Medical information processing device, medical information processing method, and program

Abstract

To enable appropriate support for individuals (children) who may have developmental disabilities. [Solution] The medical information processing device of the embodiment comprises an acquisition unit, an identification unit, an analysis unit, and a generation unit. The acquisition unit acquires behavioral information relating to the words and actions of a subject. The identification unit identifies trouble information relating to troubles that have occurred to the subject. The analysis unit analyzes the trouble information by behavioral analysis. The generation unit generates analysis result information according to the analysis results of the trouble information.

Description

Technical Field

[0001] The embodiments disclosed in this specification and the drawings relate to a medical information processing device, a medical information processing method, and a program.

Background Art

[0002] In recent years, while the number of children with potential developmental disorders has been increasing, there is a concern that they may not be receiving appropriate support such as classroom guidance. For example, when a child exhibits problem behaviors or causes trouble at school or elsewhere, the surrounding people may regard it as something transient and end up with short-term responses such as attention. It is desirable that families, educational institutions, etc. conduct applied behavior analysis and lead to appropriate support as needed.

[0003] Conventionally, various examinations related to developmental disorders may be conducted on subjects such as children in whom potential developmental disorders are suspected, and a determination may be made as to whether or not they correspond to developmental disorders. Examples of various examinations related to developmental disorders include examinations such as the ASSQ examination, the ADHD-RS examination, the DCDQ examination, the SDQ-T examination, or the OSI examination. However, even when these examinations are conducted, if no potential for developmental disorders is suspected in the subject, the examination is not carried out. Therefore, it has been difficult to recognize developmental disorders in children who potentially have developmental disorders and to provide appropriate support.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] The problem that the embodiments disclosed in this specification and drawings aim to solve is to enable appropriate support for individuals (children) who may have developmental disabilities. However, the problem that the embodiments disclosed in this specification and drawings aim to solve is not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described later can also be positioned as other problems. [Means for solving the problem]

[0006] The medical information processing device of this embodiment includes an acquisition unit, an identification unit, an analysis unit, and a generation unit. The acquisition unit acquires behavioral information relating to the subject's words and actions. The identification unit identifies trouble information relating to troubles that have occurred to the subject. The analysis unit analyzes the trouble information by behavioral analysis. The generation unit generates analysis result information corresponding to the analysis results of the trouble information. [Brief explanation of the drawing]

[0007] [Figure 1] A diagram showing an example of the configuration of the medical information processing device 100 according to the embodiment. [Figure 2] A diagram showing an example of a usage environment for the speech and behavior recognition device 10. [Figure 3] A diagram showing an example of a usage environment for the speech and behavior recognition device 10. [Figure 4] This diagram shows an example of a series of still images that represent a problem situation in text format. [Figure 5] A flowchart illustrating an example of the procedure by which the trouble occurrence status extraction function 162 extracts trouble status. [Figure 6] A diagram showing an example of frame images and the corresponding text for each frame image. [Figure 7] A flowchart showing an example of the processing performed by analysis function 143. [Figure 8] A diagram visualizing an example of the relationships between structured elements. [Figure 9] A diagram showing an example of the contents of reinforcer table 152. [Figure 10] A diagram showing an example of the contents of weakener table 153. [Figure 11] A diagram showing an example of the contents of the Behavioral Change Communication Table 154. [Figure 12] A diagram showing an example of the contents of the extracted results DB155. [Figure 13] A diagram showing an example of the contents of the Similarity DB156. [Figure 14] A diagram showing an example of the content of the scored results of this analysis and similar analyses. [Figure 15] A diagram showing an example of the contents of the score mapping data 157. [Figure 16] A diagram illustrating an example of the content of behavioral improvement information. [Modes for carrying out the invention]

[0008] The medical information processing device, medical information processing method, and program of the embodiment will be described below with reference to the drawings.

[0009] Figure 1 shows an example of the configuration of the medical information processing device 100 according to the embodiment. The medical information processing device 100 is connected to a speech and behavior recognition device 10 and an output device 20 so as to be able to communicate with each other, for example, via a network. The medical information processing device 100 is a device for recognizing developmental disorders in a target person, for example, a child, and providing support as needed, such as providing methods for dealing with developmental disorders when they are observed.

[0010] The medical information processing device 100 generates support information based on the information transmitted by the speech and behavior recognition device 10, such as whether or not the subject has a developmental disability and the support provided for the developmental disability, such as the content of communication between the subject and supporters, such as family members or school teachers, and provides the generated support information to the supporters.

[0011] The behavior recognition device 10 is a device for recognizing the behavior of a target person. The behavior recognition device 10 includes, for example, a microphone 11, a GNSS (Global Navigation Satellite System) device 12, a camera 13, an acceleration sensor 14, and a pressure sensor 15. The behavior recognition device 10 may include other devices as long as they are devices for recognizing the behavior of the target person, or may not include some of these devices.

[0012] The microphone 11 collects the speech of the target person and generates voice information. The microphone 11 transmits the generated collected sound information to the medical information processing device 100. The GNSS device 12 detects its own position by receiving GNSS signals transmitted by GNSS satellites. The GNSS device 12 transmits the detected position of itself to the medical information processing device 100 as position information indicating the position of the target person.

[0013] The camera 13 captures an image of the target person within the imaging range. The camera 13 transmits the captured image to the medical information processing device 100 as image information. The imaging range of the camera 13 may be the range around the target person that does not include the target person. The camera 13 may be able to acquire two-dimensional information or three-dimensional information as image information.

