Pressure sore risk area real-time monitoring and protection method based on air bag mattress

Abstract

The invention discloses a pressure sore risk area real-time monitoring and dynamic protection method based on an air bag mattress. The method comprises the following steps: firstly, constructing a pressure sore risk position data set, training a key point detection model, and predicting a pressure sore risk position; then, the optimal inflation height and the alternate inflation and deflation frequency of the air bag sets in the high-risk area, the medium-risk area and the low-risk area of the pressure sores are obtained through a pressure sore protection comfort experiment; finally, pressure data of the patient are collected in real time, and a pressure heat map is obtained; predicting a pressure sore risk position in real time according to the trained key point detection model, determining a pressure sore risk area according to the pressure sore risk position, when a certain area is predicted to be a high / medium / low pressure sore risk area for the first time, accumulating the continuous prediction duration of the area, and if the continuous prediction duration of the area reaches a threshold value, determining that the area is a high / medium / low pressure sore risk area. And inflating and deflating the air bag group in the area according to the optimal inflating height and the alternate inflating and deflating times, and after the inflating and deflating are completed, resetting the continuous predicted duration and re-accumulating. The pressure sore risk position is predicted in real time by monitoring pressure data in real time, the purpose of relieving the local pressure of the body of a patient is achieved through inflation and deflation of the air bag set, and precise monitoring and protection of pressure sores are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent medical care, and specifically relates to a method for real-time monitoring and protection of pressure ulcer risk areas based on an airbag mattress. Background Art

[0002] Disabled patients are prone to local pressure injuries and pressure ulcers due to prolonged bed rest, with a certain part of the body being continuously compressed, especially at the bony prominences where there is a lack of adipose tissue protection, no muscle wrapping, or a thin muscle layer.

[0003] Currently, for the pressure ulcer monitoring of long-term bedridden disabled patients, contact or non-contact sensors are usually used. For example, inertial sensors, cameras, or infrared sensors are used to monitor the patient's lying position, and the duration of maintaining the lying position is regarded as the duration of continuous compression. When the duration of maintaining the lying position reaches a certain value, it is considered that the patient has a risk of developing pressure ulcers. However, this method ignores the influence of pressure distribution on the occurrence of pressure ulcers.

[0004] Existing pressure ulcer protection methods generally adopt manual, mechanical, pneumatic, and other means. Manual means usually involve professional nursing staff regularly turning the patient over or adjusting the body position according to the patient's body position and compression situation to relieve local pressure, but this increases the labor intensity of the nursing staff. Mechanical means use a nursing bed that can change its position and posture, and perform protection through actions such as small-angle turning over, which is more convenient than manual means, but the protection method is relatively fixed, with poor flexibility and comfort. Pneumatic means use a nursing mattress to perform massage through methods such as fluctuating or small-angle turning over. However, the pressure on different body parts is different, so the risk degree of pressure ulcer occurrence is different, and the required protection intensity for different risk areas is also different, and these protection methods all ignore the distribution of risk areas.

[0005] Therefore, in view of the limitations of existing pressure ulcer monitoring and protection methods, the present invention provides a method for real-time monitoring and protection of pressure ulcer risk areas based on an airbag mattress, which performs pressure ulcer protection through real-time monitoring and dynamic regulation. Summary of the Invention

[0006] Aiming at the deficiencies of the prior art, the technical problem to be solved by the present invention is to propose a method for real-time monitoring and protection of pressure ulcer risk areas based on an airbag mattress.

