Bedside care method and system

Abstract

The present application relates to a kind of bed care method and care system, bed care method is applied to bed care system, bed care system is equipped with terminal and is arranged in the non-contact sensing component of terminal;Care method includes controlling non-contact sensing component real-time acquisition bed of the person being cared for Multidimensional vital sign signal;Based on multidimensional vital sign signal extraction each dimension state feature data;The state feature data of each dimension is compared respectively with the preset range of corresponding dimension, obtains the state type of each dimension;The state type of each dimension obtained is data fusion, generates multidimensional state description data;Multidimensional state description data includes the state type of each dimension and the duration corresponding to the state type of each dimension;Multidimensional state description data is transmitted to terminal, generates state presentation interface.The bed care method provided in the present application provides continuous, multidimensional vital sign state comprehensive perception and information presentation for bed of the person being cared for.

Description

Technical Field

[0001] This invention relates to the field of medical and health monitoring technology, and more specifically, to a method and system for bedridden care. Background Technology

[0002] In bedridden care scenarios, continuous monitoring of the vital signs of those being cared for is a core task to ensure their safety and health. The caregivers include special groups such as post-operative recovery patients and elderly people who are bedridden for extended periods. Their daily care relies on caregivers frequently observing key indicators such as physical activity, breathing patterns, and level of consciousness to promptly identify abnormalities and implement human intervention.

[0003] There are several existing methods of bedridden care, as detailed below: For example, manual rounds require caregivers to enter the room at fixed intervals for visual inspection. However, this method is highly dependent on human input and subjective judgment, especially when care resources are scarce and multiple caregivers need to be managed simultaneously. The intervals between rounds inevitably create blind spots in status information, making it impossible to detect emergencies such as respiratory arrest or unexpected body movements in time, thus creating safety hazards. Another example is wearable device monitoring, which requires the caregiver to wear sensors (such as smart bracelets, chest strap respiratory monitors, or finger clip pulse oximeters) to collect parameters such as heart rate, blood oxygen, and body movement and transmit them to a monitoring terminal. However, wearable devices have several limitations in bedridden care scenarios: long-term wear may cause skin pressure discomfort or allergies; the caregiver may inadvertently remove or move the sensors, causing data interruption; in particular, for caregivers with cognitive impairment, wearable devices are more likely to be unintentionally removed. In summary, existing bedridden care methods have at least the following shortcomings: frequent manual rounds make it impossible to perceive the caregiver's status in a timely manner; and wearable devices have poor comfort and dependency. Summary of the Invention

[0004] This invention provides a bedridden care method to solve the technical problems of existing bedridden care methods, such as frequent manual inspections, inability to promptly perceive the condition of the person being cared for, poor comfort and dependence of wearable devices, and high maintenance costs.

[0005] This invention provides a bedridden care method applied to a bedridden care system, wherein the bedridden care system is equipped with a terminal and a non-contact sensing component, the non-contact sensing component being located on the terminal; the care method includes the following steps: The steps are as follows: Vital Signs Acquisition Step: Controlling the non-contact sensing component to acquire multi-dimensional vital signs signals of the bedridden person in real time; State Feature Extraction Step: Extracting state feature data for each dimension based on the multi-dimensional vital signs signals; State Type Identification Step: Comparing the state feature data for each dimension with a preset range for the corresponding dimension to obtain the state type for each dimension; Data Fusion Step: Fusion of the obtained state types for each dimension to generate multi-dimensional state description data; The multi-dimensional state description data includes the state type for each dimension and the duration corresponding to each state type; Care State Presentation Step: Transmitting the multi-dimensional state description data to the terminal to generate a state presentation interface.

[0006] Optionally, in the step of acquiring vital signs signals, the multidimensional vital signs signals include at least two types: body movement signals, respiratory rhythm signals, voice activity signals, and visible facial state signals; wherein, the body movement signals and / or the respiratory rhythm signals are acquired by at least one of an optical camera, an infrared camera, and a depth sensor; the voice activity signals are acquired by a microphone array; and the visible facial state signals are acquired by an optical camera.

[0007] Optionally, if the multi-dimensional vital signs signal includes body movement signal; the steps of the vital signs signal acquisition step include: controlling the non-contact sensing component to acquire the body movement signal in real time; the steps of the state feature extraction step include: extracting the motion amplitude and motion frequency based on the body movement signal; the steps of the state type identification step include: comparing the motion amplitude with the corresponding first preset amplitude range and the motion frequency with the corresponding first preset frequency range one by one; if the motion amplitude falls within the first preset amplitude range and the motion frequency falls within the first preset frequency range, then the state type of the body movement signal is normal activity; if the motion amplitude is higher than the first preset amplitude range and the motion frequency is higher than the first preset frequency range, then the state type of the body movement signal is frequent activity; if the motion amplitude is lower than the first preset amplitude range and the motion frequency is lower than the first preset frequency range, determine whether the duration of the body movement exceeds a first preset duration; if yes, then the state type of the body movement signal is long-term stillness; if no, then the state type of the body movement signal is reduced activity.

