An automated nursing record creation system, as well as an automated nursing time tracking system and a nurse skill level evaluation system utilizing the same.

Abstract

This system provides an automated nursing record creation system for use in nursing settings. [Solution] An automated recording system comprising a wearable terminal equipped with communication means for transmitting recorded video data and recorded audio data to an external source, and a computer equipped with means for receiving video data and audio data, analyzing this data, and automatically recognizing nursing content, wherein the wearable terminal is configured to start recording when the name of the person being cared for, as grasped by facial recognition means, matches the name of the person being cared for, as spoken by the nurse. The computer is equipped with means for automatically recognizing nursing content from the video of the person being cared for contained in the video data, means for automatically recognizing nursing content from the words spoken by the nurse contained in the audio data, and means for transcribing the nursing content automatically recognized from the video data and audio data into text to create a nursing record.

Description

Technical Field

[0001] The present invention relates to a system that automatically recognizes the content of care services (care details) provided by a caregiver to a care recipient and automatically creates a care record. Furthermore, it relates to a system for automatically grasping care time, a support system for care needs certification, a support system for calculating care benefits, and a skill level evaluation system for caregivers using this automatic care record creation system.

Background Art

[0002] Recently, in the situation of our country, which is said to be a super-aged society, it is predicted that there will be a shortage of about 570,000 care staff in 2040. Considering this prediction, it is necessary to increase the number of care staff by 32,000 each year, and since such an increase is extremely difficult, there is concern that the shortage of care staff will become an even more serious problem. Problems such as the increase in the burden of social insurance premiums such as long-term care insurance and the shortage of caregivers (persons involved in care such as care workers and helpers) have become particularly major social problems. As the number of elderly people increases, the number of dementia patients will also increase, and there is a prediction that one in five people aged 65 or over will have dementia in 2025. The increase in dementia patients means an increase in care recipients, and it is expected that the financial problems and personnel problems related to care will become even larger social problems.

[0003] Although the problems related to care are major social problems that affect the entire nation, the progress of IT (Information Technology) and ICT (Information and Communication Technology) has been extremely slow. Due to this influence, the efficiency of various operations related to care is not sufficient, cost reduction through improved work efficiency has not been achieved, and an environment for providing care to many recipients with limited staff has not been established.

[0004] By promoting IT and ICT integration, various tasks related to caregiving can be made more efficient, thereby reducing operating costs and enabling a greater number of tasks to be performed with limited personnel.

[0005] To address these issues, the Ministry of Health, Labour and Welfare is promoting the introduction and widespread adoption of ICT among care providers and other related organizations. This ICT implementation primarily involves the digitalization of care records, and the goal is to collect this digitalized care-related information to build up big data, which will then be analyzed to help solve problems.

[0006] However, these digitized care records are data manually entered by caregivers on-site using computers or tablet PCs, and are therefore susceptible to input errors and omissions. Furthermore, even when the same care is provided, there may be differences in how it is described, and differences in judgment regarding whether or not something should be recorded as care, meaning that the content of these records can be inconsistent and subjective.

[0007] Furthermore, for caregivers who are already burdened with numerous caregiving tasks, even the task of entering care records can be a significant burden. Care records are often created over a period of one to two hours each day after completing their regular caregiving duties, and this is a major workload given the shortage of caregivers. In light of these circumstances, there is a particular need, especially in caregiving settings, for a means to automatically and accurately grasp and recognize caregiving content (eliminating variability and arbitrary elements) and to automatically create caregiving records.

[0008] The inventor focused particularly on capturing the specific details of the care services provided (also simply called "care content") and the actual time spent on these care services (also simply called "care time") within care records. This is because the content and time of care are closely related to the calculation and billing of care fees, the collection of information for determining the need for care, and the creation of care records, which are all important tasks in the care setting. In other words, if the content and time of care could be captured automatically and accurately, the workload related to these matters would be greatly reduced.

[0009] Regarding the creation of care records, Patent Document 1 discloses prior art aimed at IT / ICT implementation. The technology described in Patent Document 1 is, "A means for acquiring voice data related to the caregiving duties of a care recipient, and A speech recognition / text conversion means that recognizes the speech data acquired by the speech acquisition means and converts it into text data, A care record creation means that creates a care record based on the text data converted from the voice data by the voice recognition and text conversion means, A display means that displays the care record created by the care record creation means, enabling the caregiver to confirm it, A modification means that allows the caregiver to modify the care record displayed on the display means as needed, A means for finalizing the care record after it has been confirmed by the caregiver and corrected as necessary, The system comprises "a voice data learning means that learns the text conversion function of the voice recognition / text conversion means based on the voice data, the text data converted from the voice data, and the text data ultimately adopted in the care record determined by the determination means."

[0010] This technology has the effect of "using voice acquisition means to acquire voice data related to the caregiving work of the person being cared for, and creating care records based on this data. This makes it possible to create care records using only the caregiver's voice, eliminating the need for voice communication between the caregiver and the system, and thus reducing the burden on caregivers during caregiving work."

[0011] However, while the technology described in Patent Document 1 can save caregivers the trouble of manually entering care records, it has the problem that it does not automatically create care records because it uses voice recognition to create the records themselves. Furthermore, there were problems with the lack of clear means for identifying care recipients and for tracking the start and end of caregiving tasks, as these tracking methods were insufficient. [Prior art documents] [Patent Documents]

[0012] [Patent Document 1] Japanese Patent Publication No. 2020-140620 [Overview of the project] [Problems that the invention aims to solve]

[0013] Therefore, the object of the present invention is to provide an automated nursing record creation system that can automatically and accurately grasp the specific content of nursing services provided in a nursing setting, automatically and accurately identify the person being cared for and the start and end times of nursing work, and automatically create nursing records. Furthermore, the invention aims to provide an automated nursing time tracking system and a nurse skill level evaluation system using this automated nursing record creation system. [Means for solving the problem]

[0014] The object of the present invention described above is achieved by the following means. 1. A system that automatically recognizes the content of nursing services provided and automatically creates nursing records based on that information, A wearable terminal equipped with means for photographing and recording the nursing care provided by a nurse, means for collecting and recording the words spoken by the nurse, and communication means for transmitting the recorded video data and the recorded audio data to an external source, The system consists of a computer equipped with means for receiving the aforementioned video data and audio data, and for analyzing these data to automatically recognize nursing content. The wearable device is configured to start recording when the name of the person being cared for, as determined by facial recognition, matches the name of the person being cared for, as verbally stated by the caregiver. The computer is, means for automatically recognizing nursing care content from the video of the care recipient included in the video data; means for automatically recognizing nursing care content from the words spoken by the caregiver included in the audio data; Furthermore, a nursing care record automatic creation system characterized by comprising means for converting the nursing care content automatically recognized from this video data and audio data into characters and creating a nursing care record.

[0015] 2. The nursing care record automatic creation system according to claim 1, characterized in that the wearable terminal is configured to end recording when it detects the name of the care recipient and a keyword indicating the end of the nursing care service from the words spoken orally by the caregiver.

[0016] 3. The nursing care record automatic creation system according to claim 1, characterized by recording video data and audio data together with characters indicating nursing care content.

