Determination support program, information processing device, and determination support method

Abstract

To provide a determination support program, an information processing device, and a determination support method that enable objective evaluation and classification of child abuse risk.SOLUTION: A mobile terminal 1 comprises questionnaire means 100 which conducts a PSI questionnaire for evaluating childcare stress of a parent 4p who is a caregiver of a child 4c, facial recognition measurement means 101 which measures facial recognition of the parent 4p, and risk determination means 102 which determines a maltreatment risk of the parent 4p on the basis of a combination of a questionnaire value obtained from answers to the questionnaire and a facial recognition measurement value being a measurement result of the facial recognition.SELECTED DRAWING: Figure 2

Description

[Technical Field]

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

[0002] As a conventional technique, a judgment support method has been proposed that supports the judgment of the risk of child abuse (maltreatment) based on the reaction time to find a smiling face when the smiling face is displayed on a screen (see, for example, Patent Document 1).

[0003] The assessment support method disclosed in Patent Document 1 displays illustrations of smiling and other facial expressions on a screen, measures the reaction time to find the smiling face, and assesses the risk of child abuse based on whether the reaction speed is faster or slower than a predetermined threshold. This is based on the fact that caregivers with a high risk of child abuse are less efficient at finding happy expressions than caregivers with a low risk of child abuse. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Shimada K*, Kasaba R, Yao A, Tomoda A. Less efficient detection of positive facial expressions in parents at risk of engaging in child physical abuse, BMC Psychology, 7(1):56, 2019, doi: 10.1186 / s40359-019-0333-9 Summary of the Invention [Problem to be solved by the invention]

[0005] However, although the above-mentioned assessment support methods can evaluate a caregiver's potential child-rearing stress and allow for objective evaluation, they have the problem of difficulty in classifying child abuse risks.

[0006] An object of the present invention is to provide a judgment support program, an information processing device, and a judgment support method that enable objective evaluation and support in classifying child abuse risks. [Means for solving the problem]

[0007] In order to achieve the above object, one aspect of the present invention provides the following judgment support program, information processing device, and judgment support method.

[0008] [1] Computer, A judgment support program that functions as a judgment means for determining a caregiver's risk of maltreatment based on a combination of questionnaire values obtained from responses to a questionnaire that evaluates the caregiver's parenting stress and facial expression recognition measurements that measure the caregiver's facial expression recognition. [2] a survey means for displaying and processing the questions of the survey, accepting answers to the questions, and calculating the survey value based on the answers; The judgment support program according to claim 1, further functioning as an expression recognition measurement means that displays and processes facial expressions, accepts an answer as to whether the facial expression includes a smile and the answer time, and calculates the expression recognition measurement value based on the answer and the answer time. [3] The judgment support program according to claim 1, wherein the judgment means judges the risk of maltreatment by combining whether the questionnaire value is high or low relative to a threshold value and whether the facial expression recognition measurement value is high or low relative to a threshold value. [4] The judgment support program according to claim 1 or 3, further functioning as an output means for outputting support content based on the maltreatment risk. [5] The decision support program according to [4], further functioning as an output means for outputting the support content in accordance with the user of the output destination. [6] The judgment support program according to [3], wherein the judgment means determines the threshold of the questionnaire value and the threshold of the facial expression recognition measurement value based on the sensitivity and specificity of the ROC curve. [7] An information processing device having a judgment means for judging the risk of maltreatment of a caregiver based on a combination of a questionnaire value obtained from responses to a questionnaire that evaluates the caregiver's parenting stress and a facial expression recognition measurement value that measures the caregiver's facial expression recognition. [8] A judgment support method comprising a step of determining a caregiver's risk of maltreatment based on a combination of a questionnaire value obtained from responses to a questionnaire assessing the caregiver's parenting stress and a facial expression recognition measurement value that measures the caregiver's facial expression recognition. [Effects of the Invention]