[0008] Optionally, if the multi-dimensional vital signs signal includes a respiratory rhythm signal; the steps of the vital signs signal acquisition step include: controlling the non-contact sensing component to acquire the respiratory rhythm signal in real time; the steps of the state feature extraction step include: extracting the respiratory cycle and respiratory amplitude based on the respiratory rhythm signal; the steps of the state type identification step include: comparing the respiratory cycle with the corresponding preset range of single respiratory duration and the respiratory amplitude with the corresponding preset range of second amplitude; if the respiratory cycle falls within the preset range of single respiratory duration and the respiratory amplitude falls within the preset range of second amplitude, then the state type of the respiratory rhythm signal is normal rhythm; if the respiratory cycle is greater than the preset range of single respiratory duration and the respiratory amplitude is greater than the preset range of second amplitude, then the state type of the respiratory rhythm signal is rapid breathing; if the respiratory cycle is less than the preset range of single respiratory duration and the respiratory amplitude is less than the preset range of second amplitude, then the state type of the respiratory rhythm signal is slow breathing.

[0009] Optionally, if the multi-dimensional vital signs signal includes a voice activity signal; the steps of the vital signs signal acquisition step include: controlling the non-contact sensing component to acquire the voice activity signal in real time; the steps of the state feature extraction step include: extracting the phonation frequency and loudness based on the voice activity signal; the steps of the state type recognition step include: comparing the phonation frequency with the corresponding preset frequency range and the loudness with the corresponding preset loudness range one by one; if the phonation frequency falls within the preset frequency range and the loudness falls within the preset loudness range, then the state type of the voice activity signal is normal phonation; if the phonation frequency is less than the preset frequency range and the loudness is less than the preset loudness range, determine whether the phonation duration exceeds a second preset duration; if yes, then the state type of the voice activity signal is prolonged silence; if no, then the state type of the voice activity signal is reduced phonation.

[0010] Optionally, if the multi-dimensional vital signs signal includes a visible facial state signal; the steps of the vital signs signal acquisition step include: controlling the non-contact sensing component to acquire the visible facial state signal in real time; the steps of the state feature extraction step include: extracting the eye state based on the visible facial state signal; the steps of the state type recognition step include: recognizing the opening and closing state of the eyes; if the eyes are open, the state type of the visible facial state signal is "awake and visible"; if the eyes are closed and the closing duration exceeds a third preset duration, the state type of the visible facial state signal is "continuously closed eyes"; if the eyes are closed and the closing duration does not exceed the third preset duration, the state type of the visible facial state signal is "temporarily closed eyes"; if the opening and closing state of the eyes cannot be recognized, the state type of the visible facial state signal is "face obscured".

[0011] Optionally, in the step of presenting the care status, the status presentation interface includes visual prompts for abnormal statuses. The visual prompts include: highlighting the corresponding dimension area on the terminal; and / or generating a prompt card in the abnormal status prompt area, the prompt card including the abnormal status and the dimension name, timestamp, and duration corresponding to the abnormal status. After generating the multidimensional state description data, the care method further includes: an anomaly detection step: comparing the multidimensional state description data of the current time window with the multidimensional state description data of the corresponding previous adjacent time window; if the multidimensional state description data of any dimension changes from a normal state to an abnormal state, then a timer for the corresponding dimension is started; if the state type of the dimension is still abnormal when the timer reaches a preset duration, it is determined to be a continuous abnormality, and then the visual prompt is triggered; if the state type of the dimension changes from an abnormal state to a normal state before the timer reaches the preset duration, then the timer is cleared and no visual prompt is triggered. If any dimension of the multidimensional state description data changes from an abnormal state to a normal state, the visual cue is cleared, and the cue card is marked as restored in the abnormal state cue area.

[0012] Optionally, the visual cues further include: providing multi-dimensional state description data corresponding to the abnormal state during a preset period before the state changes from normal to abnormal, for caregivers to refer to and compare.

[0013] Optionally, in the step of presenting the care status, the status presentation interface includes at least one of a navigation bar, a multi-dimensional real-time status panel, a multi-dimensional time period statistical view, a multi-day status change trend chart, and a periodic status summary. The navigation bar is used for switching between caregivers and filtering time ranges. The multi-dimensional real-time status panel includes at least two of the following: a body movement status panel, a breathing status panel, a speech status panel, and a facial status panel. The body movement status panel displays the current status category and duration corresponding to the body movement signal; the breathing status panel displays the current status category and duration corresponding to the breathing rhythm signal; the speech status panel displays the current status category and duration corresponding to the speech activity signal; and the facial status panel displays the visible facial status signal and its duration. The multi-dimensional time-period statistical view includes a body movement status type distribution map, a breathing status type time series map, and a speech status type heatmap. The status summary is generated by aggregating the multi-dimensional status description data of all time windows of each day at preset periodic nodes and generating a periodic status summary according to a preset template.

[0014] The bedridden care method provided by this invention has at least the following beneficial technical effects: This method collects multi-dimensional vital signs signals in real time using a non-contact approach, extracts state features, identifies state types, and fuses the data. Finally, it presents the state description data of the person being cared for on the terminal. Compared with existing technologies such as manual inspections and wearable device monitoring, this method eliminates the need for frequent manual inspections, reduces or even eliminates the inability to perceive the state of bedridden individuals in a timely manner, avoids the constraints and discomfort of wearable devices, and provides continuous, multi-dimensional comprehensive perception and information presentation of the vital signs of bedridden individuals. In addition, by placing the non-contact sensing components on the terminal, the solution can adapt to the layout of different care rooms without requiring bed modifications, and its deployment location can be flexibly adjusted according to changes in care needs.

