBSA 48%" is converted to "body surface area 40-49%," and the target character string is converted to "second-degree burns on a body surface area of ​​40-49%." Then, in the expansion process, it is converted to "second-degree burns, burns on a body surface area of ​​40-49%." On the other hand, if the expansion process were performed before the character string conversion process, the string would be split into "second-degree burns" and "TBSA 48%," resulting in an interpretation contrary to the author's intention.

[0064] If there is no change in step S224 (N in step S224), the unifying unit 102 proceeds to step S226. In step S226, the unifying unit 102 performs an extraction process. In the extraction process (step S226), the unifying unit 102 searches for parentheses and the like in the target character string, and if the characters in the parentheses match the disease name, extracts it. For example, for the entry "Other (subarachnoid hemorrhage)", the characters in the parentheses are moved out of the parentheses, such as "subarachnoid hemorrhage other". In this case, the part "subarachnoid hemorrhage" is determined to be a partial match in the partial match process (step S218) executed thereafter.

[0065] If a change occurs in the extraction process (step S226), i.e., if extraction is performed (Y in step S228), the unifying unit 102 proceeds to step S202 and performs the processes from step S202 onwards with the extracted character string as the target character string.

[0066] If there is no change (N in step S228), the unifying unit 102 proceeds to step S230. In step S230, the unifying unit 102 performs a combination search. In the combination search process (step S230), the unifying unit 102 attempts partial matches by inserting separators into the target character string one character at a time.

[0067] For example, if the target character string is "depression acute drug poisoning," the unification unit 102 divides it into "u" and "tsu acute drug poisoning." The unification unit 102 treats each of the two divided character strings as target character strings and determines whether or not they match standard medical terminology. If neither character string matches standard medical terminology, the unification unit 102 changes the delimiter position. As a result, the character string is divided into "depression" and "acute drug poisoning." The unification unit 102 then treats each character string as target character strings and determines whether or not they match standard medical terminology. In this way, the unification unit 102 repeats the process while changing the delimiter position one character at a time until a character matches a standard medical terminology. If the unification unit 102 finds that some character strings match standard medical terminology in the combination search (Y in step S232), the unification unit 102 repeats the process from step S202 for the remaining character strings. If there are no matches in the combination search, the unification unit 102 completes the medical terminology unification process.

[0068] As described above, in the medical terminology unification process, the unification unit 102 first determines whether the target character string exactly matches the standard term. Then, the unification unit 102 applies multiple conversion rules, such as line splitting, numeric expansion, word conversion, partial match, expansion, extraction, and combination search, in a predetermined order. In this way, the unification unit 102 converts the target character string appropriately according to each conversion rule, and each time, it again determines whether the target character string matches the standard medical term. This allows the target character string to be correctly converted into the corresponding standard medical term.

[0069] In the above, the conversion rules for the combination search are applied after the partial match rules, which prevents unnecessary combination searches from being applied to partially matching medical terms.

[0070] Next, the standard format unification process (step S104) described with reference to Fig. 2 will be described. The standard format unification process is a process for unifying the data format of each piece of data included in the medical data into a predetermined standard format. The unification unit 102 unifies each piece of data included in the medical data into the predetermined standard format.

[0071] The standard formats include a static format that does not include time and a time-series format that includes time. Patient information is static information that does not include time and is recorded in the static format as learning data in the learning data DB 111. Treatment information and measurement results are information corresponding to time, i.e., dynamic information that includes time, and are recorded in the time-series format as learning data in the learning data DB 111.

[0072] Furthermore, time series formats are classified into point time series formats and interval time series formats. The point time series format is a data format for evaluation targets that is managed based on a predetermined point in time. Information in the point time series format includes information about oral administration by patients, information about the administration of injection drugs, etc. Here, information about oral administration is information that a certain drug was administered orally at a certain time. Information about the administration of injection drugs is information that a certain drug was administered as an injection at a certain time. Other examples of information in the point time series format include vital signs (pulse, blood pressure, respiratory rate, etc.), blood tests (white blood cell count, creatinine, urea nitrogen, etc.), blood gas analysis (pH, arterial oxygen partial pressure, lactate level, etc.), etc.

[0073] In addition, with regard to the administration of injectable drugs, the time required for administration may or may not be managed. For example, while the time required for anesthetics is managed to the extent of being slowly administered, the administration time of antibiotics may be managed precisely, such as by administering them over one hour at 6:00, 14:00, and 22:00 each day. In this case, in the case of anesthetics, the time series format is preferably a point time series format, and in the case of antibiotics, the time series format is preferably an interval time series format. Thus, with regard to the administration of injectable drugs, either a point time series format or an interval time series format is predefined depending on the type of drug to be administered and the method of administration.