[0014] The acceleration sensor 14 detects the acceleration related to the target person, for example, the acceleration when the target person moves or the acceleration of an object approaching the target person. The acceleration sensor 14 transmits the detected acceleration to the medical information processing device 100 as acceleration information. The pressure sensor 15 detects the pressure related to the target person, for example, the pressure applied to the target person or the pressure applied by the target person to an object along with the actions of the target person. The pressure sensor 15 transmits the detected pressure to the medical information processing device 100 as pressure information.

[0015] FIG. 2 and FIG. 3 are diagrams showing examples of the usage environments of the speech and action recognition device 10. As shown in FIG. 2, the subject P wears, for example, a smartwatch 16 and a card-type device 17. The smartwatch 16 includes a microphone 11, a camera 13, and an acceleration sensor 14, and the card-type device 17 includes a GNSS device 12 and a pressure sensor 15.

[0016] Each device of the speech and action recognition device 10 may be included in either one or both of the smartwatch 16 and the card-type device 17. Each device of the speech and action recognition device 10 may also be included in smart glasses, a smart ring, etc. that the subject P wears. Devices such as the smartwatch 16 and the card-type device 17 may be worn by supporters such as parents and teachers.

[0017] As shown in FIG. 3, each device of the speech and action recognition device 10 may be installed in the room where the subject P stays. In the room where the subject P stays, for example, a microphone 11, a camera 13, a pressure sensor 15, and a tablet terminal 18 are installed, and the tablet terminal 18 is provided with a GNSS device and an acceleration sensor 14.

[0018] The camera 13 is, for example, a security camera, and the pressure sensor 15 also functions as a mat laid on the floor surface. Each device of the speech and action recognition device 10 may be provided in other devices installed in the surrounding environment of the subject in the room, for example, an AI (Artificial Intelligence) speaker, or may be embedded in toys that the subject P plays with. The speech and action recognition device 10 may be other devices, for example, provided in a classroom of a school that the subject P attends, and may be a call button mainly operated by teachers and students. When the call button is operated, trouble occurrence information is transmitted to the medical information processing device 100.

[0019] The output device 20 is a device such as a terminal that outputs support information transmitted by the medical information processing device 100 to the supporter. The output device 20 includes, for example, a speaker and a display. The speaker outputs support information by sound, and the display outputs support information by image (including video and moving images). The output device 20 may also be a tablet terminal 18.

[0020] The medical information processing device 100 includes, for example, a communication interface 110, an input interface 120, a display 130, a processing circuit 140, and a memory 150. The communication interface 110 communicates with external devices such as a speech recognition device 10 and an output device 20 via a network NW such as a LAN (Local Area Network). The communication interface 110 includes, for example, a communication interface such as a NIC (Network Interface Card).

[0021] The input interface 120 receives various input operations from users such as doctors, converts the received input operations into electrical signals, and outputs them to the processing circuit 140. For example, when an input operation is performed by a user, the input interface 120 generates information corresponding to the input operation. The input interface 120 outputs the generated information corresponding to the input operation to the processing circuit 140.

[0022] The input interface 120 includes, for example, a mouse, keyboard, trackball, switch, button, joystick, touch panel, etc. The input interface 120 may also be a user interface that accepts audio input, such as a microphone. The input interface 120 may also have a display function as a display 130, such as a touch panel.

[0023] In this specification, the term "input interface" is not limited to those equipped with physical operating components such as a mouse or keyboard. For example, an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device located separately from the device and outputs this electrical signal to a control circuit is also included as an example of an input interface.

[0024] The display 130 is a display unit that displays various types of information. For example, the display 130 displays images generated by the processing circuit 140, or a GUI (Graphical User Interface) for receiving various input operations from the user. For example, the display 130 may be an LCD (Liquid Crystal Display), a CRT (Cathode Ray Tube) display, or an organic EL (Electro Luminescence) display.

[0025] The processing circuit 140 includes, for example, an acquisition function 141, a specific function 142, an analysis function 143, a generation function 144, a suggestion function 145, a similar ABC analysis extraction function 146, a behavior improvement analysis function 147, and a provision function 148. The processing circuit 140 realizes these functions, for example, by a hardware processor (computer) executing a program stored in the memory (storage circuit) 150.

[0026] Hardware processors refer to circuits such as CPUs (Central Processing Units), GPUs (Graphics Processing Units), Application Specific Integrated Circuits (ASICs), programmable logic devices (e.g., Simple Programmable Logic Devices (SPLDs) or Complex Programmable Logic Devices (CPLDs)), and Field Programmable Gate Arrays (FPGAs).

[0027] Instead of storing the program in memory 150, the system may be configured to directly incorporate the program into the hardware processor's circuitry. In this case, the hardware processor performs its function by reading and executing the program incorporated into the circuitry. The program may be stored in memory 150 beforehand, or it may be stored on a non-temporary storage medium such as a DVD or CD-ROM, and installed from the non-temporary storage medium to memory 150 when the non-temporary storage medium is mounted on the drive device (not shown) of the medical information processing device 100.

[0028] A hardware processor is not limited to being a single circuit; it may also be composed of multiple independent circuits combined to perform various functions. Alternatively, multiple components may be integrated into a single hardware processor to perform various functions.

[0029] Memory 150 can be implemented by semiconductor memory elements such as RAM and flash memory, hard disks, and optical discs. These non-transient storage media may also be implemented by other storage devices connected via a communication network, such as NAS (Network Attached Storage) or external storage server devices. Memory 150 may also include non-transient storage media such as ROM (Read Only Memory) and registers.