[0009] According to the inventions of claims 1, 7 and 8, objective evaluation is possible and support for classification of child abuse risk is possible. According to the invention of claim 2, it is possible to display and process questionnaire questions, accept answers to the questions, and calculate a questionnaire value based on the answers, and it is possible to display and process facial expressions, accept answers as to whether the facial expressions include a smile and the answer time, and calculate the facial expression recognition measurement value based on the answers and answer time. According to the invention of claim 3, the risk of maltreatment can be determined by combining whether the questionnaire value is high or low relative to a threshold value and whether the facial expression recognition measurement value is high or low relative to a threshold value. According to the invention of claim 4, it is possible to output support content based on maltreatment risk. According to the invention of claim 5, it is possible to output support contents according to the user of the output destination. According to the invention of claim 6, the threshold value of the questionnaire value and the threshold value of the facial expression recognition measurement value can be determined based on the sensitivity and specificity of the ROC curve. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a schematic diagram illustrating an example of the configuration of a judgment support system according to an embodiment. [Figure 2] FIG. 2 is a block diagram illustrating an example of the configuration of an information processing device according to an embodiment. [Figure 3] FIG. 3 is a schematic diagram showing an example of the configuration of a screen displayed by the questionnaire means. [Figure 4] FIG. 4 is a schematic diagram showing an example of the configuration of a screen displayed by the facial expression recognition measurement means. [Figure 5A] FIG. 5A is a graph illustrating a method for determining the threshold value. [Figure 5B] FIG. 5B is a table showing specific examples of risk classification. [Figure 6] FIG. 6 is a graph showing ROC curves for the cases where only the PSI questionnaire was administered and where the PSI questionnaire and facial expression recognition measurement were administered. [Figure 7] FIG. 7 is a flowchart showing an example of the operation of the mobile terminal. [Figure 8] FIG. 8 is a branching diagram for explaining the operation of the output means. DETAILED DESCRIPTION OF THE INVENTION

[0011] [Embodiment Mode] FIG. 1 is a schematic diagram illustrating an example of the configuration of a judgment support system according to an embodiment.

[0012] The assessment support system is configured by connecting mobile terminal 1, which serves as an information processing device, and terminal 2 so that they can communicate via network 3. Mobile terminal 1 is operated primarily by parent 4p, who is the caregiver of child 4c, and terminal 2 is operated primarily by doctor 5. Mobile terminal 1 supports parent 4p in classifying the parenting stress and maltreatment (child abuse) risk of child 4c based on input from parent 4p, and displays the classification results on the display unit of mobile terminal 1, as well as displaying the classification results on terminal 2 as necessary to support doctor 5 in making a diagnosis.

[0013] (Configuration of information processing device) FIG. 2 is a block diagram illustrating an example of the configuration of an information processing device according to an embodiment.

[0014] The mobile terminal 1 comprises a control unit 10 which is composed of a CPU (Central Processing Unit) and the like and controls each part and executes various programs; a memory unit 11 which is composed of a storage medium such as a flash memory and stores information; a display unit 12 which is composed of an LCD (Liquid Crystal Display) and the like and displays characters and images; an operation unit 13 which is composed of a touch panel, buttons, etc. arranged on the display unit and accepts user operations; and a communication unit 14 which communicates with the outside via a network 3.

[0015] The control unit 10 executes a judgment support program 110, which will be described later, to function as a questionnaire means 100, a facial expression recognition measurement means 101, a risk judgment means 102, an output means 103, and the like.

[0016] The questionnaire means 100 displays a questionnaire, for example, questions on the PSI (Parenting Stress Index), on the display unit 12, accepts answers from the parent 4p who is the user via the operation unit 13, and stores them in the storage unit 11 as questionnaire result information 111.

[0017] The facial expression recognition measurement means 101 displays a facial expression pattern on the display unit 12, accepts an answer from the parent 4p who is the user via the operation unit 13, and stores it in the storage unit 11 as facial expression recognition measurement result information 112.

[0018] The risk assessment means 102 compiles the questionnaire result information 111 to evaluate the parenting stress that the parent 4p is aware of, and also compiles the facial expression recognition measurement result information 112 to evaluate the parent 4p's unconscious (latent) parenting stress, and performs risk classification based on the combination of these assessment results and stores them in the memory unit 11 as risk assessment result information 113.