[0015] The present invention also provides a bedridden care system, which applies the bedridden care method described above, the system comprising: The vital signs acquisition module is used to collect multi-dimensional vital signs signals of bedridden caregivers in real time. The state feature extraction module is used to extract state feature data of each dimension based on the multi-dimensional vital sign signal; The state type identification module is used to compare the state feature data of each dimension with the preset range of the corresponding dimension to obtain the state type of each dimension; Data fusion module: used to fuse the obtained state types from each dimension to generate multi-dimensional state description data; and, Care status presentation module: used to transmit the multi-dimensional status description data to the terminal and generate a status presentation interface.

[0016] The bed rest care system provided by this invention, when applied to the above-mentioned bed rest care method, has all the advantages of the bed rest care method, which will not be elaborated here. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a bedridden care system provided in an embodiment of the present invention; Figure 2 A schematic flowchart of a bedridden care method provided in an embodiment of the present invention; Figure 3 A flowchart illustrating the body movement signal dimension in a bedridden care method provided in an embodiment of the present invention; Figure 4 A schematic diagram of the respiratory rhythm signal dimension in a bedridden care method provided in an embodiment of the present invention; Figure 5 A flowchart illustrating the voice activity signal dimension in a bedridden care method provided in an embodiment of the present invention; Figure 6 A flowchart illustrating the visible facial state signal dimension in a bedridden care method provided by an embodiment of the present invention; Figure 7 This is a flowchart illustrating the abnormality detection and triggering of visual cues in a bedridden care method provided by an embodiment of the present invention. Figure 8 This is a structural block diagram of a bedridden care system provided in an embodiment of the present invention.

[0018] Explanation of reference numerals in the attached figures: 01. Base; 02. Robotic arm; 03. Terminal; 04. Non-contact sensing component; 10. Vital sign signal acquisition module; 20. Status feature extraction module; 30. Status type recognition module; 40. Data fusion module; 50. Care status presentation module. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the following description is provided in conjunction with the accompanying drawings. Figures 1-8 Specific embodiments of the present invention will be described in detail below.

[0020] This invention provides a bedridden care method, applied to a bedridden care system, see attached document. Figure 1 and Figure 2 The bedridden care system includes a terminal 03 and a non-contact sensing component 04. The non-contact sensing component 04 is located on the terminal 03 and is communicatively connected to it. The terminal 03 is a device for display, interaction, and control, and can be a tablet computer, monitor, or smartphone. The non-contact sensing component 04 is a sensor that can collect the vital signs of the bedridden person without direct contact with their body. The non-contact sensor can collect vital signs through optical, acoustic, or other non-contact methods. The bedridden care system also includes a base 01 and a multi-degree-of-freedom robotic arm 02. The multi-degree-of-freedom robotic arm 02 is located on the base 01 and is used to adjust the position and posture of the terminal 03. The terminal 03 is located at the end of the multi-degree-of-freedom robotic arm 02. See appendix. Figure 2 The care method includes the following steps: S100, Vital Signs Acquisition Step: Control the non-contact sensing component to acquire multi-dimensional vital signs signals of the bedridden person in real time; for example, the non-contact sensing component can be a camera to acquire video images of the bedridden person; or, for example, the non-contact sensing component can be a microphone array to acquire sound signals of the bedridden person; in this step, the multi-dimensional vital signs signals include at least one of body movement signals, respiratory rhythm signals, voice activity signals, and visible facial state signals; S110, State Feature Extraction Step: Extract state feature data of each dimension based on multi-dimensional vital signs signals; for example, extract motion amplitude and motion frequency based on body movement signals; and for another example, extract respiratory cycle and respiratory amplitude based on respiratory rhythm signals. S120, State type identification step: Compare the state feature data of each dimension with the preset range of the corresponding dimension to obtain the state type of each dimension; for example, the state type of body movement signal may include normal activity, frequent activity, long-term stillness or reduced activity; for another example, the state type of respiratory rhythm signal may include normal rhythm, rapid breathing or slow breathing. S130, Data Fusion Step: The obtained state types of each dimension are fused to generate multi-dimensional state description data; the multi-dimensional state description data includes the state type of each dimension and the duration corresponding to each state type; for example, the multi-dimensional state description data includes normal activity and the duration corresponding to normal activity; another example is that the multi-dimensional state description data includes shortness of breath and the duration corresponding to shortness of breath. S140, Care Status Presentation Step: Transmit multi-dimensional status description data to the terminal to generate a status presentation interface.

[0021] The bedridden care method provided in this invention collects multi-dimensional vital signs signals in real time through a non-contact method, extracts state features, identifies state types, and fuses data from the collected signals, ultimately presenting the state description data of the person being cared for on the terminal. Compared with manual inspections and wearable device monitoring in the prior art, this method eliminates the need for frequent manual inspections, reduces or even eliminates the inability to perceive the state of the bedridden person in a timely manner, avoids the constraints and discomfort of wearable devices, and provides the bedridden person with continuous, multi-dimensional comprehensive perception and information presentation of vital signs. In addition, by placing the non-contact sensing components on the terminal, the solution can adapt to the layout of different care rooms without bed modifications, and the deployment location can be flexibly adjusted according to changes in care needs.