[0030] The memory 150 stores programs for implementing each function in the processing circuit 140, as well as, for example, ABC analysis information 151. The ABC analysis information 151 includes a reinforcer table 152, a depressor table 153, a behavioral change communication table 154, an extraction result database (hereinafter referred to as DB) 155, a similarity DB 156, and score mapping data 157. Each element of the ABC analysis information 151 will be described sequentially thereafter.

[0031] The acquisition function 141 acquires behavioral information relating to the subject's words and actions. The acquisition function 141 acquires behavioral information, including, for example, voice information, location information, image information, acceleration information, and pressure information transmitted by the behavior recognition device 10 and received by the communication interface 110. The acquisition function 141 is an example of an acquisition unit.

[0032] The identification function 142 identifies trouble information related to troubles that have occurred to the target person. The identification function 142 includes, for example, a recognition function 161 and a trouble occurrence status extraction function 162. The recognition function 161 recognizes troubles that have occurred to the target person. The trouble occurrence status extraction function 162 extracts the trouble occurrence status in which the trouble occurred as trouble information. The identification function 142 identifies the trouble occurrence status extracted by the trouble occurrence status extraction function 162 as trouble information. The identification function 142 is an example of an identification unit.

[0033] The recognition function 161 recognizes, for example, a problem that has occurred to the subject based on the behavioral information acquired by the acquisition function 141. The recognition of the problem by the recognition function 161 becomes the trigger for the occurrence of the problem. The recognition function 161 recognizes, for example, the post-behavioral elements extracted as a result of the ABC analysis performed by the analysis function 143 as a problem that has occurred to the subject. ABC analysis and post-behavioral elements will be explained later.

[0034] The recognition function 161 may, for example, syntactically analyze the audio information acquired by the acquisition function 141, extract words and various sounds (hereinafter referred to as "words, etc."), and recognize the occurrence of trouble if the extracted words, etc. contain specific keywords. Examples of specific keywords include words such as "stop it" and "give it back," loud noises such as "bang," and sounds such as crying (including the person in question). The recognition function 161 is an example of a recognition unit.

[0035] The recognition function 161 may, for example, detect the actions of subject P and the surrounding circumstances based on the image information by processing the image information transmitted by the camera 13. The recognition function 161 may recognize the occurrence of trouble by detecting, for example, a situation in which multiple children, including subject P, are gathered in one place, or a situation in which one person is separated from the group. Alternatively, the recognition function 161 may recognize the occurrence of trouble based on the posture of subject P included in the image information, for example, by detecting that the subject is in a position where they have fallen, are stomping their feet, or are crouching down.

[0036] The recognition function 161 may also recognize the occurrence of trouble when it detects that the subject P is in a place other than the children's play area (a space where children play), for example, inside a storage room, based on location information transmitted by the GNSS device 12. Alternatively, the recognition function 161 may recognize the occurrence of trouble when it detects that the subject P has fallen, based on acceleration information transmitted by the acceleration sensor 14. Alternatively, the recognition function 161 may recognize the occurrence of trouble when it detects that the pressure transmitted by the pressure sensor 15 is large and that a strong impact has been applied to a mat or the like. The recognition function 161 may also recognize the occurrence of trouble when the communication interface 110 receives trouble occurrence information transmitted by operating the call button installed as the speech and action recognition device 10.

[0037] The trouble occurrence status extraction function 162 extracts the trouble occurrence status as trouble information when a trouble occurs. For example, the trouble occurrence status extraction function 162 extracts the content of post-action elements extracted as a result of ABC analysis performed by the analysis function 143 as trouble status. The trouble occurrence status extraction function 162 is an example of a trouble occurrence status extraction unit.

[0038] The trouble occurrence status extraction function 162 may, when the recognition function 161 recognizes the occurrence of a trouble, convert the audio information syntactically analyzed by the recognition function 161 and the image information acquired by the acquisition function 141 into text, and extract the trouble situation from the text generated by the text conversion. The trouble occurrence status extraction function 162 may also generate the text using an interactive AI.

[0039] The following describes an example of how the trouble occurrence status extraction function 162 extracts trouble information by converting audio and image information into text. First, we will explain an example of how the trouble occurrence status extraction function 162 converts audio information into text. Here, we will explain the case where the acquisition function 141 acquires audio information such as, "Why did you take the ball? I was playing with it. Give it back."

[0040] In this case, first, the trouble occurrence situation extraction function 162 performs speech recognition on the acquired audio information and converts it into text. Next, the trouble occurrence situation extraction function 162 recognizes the target person P and other persons, such as the target person P's friends, based on speaker recognition, microphone volume, and secondary audio (pitching, speed, volume, etc.). For example, since the target person P is wearing a microphone and their friends are often not, the person with the loudest voice is recognized as the target person P.

[0041] Next, the trouble situation extraction function 162 may estimate the situation, which is represented by the audio information acquired by the acquisition function 141, and convert it into text. In this case, the trouble situation extraction function 162 generates a text sentence, for example, "The person in question is angry because their friend took their ball while they were playing with it," and extracts the estimated trouble situation by converting it into text.

[0042] Next, we will explain an example of how the trouble occurrence status extraction function 162 extracts trouble information by converting image information into text. Here, we will explain an example of converting trouble information into text from still images. Figure 4 is a diagram showing an example of a series of still images from which trouble information is converted into text.

[0043] The first still image SC1 includes subject P1 and subject P2's friend. Friend P2 is holding a ball, and subject P1 has noticed their friend. The trouble situation extraction function 162 analyzes the first still image SC1 to detect the actions of subject P1 and friend P2 and the position of the ball, and generates the text "Your friend is playing with a ball."