[0017] 4. The nursing care record automatic creation system according to claim 1, characterized by comprising means for collating the nursing care content automatically recognized from the video data and the nursing care content automatically recognized from the audio data to confirm whether there is no difference in the nursing care content performed at the same time.

[0018] 5. A system using the nursing care record automatic creation system according to claims 1 to 4, means for automatically determining whether the nursing care content is content to be accumulated for the nursing care time; means for accumulating the nursing care time performed for each care recipient, characterized by having a nursing care time automatic grasping system.

[0019] 6. A system using the nursing care time automatic grasping system according to claim 5, A caregiver skill level evaluation system characterized by evaluating the caregiver's skill level from the caregiver's skill score obtained by video analyzing video data and the nursing care time required for each nursing care content.

[0020] 7. In cases where a nurse administers medication or intravenous fluids to a patient, A wearable device is equipped with an image / video recognition function to identify the name of a drug or intravenous drip and to verify whether it matches the name of the drug or intravenous drip read aloud by the nurse. An automated nursing record creation system according to any one of claims 1 to 4, characterized in that it includes a configuration to issue a warning if there is a discrepancy.

[0021] 8. In cases where a nurse administers medication or intravenous fluids to a patient, A wearable device is equipped with image / video recognition capabilities to identify the name of a drug or intravenous infusion, compare it with the patient's nursing records, medical chart, or prescription, and verify that the prepared drug or intravenous infusion is correct. An automated nursing record creation system according to any one of claims 1 to 4, characterized in that it includes a configuration to issue a warning if there is a discrepancy.

[0022] Reference inventions R1 to R9 of the present invention are described below. R1. A system that automatically recognizes the content of the care services provided and automatically creates care records based on that, A wearable terminal equipped with means for filming and recording the care provided by a caregiver, means for collecting and recording the words spoken by the caregiver, and communication means for transmitting the recorded video data and the recorded audio data to an external source, The system consists of a computer equipped with means for receiving the aforementioned video data and audio data, and for analyzing these data to automatically recognize the content of care. The wearable device is configured to start recording when the name of the person being cared for, as determined by facial recognition, matches the name of the person being cared for, as verbally stated by the caregiver (e.g., when the caregiver confirms the name of the person being cared for). The aforementioned computer, A means for automatically recognizing the care provided from the video data of the person receiving care (using a facial recognition system, etc.), The system includes means for automatically recognizing the content of care provided from the words spoken by the caregiver included in the aforementioned audio data (such as confirming the name of the person being cared for), Furthermore, the system for automatically creating care records is characterized by having a means for transcribing the care content automatically recognized from the video and audio data into text and creating care records. R2. The automatic care record creation system according to claim 1, characterized in that the wearable device is configured to terminate recording when it detects the name of the person receiving care and a keyword indicating the end of care services from among the words spoken by the caregiver. R3. Regarding a means of automatically recognizing the content of care from words spoken by caregivers contained in audio data, The computer has a means of pre-memorizing specific terms that represent specific caregiving procedures. The automatic care record creation system according to claim 1, characterized in that it detects specific terms uttered by the caregiver and automatically recognizes the care content based on these specific terms. R4. The automatic care record creation system according to claim 1, characterized in that it records video data and audio data along with text indicating the care content. R5. The automatic care record creation system according to claim 1, characterized in that it includes means for comparing care content automatically recognized from video data with care content automatically recognized from audio data to confirm whether there are any discrepancies in the care content performed at the same time. R6. A system that utilizes the automatic care record creation system described in claims 1 to 5, A means to automatically determine whether the care provided should be included in the calculation of care time, An automated care time tracking system characterized by having means for accumulating the amount of care time provided to each individual care recipient. R7. A system utilizing the automatic care time tracking system described in claim 6, Regarding a means to automatically determine whether the care provided should be included in the calculation of care time, The determination is based on whether the care provided falls into one of the following five categories: direct assistance with daily living, assistance with joints, behavioral problems, functional training, or medical care. For care services that fall under any of the five aforementioned categories, the time spent on such care is accumulated as care time. A support system for assessing the need for long-term care, characterized in that the accumulated care hours are used as the standard care hours for assessing the need for long-term care. R8. A system utilizing the automatic care time tracking system described in claim 6, Regarding a means to automatically determine whether the care provided should be included in the calculation of care time, The determination is based on whether or not the care provided is eligible for reimbursement under the care fee system. For each care service that falls under the aforementioned claimable items, the time spent on each individual care service is accumulated as care time. A system for supporting the calculation of care fees, characterized in that the care details and the time spent on each care item obtained are used for calculating care fees. R9. A system utilizing the automatic care time tracking system described in claim 6, A caregiver skill level evaluation system characterized by evaluating the caregiver's skill level based on a skill score obtained by analyzing video data and the care time required for each care task. [Effects of the Invention]

[0023] According to the above-mentioned points 1 to 6 of the present invention, recording can be started and stopped automatically without bothering the nurse, and the nurse can maintain a hands-free state. Furthermore, patient identification can be double-checked using both facial recognition and verbal confirmation by nurses, enabling automatic and accurate patient identification. This prevents patient mix-ups and errors such as incorrect entries in other patients' nursing records. Furthermore, according to the above-mentioned 7-8 of the present invention, it is possible to double-check the contents of drugs and intravenous infusions, thereby preventing errors in administering the wrong drugs or intravenous infusions to patients.

[0024] The effects of the reference invention of this present invention are described below. The invention described in R1 above is a system that automatically recognizes the content of the care services actually provided from both audio and video perspectives, and automatically creates care records based on this. According to the invention described in R1 above, the actual and specific details of the care services provided (hereinafter also simply referred to as "care content") can be automatically and accurately grasped, and care records can be automatically created based on that content without the caregiver having to consciously record it. The specific effects and benefits are as follows: First, if the system has a means of recording the words spoken by the caregiver (for example, "Mr. / Ms. XX, I'll help you with meals now," "Mr. / Ms. XX, we're going to take a bath now," "That's all for today," "You've finished your bath, let's dry your hair," etc.) and a means of automatically analyzing the care provided from this recorded audio data, then the system can automatically recognize from the results of this analysis what kind of care service was provided to which care recipient at a given time.

[0025] Technologies with such effects are known as speech recognition and speech analysis, and are widely used in smartphones and smart speakers. By analyzing the caregiver's voice data using known speech recognition and speech analysis technologies, the content of care can be understood concretely and automatically.

[0026] Next, if the system includes means for filming and recording the care provided by the caregiver, and means for automatically analyzing the care provided from this recorded video data, then the system can automatically recognize from the results of this analysis what kind of care service was provided to which care recipient at a given time.

[0027] Technologies with such effects are known as video analysis and are widely used in security cameras, medical diagnostics, and other applications. By analyzing video data of a person receiving care using known image analysis techniques, the details of the care provided can be understood concretely and automatically.

[0028] In the invention described in R1 above, the analysis means based on audio data or video data is entirely automated, meaning that the analysis is performed automatically by a computer or the like. When humans review and analyze audio and video, there are risks and disadvantages associated with infringing on the privacy of the person receiving care. The configuration according to the present invention eliminates these risks and disadvantages. Therefore, from the standpoint of privacy protection, the present invention is effective.