[0022] In this embodiment of the invention, in step S100, the multi-dimensional vital sign signals include at least two types: body movement signals, respiratory rhythm signals, voice activity signals, and visible facial state signals. It should be noted that the body movement signals reflect the frequency and amplitude of the caregiver's limb movements; the respiratory rhythm signals reflect the rhythmic pattern of the caregiver's chest and abdomen movements; the voice activity signals reflect the frequency, loudness, and temporal distribution of the sounds emitted by the caregiver; and the visible facial state signals reflect characteristics such as the caregiver's eye condition. These multi-dimensional vital sign signals collectively constitute a comprehensive perception of the caregiver's overall condition. Body movement signals and / or respiratory rhythm signals can be acquired using at least one of an optical camera, an infrared camera, and a depth sensor; voice activity signals can be acquired using a microphone array; and visible facial state signals can be acquired using an optical camera. This setup has two advantages. First, by introducing at least two of the following multi-dimensional vital sign signals—body movement signals, respiratory rhythm signals, voice activity signals, and visible facial state signals—it enables the monitoring of physiological activities, vital signs, and voice and facial states across multiple dimensions. This provides a rich and reliable raw data foundation for subsequent state feature extraction, state type identification, and data fusion, thereby significantly improving the perception and accuracy of the overall state of the person being cared for. It provides caregivers with more comprehensive and timely information support, helping to detect abnormalities early and intervene promptly. Second, by using non-contact sensing components such as optical cameras, infrared cameras, depth sensors, and microphone arrays for data collection, it avoids direct contact and interference with the person being cared for, improving the accuracy and continuity of data collection.

[0023] In this embodiment of the invention, see appendix. Figure 3 If the multidimensional vital signs signal includes body movement signal; S100, the steps of the vital sign signal acquisition step include: S201, controlling the non-contact sensing component to acquire the body movement signals of the bedridden person being cared for in real time; the body movement signals can be acquired by at least one of an optical camera, an infrared camera, or a depth sensor; wherein, the optical camera can obtain the number of moving pixels and the amplitude of movement in the body area of ​​the person being cared for by analyzing the changes in pixels or optical flow information in the video frame; the infrared camera can capture the movement of the heat source; the depth sensor can obtain accurate three-dimensional motion data by measuring distance changes or performing skeletal tracking; S110, the steps of the state feature extraction step include: S202, extracting motion amplitude and motion frequency based on body movement signals; it should be noted that motion amplitude refers to the degree or range of displacement of the body part or the whole body of the person being cared for. Motion amplitude includes the maximum distance of limb swing, the angle of trunk turning, etc., and can be obtained by processing the motion amplitude signal, for example, by statistically analyzing the variance of body position coordinates; motion frequency refers to the number of body movement events or the rate of change of body movement of the person being cared for within a preset time period, for example, the number of times the person turns over or moves their limbs per minute; motion frequency can be extracted by processing the motion frequency signal, for example, by peak detection, period analysis, or Fourier transform; the first preset duration is a time threshold used to distinguish between a brief reduction in activity and a long period of stillness with potential risks. This duration can be configured according to clinical recommendations or care experience, for example, set to 30 minutes, 1 hour, etc., to effectively prevent the occurrence of complications such as pressure sores.

[0024] S120, the steps of the state type identification step include: S203, compare the amplitude of the movement with the corresponding first amplitude preset range and the frequency of the movement with the corresponding first frequency preset range one by one; wherein, the first amplitude preset range and the first frequency preset range are set based on clinical data or experience to reflect normal physiological activities under the care needs of different ages. S204, if the amplitude of the movement falls within the first amplitude preset range and the frequency of the movement falls within the first frequency preset range, then the state type of the body movement signal is normal activity. S205, if the amplitude of the movement is higher than the first amplitude preset range and the frequency of the movement is higher than the first frequency preset range, then the state type of the body movement signal is frequent activity. S206, if the amplitude of movement is lower than the first preset range and the frequency of movement is lower than the first preset range, S207, determine whether the duration of body movement exceeds the first preset duration. The first preset duration is set based on clinical recommendations or care experience; for example, it can be set to 30 minutes or 1 hour. If yes, the state type of the body movement signal is prolonged stillness; otherwise, the state type is reduced activity. It should be noted that normal activity is the normal state, while excessive activity and prolonged stillness are abnormal or suspected abnormal states. This setting, by extracting the amplitude and frequency of movement and comparing them with their respective preset ranges, identifies the state types of normal and frequent activity. Furthermore, the determination of the duration of body movement is compared with the first preset time to identify the temporary state of reduced activity and the abnormal state of prolonged stillness. In other words, through multi-dimensional judgment, the accuracy and precision of body movement state recognition are significantly improved.