[0044] The second still image SC2, following the first still image SC1, contains an image of subject P1 taking the ball from friend P2. The trouble occurrence situation extraction function 162 analyzes the second still image SC2 to detect the actions of subject P1 and friend P2 and the movement of the ball, and generates the text "User took the ball from friend." The "user" in the text generated by the trouble occurrence situation extraction function 162 represents subject P1.

[0045] The third still image, following the second still image SC2, shows subject P1 playing with a ball that he took from his friend P2. Friend P2 is watching subject P1 playing with the ball from a distance. The trouble situation extraction function 162 detects the state of subject P1, friend P2, and the ball, and generates the text "He started playing with the ball." In this way, the trouble situation extraction function 162 extracts the trouble situation by converting the image information of the still images into text.

[0046] Next, we will explain an example of how the trouble occurrence status extraction function 162 extracts trouble information by converting the image information into text when the image is a video. Figure 5 is a flowchart showing an example of the procedure by which the trouble occurrence status extraction function 162 extracts trouble information.

[0047] The trouble occurrence status extraction function 162 first acquires the video acquired by the acquisition function 141 as is (step S101). Next, the trouble occurrence status extraction function 162 extracts frame images from the acquired video (step S103). Next, the trouble occurrence status extraction function 162 converts each of the extracted frame images into text (step S105).

[0048] Figure 6 shows an example of frame images and corresponding text sentences for each frame image. As shown in Figure 6, for example, suppose the video contains eight frame images, from the first frame image CU1 to the eighth frame image CU8. In this case, the trouble occurrence situation extraction function 162 analyzes the first frame image CU1 and generates a first text sentence from the results of the image analysis, thereby converting the first frame image CU1 into text.

[0049] For example, the trouble occurrence detection function 162 detects a person resembling a child in the first frame image CU1. In this case, the trouble occurrence detection function 162 generates the text "There is a child" as the first text. In the next second frame image CU2, the trouble occurrence detection function 162 detects that the same person resembling a child detected in the first frame image CU1 has changed direction. In this case, the trouble occurrence detection function 162 generates the same text "There is a child" as the first text as the second text based on the second frame image CU2.

[0050] The trouble occurrence detection function 162 detects in the next third frame image CU3 that the person resembling a child, detected in the second frame image CU2, has raised their hand. In this case, the trouble occurrence detection function 162 generates the text "A child is raising their hand" as a third text sentence based on the third frame image CU3.

[0051] The trouble occurrence status extraction function 162 then repeats the same process to generate text documents for each frame image from the 4th frame image CU4 to the 8th frame image CU8, thereby converting the 4th frame image CU4 to the 8th frame image CU8 into text. The converted text documents may be the same or different between adjacent frame images.

[0052] Returning to Figure 5, the trouble occurrence extraction function 162 determines whether the generated text is different from the text based on the previous frame image (step S107). In the example shown in Figure 6, the text based on the second frame image CU2 and the sixth frame image CU6 is the same as the text based on the previous frame image, while the text based on the other frames is different from the text based on the previous frame image.

[0053] The trouble occurrence status extraction function 162, if it determines that the generated text is different from the text based on the previous frame image, adds the situation indicated in the generated text (step S109). In the example in Figure 6, the trouble occurrence status extraction function 162 first creates the sentence "There is a child" when it generates the first text, then adds the sentence "The child has raised their hand" when it generates the third text, and then adds the sentence "The child has received a balloon" when it generates the fourth text. Thereafter, the trouble occurrence status extraction function 162 generates and adds text in the same manner.

[0054] In step S107, if the generated text is determined to be the same as the text based on the previous frame image, the trouble occurrence status extraction function 162 skips the process in step S109. Next, the trouble occurrence status extraction function 162 determines whether or not the text conversion of all frame images has been completed (step S111).

[0055] If the trouble occurrence status extraction function 162 determines that the text conversion of all frame images is not yet complete, it returns to step S105 and proceeds with the text conversion of the next frame image. If the trouble occurrence status extraction function 162 determines that the text conversion of all frame images is complete, it terminates the process shown in Figure 5. If the image is a video, the trouble occurrence status extraction function 162 converts the image information into text using this procedure and extracts the trouble status.

[0056] The analysis function 143 analyzes trouble information using behavioral analysis, such as applied behavior analysis. The analysis function 143 includes, for example, a post-behavior element extraction function 163 and a reinforcer / punisher extraction function 164. The post-behavior element extraction function 163 uses ABC analysis, a method of applied behavior analysis, on the verbal and behavioral information acquired by the acquisition function 141 to extract structured elements that include post-behavior elements corresponding to events after the behavior. The reinforcer / punisher extraction function 164 extracts reinforcers and punishers related to post-behavior elements. The analysis function 143 is an example of an analysis unit. Although the analysis method using applied behavior analysis is shown as a specific example of the behavioral analysis method used by the analysis function 143 to analyze trouble information, other behavioral analysis methods may also be used.

[0057] The post-action element extraction function 163 performs ABC analysis based, for example, on text documents generated by the trouble occurrence status extraction function 162 in the specific function 142, or text documents generated by a method similar to the method used to extract text documents by the trouble occurrence status extraction function 162. The post-action element extraction function 163 is an example of a post-action element extraction unit.

[0058] The post-action element extraction function 163, as an ABC analysis, first structures the text and generates structured elements. The post-action element extraction function 163 then extracts post-action elements (events after the action, Consequence) from the structured elements. Structured elements that are strongly related to post-action elements are extracted as behavioral elements (actions, Behavior), and structured elements that are weakly related to post-action elements are extracted as pre-action elements (events before the action, Antecedent).