[0029] As described above, this system automatically recognizes the details of care services provided, using both audio and video. While automatic recognition of care content based on either audio or video alone could lead to errors in audio recognition or image / video analysis, this system automatically recognizes care content from both audio and video, thus improving the accuracy of automatic recognition. Furthermore, by transcribing this accurately recognized care content into text, accurate care records can be automatically created.

[0030] Furthermore, according to the invention described in R1 above, by using a wearable device in implementing the present invention, it is not necessary to install microphones or cameras in each location where care is provided, such as the living room, dining room, or bathroom, and initial costs can be reduced. In addition, since the care service can be filmed from the caregiver's perspective, the person being cared for and the care provided can be clearly captured, and problems such as blind spots that prevent filming can also be resolved.

[0031] The invention described in R1 above is ideal for understanding the content and duration of care in a one-on-one care environment where one caregiver provides care to one person receiving care. In particular, in home care settings, where it is difficult to install microphones and cameras in the homes being visited, a configuration in which the caregiver wears a wearable device to record and videotape is effective.

[0032] Furthermore, according to the invention described in R1 above, the wearable terminal is configured to start recording (which may also include audio recording; the same applies hereinafter) when the name of the person being cared for, as determined by the facial recognition means, matches the name of the person being cared for, as verbally spoken by the caregiver (the name spoken when confirming the name of the person being cared for). As such, recording can be started automatically without requiring any effort from the caregiver, and the caregiver can maintain a hands-free state. The start time of caregiving tasks for a specific person being cared for can also be automatically and accurately determined and recorded. Furthermore, the identification of care recipients can be double-checked using both facial recognition and the caregiver's statements, enabling automatic and accurate identification. This prevents misidentification of care recipients and avoids errors such as incorrect entries in the care records of other care recipients.

[0033] According to the invention described in R2 above, recording can be automatically terminated by detecting the name of the person receiving care and a keyword indicating the end of care services. This allows recording to be automatically terminated without requiring any action from the caregiver, and also enables the automatic and accurate determination and recording of the end time of care.

[0034] The invention described in R3 above is a configuration for further improving the reliability of automatic speech recognition. If different caregivers use different terminology to describe a particular care service, the care provided may not be correctly recognized. Therefore, if caregivers agree to use specific terminology (specific terms) to describe the content of care services, and if this specific terminology is also stored in a computer, the content of the care service can be automatically and correctly recognized.

[0035] Furthermore, the use of this specific term triggers the indication that the care provided should be recorded in the care record, ensuring that all necessary information is reliably included in the care record. Therefore, according to the invention described in item 2 above, by pre-determining terms (specific terms) related to the content of care services used by caregivers, and by having a computer detect these specific terms, the content of the care services provided can be accurately understood.

[0036] According to the invention described in R4 above, by recording video data and audio data along with text indicating the care provided, the care provided is recorded using video, audio, and text. The care record is created automatically by a computer without human intervention, but if, for example, doubts arise about the content of the care record or if any trouble occurs during the care service, it can be verified retrospectively using human hearing and sight rather than a computer.

[0037] According to the invention described in R5 above, by providing a means to compare the caregiving content automatically recognized from video data with the caregiving content automatically recognized from audio data and to confirm whether there are any discrepancies in the caregiving content performed at the same time, it is possible to retrospectively confirm the scene using video and audio when a discrepancy occurs in the automatically recognized caregiving content. This makes it possible to create more accurate caregiving records and to increase the reliability of these caregiving records.

[0038] The invention described in R6 above is a system that utilizes the automatic care record creation system described in R1 to R5 above, and is a system for automatically determining the time (care time) spent on the care services actually provided. By accurately tracking the actual time spent on care services, it becomes possible to quantitatively understand the level of care service needs for each individual care recipient. In other words, it becomes possible to quantify and understand the need for care services for each individual care recipient.

[0039] More specifically, the invention described in R6 above has a configuration that has means for automatically determining from analysis data obtained from audio data and video data whether or not the care content should be included in the care time. In other words, it can automatically determine and include in the care time whether or not it should be included, depending on the purpose of understanding the care content. For example, in a contract for care services used, if meal assistance is included in the monthly fee, but bathing assistance is not included in the monthly fee and requires an additional fee, then meal assistance and bathing assistance should not be included in the same care time range, but rather the care time included in the monthly fee and the care time not included should be included separately. The configuration of the present invention makes it possible to meet such requirements for categorization and classification.

[0040] As described above, according to the invention shown in R6, it is possible to automatically and accurately grasp the specific content of the care services provided and the actual time required for those care services, making it possible to "quantify" the care services actually provided, and also to "quantitatively" grasp the degree of care service required for each individual care recipient. All components of the system can operate automatically without any operation by the caregiver, and the caregiver can remain hands-free, so it does not interfere with providing care to the person being cared for, allowing them to concentrate on their work. Furthermore, it frees caregivers from indirect tasks such as recording care details and care time, reducing the overall burden on the caregiver.

[0041] According to the invention described in R7 above, by automatically and accurately determining the care time (standard time for care needs assessment), which is one of the criteria for determining the level of care needed in care needs assessment, from the actual care provided, it is possible to support the assessment of the level of care needed with high accuracy. Furthermore, it can reduce the burden on caregivers, visiting investigators, and care needs assessment committees involved in the care needs assessment process. Moreover, because it can automatically and accurately "quantify" the actual care time used as the basis for determining the care needs assessment, it is expected that some or all of the procedures and assessment tasks involved in the care needs assessment process can be automated, and fairness can be enhanced by eliminating subjective and arbitrary judgments by humans. As it is a means of determining care time from the actual care provided, it is particularly effective when reassessing the care needs level of individuals who are already receiving care.

[0042] Further details are as follows: According to the certification standards for long-term care stipulated by the Ministry of Health, Labour and Welfare, the level of care needed (support level 1-2 or care level 1-5) is determined by the amount of care time (standard time for long-term care certification) for five categories of care: direct assistance with daily living, assistance with joints, behavioral problems, functional training, or medical care. For example, if the amount of care time for the above five categories is 32 minutes or more but less than 50 minutes, the person will be judged as "care level 1".

[0043] The standard care needs assessment time used here is a measure of the effort required for care, and is an estimated time based on statistical data from the care recipient's "abilities," "methods of assistance," and "presence or absence of (disabilities or conditions)," and is not a sum of the actual time spent on care. The care recipient's "abilities," "methods of assistance," and "presence or absence of (disabilities or conditions)" are based on interviews conducted by visiting investigators, and there are potential problems such as differences in the content of responses from the care recipient or their family, and differences in how the visiting investigator interprets the information, which may include subjective and arbitrary elements.

[0044] Furthermore, given the increasing number of people requiring care, this home visit survey will become a significant burden on surveyors, care providers, and municipalities as the number of care recipients grows. This is likely to further increase the burden on care facilities, which are already facing shortages and a lack of personnel.

[0045] While the current system for assessing the need for long-term care has the aforementioned problems, the invention described in item 6 above allows for the use of "care time," which is the actual time spent on care services, as a basis for assessing the need for long-term care. In other words, it becomes possible to assess the need for long-term care based on "quantitative" information, whereas previously it was judged "qualitatively." This eliminates subjective and arbitrary elements by humans, allowing for a more fair assessment. Moreover, since this "care time" can be automatically and accurately collected, the burden on visiting investigators and other personnel involved in assessing the need for long-term care can be reduced.