[0025] In this embodiment of the invention, see appendix. Figure 4 If multidimensional vital signs include respiratory rhythm signals; S100, the steps of the vital signs signal acquisition step include: S301, controlling the non-contact sensing component to acquire the respiratory rhythm signal of the bedridden person in real time; the respiratory rhythm signal can be acquired by at least one of an optical camera, an infrared camera, or a depth sensor; for example, the optical camera can acquire a video stream of respiratory movements by monitoring the rise and fall of the person's chest and abdomen, and then analyze the respiratory changes; the infrared camera depth sensor can acquire the body surface displacement caused by breathing by sensing changes in body surface temperature, and generate corresponding respiratory waveform data; the depth sensor can capture the body surface displacement caused by breathing by sensing changes in distance, and generate corresponding respiratory waveform data. S110, the steps of the state feature extraction step include: S302, extracting the respiratory cycle and respiratory amplitude based on the respiratory rhythm signal; wherein, the respiratory cycle refers to the time required for a complete inhalation and exhalation process, which can be accurately calculated by analyzing signal processing methods such as the time interval between peaks and troughs of the respiratory waveform; the respiratory amplitude refers to the degree of chest and abdominal undulation or the change in respiratory air flow during a breath, which can be quantified by measuring the peak-to-trough difference of the respiratory waveform, for example, the vertical distance between the peak of inhalation and the trough of exhalation can be used as a quantitative indicator of respiratory amplitude; S120, the steps of the state type identification step include: S303, compare the respiratory cycle with the corresponding preset range of single breath duration and the respiratory amplitude with the corresponding preset range of second amplitude; wherein, the preset range of second amplitude and the preset range of second frequency are set based on clinical data or experience to reflect the normal respiratory cycle and respiratory amplitude under the care needs of different ages. S304, if the respiratory cycle falls within the preset range of single respiratory duration and the respiratory amplitude falls within the preset range of second amplitude, then the state type of the respiratory rhythm signal is normal rhythm. S305, if the respiratory cycle is greater than the preset range of single breath duration and the respiratory amplitude is greater than the preset range of second amplitude, then the state type of the respiratory rhythm signal is rapid breathing. S306: If the respiratory cycle is less than the preset range for single breath duration and the respiratory amplitude is less than the preset range for the second amplitude, then the respiratory rhythm signal status type is slow breathing. This setting, by extracting the respiratory cycle and respiratory amplitude and comparing them with their corresponding preset ranges, accurately identifies different respiratory states such as normal rhythm, rapid breathing, and slow breathing. This improves the precision and accuracy of monitoring the respiratory status of bedridden individuals. Caregivers can promptly detect abnormal respiratory states based on the respiratory-related status types, thereby taking appropriate intervention measures and effectively ensuring the life safety and health of the individuals being cared for.

[0026] In this embodiment of the invention, see appendix. Figure 5 If the multidimensional vital signs signal includes voice activity signal; S100, the steps of the vital signs signal acquisition step include: S401, controlling the non-contact sensing component to acquire the voice activity signal of the bedridden caregiver in real time; the voice activity signal reflects the caregiver's vocalization. S110, the steps of the state feature extraction step include: S402, extracting vocal frequency and loudness based on the speech activity signal; where, vocal frequency refers to the number or frequency of sounds emitted by the caregiver within a certain time window, which can reflect the level of activity in their language communication; loudness refers to the intensity or volume of the sound, which can reflect the energy level of the caregiver's vocalization. Vocal frequency can be extracted by performing short-time energy analysis, zero-crossing rate analysis, etc. on the speech activity signal, and loudness information can be extracted by calculating the root mean square value or peak value of the speech signal; S120, the steps of the state type identification step include: S403, compare the vocal frequency with the corresponding preset frequency range and the loudness with the corresponding preset loudness range one by one; wherein, the preset frequency range and the preset loudness range are set based on clinical data or experience to reflect the normal vocalization under the care needs of different ages. S404, if the frequency of utterance falls within the preset frequency range and the loudness falls within the preset loudness range, then the state type of the speech activity signal is normal utterance; S405, if the frequency of vocalization is less than a preset frequency range and the loudness is less than a preset loudness range, determine whether the duration of vocalization exceeds a second preset duration. The second preset duration distinguishes between a brief decrease in vocalization and a prolonged period of silence. If so, the state type of the speech activity signal is prolonged silence, indicating that the person being cared for may be asleep, comatose, or depressed, leading to an inability to vocalize. If not, the state type of the speech activity signal is decreased vocalization, indicating that the person being cared for may be resting or thinking briefly, usually requiring no immediate intervention. This setting, by extracting the frequency and loudness parameters and comparing them with their respective preset ranges, identifies normal vocalization states. Furthermore, it introduces the judgment of vocalization duration, comparing it with the second preset time to identify temporary states of decreased vocalization and abnormal states of prolonged silence. In other words, through multi-dimensional judgment, the accuracy and precision of speech activity state recognition are significantly improved.