[0059] The reinforcer / punisher extraction function 164 extracts reinforcers and punishes contained in the behavioral elements and post-behavioral elements. The reinforcer / punisher extraction function 164 is an example of a reinforcer / punisher extraction unit. The procedure for processing performed by the post-behavioral element extraction function 163 and the reinforcer / punisher extraction function 164 in the analysis function 143 will be explained below with reference to Figures 7 and 8.

[0060] Figure 7 is a flowchart showing an example of the processing of the analysis function 143, and Figure 8 is a diagram visualizing an example of the relationship between structured elements. First, the analysis function 143 structures the text generated based on the verbal and physical information acquired by the acquisition function 141 or the text extracted by the trouble occurrence situation extraction function 162 in the post-action element extraction function 163 (step S201).

[0061] For example, suppose the following four text sentences are generated. The content of the first text sentence is, "A-kun approached his friend with the ball." Note that the subject is referred to as "A-kun" in the text sentence. The content of the second text sentence is, "A-kun took the ball from his friend." The content of the third text sentence is, "A-kun started playing with the ball." The content of the fourth text sentence is, "The friend whose ball was taken said, 'Give it back.'" Of these, the fourth text sentence is the text sentence (text information including post-action elements) that triggered the occurrence of the trouble, which was recognized by the recognition function 161.

[0062] The post-action element extraction function 163 structures the generated first to fourth text sentences. Specifically, as shown in Figure 8, the post-action element extraction function 163 structures the first to fourth text sentences to generate the first structured element MT1 to the fourth structured element MT4. The content of the first structured element MT1 is "approach your friend", the content of the second structured element MT2 is "take the ball", the content of the third structured element MT3 is "play with the ball", and the content of the fourth structured element MT4 is "give it back".

[0063] Next, the post-action element extraction function 163 extracts the fourth structured element MT4, which is a structured element containing the action that triggered the occurrence of the trouble and was recognized as a trouble by the recognition function 161, from among the generated structured elements as a post-action element (step S203). Subsequently, the post-action element extraction function 163 compares the relationship between the fourth structured element MT4 (post-action element) and each of the first structured elements MT1 to the third structured elements MT3 other than the fourth structured element MT4 (step S205). For comparison between structured elements, knowledge created in advance using, for example, natural language processing is used.

[0064] Next, the post-action element extraction function 163 determines whether the first structured element MT1 to the third structured element MT3 each have a strong relationship with the fourth structured element MT4 (step S207). The post-action element extraction function 163 extracts the structured elements that it determined to have a strong relationship with the fourth structured element MT4 as action elements (step S209). On the other hand, the post-action element extraction function 163 extracts the structured elements that it determined to have a weak relationship with the fourth structured element MT4 as pre-action elements (step S211).

[0065] In the example shown in Figure 8, the post-action element extraction function 163 extracts the second structured element MT2 and the third structured element MT3, which have a high correlation RS with the fourth structured element MT4, as action elements. On the other hand, the post-action element extraction function 163 extracts the first structured element MT1, which has a low correlation RW with the fourth structured element MT4, as a pre-action element. The post-action element extraction function 163 generates ABC analysis extraction results that include the extracted pre-structured elements, action elements, and post-action elements.

[0066] Next, the reinforcer / punisher extraction function 164 extracts reinforcers and punishes from the behavioral elements (second structured element MT2 and third structured element MT3) and the post-behavioral elements (fourth structured element MT4) (step S213). The reinforcer / punisher extraction function 164 extracts reinforcers and punishes from the behavioral elements and post-behavioral elements by referring, for example, to the reinforcer table 152 and the punisheerer table 153 contained in the ABC analysis information 151 stored in memory 150.

[0067] Figure 9 shows an example of the contents of the reinforcer table 152. Figure 10 shows an example of the contents of the punisher table 153. The reinforcer table 152 lists terms that correspond to reinforcers and the reinforcer IDs associated with those terms. The punisher table 153 lists terms that correspond to punishers and the reinforcer IDs associated with those terms. Terms include sentences, words, verbs, etc. Reinforcer terms are, for example, terms that are considered to be positive and lead to repeated behavior. Punisher terms are, for example, terms that are considered to be less likely to occur as a result of the behavior.

[0068] The reinforcer / punisher extraction function 164 determines whether a term contained in the sentences of the behavioral elements and behavioral elements is among the terms listed in the reinforcer table 152 and the punisher table 153. If there is a term in the sentences of the behavioral elements and behavioral elements that is listed in the reinforcer table 152, the reinforcer / punisher extraction function 164 extracts that term as a reinforcer along with its reinforcer ID. If there is a term in the sentences of the behavioral elements and behavioral elements that is listed in the punisher table 153, the reinforcer / punisher extraction function 164 extracts that term as a punisher along with its punisher ID.

[0069] For example, the second and third structured elements MT2 and MT3, which are behavioral elements, contain the terms "stole" and "started playing," respectively, while the fourth structured element MT4, which is a post-action element, contains the term "got angry." These terms correspond to the weakener ID 0005's weakener "stole," the reinforcer ID 0001's "was able to play," and the weakener ID 0001's "got angry," respectively. The reinforcer / weakening element extraction function 164 extracts these reinforcers and weakeners along with their reinforcer IDs and weakener IDs. In this way, the medical information processing device 100 completes the process shown in Figure 7.