[0046] According to the invention shown in R8 above, the content of the care services actually provided and the time required for those services can be automatically and accurately grasped. By utilizing the invention shown in 6 above, the calculation and billing of care fees (the sum of care benefits and user contributions) can be automated or supported.

[0047] The calculation and billing of caregiving fees is a significant burden for caregivers and care providers. Therefore, implementing this invention would significantly reduce this burden by automatically and accurately grasping the content and duration of caregiving and applying them to the corresponding fee categories. Furthermore, it would prevent billing errors, such as failing to bill for caregiving services actually provided. Moreover, since calculations are automated and audio and video data can be reviewed later, it is expected to have a deterrent effect on preventing and suppressing fraudulent billing for both care recipients and insurers (such as municipalities).

[0048] According to the invention described in R9 above, the actual caregiving content and caregiving time can be automatically and accurately grasped. By utilizing the invention described in 6 above, the skill level of the caregiver can be automatically evaluated by analyzing video data of the caregiving process and grasping the caregiving time required for each caregiving task.

[0049] This skill level evaluation is based on automated analysis of video data and automated tracking of care time, providing a fair and objective evaluation system free from arbitrary elements. Furthermore, by accurately understanding the skill level of caregivers, it is possible to fairly evaluate their working conditions, which is expected to improve their sense of fulfillment and motivation, and is beneficial for care providers in securing excellent personnel.

[0050] Furthermore, by linking with the care needs assessment support system shown in R7 above, it becomes possible to make judgments that also take into account the skill level of the caregiver in charge, which can be used to make appropriate judgments in care needs assessment. Specifically, when using the actual care time obtained from the care needs assessment support system shown in 7 above as the standard care needs assessment time, for example, when the same care service is provided, the care time tends to be shorter if the caregiver has a high level of care skills and longer if the caregiver has a low level of care skills. It is possible to perform an evaluation that adjusts and corrects for this difference in the level of care skills among caregivers. [Brief explanation of the drawing]

[0051] [Figure 1] A diagram illustrating one example of an automated care record creation system. [Figure 2] A schematic diagram illustrating one embodiment of a wearable device. [Figure 3] A diagram illustrating one example of an automated care time tracking system. [Figure 4] A diagram illustrating one example of a support system for long-term care certification. [Figure 5] A diagram illustrating one example of a system for supporting the calculation of long-term care service fees. [Figure 6] A diagram illustrating one example of a caregiver skill level evaluation system. [Modes for carrying out the invention]

[0052] The automatic care record creation system 1 according to the present invention (hereinafter also referred to as "automatic care record creation system 1") will be described with reference to the drawings. In this specification, "care recipient Y" refers to an elderly person, person with a disability, etc., who needs care or assistance and receives services, and "caregiver Z" refers to a person who provides care and assistance services to care recipient Y, such as a certified care worker or a home helper, who is engaged in care and assistance work.

[0053] Figure 1 is a diagram illustrating the basic configuration of the automated care record creation system 1. The automated care record creation system 1 is as shown in Figure 1. (1) Means for identifying the person who is the target of care (means for identifying the target person), (2) Means for collecting and recording the care content spoken orally by the caregiver (recording means), (3) Means for photographing or recording the care provided by the caregiver to the person being cared for (recording means), (4) Means for transmitting recorded audio data and recorded video data to a computer (communication means), (5) Means for automatically recognizing the content of the care services actually performed from recorded audio data and recorded video data (recognition means), (6) Means for transcribing automatically recognized care information into text and creating care records (creation means) It has a configuration that includes [this].

[0054] The methods described above (1) to (6) will be explained in order. The means of identifying the target person described in (1) above is a means of identifying the person who will receive care. Since it is necessary to understand the content and duration of care for each person receiving care, it is essential to specifically and accurately identify the target person for care services. In particular, this function is indispensable when using the present invention for the automatic creation of care records, the tracking of care time, support for care needs assessment, and support for calculating care fees, as described later.

[0055] One specific method involves caregiver Z verbally uttering the name of the person being cared for Y, along with an identification tag assigned to each individual person Y, and then automatically analyzing that voice data to identify the person. By utilizing conventionally known technologies such as speech recognition, it is possible to automatically and accurately identify the person being cared for, Y. The identification identifier assigned to each individual care recipient Y consists of numbers, symbols, etc., and is a unique identification identifier assigned to an individual, similar to a patient ID number or medical card number in a hospital, or a student ID number or student number in a school.

[0056] Other methods can also be employed as specific means of identifying the target individuals. For example, a band attached to the wrist of the person being cared for, Y, displays a two-dimensional code or similar, and caregiver Z can read this code to identify the person being cared for during caregiving tasks. Alternatively, methods such as RFID (radio frequency identifier) ​​or Bluetooth can be used to automatically identify the person receiving care, Y.

[0057] Facial recognition systems can also be used as a means of identifying individuals. A facial recognition system is a system that detects human faces from images captured by cameras, converts the information obtained from those faces into numerical data, and then identifies individuals by comparing that facial data with facial data registered in a database. In this invention, the facial data of the person receiving care is registered in a database in advance, and the facial data of the person receiving care is obtained from video footage captured by a wearable device and compared with it to identify the person receiving care (the target person). In this invention, the means for identifying a target person using this facial recognition system is referred to as the facial recognition means.

[0058] There are no specific limitations on the configuration of the facial recognition system used in this invention, and publicly known and publicly available technologies such as visual methods (2D authentication) and IR methods (3D authentication) can be used without any special restrictions.

[0059] Furthermore, a method can be employed to confirm the subject if the name of the care recipient obtained by the facial recognition method described above matches the name of the care recipient spoken verbally by the caregiver (the name spoken when confirming the care recipient's name). This method allows for a double-check of the care recipient's name, effectively preventing misidentification and recording errors.

[0060] The recording means described in (2) above is a means of collecting and recording / storing on a recording medium the voice (terms indicating the content of care, etc.) spoken by caregiver Z. There are no restrictions on the specific method of sound collection and recording, and any publicly known and publicly available recording means can be adopted without any special restrictions.

[0061] The content to be recorded is the voice spoken by caregiver Z, but it is especially important to record the voice describing the care provided. As will be explained later, this is so that the content of the care provided can be automatically understood by analyzing the voice recording of the care provided.

[0062] The caregiving tasks that caregiver Z verbally communicates include, for example, "I will now assist with meals," "I will now assist with bathing," "I have finished assisting with meals," and "I have finished assisting with bathing." By verbally stating the caregiving tasks that are about to be performed or have been performed, the caregiving content can be automatically understood using technologies such as voice recognition. Furthermore, for caregiver Z, verbally confirming the caregiving tasks helps reduce errors, similar to confirmation calls and point-and-call procedures, and also allows for communication with the person being cared for, Y. In addition, for tasks necessary for treatment, such as administering medication or intravenous drips, confirmation of whether medication has been administered, and confirmation of whether the medication should be given to the person being cared for, Y, can be confirmed on-site by reading the information aloud. This can also be confirmed retrospectively by playing back or analyzing the audio data and viewing the caregiving records.

[0063] The recording means described in (3) above is a means of recording video footage of the caregiving process onto a recording medium. There are no restrictions on the specific method of filming and recording, and any publicly known and publicly available recording means can be adopted without any special limitations.