[0019] The output means 103 displays the risk assessment result information 113 on the display unit 12 of the mobile terminal 1 and also outputs it to the terminal 2 via the communication unit 14 .

[0020] The storage unit 11 stores a judgment support program 110 that causes the control unit 10 to operate as each of the above-mentioned means 100-103, questionnaire result information 111, facial expression recognition measurement result information 112, risk judgment result information 113, and the like.

[0021] (Operation of information processing device) Next, the operation of this embodiment will be explained by dividing it into (1) input reception operation and (2) determination and output operation.

[0022] (1) Input acceptance behavior First, a parent 4p as a user operates the mobile terminal 1 to receive support for assessing their own stress in child rearing and risk of maltreatment. Specifically, they input answers to the PSI questionnaire and answers to measure facial expression recognition.

[0023] The mobile terminal 1 executes the judgment support program 110 in accordance with the operation content of the parent 4p, and functions as the means 100 to 103, etc., thereby performing the following operations.

[0024] FIG. 7 is a flowchart showing an example of the operation of the mobile terminal 1.

[0025] First, the survey means 100 displays a survey, for example, PSI questions, on the display unit 12 (S1).

[0026] FIG. 3 is a schematic diagram showing an example of the configuration of a screen displayed by the questionnaire means 100. As shown in FIG.

[0027] The screen 120 has PSI survey questions 1201 and answer options 1202 for the questions 1201. A plurality of questions (for example, 19 questions) are prepared, and answer options (for example, five levels) are prepared for each question.

[0028] The parent 4p checks the content of the question 1201 and selects an answer to the question 1201 from the answer options 1202. The parent 4p also selects answer options for other questions.

[0029] The survey means 100 receives responses from the parent 4p, who is the user, via the operation unit 13 and stores them in the storage unit 11 as survey result information 111 (S2).

[0030] Next, the facial expression recognition measurement means 101 displays the facial expression pattern on the display unit 12 (S3).

[0031] FIG. 4 is a schematic diagram showing an example of the configuration of a screen displayed by the facial expression recognition measurement means 101.

[0032] The screen 120a has an expression display field 1203 in which expressions such as a blank expression and a smile are displayed, a found button 1204 that the parent 4p presses when they find a smile in the expression display field 1203, and a not found button 1205 that the parent 4p presses when they cannot find a smile.

[0033] Furthermore, the number of facial expressions displayed in the facial expression display field 1203 on screen 120a is one, on screen 120b six, and on screen 120c twelve. The number of facial expressions displayed on the screen is not limited to those shown in the example, and the type and number of screens can be changed as appropriate. Furthermore, the facial expressions may be illustrations or photographs. In the case of photographs, a photo of the face of child 4c may be used. Furthermore, the facial expressions are not limited to a blank expression or a smiling face, and may also be a crying face.

[0034] The parent 4p checks the contents of the expression display field 1203, and if he finds a smile, he presses the found button 1204, and if he cannot find a smile, he presses the not found button 1205.

[0035] The facial expression recognition measurement means 101 receives an answer from the parent 4p who is the user via the operation unit 13, measures the answer time, and stores it in the storage unit 11 as facial expression recognition measurement result information 112 (S4).

[0036] Before conducting the questionnaire and facial expression recognition measurement, an explanation of how to answer the questionnaire or a practice step may be provided.

[0037] (2) Judgment and output operations Next, the risk assessment means 102 compiles the questionnaire result information 111 to evaluate the parenting stress that the parent 4p is aware of, and also compiles the facial expression recognition measurement result information 112 to evaluate the parent 4p's unconscious (latent) parenting stress (S5).

[0038] The risk assessment means 102 quantifies the questionnaire result information 111 using a method defined in the PSI questionnaire. The risk assessment means 102 also calculates, for example, the average answer time of correct answers as a method of aggregating the facial expression recognition measurement result information 112. The average may be calculated using a simple average, or other methods such as a weighted average.