[0070] The generation function 144 generates analysis result information according to the analysis results of the trouble information. For example, the generation function 144 generates analysis result information by including the reinforcers and weakeners extracted by the reinforcer / weakening element extraction function 164 of the analysis function 143 in the ABC analysis extraction results. The generation function 144 is an example of a generation unit. The analysis result information may also include information containing post-behavior elements extracted by the post-behavior element extraction function 163.

[0071] The suggestion function 145 proposes communication content for the subject based on the analysis results information generated by the generation function 144. For example, the suggestion function 145 generates communication content for the subject based on the analysis results information generated by the generation function 144 and proposes it to the supporter. The suggestion function 145 is an example of a suggestion unit.

[0072] The proposed function 145, for example, refers to the reinforcers and punishers extracted by the reinforcer / punisher extraction function 164 in the behavior change communication table 154 stored in memory 150, and extracts behavior change communication based on the reinforcers and punishers. Behavior change communication is a part of the communication content.

[0073] Figure 11 shows an example of the contents of the behavior change communication table 154. The behavior change communication table 154 shows, for example, the relationship between the contents of the behavior change communication proposals corresponding to the reinforcer ID and the punisher ID. For example, if the reinforcer ID is reinforcer 001 and its content is "I was able to play", and the punisher ID is punisher 005 and its content is "I took it", the contents of the behavior change communication would be "Teach communication such as 'Can I borrow it?', teach to return it..."

[0074] The suggestion function 145 generates communication content that includes behavioral change communications extracted by referencing the behavioral change communication table 154, which contains reinforcers and punishers. The suggestion function 145 proposes behavioral change communications to the supporter by transmitting the generated communication content to the output device 20.

[0075] The behavioral change communication suggested by the suggestion function 145 may be suggested using the contents of the behavioral change communication table 154 as is, or it may be suggested by adding embellishments to the contents of the behavioral change communication table 154. For example, the suggestion function 145 may use the contents of the behavioral change communication table 154 as is and suggest behavioral change communication such as "Try the following communication," along with "Teach communication such as 'Can I borrow it?', teach to return, teach how to play together, teach how to play with alternative toys."

[0076] Suggestion function 145 may also propose behavioral change communication, such as adding embellishments to the content of the behavioral change communication table 154, along with the statement, "The following measures can be considered," including: "Teach them to play with their friends instead of taking the ball. When doing so, emphasize that they can make friends and have fun playing together." "Encourage them to take the ball back and return it to their friends. When doing so, explain that they may be scolded or left out by their friends." "Introduce alternative ways to play or toys other than balls. When doing so, explain that they can try something more interesting or new than playing with a ball."

[0077] The generation function 144 generates an extraction result DB 155, which is a database of the ABC analysis extraction results generated by the post-action element extraction function 163. When the generation function 144 generates new ABC analysis extraction results, it adds the generated ABC analysis results to the extraction result DB 155 and updates the extraction result DB 155. The extraction result DB 155 is a database of ABC analysis extraction results extracted by the post-action element extraction function 163.

[0078] The extraction results DB155 may be updated using the post-action element extraction function 163. The extraction results DB155 may contain data only from the subject, or it may include data from people other than the subject. If it includes data from people other than the subject, the attributes of the vehicle from which the data is collected may be set according to the subject. For example, if the subject is a child, only information about children may be included, or attributes may be set according to the subject's gender, preferences (how they play), number of friends, etc.

[0079] Figure 12 shows an example of the contents of the extracted results DB155. The extracted results DB155 includes, for example, the extracted result ID, the extraction date and time, and the contents of the pre-action element indicated as "(A) Previous event", the action element indicated as "(B) Action", and the post-action element indicated as "(C) Post-action event". For example, the pre-action element of the ABC analysis extracted result for extracted result ID 00001 is "He was trying to open a locked box containing a toy he wanted to play with", the action element is "He broke the box and took out the toy", and the post-action element is "He played with the toy".

[0080] The extracted results DB155 is used, for example, by the similar ABC analysis extraction function 146 described below. The generation function 144 may include other information in the extracted results DB155. For example, the generation function 144 may detect the location where the trouble occurred based on location information and include it in the extracted results DB155.

[0081] The Similar ABC Analysis Extraction Function 146 extracts similar ABC analysis results that are similar to the ABC analysis results extracted by the ABC analysis. For example, the Similar ABC Analysis Extraction Function 146 extracts similar ABC analysis results (hereinafter referred to as similar analysis results) that are similar to each of the multiple ABC analysis results contained in the Extraction Results DB 155 to generate the Similarity DB 156. The Similar ABC Analysis Extraction Function 146 is an example of a Similar ABC Analysis Extraction Unit.

[0082] The Similar ABC Analysis Extraction Function 146 compares the ABC analysis result (hereinafter referred to as the current analysis result) generated by the generation function 144 with each of the multiple similar analysis results. When comparing the current analysis result with the similar analysis results, the Similar ABC Analysis Extraction Function 146 calculates the similarity of pre-action elements, action elements, and post-action elements in both the similar analysis results and the current analysis result. The Similar ABC Analysis Extraction Function 146 calculates the average value of the calculated similarity of pre-action elements, action elements, and post-action elements as the similarity between the similar analysis result and the current analysis result.

[0083] The Similarity ABC Analysis Extraction Function 146 calculates the similarity between the similarity analysis results and the current analysis results using methods such as pattern matching, cosine similarity, and word carrying distance. The Similarity ABC Analysis Extraction Function 146 then generates a Similarity DB 156 by creating a database of the calculated similarity scores. Figure 13 shows an example of the contents of the Similarity DB 156.