[0064] The content to be recorded is the care services that caregiver Z provides to care recipient Y. The purpose is to analyze the video footage of this caregiving process to automatically understand the content of the care. By analyzing the caregiving process using methods such as video analysis, the content of the care can be understood concretely and automatically, and the time required for it can also be easily determined.

[0065] Next, we will explain how to turn the recording means (2) and the video recording means (3) on and off. Regarding the recording means (2) and video recording means (3) above, in addition to means for continuous recording and video recording, means for starting and ending recording and video recording only when providing care services may be adopted. First, regarding the start of recording and video recording, the wearable device W starts recording and video recording when it finds that the name of the person being cared for, as determined by the facial recognition means described above, matches the name of the person being cared for that caregiver Z has verbally stated (the name spoken when confirming the person's name). The facial recognition system is configured to continuously detect human faces (the faces of care recipient Y) in the video captured by the wearable device W, and identify care recipient Y by comparing them with a database. Furthermore, it is a normal practice for caregiver Z to state the name of care recipient Y at the start of providing care services. In other words, with this configuration, recording and video recording are initiated by the caregiver X uttering (reading aloud, etc.) the name of the person being cared for, which is a normal action for the caregiver X. Therefore, caregiver X does not need to take any action to start recording or video recording, and can maintain a hands-free state. Thus, the recording and video recording function can be started automatically without bothering caregiver X, and caregiver X can maintain a hands-free state.

[0066] In other words, the recording / video recording initiated by the wearable device W will begin when both (a) and (b) below are met. (a) Name of the person receiving care obtained through facial recognition (b) The name of the person receiving care as spoken by the caregiver.

[0067] Next, regarding the termination of recording and video, the wearable device W will terminate recording and video when it detects the name of the person being cared for Y and a keyword indicating the end of care services from among the words spoken by the caregiver X. Examples of keywords that indicate the end of care services include "Thank you for your hard work" and "I'll come again." Regarding these keywords, a method can be employed to train a computer using voice data obtained from a wearable device W, so that it can understand which keywords indicate the end of a predetermined care service. As mentioned above, the recording / video recording will end when the caregiver X detects the name of the person being cared for (Y) and a keyword indicating the end of the care service. For example, the recording / video recording will end when it recognizes phrases such as, "Thank you for your hard work, Mr. / Ms. XX," or "I'll come again, Mr. / Ms. XX."

[0068] To put the above-mentioned termination of recording / video recording into other words, the configuration is such that recording / video recording by the wearable terminal W is terminated when (c) and (d) below are issued consecutively. (c) Keywords indicating the termination of care services (d) The name of the person receiving care as spoken by the caregiver. Furthermore, there is no restriction on the order in which (c) and (d) above are issued; they may be issued in the order of (c) to (d), or in the order of (d) to (c). There may be a gap of a few seconds between (d), and other statements may be made during this gap of a few seconds. This gap of a few seconds is preferably 10 seconds or less, and more preferably 5 seconds or less.

[0069] The communication function described in (4) above is a means of transmitting recorded audio data and recorded video data to an external computer. External computers refer to, for example, computers or servers on the cloud. To transmit audio and video data to these cloud computers, it is sufficient to use wireless or wired networks to send and receive data. There are no restrictions on the specific methods of communication; publicly known and publicly available communication methods can be adopted without any special limitations.

[0070] For example, a wireless communication system used in mobile phones is preferred, and it is especially preferable to use the fifth-generation mobile communication system (commonly known as "5G"). With a communication method using 5G, large amounts of data can be sent and received at high speed, and furthermore, the communication method can be used even in environments where care is provided but where communication equipment such as wireless or wired LAN is not available. If it is difficult to use mobile phone communication in remote areas or areas with radio wave blockage, the audio and video data recorded and stored on the wearable device W may be transferred to the computer C using a recording medium M such as a USB flash drive.

[0071] As the device equipped with the above-mentioned (2) recording means, (3) video recording means, and (4) communication means, it is preferable to use a wearable terminal W worn by caregiver Z, as shown in Figure 2. The same applies when the (1) subject identification means employs a method in which caregiver Z verbally pronounces the name of the person being cared for Y or an identification mark assigned to each person being cared for Y, and the voice data is automatically analyzed to identify the person.

[0072] A wearable device W is an information terminal worn by a person. Common types include wristbands and watches worn on the wrist, armbands worn on the arm, and glasses, headbands, or headsets worn on the head and face. There are no specific limitations on the configuration of the wearable device W used in this invention, but it is required to have at least a microphone function W1, a camera function W2, and a communication function (not shown), and to be capable of recording, video recording, and communication (sending and receiving data).

[0073] As the wearable terminal W used in this invention, from the viewpoint that it is necessary to record the voice of the caregiver and to film the caregiving process (in particular, to film the person being cared for Y from the perspective of the caregiver Z), a glasses-type, headband-type, or headset-type device that is worn on the head and face is preferred. It is especially preferred to use smart glasses, which are widely used as devices equipped with the above-mentioned (2) recording means, (3) video recording means, and (4) communication means.

[0074] Currently, as a measure against infectious diseases such as COVID-19, caregivers Z may wear face shields in nursing care settings. Therefore, it is particularly preferable that a face shield W3 is attached to or can be attached to a wearable terminal W that is worn on the head and face, such as glasses, a headband, or a headset (see Figure 2(b)). Previously, caregivers Z tended to find it bothersome to have any kind of device attached to their body, but due to the fact that wearing face shields has become mandatory in some nursing care settings due to the spread of COVID-19, their resistance to wearing a wearable terminal along with a face shield has decreased. It would be convenient if caregivers Z could take this opportunity to wear the wearable terminal W and adopt the system according to the present invention.

[0075] Figure 2 shows an example of a wearable device W. Figure 2(a) shows a configuration in which a microphone function W1 and a camera function W2 are attached to the side of the caregiver Z's face. This configuration is generally referred to as smart glasses. Figure 2(b) shows a configuration in which the microphone function W1 is attached to the side of the face and the camera function W2 is attached to the forehead. In either configuration, it is preferable that the microphone function W1 be located close to the caregiver Z's mouth, and the camera function W2 be located close to the caregiver Z's eyes. Furthermore, in the embodiment shown in Figure 2(b), a face shield W3 is provided in a manner that covers the caregiver Z's face.

[0076] The recognition means described in (5) above is a means that automatically analyzes the audio data recorded by the recording means described in (2) above and the video data recorded by the video recording means described in (3) above, and automatically recognizes the care provided. This recognition mechanism is performed by the computer C.

[0077] Analysis from voice data is a method of analyzing the content of caregiving tasks verbally spoken by caregiver Z using technologies such as speech recognition. Technologies such as speech recognition are already used as functions in smartphones and smart speakers, and by adapting these technologies for caregiving, they can be used as this analysis method. For example, if caregiver Z says, "I will now assist with feeding," this voice data is analyzed using speech recognition technology to recognize that the caregiving task being performed is "feeding assistance."

[0078] If different caregivers use different terminology to describe a particular care service, the care provided may not be correctly recognized. Therefore, if caregivers agree to use specific terminology (specific terms) to describe the content of care services, and if computer C also stores this specific terminology, the content of the care service can be automatically and correctly recognized. Therefore, by pre-determining terms (specific terms) related to the content of care services used by caregivers, and by having computer C detect these specific terms, the content of the care services provided can be accurately understood.