[0007] The present invention solves the above-mentioned technical problem by adopting the following technical solutions:

[0008] A real-time monitoring and dynamic protection method for pressure ulcer risk areas based on an airbag mattress, characterized in that the airbag mattress is divided into 12 pressure ulcer risk areas according to the prone positions of pressure ulcers on the human body, namely the left scapular area, the right scapular area, the left elbow area, the right elbow area, the left vertebral body prominence area, the right vertebral body prominence area, the left sacral area, the right sacral area, the left thigh area, the right thigh area, the left heel area and the right heel area. All the airbags in each area are connected as a whole to form an airbag group, and the inflation and deflation of the airbag groups in each area are independent of each other; among them, the heel area and the sacral area are high-risk areas for pressure ulcers, the thigh area and the vertebral body prominence area are medium-risk areas for pressure ulcers, and the elbow area and the scapular area are low-risk areas for pressure ulcers; the method includes the following steps:

[0009] Step 1: Collect the pressure data of the subject in different lying postures, and generate a pressure heat map according to the pressure data; mark the pressure ulcer risk positions on the pressure heat map according to the prone positions of pressure ulcers on the human body; several marked pressure heat maps form a pressure ulcer risk position data set, and train the key point detection model to make the model predict the pressure ulcer risk positions;

[0010] Step 2: Through the pressure ulcer protection comfort experiment, obtain the optimal inflation height and the optimal number of alternating inflation and deflation times of the airbag groups in the high, medium, and low-risk areas of pressure ulcers;

[0011] Step 3: Real-time collect the pressure data of the patient lying on the airbag mattress to obtain a pressure heat map; input the pressure heat map into the trained key point detection model to predict the pressure ulcer risk positions in real time; determine the pressure ulcer risk areas according to the pressure ulcer risk positions. When a certain area is first predicted to be a high / medium / low-risk area for pressure ulcers, accumulate the continuous prediction duration of this pressure ulcer risk area. If the continuous prediction duration of this pressure ulcer risk area reaches the threshold, inflate and deflate the airbag group of this pressure ulcer risk area according to the optimal inflation height and the alternating inflation and deflation times. After the inflation and deflation are completed, clear the continuous prediction duration of this pressure ulcer risk area and start accumulating again;

[0012] If the continuous prediction of this pressure ulcer risk area is interrupted, and this area is predicted to be a high / medium / low-risk area for pressure ulcers again within the interruption time threshold, then continue to accumulate on the basis of the original continuous prediction duration. When the continuous prediction duration reaches the threshold, inflate and deflate the airbag group of this pressure ulcer risk area according to the optimal inflation height and the alternating inflation and deflation times. After the inflation and deflation are completed, clear the continuous prediction duration of this pressure ulcer risk area and start accumulating again;

[0013] If the continuous prediction of this pressure ulcer risk area is interrupted, and this area is not predicted to be a high / medium / low-risk area for pressure ulcers again within the interruption time threshold, then clear the continuous prediction duration of this pressure ulcer risk area and start accumulating again.

[0014] Furthermore, the experimental process of pressure ulcer prevention comfort is as follows: The subject lies flat on the airbag mattress, with both arms placed on both sides of the torso, and the body is in a relaxed and comfortable state. When the airbag mattress is in a deflated state, the airbag groups in each pressure ulcer risk area are inflated to different heights respectively, and the comfort is evaluated according to the subjective feelings of the subject to obtain the optimal inflation height. On the basis of the optimal inflation height, the airbag groups in the pressure ulcer risk area are alternately inflated and deflated, and the comfort is evaluated according to the subjective feelings of the subject to obtain the optimal number of alternate inflations and deflations. The optimal inflation height of the airbag group in the low-pressure ulcer risk area is 6 cm - 8 cm, and the optimal number of alternate inflations and deflations is 8 - 10 times. The optimal inflation height of the airbag group in the medium-pressure ulcer risk area is 8 - 10 cm, and the optimal number of alternate inflations and deflations is 10 - 12 times. The optimal inflation height of the airbag group in the high-pressure ulcer risk area is 10 - 12 cm, and the optimal number of alternate inflations and deflations is 12 - 14 times.

[0015] Furthermore, the threshold of the continuous prediction duration is two hours, and the threshold of the intermittent time is 10 s.