[0027] In this embodiment of the invention, see appendix. Figure 6 If the multidimensional vital signs signal includes visible facial state signals; S100, the steps of the vital signs signal acquisition step include: S501, controlling the non-contact sensing component to acquire the visible facial state signals of the bedridden person in real time; the visible facial state signals can be acquired by an optical camera to reflect the visual signals of the facial area of ​​the bedridden person. S110, the steps of the state feature extraction step include: S502, extracting eye state based on visible facial state signals; eye state refers to extracting information about the opening and closing of the eyes and eye movements of the person being cared for by analyzing visible facial state signals, which is used to determine whether the person being cared for is in different physiological or environmental states such as being awake, asleep, with eyes closed or with their face covered. S120, the steps of the state type identification step include: S503, recognizes the opening and closing state of the eyes; S504, If the eyes are open, the state type of the visible facial state signal is "awake and visible"; "awake and visible" refers to the visible facial state type when the eyes of the person being cared for are open, indicating that the person being cared for is awake or semi-awake and can perceive the surrounding environment. S505, if both eyes are closed and the duration of closure exceeds the third preset duration, the state type of the visible facial state signal is continuous eye closure, indicating that the person being cared for may be asleep or may be unconscious for a long time. S506, if both eyes are closed and the duration of closure does not exceed the third preset duration, the state type of the visible facial state signal is brief eye closure, indicating that the person being cared for may be in a state of blinking, brief rest or light sleep. S507: If the opening and closing of the eyes cannot be detected, the facial visibility signal will be classified as "face obscured," indicating that the person being cared for may be turning over, covered by blankets, or the camera's field of view may be limited, reminding the caregiver to pay attention to the person. This setup, by recognizing the person's eye status, can identify various states, including awake and visible, continuously closed eyes, briefly closed eyes, and face obscured. This allows for monitoring the person's sleep patterns and potential risks (e.g., prolonged eye closure may indicate a risk of loss of consciousness, while face obscuration may indicate a risk of suffocation due to blankets), providing guidance to caregivers, improving the safety of bedridden care, and avoiding misjudgments or delayed intervention.

[0028] In this embodiment of the invention, in step S140, the status presentation interface includes visual cues for abnormal states, which are used to convey to the caregiver that the person being cared for may be in an abnormal state, in order to quickly attract the caregiver's attention and enable the caregiver to understand the situation and take action in a timely manner. The visual cues may include: highlighting the corresponding dimension area of ​​the terminal, such as changing the background color, border color, font color, or adding flashing, etc.; and / or generating a prompt card in the abnormal state prompt area, the prompt card including the abnormal state and the dimension name, timestamp, and duration corresponding to the abnormal state, for example, the prompt card includes long-term stillness, body movement signal, body movement timestamp, and body movement duration.

[0029] In this embodiment of the invention, see appendix. Figure 7 S130, Data Fusion Step: After fusing the obtained state types from various dimensions to generate multidimensional state description data, the care method also includes: S151, Anomaly detection step: Compare the multidimensional state description data of the current time window with the multidimensional state description data of the corresponding previous adjacent time window; S152, if the multidimensional state description data of any dimension changes from a normal state to an abnormal state, then the timer of the corresponding dimension is started; for example, an abnormal state can be frequent body movement, prolonged stillness, rapid breathing, prolonged silence, face being obscured, etc. S153, If the state type of this dimension is still abnormal when the timer reaches the preset duration, it is determined to be a continuous abnormality, and a visual prompt is triggered. S154. If the timer has not reached the preset duration, the state type of this dimension changes from abnormal state to normal state, indicating that it is an instantaneous or short-term fluctuation. Then the timer is cleared and no visual prompt is triggered. S155, if the multidimensional state description data of any dimension changes from an abnormal state to a normal state, the visual cue is cleared, and the cue card is marked as restored in the abnormal state cue area. This setting can distinguish between instantaneous or short-term fluctuations and continuous abnormalities, avoiding misjudgments caused by instantaneous or short-term fluctuations and improving the accuracy of abnormality identification; using visual cuees can intuitively represent the situation, quickly attract the caregiver's attention, enable them to promptly detect abnormalities and take action, and improve the caregiver's response efficiency.

[0030] In this embodiment of the invention, the visual cues further include: providing multi-dimensional state description data corresponding to the abnormal state during a preset period before the change from a normal state to an abnormal state, for caregivers to refer to and compare. That is, if the multi-dimensional state description data of any dimension changes from a normal state to an abnormal state, in addition to triggering the current visual cues, multi-dimensional state description data corresponding to the abnormal state is also provided during the preset period before the change. For example, if the body movement signal changes from normal activity to prolonged stillness, the screen will display whether the person being cared for was in normal activity, reduced activity, or other body movement state during the preset period before the stillness occurred (e.g., within a few minutes), and how long each of these states lasted. This setup allows caregivers to intuitively understand the precursors and changes of abnormal states, which helps in analyzing the causes of these abnormal states and guiding the development of intervention strategies. For example, when prompted with "prolonged stillness," caregivers can review the caregiver's physical activity over a previous period to determine whether activity is gradually reduced until stillness or a sudden cessation of activity. This significantly improves the precision and proactivity of care, avoiding misjudgments or delays that may result from relying solely on current status information, thereby enhancing the quality and efficiency of bedridden care.