[0079] Some care and assistance facilities have established specific terminology to describe certain types of care and instruct caregivers to use this terminology. If the terminology used in each care and assistance facility is stored in computer C in advance as specific terminology, it will be possible to recognize the specific terminology spoken by caregiver Z and accurately understand the type of care being provided.

[0080] Analysis from video data is a method of analyzing the caregiving services performed by caregiver Z for care recipient Y using technologies such as video recognition and video analysis. These technologies, known as video recognition and video analysis, are widely used in various fields such as security cameras and medical diagnosis, and can be adapted for caregiving purposes and used in this analysis. For example, a computer can be trained on a large amount of video data of caregiving services, and then the content of caregiving can be analyzed from video data of caregiving tasks based on this accumulated data. It is preferable to use technologies such as deep learning and artificial intelligence (AI) for such training and analysis.

[0081] Furthermore, regarding the recognition means described in (5) above, a configuration can be adopted in which the video data of the caregiving situation, which includes images of a person (mainly the person being cared for, or both the person being cared for, Y, and the caregiver, Z), is converted into a stick figure, and the content of the caregiving is automatically recognized from the movements of this converted stick figure. A stick figure is a representational technique that drastically deforms the human body, depicting the head as a circle and the limbs as sticks. For example, it is used in the field of computer graphics (CG) to convert the movements of real people into stick figures and generate new computer graphics based on these figures, and in the field of sports science to convert the movements of athletes into stick figures and scientifically analyze them based on these figures. Stick figures are used in a variety of fields.

[0082] Converting people in video data into stick figures allows for accurate tracking of their movements in computer C, and also helps protect the privacy of those being filmed.

[0083] Furthermore, regarding the recognition means described in (5) above, a configuration can be adopted that uses musculoskeletal analysis technology to analyze the video data of people (especially those receiving care) included in the video data of the caregiving process, and automatically recognize the content of the care. Musculoskeletal analysis is a technique that uses a human body model (musculoskeletal model) composed of muscles and bones to reproduce human movements and calculates muscle activity and the forces generated in each part of the model during those movements. By using this technique, it is possible to estimate the load on a person's muscles and joints from video footage and perform motion analysis, thereby accurately understanding the condition and movements of the person receiving care and accurately determining what kind of care is being provided.

[0084] The recognition means described in (5) above can also be configured to analyze either the audio data recorded by the recording means described in (2) above or the video data recorded by the video recording means described in (3) above. For example, only the audio data can be analyzed and the analysis data output can be used for processing in the judgment means and summation means described later. In this case, the video data can be used as verification material when any trouble occurs or when doubts arise about the results of understanding the care content or care time.

[0085] On the other hand, in order to improve the accuracy of automatic recognition of care services, it is preferable to automatically recognize the actual care services provided from both audio and video. If care services are automatically recognized from either audio or video, errors in audio recognition or image / video analysis are possible. However, since the system automatically recognizes care services from both audio and video, the accuracy of automatic recognition of care services can be improved. Furthermore, by transcribing this accurately recognized care service content into text, accurate care records can be automatically created.

[0086] Furthermore, it is preferable to have a means to compare the caregiving content automatically recognized from video data with the caregiving content automatically recognized from audio data to confirm whether there are any discrepancies in the caregiving content performed at the same time. If a discrepancy occurs in the automatically recognized caregiving content, the scene can be reviewed retrospectively using video and audio, thereby enabling the creation of more accurate caregiving records.

[0087] The creation method described in (6) above is a method for creating a care record by transcribing the care content automatically recognized from audio data and video data into text.

[0088] Care records are daily records created for each individual receiving care, detailing the care services provided and stored as paper or electronic data. Specifically, they include information such as the details of the care services provided by the caregiver, the location and date of care, vital data such as blood pressure and temperature, and the care recipient's condition.

[0089] As described in (1) to (5) above, the care recipient and the care provided can be identified, and this information is then transcribed and recorded as text. Video data and images extracted from this video data may also be recorded and posted. The location where care was provided can be automatically recognized from audio data spoken by the caregiver, or, if the wearable device W worn by the caregiver has a built-in location-determining system such as GPS, this can be used. The date and time of care provided are generally included in the audio and video data, and can be identified using these. Vital data such as blood pressure and body temperature can be automatically recognized from audio data recorded when the caregiver reads out the measurement or temperature reading results, or the results can be automatically recognized and recorded from video data recorded by taking a picture of the screen displaying the measurement or temperature reading results.

[0090] The means for creating the above (6) is preferably configured to be linked with the Scientific Care Information System (LIFE). The Scientific Care Information System (LIFE) is a system created and operated by the Ministry of Health, Labour and Welfare with the aim of promoting the provision of high-quality services that contribute to independent living support and prevention of worsening conditions, with scientifically supported effectiveness. It also aims to promote the PDCA cycle using LIFE and initiatives to improve the quality of services.

[0091] In short, LIFE is an initiative to centrally collect caregiving data from all over Japan. Specifically, each caregiver and care provider will regularly provide data they collect as the basis for their caregiving practices to LIFE, a centralized national management system. LIFE will then analyze this information (big data), provide analytical information to caregiving facilities, and is expected to play a role in guiding the direction in which welfare in Japan should progress. Therefore, each caregiver and care provider is required to input and submit daily care records into LIFE, and this input work is a burden on those working on the front lines.

[0092] The creation of care records and the input of data into the Scientific Care Information System (LIFE) are generally performed by caregivers after they have finished providing care services to the person they are caring for, which places a considerable burden on them. If the creation of care records and the input of data into the Scientific Care Information System (LIFE) are automated using the creation method described in (6) above, it will be possible to reduce the workload on caregivers.

[0093] Each of the above methods (5) to (6) is processed by a computer (including any form of electronic computer such as a personal computer, general-purpose server, mobile information terminal, tablet terminal, or smartphone).

[0094] On the other hand, as mentioned above, it is preferable to use a wearable terminal W for each of the means (1) to (4) described above. The subject data, voice data and / or video data obtained by each of the means (1) to (3) described above are transferred to the computer C or mobile information terminal S used by each of the means (5) to (6) described above via the communication means (4) described above.

[0095] There are no limitations on the means of data migration, but examples include wired connections such as USB (Universal Serial Bus) and wired LAN (Ethernet), wireless connections such as wireless LAN and mobile phone communication networks (Network W), and storage media such as USB flash drives (Storage Medium M). Furthermore, for example, it is preferable to use a wireless communication system used in mobile phones, and in particular, to use the fifth-generation mobile communication system (commonly known as "5G"). With a communication method using 5G, large amounts of data can be sent and received at high speed, and the communication method can be used even in environments where care is provided but where communication equipment such as wireless or wired LAN is not available.

[0096] The automated care record creation system 1 consists of the above-mentioned means (1) to (6), and is characterized in that all means are performed automatically without requiring any operation by caregivers or other persons. The above-mentioned means (1) to (6) may be processed in any order.

[0097] <Automatic system for tracking caregiving time> As an example of an application using the automated care record creation system 1, we will explain the automated care time tracking system 2. Caregiving time refers to the actual time spent providing care services. Specifically, it is the time calculated from the start time and end time of each individual care service.