[0016] Furthermore, in step one, the subjects are healthy adults of different body types, including three heights: tall, medium, and short, and three body postures: fat, medium, and thin. There are a total of 9 body types obtained by combining height and body posture in pairs. The lying postures include three types: lying flat in the middle, lying flat on the right side, and lying flat on the left side. Lying flat in the middle means that the subject lies flat in the middle area of the airbag mattress and the sagittal plane of the subject is parallel to the left-right bisecting plane of the airbag mattress. Lying flat on the right / left side means that the subject lies flat on the right / left side of the airbag mattress and the sagittal plane of the subject intersects with the left-right bisecting plane of the airbag mattress.

[0017] Furthermore, the key point detection model is a YOLO series model for key point detection tasks.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0019] According to the common pressure ulcer prone areas in the human lying posture, the array airbag mattress is partitioned to obtain 12 pressure ulcer risk areas, and these areas are divided into three categories: high, medium, and low risk areas according to the risk degree. The high-density flexible pressure sensor is used to collect pressure data in real time, and the pressure heat map is used to predict the pressure ulcer risk areas. The airbag groups in each pressure ulcer risk area are independent of each other, and the purpose of massaging the patient's body parts is achieved by inflating and deflating the airbag groups, so as to effectively relieve the local pressure and realize the accurate monitoring and prevention of pressure ulcers.

[0020] To improve the effectiveness and comfort of pressure ulcer prevention, the optimal inflation and deflation heights and the optimal number of alternating inflation and deflation cycles for the airbag groups in high, medium, and low pressure ulcer risk areas are obtained based on pressure ulcer prevention comfort experiments. Since the pressure data is collected in real time, to avoid the influence of patient body movement on the prediction result, when the continuous prediction duration of the pressure ulcer risk area reaches the set threshold, the airbag group in that pressure ulcer risk area is alternately inflated and deflated, improving the prevention accuracy. Brief Description of the Drawings

[0021] Figure 1 It is a zoning diagram of the airbag mattress of the present invention;

[0022] Figure 2 It is a control system diagram of the present invention;

[0023] Figure 3 It is an overall flowchart of the present invention;

[0024] In the figure, 1 - nursing pillow; 2 - array airbag mattress; 3 - high - density flexible pressure sensor; 4 - pressure data acquisition module; 5 - host computer; 6 - air pump; 7 - solenoid valve; 8 - relay; 9 - airbag pressure acquisition module;

[0025] 21 - rectangular airbag; 22 - fecal orifice airbag; 23 - wedge - shaped airbag; 24 - left scapular region; 25 - right scapular region; 26 - left elbow region; 27 - right elbow region; 28 - left vertebral spinous process region; 29 - right vertebral spinous process region; 210 - left sacral region; 211 - right sacral region; 212 - left thigh region; 213 - right thigh region; 214 - left heel region; 215 - right heel region. Detailed Embodiment

[0026] The following provides specific embodiments in conjunction with the drawings. The specific embodiments are only used to introduce the technical solution of the present invention in detail and do not limit the protection scope of this application.

[0027] The method of the present invention relies on the high - density flexible pressure sensor 3 and the array airbag mattress 2; among them, the high - density flexible pressure sensor 3 is laid on the array airbag mattress 2, and its upper end is closely attached to the lower end of the nursing pillow 1; the high - density flexible pressure sensor 3 uses a piezoresistive sensor, the pressure data acquisition density is 3mm×3mm, the overall size is 1.64m×1.2m, and the pressure data is collected at high speed by the pressure data acquisition module 4 and sent to the host computer 5 through the serial port for pressure data processing.