[0031] In this embodiment of the invention, in step S140, the status presentation interface includes at least one of a navigation bar, a multi-dimensional real-time status panel, a multi-dimensional time period statistical view, a multi-day status change trend chart, and a periodic status summary; wherein, the navigation bar is used for switching the person being cared for and filtering the time range; the multi-dimensional real-time status panel includes at least two of a body movement status panel, a breathing status panel, a language status panel, and a facial status panel, the body movement status panel is used to display the current status category and duration corresponding to the body movement signal, the breathing status panel is used to display the current status category and duration corresponding to the breathing rhythm signal, the language status panel is used to display the current status category and duration corresponding to the speech activity signal, and the facial status panel is used to display the visible facial status signal and duration; the multi-dimensional time period statistical view includes a body movement status type distribution chart, a breathing status type time series chart, and a speech status type heatmap; the status summary is generated by aggregating the multi-dimensional status description data of all time windows of each day at a preset periodic node and generating a periodic status summary according to a preset template. It should be noted that the navigation bar can be located at the top of the status presentation interface, the multi-dimensional real-time status panel can be placed in the middle of the status presentation interface, and the multi-day status change trend chart can be placed at the bottom of the status presentation interface. This setting can significantly improve the efficiency and quality of bedridden care information presentation. For example, by introducing a navigation bar, caregivers can easily manage the data of multiple caregivers and flexibly filter the required time range, greatly improving the efficiency of data retrieval and analysis. By introducing a multi-dimensional real-time status panel, an immediate and intuitive current status is provided, enabling caregivers to quickly identify abnormal situations and take timely intervention measures to ensure the safety and comfort of caregivers. By introducing multi-dimensional time period statistical views and multi-day status change trend charts, the behavioral patterns and physiological changes of caregivers can be revealed from different time granularities, helping caregivers to deeply understand the dynamic evolution of their health status, thereby developing more scientific and personalized care plans. In addition, by generating periodic status summaries, it is not only convenient for caregivers to conduct periodic assessments and summaries, but also provides a strong basis for communication with medical staff, comprehensively improving the intelligence level and quality of bedridden care.

[0032] This invention also provides a system for bedridden care, the system including a terminal 03 and a non-contact sensing component 04, the non-contact sensing component 04 being disposed on the terminal 03; for the application of the above-described bedridden care method, see the appendix. Figure 8 The care system also includes: The vital signs acquisition module 10 is used to acquire multi-dimensional vital signs signals of bedridden caregivers in real time. State feature extraction module 20 is used to extract state feature data of each dimension based on multi-dimensional vital signs signals; The state type identification module 30 is used to compare the state feature data of each dimension with the preset range of the corresponding dimension to obtain the state type of each dimension. Data fusion module 40: used to fuse the obtained state types from various dimensions to generate multi-dimensional state description data; and, Care status presentation module 50: Used to transmit multi-dimensional status description data to the terminal and generate a status presentation interface.

[0033] The bed rest care system provided in this embodiment of the invention is applied to the above-mentioned bed rest care method and has all the advantages of the above-mentioned bed rest care method, which will not be repeated here.

[0034] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A method for bedridden care, characterized in that, An application is made in a bedridden care system, the bedridden care system including a terminal (03) and a non-contact sensing component (04), the non-contact sensing component (04) being disposed on the terminal (03); the care method includes the following steps: Vital sign signal acquisition step: Control the non-contact sensing component to acquire multi-dimensional vital sign signals of the bedridden person in real time; State feature extraction step: Extract state feature data for each dimension based on the multi-dimensional vital signs signal; State type identification step: Compare the state feature data of each dimension with the preset range of the corresponding dimension to obtain the state type of each dimension; Data fusion step: The obtained state types of each dimension are fused to generate multi-dimensional state description data; the multi-dimensional state description data includes the state type of each dimension and the duration corresponding to each state type. Care status presentation step: The multi-dimensional status description data is transmitted to the terminal to generate a status presentation interface.

2. The bedridden care method according to claim 1, characterized in that, In the step of acquiring vital signs signals, the multidimensional vital signs signals include at least two types: body movement signals, respiratory rhythm signals, voice activity signals, and visible facial state signals; wherein, the body movement signals and / or the respiratory rhythm signals are acquired through at least one of an optical camera, an infrared camera, and a depth sensor; the voice activity signals are acquired through a microphone array; and the visible facial state signals are acquired through an optical camera.

3. The bedridden care method according to claim 2, characterized in that, If the multidimensional vital signs signal includes body movement signal; The steps of the vital sign signal acquisition step include: controlling the non-contact sensing component to acquire the body movement signal in real time; The steps of the state feature extraction step include: extracting motion amplitude and motion frequency based on the body motion signal; The steps of the state type identification step include: comparing the motion amplitude with the corresponding first preset amplitude range and the motion frequency with the corresponding first preset frequency range one by one; if the motion amplitude falls within the first preset amplitude range and the motion frequency falls within the first preset frequency range, then the state type of the body movement signal is normal activity; if the motion amplitude is higher than the first preset amplitude range and the motion frequency is higher than the first preset frequency range, then the state type of the body movement signal is frequent activity; if the motion amplitude is lower than the first preset amplitude range and the motion frequency is lower than the first preset frequency range, determine whether the duration of the body movement exceeds a first preset duration; if yes, then the state type of the body movement signal is prolonged stillness; if no, then the state type of the body movement signal is reduced activity.