[0098] The specific configuration of the automated care time tracking system 2 is as follows: Figure 3 is a diagram illustrating the basic configuration of the automated care time tracking system 2. Of these, methods (1) to (6) are the same as those used in the automated care record creation system 1 shown in Figure 1.

[0099] The automated care time tracking system 2 is as shown in Figure 3. (1) Means for identifying the person who is the target of care (means for identifying the target person), (2) Means for collecting and recording the care content spoken orally by the caregiver (recording means), (3) Means for photographing or recording the care provided by the caregiver to the person being cared for (recording means), (4) Means for transmitting recorded audio data and recorded video data to a computer (communication means), (5) Means for automatically recognizing the content of the care services actually performed from recorded audio data and recorded video data (recognition means), (6) Means for transcribing automatically recognized care information into text and creating care records (creation means) (7) A means (determination means) for automatically determining from the analysis data whether the care content should be included in the care time calculation, (8) Means for calculating the amount of care time provided to a specific person receiving care (calculation means), It has a configuration that includes [this].

[0100] The judgment means described in (7) above is a means for automatically determining whether the content of the care work (care content) analyzed from the analysis data obtained by the recognition means described in (5) above, namely the audio data and video data, is content that should be accumulated as care time. For example, if the care activity analyzed as "meal assistance" should be counted as part of the care time, the answer is "YES". Conversely, if the care activity analyzed as "transfer (moving from bed to wheelchair, from wheelchair to dining chair, etc.)" should not be counted as part of the care time, the answer is "NO".

[0101] It is preferable that the criteria for determining whether or not a period of time should be counted as care time be arbitrarily set. For example, if understanding the content and duration of care is necessary as a basis for determining the need for long-term care, then the criteria should be set to match the requirements for long-term care certification. Similarly, if understanding the content and duration of care is necessary for calculating and billing care fees, then the criteria should be set to match the billing conditions for care fees (more details will be provided later).

[0102] The calculation means described in (8) above is the process of automatically calculating the time based on the judgment result in (7) above, which determines whether or not the content should be calculated as care time. The time required for each care item may be calculated separately, or the total care time for the entire day's care services may be calculated.

[0103] <Support system for long-term care certification> As an application example of the automated care record creation system 1, we will describe the care needs assessment support system 3. According to the Ministry of Health, Labour and Welfare, "Under the long-term care insurance system, if a person becomes bedridden or suffers from dementia and requires constant care (a state requiring long-term care), or if they require assistance with daily activities such as housework and personal grooming, and are in a state where preventive care services are particularly effective (a state requiring support), they can receive long-term care services. Long-term care certification (including support certification; the same applies hereinafter) determines whether a person is in a state requiring long-term care or support, and to what extent, and is determined by a long-term care certification review committee established in the municipality, which is the insurer. Since long-term care certification is linked to the amount of long-term care services provided, the standards are set objectively and uniformly nationwide."

[0104] Furthermore, the process for determining the need for long-term care is explained as follows: "A physical and mental condition survey (certification survey) is conducted by a municipal certification surveyor (which can be outsourced to a designated home care support service provider, etc.) and a computer assessment (primary assessment) is made based on the attending physician's opinion. Next, a long-term care certification review committee, composed of experts in health, medical care, and welfare, conducts a review and assessment (secondary assessment) based on the results of the primary assessment, the attending physician's opinion, etc."

[0105] In the initial assessment described above, the level of care required (Care Level 1-2 or Care Level 1-5) is determined based on the care time (Care Level Standard Time) for the five care categories defined as direct daily living assistance, joint daily living assistance, problem behavior-related activities, functional training-related activities, or medical-related activities. For example, if the care time for the above five categories is 32 minutes or more but less than 50 minutes, the person will be judged as "Care Level 1". The Care Level Standard Time used here is a measure of the effort required for care, and is an estimated time based on statistical data from the care recipient's "abilities," "method of assistance," and "presence or absence (of disabilities or phenomena)," and does not reflect the actual time spent on care.

[0106] If the estimated standard time for long-term care assessment can be replaced with the actual time spent providing care (care time), it will be possible to base long-term care assessments, which were previously judged qualitatively, on quantitative information. By operating a system that enables this quantitative judgment, subjective and arbitrary judgment elements by assessment investigators can be eliminated, resulting in a more fair assessment. Moreover, since this "care time" can be automatically and accurately grasped, the burden on assessment investigators in long-term care assessment can be reduced.

[0107] The specific configuration of the support system 3 for long-term care certification is as follows: Figure 4 is a diagram illustrating the basic structure of the support system 3 for long-term care certification. Of these, methods (1) to (6) are the same as those used in the automated care record creation system 1 shown in Figure 1.

[0108] The long-term care certification support system 3 is a system that utilizes the aforementioned automatic care record creation system 1 and automatic care time tracking system 2. In other words, it is a system that aims to support long-term care certification using the information on care content and care time obtained from these systems.

[0109] In the support system 3 for long-term care certification, the judgment means described in (7) above in the automatic care time tracking system 2 is specifically defined as follows: First, the determination of whether or not a care service should be included in the calculation of care time is based on whether or not each care service falls under one of the five care classifications used in the initial assessment of care needs: direct assistance with daily living, assistance with joints, behavioral problems, functional training, or medical treatment. Next, in the calculation method described in (8) above, the care time is calculated for each of the five care services that fall under one of the categories. This calculated care time is then used as the standard care time for the initial assessment of the care needs assessment.

[0110] <Support system for calculating long-term care service fees> As an example of an application of the automated care record creation system 1, we will describe the care fee calculation support system 4. Figure 5 is a diagram illustrating the basic configuration of the long-term care fee calculation support system 4. Of these, methods (1) to (6) are the same as those used in the automated care record creation system 1 shown in Figure 1.

[0111] Care service fees refer to the remuneration paid to care service providers as compensation for providing care services to users (those requiring care or support). Care service fees are determined for each type of care service, taking into account the average cost based on the service content, the level of care required, the location of the service provider or facility, etc. The calculation and billing of long-term care fees requires matching the content of the care services provided to each care recipient to predetermined points (units) for long-term care fees, a task that places a considerable burden on caregivers and care providers. The Long-Term Care Fee Calculation Support System 4 is a system that automatically grasps the content of the care services actually provided and automatically applies it to the points (units).

[0112] Caregiving fees are determined by the content of the caregiving service (caregiving content) and the time spent providing the caregiving service (caregiving time). Therefore, it is necessary to accurately grasp the actual caregiving content and time spent providing care. The content and duration of care provided can be determined by the automated care content and duration tracking system 1 described above. In the (7) judgment means of this automated care duration tracking system 2, the criterion for judgment is whether or not the care service is eligible for claim as care service fees. Furthermore, if the care service is eligible for claim as care service fees, the system is configured to apply the applicable units.

[0113] If the units are known, the unit price can be determined based on factors such as regional classification and labor cost ratio, thus enabling the determination of care service fees. The care service fee calculation support system 4 is configured to automatically perform the process of determining these units, calculating unit prices, and calculating care service fees.