[0028] The array airbag mattress 2 includes 24 rectangular airbags 21, 20 wedge-shaped airbags 23 and 2 vent airbags 22, see the invention application with publication number CN117180025A for details; in order to implement the method of the present invention, the airbag area of ​​the array airbag mattress 2 is uniformly divided into 12 pressure sore risk areas according to the parts of the human body where pressure sores are prone to occur, namely, the left shoulder blade area 24, the right shoulder blade area 25, the left elbow area 26, the right elbow area 27, the left vertebral protuberance area 28, the right vertebral protuberance area 29, the left sacral area 210, the right sacral area 211, the left thigh area 212, the right thigh area 213, the left heel area 2 14 and the right heel area 215, all the airbags in each area are connected as a whole to form an airbag group. The airbag groups in each area are independent of each other. Each airbag group is connected to the air pump 6 through an air pipe and a two-position two-way solenoid valve 7. The air pump 6 and the solenoid valve 7 are both connected to a relay 8, and the relay 8 is connected to the host computer 5 through a serial port. The host computer 5 sends instructions to the relay 8 through the serial port to control the opening and closing of the air pump 6 and the solenoid valve 7, thereby controlling the inflation and deflation of each airbag group; at the same time, all the airbag groups are connected to the airbag pressure acquisition module 9, and the airbag pressure acquisition module 9 is connected to the host computer 5 through a serial port to monitor the internal pressure of each airbag group.

[0029] The high-density flexible pressure sensor 3 adopts a sandwich structure, consisting of a bottom base, a middle sensing layer and a top covering layer. It has high stability and durability, and has a response time of 25ms and a recovery time of 50ms. It can respond quickly to tiny external stimuli and can withstand a maximum pressure of 20kpa.

[0030] A method for real-time monitoring and dynamic protection of pressure sore risk areas based on an airbag mattress comprises the following steps:

[0031] Step 1: Construct a pressure ulcer risk location dataset and use it to train the key point detection model so that the model can predict the pressure ulcer risk location;

[0032] High-density flexible pressure sensors are used to collect pressure data of subjects in different lying positions, and the pressure data is processed to generate a pressure heat map. According to the prone locations of pressure sores on the human body, the pressure heat map is annotated using the annotation tool Labelme. The annotation information is the pressure sore risk location. Several annotated pressure heat maps constitute a pressure sore risk location data set.

[0033] The subjects are healthy adults of different body types, including tall, medium, and short in height, and fat, medium, and thin in body build. There are a total of 9 body types obtained from the pairwise combination of height and body build, and multiple subjects are selected for each body type. The lying postures include lying flat in the middle, lying flat on the right side, and lying flat on the left side. Lying flat in the middle means that the subject lies flat in the middle area of the airbag mattress and the sagittal plane of the subject is parallel to the left-right bisecting plane of the airbag mattress. Lying flat on the right side means that the subject lies flat on the right side of the airbag mattress and the sagittal plane of the subject intersects with the left-right bisecting plane of the airbag sheet. The same applies to lying flat on the left side. In this embodiment, the pressure sore risk position dataset includes a total of 9000 annotated pressure heatmaps.

[0034] Use the annotated pressure heatmaps to train the key point detection model and predict the pressure sore risk positions; the key point detection model adopts the YOLO series models with the key point detection task, such as YOLOv8-Pose, YOLOv11-Pose, etc.

[0035] Step 2: According to the pressure sore prevention comfort experiment, obtain the optimal inflation height and the optimal number of alternating inflation and deflation for the airbag groups in the high, medium, and low pressure sore risk areas.

[0036] The risks of pressure sores occurring in different body parts of disabled patients are different. According to the risk level, the pressure sore risk areas are divided into three categories: high pressure sore risk areas: heels, sacrum; medium pressure sore risk areas: thighs, vertebral prominences; low pressure sore risk areas: elbows, scapulae. Conduct pressure sore prevention comfort experiments on the three types of areas: The subject lies flat on the array airbag mattress, with both arms placed on both sides of the torso, and the body is in a relaxed and comfortable state; when the array airbag mattress is in the deflated state, inflate the airbag groups in each pressure sore risk area to different heights respectively, and evaluate the comfort according to the subjective feelings of the subject to obtain the optimal inflation height; on the basis of the optimal inflation height, perform alternating inflation and deflation on the airbag groups in the pressure sore risk areas, which is equivalent to massaging the corresponding body parts of the subject, and evaluate the comfort according to the subjective feelings of the subject to obtain the optimal number of alternating inflation and deflation; comprehensively analyze the experimental data of all subjects to obtain that the optimal inflation height of the airbag group in the low pressure sore risk area is 6 cm - 8 cm, and the optimal number of alternating inflation and deflation is 8 - 10 times; the optimal inflation height of the airbag group in the medium pressure sore risk area is 8 - 10 cm, and the optimal number of alternating inflation and deflation is 10 - 12 times; the optimal inflation height of the airbag group in the high pressure sore risk area is 10 - 12 cm, and the optimal number of alternating inflation and deflation is 12 - 14 times. The inflation height reflects the massage intensity, and the number of alternating inflation and deflation reflects the number of massages.