[0084] For example, when comparing the similarity analysis results of extraction result ID 00001 with the current analysis results generated based on the trouble situation shown in Figure 6, the similarity of pre-action elements is "0.303", the similarity of action elements is "0.183", and the similarity of post-action elements is "0.033". The similarity ABC analysis extraction function 146 calculates the similarity between the similarity analysis results and the current analysis results in this case as "0.173".

[0085] The similarity ABC analysis extraction function 146 calculates the similarity between the current analysis result and all ABC analysis results included in the extracted results DB 155 using a similar procedure. A higher similarity score indicates greater similarity, with a maximum value of 1 and a minimum value of 0. The similarity is calculated using the mean values ​​of the pre-action elements, action elements, and post-action elements. However, each element may be weighted according to a predetermined criterion, or indicator values ​​other than the mean (representative values), such as the median, may be used.

[0086] The behavioral improvement analysis function 147 analyzes the improvement of the subject's behavior based on the results of comparing the pre-behavioral elements and behavioral elements included in the ABC analysis extraction results with similar pre-behavioral elements and similar behavioral elements included in the similar ABC analysis extraction results. For example, the behavioral improvement analysis function 147 compares the ABC analysis extraction results extracted by the analysis function 143 with the similar ABC analysis extraction results extracted by the similar ABC analysis extraction function 146. Based on the results of comparing the ABC analysis extraction results with the similar ABC analysis extraction results, the behavioral improvement analysis function 147 analyzes the trend of improvement in the subject's behavior, such as whether or not improvement in the subject's behavior was observed. The behavioral improvement analysis function 147 is an example of the behavioral improvement analysis unit.

[0087] The behavioral improvement analysis function 147 analyzes whether there has been a change or improvement in the behavioral and post-behavioral elements, for example, based on similar analysis results that are similar to the current analysis result, such as similar pre-behavioral elements to the current analysis result. Therefore, the behavioral improvement analysis function 147 assigns scores to the behaviors (events) included in the post-behavioral elements for both the current analysis result and the similar analysis result.

[0088] Figure 14 shows an example of the content of the scored analysis results and similar analysis results, and Figure 15 shows an example of the content of the score mapping data 157. Here, it is assumed that the similar analysis results with extraction result IDs "0005", "0042", "0452", and "0646" were extracted as similar analysis results similar to the analysis results of the current analysis.

[0089] As shown in Figure 14, the analysis results in this case were assigned a score of "+4". The similarity analysis result for extraction result ID "0005" was assigned "-4", the similarity analysis result for extraction result ID "0042" was assigned "-4", the similarity analysis result for extraction result ID "0452" was assigned "+1", and the similarity analysis result for extraction result ID "0645" was assigned "-4".

[0090] The behavioral improvement analysis function 147 calculates the scores to be assigned to the current analysis result and similar analysis results by referring to the score mapping data 157, for example. For example, if the post-action events of the current analysis result or similar analysis result (hereinafter referred to as the target analysis result) that are subject to score assignment include the action of "taking something," a score of "-4" will be assigned.

[0091] Similarly, if the post-action events in the subject analysis results include actions (events) such as "pulling a friend hard" or "staring intently," points of "-5" and "+1" are assigned, respectively. The points assigned for each action (event) are set within a range, for example, from "-5" to "+5". The behavioral improvement analysis function 147 determines that an improvement has been observed in the subject's behavior if the assigned points are high, for example, above a predetermined threshold (for example, +1 or higher).

[0092] The behavioral improvement analysis function 147 may perform scoring based on the analysis results using methods other than the method of referring to the mapping table shown in Figure 15. The behavioral improvement analysis function 147 may, for example, utilize the output of a model created by machine learning or artificial intelligence based on the results of human scoring in the past. The behavioral improvement analysis function 147 may also perform qualitative evaluations (e.g., "improvement is being seen," "it is getting a little worse," etc.) using artificial intelligence or natural language processing in place of or in addition to scoring.

[0093] The provision function 148 provides behavioral improvement information based on the analysis results of the subject's behavioral improvement analyzed by the behavioral improvement analysis function 147. The provision function 148 provides the generated behavioral improvement information to the supporter by transmitting it to the output device 20 using the communication interface 110. For example, the provision function 148 provides the supporter with similar pre-behavioral elements and similar behavioral elements included in similar ABC analysis extraction results that are similar to the ABC analysis extraction results extracted from the subject's post-behavioral elements. The provision function 148 may also include other information in the behavioral improvement information.

[0094] The provided function 148 provides supporters with information on behavioral improvement as a warning, regardless of whether a trouble situation has occurred, for example, before a trouble situation occurs. It may also provide behavioral improvement information to supporters when a situation prone to trouble arises near the person in question. Provided function 148 is an example of a provisioning unit.

[0095] Figure 16 shows an example of the content of behavioral improvement information. The provisioning function 148 generates data to display the content shown in Figure 16 as behavioral improvement information and transmits it to the output device 20. The output device 20 provides the information transmitted by the provisioning function 148 to the supporter by displaying it on a screen or other means.

[0096] As shown in Figure 14, the display screen of the tablet terminal 18, which is the output device 20, displays information in a first display area GA1 and a second display area GA2. For example, the first display area GA1 displays "Current Analysis Results" and "Past Similar Analysis Results," and the second display area GA2 displays a graph showing the temporal changes in each score when the current analysis results were extracted. In addition to these, the display screen may also display the name of the person being reported to (e.g., Mr. A) and the title "Behavioral Report." Furthermore, comments such as "AI-generated trend analysis comments" may be displayed. In this case, progress information such as "Improvement was observed from February to April 2023, but has gradually worsened since then" and advice information such as "Please pay a little attention to how you communicate with your friends" may also be included.