4. The bedridden care method according to claim 2, characterized in that, If the multidimensional vital signs signal includes respiratory rhythm signal; The steps of the vital sign signal acquisition step include: controlling the non-contact sensing component to acquire the respiratory rhythm signal in real time; The steps of the state feature extraction step include: extracting the respiratory cycle and respiratory amplitude based on the respiratory rhythm signal; The steps of the state type identification step include: comparing the respiratory cycle with the corresponding preset range of single breath duration and the respiratory amplitude with the corresponding preset range of second amplitude; if the respiratory cycle falls within the preset range of single breath duration and the respiratory amplitude falls within the preset range of second amplitude, then the state type of the respiratory rhythm signal is normal rhythm; if the respiratory cycle is greater than the preset range of single breath duration and the respiratory amplitude is greater than the preset range of second amplitude, then the state type of the respiratory rhythm signal is rapid breathing; if the respiratory cycle is less than the preset range of single breath duration and the respiratory amplitude is less than the preset range of second amplitude, then the state type of the respiratory rhythm signal is slow breathing.

5. The bedridden care method according to claim 2, characterized in that, If the multidimensional vital signs signal includes a voice activity signal; The steps of the vital sign signal acquisition step include: controlling the non-contact sensing component to acquire the voice activity signal in real time; The steps of the state feature extraction step include: extracting the frequency and loudness of vocalizations based on the speech activity signal; The steps of the state type identification step include: comparing the phonation frequency with the corresponding frequency preset range and the loudness with the corresponding loudness preset range one by one; if the phonation frequency falls within the frequency preset range and the loudness falls within the loudness preset range, then the state type of the voice activity signal is normal phonation; if the phonation frequency is less than the frequency preset range and the loudness is less than the loudness preset range, determining whether the phonation duration exceeds a second preset duration; if yes, then the state type of the voice activity signal is long-term silence; if no, then the state type of the voice activity signal is reduced phonation.

6. The bedridden care method according to claim 2, characterized in that, If the multidimensional vital signs signal includes visible facial state signals; The steps of the vital sign signal acquisition step include: controlling the non-contact sensing component to acquire the visible facial state signals in real time; The steps of the state feature extraction step include: extracting the eye state based on the visible facial state signal; The steps of the state type identification step include: The system identifies the opening and closing state of the eyes. If the eyes are open, the state type of the visible facial state signal is "awake and visible". If the eyes are closed and the closing duration exceeds a third preset duration, the state type of the visible facial state signal is "continuously closed eyes". If the eyes are closed and the closing duration does not exceed the third preset duration, the state type of the visible facial state signal is "briefly closed eyes". If the opening and closing state of the eyes cannot be identified, the state type of the visible facial state signal is "face obscured".

7. The method of bedridden care according to any one of claims 1-6, characterized in that, In the care status presentation step, the status presentation interface includes visual prompts for abnormal statuses. The visual prompts include: highlighting the corresponding dimension area on the terminal; and / or generating a prompt card in the abnormal status prompt area. The prompt card includes the abnormal status and the corresponding dimension name, timestamp, and duration of the abnormal status. After generating the multidimensional state description data, the care method further includes: Anomaly detection step: Compare the multidimensional state description data of the current time window with the multidimensional state description data of the corresponding previous adjacent time window; If the multidimensional state description data of any dimension changes from a normal state to an abnormal state, the timer for the corresponding dimension is started; if the state type of the dimension is still abnormal when the timer reaches the preset duration, it is determined to be a continuous abnormality, and the visual prompt is triggered; if the state type of the dimension changes from an abnormal state to a normal state before the timer reaches the preset duration, the timer is cleared and the visual prompt is not triggered. If any dimension of the multidimensional state description data changes from an abnormal state to a normal state, the visual cue is cleared, and the cue card is marked as restored in the abnormal state cue area.

8. The bedridden care method according to claim 7, characterized in that, The visual cues also include: providing multi-dimensional state description data corresponding to the abnormal state during a preset period before the state changes from normal to abnormal, for caregivers to refer to and compare.

9. The bedridden care method according to any one of claims 2-6, characterized in that, In the care status presentation step, the status presentation interface includes at least one of the following: a navigation bar, a multi-dimensional real-time status panel, a multi-dimensional time period statistical view, a multi-day status change trend chart, and a periodic status summary. The navigation bar is used for switching between caregivers and filtering time ranges. The multi-dimensional real-time status panel includes at least two of the following: a body movement status panel, a breathing status panel, a speech status panel, and a facial status panel. The body movement status panel displays the current status category and duration corresponding to the body movement signal; the breathing status panel displays the current status category and duration corresponding to the breathing rhythm signal; the speech status panel displays the current status category and duration corresponding to the speech activity signal; and the facial status panel displays the visible facial status signal and its duration. The multi-dimensional time-period statistical view includes a body movement status type distribution map, a breathing status type time series map, and a speech status type heatmap. The status summary is generated by aggregating the multi-dimensional status description data of all time windows of each day at preset periodic nodes and generating a periodic status summary according to a preset template.

10. A bedridden care system, characterized in that, The bedridden care method according to any one of claims 1-9, wherein the system comprises: The vital signs acquisition module (10) is used to acquire multi-dimensional vital signs signals of bedridden caregivers in real time. The state feature extraction module (20) is used to extract state feature data of each dimension based on the multi-dimensional vital signs signal; The state type identification module (30) is used to compare the state feature data of each dimension with the preset range of the corresponding dimension to obtain the state type of each dimension; Data fusion module (40): used to fuse the obtained state types of each dimension to generate multi-dimensional state description data; and, Care status presentation module (50): used to transmit the multi-dimensional status description data to the terminal and generate a status presentation interface.

Images