[0074] The interval time series format is a data format for evaluation targets that are managed based on a predetermined interval. Information in the interval time series format includes information about administration by intravenous injection, information about treatments for which a period is specified by a medical professional, and the like. Information about administration by intravenous injection is information that administration by intravenous injection was performed continuously over a certain interval (a period from a first time point to a second time point). Information about treatments for which a period is specified is information that an artificial respirator was used during a certain interval. Examples of treatments include the use of medical devices such as an artificial respirator. Examples of such medical devices include IABP (Intra Aortic Balloon Pumping) and ECMO (Extracorporeal Membrane Oxygenation). Other examples of information in interval time series format include infusion information (e.g., saline, noradrenaline, aminoglycoside), ventilator settings (e.g., inspired oxygen concentration, pressure support, positive end-expiratory pressure), non-invasive positive pressure ventilation (e.g., inspired oxygen concentration, pressure support, positive end-expiratory pressure), etc. In this way, the evaluation target information indicating each evaluation target related to the patient contained in the medical data is standardized into a standard format that is preset for each type of evaluation target information.

[0075] In the standard format unification process (S104), for example, when the numerical information of the artificial respirator is given in a point time series format, the unification unit 102 rewrites this into information in an interval time series format. On the other hand, the unification unit 102 manages information related to administration by oral administration or injection in a point time series format.

[0076] For example, with regard to drug administration, one medical institution's information system records the hourly drug dosage as data, while another medical institution's information processing system records the start and end times of drug administration and the administration rate as data. However, when administering drugs, it is preferable to record information that allows medical professionals to reproduce similar procedures, and such information is data over a certain period, such as the type of drug administered and the duration. Therefore, such information regarding drug administration is managed in an interval time series format. On the other hand, for example, information regarding oral administration of drugs, administration by injection, blood pressure measurements, and vital signs requires data at a certain point in time, and therefore, such information is managed in a point time series format.

[0077] For example, the data groups shown in Figures 6 and 7 are treatment information related to intravenous administration of medication contained in medical data sent from Hospital A and Hospital B, respectively. The treatment information from Hospital A shown in Figure 6 includes patient ID, time, order ID, bottle ID, condition, medication, and dosage (ml), with the dosage recorded on a flow rate basis. Patient ID is identification information for the patient. Order ID is identification information for the order. Bottle ID is identification information for the medication bottle. Status is the status of the treatment.

[0078] On the other hand, the treatment information of Hospital B shown in Figure 7 includes patient ID, patient name, order ID, time, administration type, bottle ID, order type, drug ID, drug, solution amount, drug amount, and quantity, and only the administration amount is recorded, and no data on flow rate is included. The administration type is the type of administration. The order type is the type of treatment. The drug ID is drug identification information. The two data groups also contain different data, such as whether or not the patient name is included and whether or not the administration type is included. Furthermore, the treatment information of Hospital A and Hospital B is recorded in a point time series format, with the drug amount at that time associated with the time. Therefore, the unifying unit 102 unifies all of this treatment information into an interval time series format.

[0079] 8 is a diagram showing an example of the data structure of treatment information related to intravenous administration of medicines recorded in an interval time series format. The unifying unit 102 of this embodiment unifies the treatment information related to intravenous administration of medicines, among the numerical data included in the medical data acquired by the acquiring unit 101, into an interval time series format.

[0080] In the interval time series format data structure, treatment information related to the infusion of a drug includes an administration ID, a patient ID, a start time, an end time, an administration rate, whether or not it is a bolus administration, the drug name, and the drug ratio. The administration ID is identification information for the administration. The start time and end time are the start and end times of the treatment, respectively. Thus, in the interval time series format data structure, for the infusion of a drug, information such as the administration rate is managed in the interval from the start time to the end time. Thus, in the interval time series format, information related to the evaluation object, i.e., the infusion of a drug, is recorded on an interval basis. The unifying unit 102 unifies the treatment information related to the infusion of a drug into the interval time series format data structure shown in Figure 8.

[0081] The data groups shown in Figures 9 and 10 are treatment information related to the use of a ventilator, included in the medical data sent from Hospital A and Hospital B, respectively. The treatment information from Hospital A shown in Figure 9 includes a patient ID, time, instruction ID, status, item, and value. The instruction ID is identification information for instructions related to treatment. The status indicates the device to be used. The item indicates the setting item, and the value indicates the value for the setting item.