[0037] Step 3: Real-time collect the pressure data of the patient lying on the airbag mattress to obtain a pressure heat map; input the pressure heat map into the trained key point detection model to predict the pressure ulcer risk location in real time; determine the pressure ulcer risk area according to the pressure ulcer risk location. When a certain area is first predicted to be a high / medium / low risk area of pressure ulcer, use a timer to accumulate the continuous prediction duration of this pressure ulcer risk area. If the continuous prediction duration of this pressure ulcer risk area reaches the threshold, inflate and deflate the airbag group of this pressure ulcer risk area according to the optimal inflation height and the optimal number of alternating inflation and deflation to achieve the purpose of massaging the corresponding body part of the patient. After the inflation and deflation are completed, clear the continuous prediction duration of this pressure ulcer risk area and start accumulating again;

[0038] If the continuous prediction of this pressure ulcer risk area is interrupted and it is predicted again that this area is a high / medium / low risk area of pressure ulcer within the interruption time threshold, it indicates that the patient's body has moved and returned to the original position. Then continue to accumulate on the basis of the original continuous prediction duration. When the continuous prediction duration reaches the threshold, inflate and deflate the airbag group of this pressure ulcer risk area according to the optimal inflation height and the optimal number of alternating inflation and deflation. After the inflation and deflation are completed, clear the continuous prediction duration of this pressure ulcer risk area and start accumulating again;

[0039] If the continuous prediction of this pressure ulcer risk area is interrupted and it is not predicted again that this area is a high / medium / low risk area of pressure ulcer within the interruption time threshold, it indicates that the patient's body has not returned to the original position after moving. Then clear the continuous prediction duration of this area and start accumulating again.

[0040] The threshold of the continuous prediction duration is two hours, and the interruption time threshold is 10s.

[0041] In summary, the present invention uses a high-density flexible pressure sensor to collect the patient's pressure data in real time and generate a pressure heat map, and the key point detection model predicts the pressure ulcer risk area in real time according to the pressure heat map; since the pressure ulcer risk area is unified with the partition of the array airbag mattress, it can achieve precise monitoring and protection of pressure ulcers, ensuring the effectiveness and comfort of pressure ulcer protection. The parts not described in the present invention are applicable to the prior art.