[0039] In addition, the risk assessment means 102 performs risk classification as shown in Figure 5B based on the combination of the evaluation results of the aggregated results (questionnaire values) of the questionnaire result information 111 and the aggregated results (facial expression recognition measurement values) of the facial expression recognition measurement result information 112, and stores this in the memory unit 11 as risk assessment result information 113 (S6).

[0040] FIG. 5B is a table showing specific examples of risk classification.

[0041] First, the risk assessment means 102 classifies the questionnaire value as either "high" or "low" depending on whether it is above or below a predetermined threshold. The threshold for the questionnaire value may be predetermined, or, for example, if there are other samples, a value above the 20%ile may be "high" and a value below the 20%ile may be "low." Note that the values may be classified as "high," "normal," and "low" in addition to "high" and "low."

[0042] Next, the risk assessment means 102 classifies the facial expression recognition measurement value as either "high" or "low" depending on whether it is above or below a predetermined threshold. The threshold for the facial expression recognition measurement value may be predetermined, or, for example, if there are other samples, a value above the 20%ile may be considered "high," and a value below the 20%ile may be considered "low." Note that the values may be classified as "high," "normal," and "low" in addition to "high" and "low."

[0043] The threshold value is determined as follows.

[0044] 5A(a) and (b) are graphs for explaining a method for determining the threshold value.

[0045] The 26 subjects completed the PSI questionnaire and evaluated their PSI scale scores for the following three thresholds. When the 15%ile was used as the threshold to determine whether the PSI scale score was high or low, the ROC curve was as shown in Figure 5A(a).

[0046] Furthermore, when the 20th percentile was used as the threshold to judge whether the PSI scale score was high or low, the ROC curve was as shown in Figure 6, which will be described later.

[0047] Furthermore, when the 25th percentile was used as the threshold to judge whether the PSI scale score was high or low, the ROC curve was as shown in Figure 5A(b).

[0048] As described above, in the example embodiment, the ROC curve with the highest sensitivity and specificity was obtained when the threshold was set to 20%ile. Therefore, when different subjects are used, the ROC curve is evaluated using the same method to select a threshold with high sensitivity and specificity. On the other hand, a threshold that can be expected to have a certain level of sensitivity and specificity may be set as a predetermined threshold.

[0049] Next, the risk assessment means 102 determines that there is both aware child-rearing stress and unconscious (latent) child-rearing stress when the questionnaire value is high and the facial expression recognition measurement value is high (1). Also, when the questionnaire value is high and the facial expression recognition measurement value is low, the risk assessment means 102 determines that there is aware child-rearing stress and unconscious (latent) child-rearing stress is low (2). Also, when the questionnaire value is low and the facial expression recognition measurement value is high, the risk assessment means 102 determines that there is low aware child-rearing stress and unconscious (latent) child-rearing stress (3). Also, when the questionnaire value is low and the facial expression recognition measurement value is low, the risk assessment means 102 determines that there is low aware child-rearing stress and unconscious (latent) child-rearing stress (4).

[0050] (1) is a case requiring protection, and requires all of the following observation requirements A1, A2, and B. Needs observation A1: Re-examine cognitive function of parenting stress If the depression is severe, there is a possibility that the mother may not be aware of her own child-rearing stress, or may fill out the questionnaire carelessly due to reasons such as not being able to understand it. Therefore, it is necessary to refer the mother to medical care such as a psychiatrist or psychosomatic medicine specialist. A2: Recheck whether the answers about childcare stress levels are true Caregivers who have experienced maltreatment may refuse support or may give false answers because they do not expect support. Therefore, it is necessary to connect them to interviews and counseling by doctors or psychologists. Observation required B: Identify factors causing high parenting stress It is necessary to address the underlying causes, that is, to provide appropriate advice on environmental adjustments and ways to deal with the situation, and to refer the child to parent training.

[0051] (2) requires observation A1 and A2, and (3) requires observation B.

[0052] (4) requires monitoring. Both self-awareness and unconscious (latent) parenting stress are considered to be low risk, but it is necessary to monitor them while keeping in mind the possibility that they may be lying about their answers to the PSI questionnaire or that they may be experiencing worries or parenting stress.