[0082] The treatment information of Hospital B shown in Figure 10 includes a patient ID, time, item name, and value. In each treatment information, information such as ventilation mode and oxygen concentration is managed for the section from the start time to the end time.

[0083] 11 is a diagram showing an example of the data structure of treatment information related to the use of a ventilator, recorded in an interval time series format. The unifying unit 102 of this embodiment unifies the treatment information related to the use of a ventilator, among the numerical data included in the medical data acquired by the acquiring unit 101, into an interval time series format.

[0084] In the interval time series data structure, the treatment information related to the use of a ventilator includes the ventilator ID, patient ID, start time, end time, ventilation mode, oxygen concentration, positive end-expiratory pressure, inspiratory pressure, tidal volume, inspiratory time, respiratory rate, and inspiratory assist pressure. In this way, the unifying unit 102 unifies the treatment information related to the use of a ventilator into the interval time series data structure shown in FIG.

[0085] As described above, in the information processing system 1 according to this embodiment, it is determined whether a character string matches a pre-registered standard medical term, and if it does not match, the conversion rules are applied to the character string in order, thereby unifying the character string into a standard medical term.

[0086] The above-described embodiment is an example for carrying out the present invention, and various other embodiments are possible. For example, various modifications and changes are possible within the scope of the gist of the present invention as defined in the claims, such as applying one modified example to another modified example. For example, some of the components of the above-described embodiment may be omitted, and the order of processing may be changed or omitted.

[0087] As such a modified example, the processing by the model management device 10 of this embodiment may be realized by multiple devices. That is, some functions of the model management device 10 may be realized by a first device, and other functions of the model management device 10 may be realized by a second device.

[0088] For example, a first device may perform processing up to generating a sudden change prediction model, and a second device may acquire the sudden change prediction model and perform sudden change prediction using the sudden change prediction model. As another example, the model management device 10 may be integrated with a single hospital server 30. Furthermore, in this case, the sudden change prediction model may be generated using medical data managed by the hospital server 30 as learning data.

[0089] According to the program, information processing system, and information processing method of this embodiment configured as described above, the acquisition unit 101 acquires a character string, and the unification unit 102 determines whether the character string matches a pre-registered standard medical terminology. If the character string does not match the standard medical terminology, the unification unit 102 applies a plurality of character string conversion rules, the order of application of which is predetermined, to the character string in order, determines whether the converted character string obtained by applying the conversion rules matches the standard medical terminology, and if the converted character string matches the standard medical terminology, unifies the character string into the standard medical terminology. This makes it possible to unify character strings having the same content into a predetermined terminology.

[0090] Furthermore, according to the program, information processing system, and information processing method of this embodiment, when a character string is changed by applying a conversion rule, the unification unit may apply the conversion rules to the changed character string in the order of application, thereby enabling accurate unification of terms.

[0091] Furthermore, according to the program, information processing system, and information processing method of this embodiment, the learning unit 103 may use training data including the unified character string to train a prediction model that predicts the condition of a patient. Furthermore, according to the program, information processing system, and information processing method of this embodiment, the learning unit 103 may use training data including the unified character string to train a prediction model that predicts a sudden change in the patient's condition. This makes it possible to train a prediction model using training data suitable for machine learning.

[0092] According to the program, information processing system, and information processing method of the present embodiment, the plurality of conversion rules may include an expansion rule for expanding a notation that collectively represents a plurality of disease names into a plurality of notations that correspond to the plurality of disease names, thereby enabling the plurality of disease names to be converted into each of the disease names.

[0093] Furthermore, according to the program, information processing system, and information processing method of the present embodiment, the combined notation of multiple disease names may include numerical notation, thereby enabling the combined notation including numerical notation to be converted into each disease name.

[0094] Furthermore, according to the program, information processing system, and information processing method of this embodiment, the plurality of conversion rules may include a character string conversion rule for converting a predetermined character string into a standard medical term, thereby enabling the character string to be standardized into a predetermined term.

[0095] Furthermore, according to the program, information processing system, and information processing method of this embodiment, the plurality of conversion rules may include a partial matching rule that deletes predetermined character strings, thereby enabling unification to predetermined terms that do not include unnecessary terms.

[0096] Furthermore, according to the program, information processing system, and information processing method of this embodiment, the multiple conversion rules include an expansion rule that expands a notation that combines multiple disease names into multiple disease names, and a partial matching rule that compares part of a character string with standard medical terminology, and the partial matching rule may be applied before the expansion rule, thereby enabling more accurate character string unification.