Claims

1. A method for real-time monitoring and dynamic protection of pressure sore risk areas based on an airbag mattress, characterized in that: The airbag mattress is divided into 12 pressure sore risk areas according to the prone positions of pressure sores on the human body, namely the left scapular area, the right scapular area, the left elbow area, the right elbow area, the left vertebral protuberance area, the right vertebral protuberance area, the left sacral area, the right sacral area, the left thigh area, the right thigh area, the left heel area and the right heel area. All the airbags in each area are connected as a whole to form an airbag group, and the airbag groups in each area are inflated and deflated independently; among them, the heel area and the sacral area are high-risk areas for pressure sores, the thigh area and the vertebral protuberance area are medium-risk areas for pressure sores, and the elbow area and the scapular area are low-risk areas for pressure sores; the following steps are included: Step 1: Collect the pressure data of the subjects in different lying positions, and generate a pressure heat map based on the pressure data; mark the pressure sore risk locations on the pressure heat map according to the prone locations of pressure sores on the human body; several marked pressure heat maps form a pressure sore risk location data set, and train the key point detection model to enable the model to predict the pressure sore risk location; Step 2: Through the pressure sore protection comfort experiment, the optimal inflation height and optimal alternating inflation and deflation times of the airbag group in the high, medium and low risk areas of pressure sores are obtained; Step 3: Collect the pressure data of the patient lying on the airbag mattress in real time to obtain a pressure heat map; input the pressure heat map into the trained key point detection model to predict the pressure sore risk position in real time; determine the pressure sore risk area according to the pressure sore risk position, and when a certain area is predicted to be a high / medium / low risk area for pressure sores for the first time, accumulate the continuous prediction time of the pressure sore risk area. If the continuous prediction time of the pressure sore risk area reaches a threshold, inflate and deflate the airbag group in the pressure sore risk area according to the optimal inflation height and alternating inflation and deflation times. After the inflation and deflation are completed, the continuous prediction time of the pressure sore risk area is reset to zero and accumulated again; If the continuous prediction of the pressure sore risk area is interrupted, and the area is predicted to be a high / medium / low risk area for pressure sores again within the interruption time threshold, the accumulation will continue on the basis of the original continuous prediction time. When the continuous prediction time reaches the threshold, the airbag group in the pressure sore risk area will be inflated and deflated according to the optimal inflation height and alternating inflation and deflation times. After the inflation and deflation are completed, the continuous prediction time of the pressure sore risk area will be reset to zero and accumulated again; If the continuous prediction of the pressure sore risk area is interrupted, and the area is not predicted as a high / medium / low risk area for pressure sores again within the interruption time threshold, the continuous prediction time of the pressure sore risk area will be reset to zero and accumulated again.

2. The method for real-time monitoring and dynamic protection of pressure sore risk areas based on an airbag mattress according to claim 1 is characterized in that: The process of the comfort experiment for pressure sore protection is as follows: the subject lies flat on the airbag mattress with his arms placed on both sides of the torso, and the body is in a relaxed and comfortable state; when the airbag mattress is in a deflated state, the airbag group in each pressure sore risk area is inflated to different heights, and the comfort is evaluated according to the subjective feelings of the subjects to obtain the optimal inflation height; on the basis of the optimal inflation height, the airbag group in the pressure sore risk area is alternately inflated and deflated, and the comfort is evaluated according to the subjective feelings of the subjects to obtain the optimal number of alternating inflation and deflation times; the optimal inflation height of the airbag group in the pressure sore low risk area is 6cm-8cm, and the optimal number of alternating inflation and deflation times is 8-10 times; the optimal inflation height of the airbag group in the pressure sore medium risk area is 8-10cm, and the optimal number of alternating inflation and deflation times is 10-12 times; the optimal inflation height of the airbag group in the pressure sore high risk area is 10-12cm, and the optimal number of alternating inflation and deflation times is 12-14 times.

3. The method for real-time monitoring and dynamic protection of pressure sore risk areas based on an airbag mattress according to claim 1 or 2, characterized in that: The threshold for continuous prediction duration is two hours, and the threshold for intermittent prediction is 10 seconds.

4. The method for real-time monitoring and dynamic protection of pressure sore risk areas based on an airbag mattress according to claim 1 is characterized in that: In step 1, the subjects are healthy adults of different body types, including three types of height, including tall, medium and short, and three types of body shapes, including fat, medium and thin. The heights and body shapes are combined in pairs to obtain a total of 9 body types; the lying positions include three types of middle flat lying, right side flat lying and left side flat lying. Middle flat lying means that the subject lies flat on the middle area of ​​the airbag mattress and the subject's sagittal plane is parallel to the left-right bisector of the airbag mattress. Right / left side flat lying means that the subject lies flat to the right / left side on the airbag mattress and the subject's sagittal plane intersects with the left-right bisector of the airbag mattress.

5. The method for real-time monitoring and dynamic protection of pressure sore risk areas based on an airbag mattress according to claim 1 or 4, characterized in that: The key point detection model adopts the YOLO series model with key point detection tasks.

Images