[0053] Next, the output means 103 displays the risk assessment result information 113 on the display unit 12 of the mobile terminal 1 and outputs it to the terminal 2 via the communication unit 14 (S7). Note that the output means 103 may display, as the risk assessment result information 113, the presence or absence of self-aware parenting stress and the presence or absence of unconscious (latent) parenting stress, as well as support content such as comments on the parenting stress and how to deal with it. The output means 103 may also transmit the risk assessment result information 113 to the terminal 2 operated by the doctor 5, so that the doctor 5 can make a formal diagnosis based on the displayed content. Furthermore, the above-mentioned assessment support may be provided during online medical consultation, and part or all of the above-mentioned risk assessment result information 113 may be used as the result of the online medical consultation, including the doctor 5's diagnosis, or may be incorporated into a digital maternal and child health handbook, for example.

[0054] The parent 4p may check the content displayed on the display unit 12 of the mobile terminal 1, and may further receive a diagnosis and / or treatment from the doctor 5 as necessary.

[0055] Furthermore, the output means 103 changes the output contents according to the output destination and the determination result as follows.

[0056] FIG. 8 is a branching diagram for explaining the operation of the output means.

[0057] First, the output means 108 identifies the user (S70). Specifically, it determines whether the user is a medical professional, a supporter, or a caregiver based on the user information of the mobile terminal 1 and the location of use (at home, at a medical institution, or at a support site). Here, a medical professional is someone who diagnoses, treats, and supports caregivers from a medical perspective, such as a doctor or psychologist. A supporter is someone involved in government, child welfare, maternal and child health, etc., that is, someone who supports and intervenes in multi-disciplinary collaboration with caregivers from a supporter perspective, such as a midwife, public health nurse, or staff member at a child-rearing support organization or child consultation center. A caregiver is someone involved in child-rearing, such as a parent currently raising a child.

[0058] If the user is a medical professional (S71), output including the following content is performed according to the judgment result (S72) (S73). If the output result is (1), content is output requesting that the child consultation center be notified, or that the person be connected to a child and family support center, a local council for protection measures, etc. If the output result is (2), content is output requesting that an interview or counseling session be conducted by a doctor, etc., or that the relevant organization be contacted. If the output result is (3), content is output requesting that appropriate advice be given on how to deal with the cause (environmental adjustments and response). If the output result is (4), content is output requesting that regular outpatient follow-up be conducted and that the person be monitored.

[0059] If the user is a supporter (S71), output including the following content is performed according to the determination result (S74) (S75). If the output result is (1), a request is made to connect the user to medical care such as a psychiatrist or psychosomatic medicine specialist, and content to introduce social resources is output. If the output result is (2), a request is made to connect the user to an interview by a child welfare worker or public health nurse, individual parent training, etc., and content to introduce and introduce social resources is output. If the output result is (3), a request is made to connect to a local parenting center or peer support, and content to introduce and introduce social resources is output. If the output result is (4), content to request regular home visits, etc. to monitor the user is output.

[0060] If the user is a caregiver (S71), output including the following content is performed according to the determination result (S76) (S77). If the output result is (1), for example, a message is output saying, "You and your child are in a very difficult situation. Please call the nearest child consultation center or child-rearing support organization immediately and ask for help." If the output result is (2), for example, a message is output saying, "Are you or your child having any problems? Please consult the nearest child consultation center or child-rearing support organization." If the output result is (3), for example, a message is output saying, "Is there anyone nearby who can help or consult with you? Please first try calling a child-rearing support organization or talking to someone you feel comfortable talking to, such as a teacher at your child's kindergarten or school." If the output result is (4), for example, a message is output saying, "There are various services and consultation organizations in your area. Please feel free to use them at any time."

[0061] FIG. 6 is a graph showing ROC curves (Receiver Operating Characteristic Curves) when only the PSI questionnaire was administered and when the PSI questionnaire and facial expression recognition measurement were administered.