[0114] <Evaluation system for caregiving skills> As an example of an application of the automated care record creation system 1, we will describe the caregiver skill level evaluation system 5. Figure 6 is a diagram illustrating the basic structure of the caregiver skill level evaluation system 5. Of these, methods (1) to (6) are the same as those used in the automated care record creation system 1 shown in Figure 1.

[0115] The skill level of a caregiver is a numerical representation or rank that evaluates and quantifies their professional abilities as a caregiver. In other words, it is an evaluation of the skills necessary to perform caregiving tasks such as care services and caregiving work, expressed as a numerical value or rank. The Caregiver Skill Level Evaluation System 5 is a system that automatically evaluates or supports this evaluation of a caregiver's skill level. By properly evaluating the skill levels of caregivers, it is possible to foster and retain personnel in caregiving settings.

[0116] Specifically, skill levels are evaluated through the following process: First, video data is obtained by using the recording means in the automated care record creation system 1 to capture the caregiver performing caregiving tasks.

[0117] Next, the captured video data is compared against a database of caregiving skills. This database is a collection of information obtained by using deep learning on a computer to analyze numerous videos of caregiving tasks. It is preferable to organize good and bad examples for each type of caregiving task.

[0118] As a means of matching, a computer equipped with artificial intelligence is used to compare video data of caregiving tasks with a database of caregiving skills to determine the skill level of the caregiving tasks. For example, in the case of meal assistance, it is preferable to judge from multiple perspectives, such as whether the movements and speed of bringing food to the care recipient's mouth are appropriate, whether the care recipient is chewing correctly, and whether the care recipient is eating in the correct posture. Similarly, in the case of walking assistance, it is preferable to judge from multiple perspectives, such as whether the assistance is being provided in accordance with the care recipient's condition, whether there is a risk of injuring the care recipient, and whether the caregiver is paying attention to the surrounding situation.

[0119] After assessing the caregiving skills using the method described above, the skill level is determined based on the assessment results, conforming to pre-set levels and ranks. The skill level determined here can be arbitrarily set by each caregiving provider.

[0120] This invention can be applied in nursing settings, specifically to provide an automated system for creating nursing records. The problem that this invention aims to solve is to provide an automated nursing record creation system that can automatically and accurately grasp the specific content of nursing care provided in a nursing setting, automatically and accurately identify the person being cared for (patient), and automatically grasp the start and end times of nursing work, and automatically create nursing records. Furthermore, using this automated nursing record creation system, the invention aims to provide an automated nursing time tracking system, a nursing fee calculation support system, and a nurse skill level evaluation system.

[0121] In this case, "caregiver" in the explanation of the automated care record creation system refers to "nurse," and "person receiving care" refers to "patient."

[0122] The configuration of the automated nursing record creation system is the same as that of the automated care record creation system, with nurses wearing wearable devices while providing care. While there are differences in format and required information between care records and nursing records, requiring adjustments as needed, the system can be adapted to the format and required information of publicly available nursing records.

[0123] In the automated nursing record creation system, the wearable device is configured to start recording when the patient's name, as determined by facial recognition, matches the patient's name spoken verbally by the nurse. Furthermore, the wearable device is configured to stop recording when it detects the patient's name and a keyword indicating the end of nursing services from the nurse's spoken words.

[0124] With these configurations, recording can be started and stopped automatically without requiring any intervention from the nurse, and the nurse can remain hands-free. Furthermore, patient identification can be double-checked using both facial recognition and verbal confirmation by nurses, enabling automatic and accurate patient identification. This prevents patient mix-ups and errors such as incorrect entries in other patients' nursing records.

[0125] As an example of applying the above configuration, the following configuration can be given. In cases where a nurse administers medication or intravenous fluids to a patient, the wearable device uses image / video recognition to identify the name of the medication or intravenous fluid and verifies whether it matches the name read aloud by the nurse. If there is a discrepancy, an alert or other warning can be added. Furthermore, the wearable device may employ a method to identify the names of medications and intravenous infusions using image and video recognition functions, and then compare them with the patient's nursing records, medical charts, or prescriptions to confirm that there are no errors in the prepared medications or intravenous infusions.

[0126] This configuration allows for a double-check of the contents of medications and intravenous infusions, thus preventing errors in administering the wrong medication or infusion to a patient. [Explanation of symbols]

[0127] 1. Automatic care record creation system 2. Automatic system for tracking caregiving time 3. Support system for long-term care certification 4. Support system for calculating long-term care service fees 5. Caregiver Skill Level Evaluation System W Wearable Device W1 Microphone function W2 Camera Function W3 Face Shield C Computer S Mobile Information Terminal M storage medium N Network (Wired / Wireless) Y Care recipient Z Caregiver

Claims

1. A system that automatically recognizes the content of nursing services provided and automatically creates nursing records based on that information, A wearable terminal equipped with means for photographing and recording the nursing care provided by a nurse, means for collecting and recording the words spoken by the nurse, and communication means for transmitting the recorded video data and the recorded audio data to an external source, The system consists of a computer equipped with means for receiving the aforementioned video data and audio data, and for analyzing these data to automatically recognize nursing content. The wearable device is configured to start recording when the name of the person being cared for, as determined by facial recognition, matches the name of the person being cared for, as verbally stated by the caregiver. The aforementioned computer, A means for automatically recognizing nursing care content from the video of the patient contained in the aforementioned video data, The system includes means for automatically recognizing nursing care content from words spoken by a nurse included in the aforementioned audio data. Furthermore, the system for automatically creating nursing records is characterized by having means for transcribing nursing content automatically recognized from this video data and audio data into text to create nursing records.

2. The automatic nursing record creation system according to claim 1, characterized in that the wearable device is configured to terminate recording when it detects the name of the person being cared for and a keyword indicating the end of nursing services from among the words spoken by the nurse.

3. The automated nursing record creation system according to claim 1, characterized in that it records video data and audio data along with text indicating nursing content.

4. The automatic nursing record creation system according to claim 1, characterized in that it includes means for comparing nursing content automatically recognized from video data with nursing content automatically recognized from audio data to confirm whether there are any discrepancies in the nursing content performed at the same time.

5. A system utilizing the automated nursing record creation system described in claims 1 to 4, A means to automatically determine whether or not nursing care content should be included in nursing time calculations, An automated nursing time tracking system characterized by having means for accumulating nursing time provided to individual patients.

6. A system utilizing the automatic nursing time tracking system described in claim 5, A nursing skill level evaluation system characterized by evaluating the skill level of nurses based on skill scores obtained from video analysis of video data and the nursing time required for each nursing procedure.

7. In cases where a nurse administers medication, drugs, or intravenous fluids to a patient, A wearable device is equipped with an image / video recognition function to identify the name of a drug or intravenous drip and to verify whether it matches the name of the drug or intravenous drip read aloud by the nurse. An automated nursing record creation system according to any one of claims 1 to 4, characterized in that it includes a configuration to issue a warning if there is a discrepancy.

8. In cases where a nurse administers medication, drugs, or intravenous fluids to a patient, A wearable device is equipped with image / video recognition capabilities to identify the name of a drug or intravenous infusion, compare it with the patient's nursing records, medical chart, or prescription, and verify that the prepared drug or intravenous infusion is correct. An automated nursing record creation system according to any one of claims 1 to 4, characterized in that it includes a configuration to issue a warning if there is a discrepancy.

Images