[0097] Furthermore, according to the program, information processing system, and information processing method of this embodiment, the multiple conversion rules include a character string conversion rule that converts a predetermined character string into a standard medical term, and a partial matching rule that compares a part of the character string with the standard medical term, and the character string conversion rule may be applied before the partial matching rule, thereby enabling more accurate character string unification.

[0098] Furthermore, according to the program, information processing system, and information processing method of this embodiment, the plurality of conversion rules include an expansion rule for expanding a notation that combines a plurality of disease names into the plurality of disease names, and a character string conversion rule for converting a predetermined character string into the standard medical term, and the character string conversion rule may be applied before the expansion rule, thereby enabling more accurate character string unification.

[0099] Furthermore, according to the program, information processing system, and information processing method of this embodiment, the plurality of conversion rules may include a conversion rule that converts a first medical term into a second medical term that includes the first medical term, thereby enabling more accurate unification of character strings.

[0100] 1. Information Processing Systems 10 Model Management Device 20 User terminal 30 Hospital Server 100 control section 101 Acquisition Department 102 Ministry of Unification 103 Learning Department 104 Prediction Department 110 Storage section 111 Learning Data DB 112 Standard Medical Terminology Database 120 Communications Department 200 Communications Department 210 Control Unit 220 Display section

Claims

1. A program for causing a computer to function as an acquisition unit and a unification unit, The acquisition unit acquires a character string, The Ministry of Unification determining whether the character string matches a pre-registered standard medical term; If the string does not match the standard medical terminology, applying a plurality of character string conversion rules, the order of application of which is predetermined, to the character string in order; A program that determines whether the converted character string obtained matches the standard medical terminology each time the conversion rule is applied, and if it matches the standard medical terminology, unifies the character string into the standard medical terminology.

2. The Ministry of Unification 2. The program according to claim 1, wherein, when the character string is changed by applying the conversion rule, the conversion rule is applied to the changed character string in the order of application.

3. the program further causes the computer to function as a learning unit, The program according to claim 1 , wherein the learning unit uses training data including the unified character string to learn a prediction model that predicts a patient's condition.

4. the program further causes the computer to function as a learning unit, The program according to claim 1 , wherein the learning unit uses training data including the unified character string to learn a prediction model for predicting a sudden change in a patient's condition.

5. The program according to claim 1 , wherein the plurality of conversion rules include a character string conversion rule for converting a predetermined character string into the standard medical terminology.

6. The program according to claim 1 , wherein the plurality of conversion rules include a partial matching rule that deletes a predetermined character string.

7. The plurality of conversion rules include: The expansion rules for expanding a notation that combines multiple disease names into multiple disease names, a partial match rule that compares a portion of the string with the standard medical term; Including, The program according to claim 1 , wherein the partial matching rule is applied before the expansion rule.

8. The plurality of conversion rules include: a character string conversion rule for converting a predetermined character string into the standard medical terminology; a partial match rule that compares a portion of the string with the standard medical term; Including, The program according to claim 1 , wherein the character string conversion rule is applied before the partial matching rule.

9. The plurality of conversion rules include: The expansion rules for expanding a notation that combines multiple disease names into multiple disease names, a character string conversion rule for converting a predetermined character string into the standard medical terminology; Including, The program according to claim 1 , wherein the character string conversion rule is applied before the expansion rule.

10. The plurality of conversion rules include:

2. The program of claim 1, further comprising a conversion rule for converting a first medical term into a second medical term that encompasses the first medical term.

11. An information processing system including an acquisition unit and a unification unit, The acquisition unit acquires a character string, The Ministry of Unification determining whether the character string matches a pre-registered standard medical term; If the string does not match the standard medical terminology, applying a plurality of character string conversion rules, the order of application of which is predetermined, to the character string in order; An information processing system that determines whether the converted character string obtained by applying the conversion rule matches the standard medical terminology, and if it matches the standard medical terminology, unifies the character string into the standard medical terminology.

12. An information processing method performed by a computer having a control unit, a step of the control unit acquiring a character string; the control unit determines whether the character string matches a pre-registered standard medical term, and if the character string does not match the standard medical term, applies a plurality of character string conversion rules, the order of application of which is pre-determined, to the character string in order, determines whether the converted character string obtained by applying the conversion rules matches the standard medical term, and if the character string matches the standard medical term, unifies the character string into the standard medical term; An information processing method, including:

Images