[0062] A PSI questionnaire and facial expression recognition test were administered to a total of 28 subjects. When assessed using only the PSI questionnaire, the AUC (Area Under the Curve) was 62%, while when assessed using the PSI questionnaire and facial expression recognition test, the AUC was 75%. The AUC, or accuracy of assessment, was improved when the PSI questionnaire and facial expression recognition test were administered. Specifically, when the PSI questionnaire and facial expression recognition test were administered, the abuse risk group consisted of 12 people, of whom 5 were correctly identified (41.7%). On the other hand, the non-abuse risk group consisted of 16 people, of whom 14 were correctly identified (87.5%). This shows a high accuracy rate.

[0063] (Effects of the embodiment) According to the above-described embodiment, not only the PSI questionnaire but also facial recognition measurements are conducted, and the maltreatment risk is classified by combining these aggregated results, allowing for objective evaluation and classification of child abuse risk. Furthermore, by displaying comments, response methods, etc. in accordance with the classification results and enabling diagnosis and / or treatment by a doctor 5, the number of steps required for diagnosis and treatment can be reduced, increasing the opportunities for more appropriate diagnosis and treatment. Furthermore, by changing the output content depending on the output destination of the classification results (supporter, medical professional, person involved (caregiver), etc.), support can be provided in collaboration with multiple professions, and rapid response to personnel and material shortages of support personnel can be made.

[0064] [Other embodiments] The present invention is not limited to the above-described embodiment, and various modifications are possible without departing from the spirit of the present invention.

[0065] For example, the above example shows that the questionnaire result information 111, facial expression recognition measurement result information 112, and risk assessment result information 113 are recorded and classified only once, but it is also possible to record and analyze this information over time and assess risk based on changes in the information.

[0066] In the above embodiment, the functions of the means 100 to 103 of the control unit 10 are realized by a program, but all or part of the means may be realized by hardware such as an ASIC. The program used in the above embodiment may also be stored on a recording medium such as a CD-ROM and provided. The steps described in the above embodiment may be replaced, deleted, or added without departing from the spirit and scope of the present invention. [Explanation of symbols]

[0067] 1: Mobile device 2: Terminal 3: Network 4p: Parent 4c: Child 5: Doctor 10: Control section 11: Storage section 12: Display section 13:Operation unit 14: Communications Department 100: Survey method 101: Facial expression recognition measurement means 102: Risk assessment tools 103: Output means 110: Judgment support program 111: Survey result information 112: Facial expression recognition measurement result information 113: Risk assessment result information

Claims

1. Computer, A judgment support program that functions as a judgment means for determining a caregiver's risk of maltreatment based on a combination of questionnaire values obtained from responses to a questionnaire that evaluates the caregiver's parenting stress and facial expression recognition measurements that measure the caregiver's facial expression recognition.

2. a survey means for displaying and processing the questions of the survey, accepting answers to the questions, and calculating the survey value based on the answers; The judgment support program according to claim 1, further functioning as an expression recognition measurement means that displays and processes facial expressions, accepts an answer as to whether the facial expression includes a smile and the answer time, and calculates the expression recognition measurement value based on the answer and the answer time.

3. The judgment support program according to claim 1 , wherein the judgment means judges the risk of maltreatment based on a combination of whether the questionnaire value is high or low relative to a threshold and whether the facial expression recognition measurement value is high or low relative to a threshold.

4. The judgment support program according to claim 1 or 3, further functioning as an output means for outputting support content based on the maltreatment risk.

5. 5. The decision support program according to claim 4, further functioning as an output unit that outputs the support content in accordance with a destination user.

6. 4. The judgment support program according to claim 3, wherein the judgment means determines the threshold of the questionnaire value and the threshold of the facial expression recognition measurement value based on sensitivity and specificity of an ROC curve.

7. An information processing device having a judgment means for judging a caregiver's risk of maltreatment based on a combination of a questionnaire value obtained from responses to a questionnaire that evaluates the caregiver's parenting stress and a facial expression recognition measurement value that measures the caregiver's facial expression recognition.

8. A judgment support method comprising a step of judging a caregiver's risk of maltreatment based on a combination of a questionnaire value obtained from responses to a questionnaire that evaluates the caregiver's parenting stress and a facial expression recognition measurement value that measures the caregiver's facial expression recognition.

Images