[0097] The medical information processing device 100 of this embodiment identifies trouble information related to troubles experienced by the subject in its identification function 142, and analyzes the trouble information using applied behavior analysis in its analysis function 143, extracting post-behavioral elements corresponding to the trouble information. Based on the extracted post-behavioral elements, the medical information processing device 100 extracts pre-behavioral elements and behavioral elements and generates ABC analysis results. The medical information processing device 100 extracts reinforcers and depressors from the behavioral elements and post-behavioral elements and generates analysis result information. Therefore, it is possible to provide appropriate support for developmental disorders. Based on the generated analysis result information, the medical information processing device 100 proposes communication content for the subject. Therefore, it is possible to provide more appropriate support for developmental disorders.

[0098] In this embodiment, the medical information processing device 100 extracts similar ABC analysis results in the similar ABC analysis extraction function 146 and compares the current analysis result with the similar analysis results. In the behavior improvement analysis function 147, the medical information processing device 100 analyzes the improvement of the subject's behavior based on the results of comparing the current analysis result with the similar analysis results, and in the provision function 148 generates behavior improvement information and provides it to the support provider. As a result, behavior improvement information can be provided to the subject before problems occur, thus preventing problems from happening in the first place.

[0099] According to at least one embodiment described above, by having an acquisition unit that acquires behavioral information relating to the subject's words and actions, an identification unit that identifies trouble information relating to troubles that have occurred to the subject, an analysis unit that analyzes the behavioral information by behavioral analysis and extracts structured elements corresponding to the trouble information, and a generation unit that generates analysis result information corresponding to the analysis results of the behavioral information, appropriate support can be provided for developmental disorders.

[0100] While several embodiments have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]

[0101] 10. Speech and Behavior Recognition Devices 11 Mike 12 GNSS equipment 13 Cameras 14. Accelerometer 15. Pressure Sensor 16 Smartwatches 17 Card-type devices 18 Tablet devices 20 Output Devices 100 Medical Information Processing Devices 110 Communication Interface 120 Input Interfaces 130 displays 140 Processing Circuits 141 Acquisition function 142 Specific Functions 143 Analysis functions 144 Generation function 145 Suggestion Function 146 Similar ABC analysis extraction function 147 Behavior improvement analysis function 148 Provided functions 150 memory 151 ABC analysis information 152 Reinforcement Table 153 Weakening element table 154 Behavioral Change Communication Table 155 Extraction result DB 156 Similarity DB 157-point mapping data 161 Recognition function 162 Troubleshooting Status Extraction Function 163 Post-Action Element Extraction Function 164 Reinforcer / weakener extraction function GA1 1st display area GA2 2nd display area NW Network P Target Persons P1 Target audience P2 Friends SC1 First still image SC2 Second still image

Claims

1. An acquisition unit that acquires behavioral information regarding the subject's words and actions, A unit for identifying trouble information related to troubles that occurred to the aforementioned target person, An analysis unit analyzes the aforementioned verbal and behavioral information through behavioral analysis and extracts structured elements corresponding to the trouble information, The system includes a generation unit that generates analysis result information corresponding to the analysis results of the aforementioned verbal and behavioral information. Medical information processing device.

2. The aforementioned subject is a child. The medical information processing device according to claim 1.

3. The system further comprises a proposal unit that proposes the content of the subject's communication based on the generated analysis result information. The medical information processing device according to claim 1.

4. The aforementioned identification unit includes a recognition unit that recognizes the trouble that occurred to the subject, The system includes a trouble occurrence status extraction unit that extracts the trouble occurrence status as trouble information when the aforementioned trouble occurs. The medical information processing device according to claim 1.

5. The analysis unit includes a post-action element extraction unit that extracts post-action elements corresponding to events after the action in the ABC analysis based on the trouble information, Includes a reinforcer / punisher extraction unit that extracts reinforcers and punishers related to the aforementioned post-action elements, The medical information processing device according to claim 1.

6. The system further comprises a proposal unit that proposes the content of the subject's communication based on the generated analysis result information, The proposal unit proposes the content of the target person's communication based on the reinforcer and the weakener. The medical information processing device according to claim 5.

7. The system further includes a similar ABC analysis extraction unit that extracts similar ABC analysis results to the ABC analysis results extracted by the aforementioned ABC analysis. The medical information processing device according to claim 5.

8. A behavioral improvement analysis unit analyzes the improvement of the subject's behavior based on the results of comparing the pre-behavioral elements and behavioral elements included in the ABC analysis extraction results with similar pre-behavioral elements and similar behavioral elements included in the similar ABC analysis extraction results. The system further comprises a provisioning unit that provides information based on the results of an analysis of the improvement of the subject's behavior, The medical information processing device according to claim 7.

9. Computers We obtain behavioral information regarding the subject's words and actions. Identify trouble information related to the troubles that occurred to the aforementioned persons, The aforementioned behavioral information is analyzed through behavioral analysis, and structured elements corresponding to the trouble information are extracted. The system generates analysis result information based on the analysis results of the aforementioned behavioral information. Medical information processing method.

10. On the computer, We obtain behavioral information regarding the subject's words and actions. Identify trouble information related to the troubles that occurred to the aforementioned persons, The aforementioned behavioral information is analyzed through behavioral analysis, and structured elements corresponding to the trouble information are extracted. The system generates analysis result information corresponding to the analysis results of the aforementioned verbal and behavioral information. program.

Images