Individualized multifunctional integrated delivery bed management method and system for pregnant women

Abstract

The application provides an individualized multifunctional integrated delivery bed pregnant woman management method and system, and relates to data processing technology, wherein the method comprises: recognizing a delivery bed view obtained, determining each limb feature area corresponding to different target limb features; when it is determined based on any limb feature area that the corresponding target limb feature is in a preset auxiliary posture, the target limb feature is determined as an auxiliary limb feature, and a target delivery bed is determined as an auxiliary array grid based on the delivery bed view, the target delivery bed comprising a delivery bed array grid having an auxiliary correlation relationship with the auxiliary limb feature; a bed plate partition where the auxiliary array grid is located is determined based on the target delivery bed, and an angle adjustment value and an inflation value of the auxiliary array grid corresponding to the bed plate partition are determined based on the auxiliary limb feature; the bed plate partition is controlled to perform angle adjustment based on the angle adjustment value, and the auxiliary array grid is controlled to perform inflation adjustment based on the inflation value. The application at least improves the adaptation efficiency.

Description

Technical Field

[0001] This invention relates to data processing technology, and more particularly to a method and system for managing pregnant women using an individualized, multifunctional, integrated delivery bed. Background Technology

[0002] In the field of obstetric and gynecological clinical nursing and maternal management, integrated maternity beds have been widely used during the labor and routine examinations of pregnant women. Traditional maternity bed management and support adjustment techniques usually rely on the manual experience of medical staff to determine the support points and adjustment range, and operate the maternity bed by manual adjustment.

[0003] As pregnant women continue to lie down and wait for labor, their lying posture and stress state will constantly change. Traditional methods of manually positioning or fixed preset angle adjustment cannot adapt to the changes in the pregnant woman's body in real time, which may lead to insufficient support or imbalance of force in certain parts of the pregnant woman's body, resulting in decreased comfort and safety hazards.

[0004] Most existing birthing bed support systems rely on a single, universal benchmark for assessment or adjustment based solely on fixed angles, failing to comprehensively consider the physiological morphology and real-time stress states of different limb regions in pregnant women. For postural abnormalities such as the overall suspension of the lower back and lateral imbalance caused by abdominal protrusion, uniform support is often applied without differentiation, resulting in poor individualized adaptation of the birthing bed and inconsistent support effects due to differences in human experience.

[0005] In addition, the existing solutions have a relatively simple adjustment method when performing support, usually relying only on the overall angle deflection of the bed board. This makes it impossible to achieve precise support compensation for local points and to accurately fill the support gaps left after angle adjustment. As a result, the support strength and position cannot fully conform to the actual body shape of the pregnant woman, which increases the operational burden and time consumption of clinical nursing.

[0006] Therefore, how to rationally plan the individualized support adjustment of the delivery bed based on the pregnant woman's limb characteristics and real-time posture, so as to improve the force imbalance and enhance the adaptation efficiency and automation level of delivery bed management, has become an urgent problem to be solved. Summary of the Invention

[0007] Based on the above problems, the present invention is proposed to provide a method and system for managing pregnant women using an individualized multifunctional integrated delivery bed that overcomes or at least partially solves the above problems.

[0008] According to one aspect of the present invention, a method for managing pregnant women using an individualized multifunctional integrated delivery bed is provided, comprising the following steps: Once the target pregnant woman is identified as being placed on the target delivery bed, the delivery bed acquisition unit is controlled to acquire data from the target delivery bed and to identify the obtained view of the delivery bed, thereby determining the limb feature regions that indicate the target pregnant woman and correspond to different target limb features. When the target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, the target limb feature is determined as an auxiliary limb feature, and the birthing bed array grids included in the target birthing bed that have an auxiliary relationship with the auxiliary limb feature are determined as auxiliary array grids based on the view of the birthing bed. Based on the target delivery bed, determine the bed board partition where the auxiliary array grid is located, and based on the characteristics of the auxiliary limbs, determine the corresponding bed board partition, the angle adjustment value of the auxiliary array grid, and the inflation value; The control bed board partition performs angle adjustment based on the angle adjustment value, and the control auxiliary array grid performs inflation adjustment based on the expansion inflation value.

[0009] Optionally, in the method according to the present invention, when it is determined that the target pregnant woman is placed on the target delivery bed, the delivery bed acquisition unit is controlled to acquire data from the target delivery bed, including: When it is determined that the pressure acquisition value output by any pressure sensing unit located under each birthing bed array cell in the target birthing bed is greater than the preset pressure threshold, the birthing bed array cell corresponding to the pressure acquisition value is determined to have pressure-bearing properties. The birthing bed compression shape, which is composed of all birthing bed array grids with compression properties, is determined, and the birthing bed compression shape is compared with the retrieved reference compression shape to determine a first evaluation value based on the obtained shape similarity. Each pre-marked array cell corresponding to the reference pressure shape is determined, and in response to any delivery bed array cell with pressure attribute whose corresponding pressure position is the same as the marking position of the corresponding mark array cell, the initial number of marks is accumulated to determine a second evaluation value based on the obtained current number of marks. The first evaluation value and the second evaluation value are weighted and summed. If the resulting comprehensive evaluation value is greater than the preset evaluation value, the target pregnant woman is determined to be placed in the target delivery bed, so as to control the delivery bed acquisition unit to acquire data from the target delivery bed.

[0010] Optionally, in the method according to the invention, determining each pre-marked array of markers corresponding to the reference compression shape includes: When it is determined that the target pregnant woman has historical examination data based on her identity information, the historical examination data is determined as reference examination data; otherwise, the retrieved medical record database is traversed based on the physical data determined by the identity information in order to determine the historical examination data of other pregnant women that match the physical data as reference examination data. Based on the reference inspection data, the compression shape of each delivery bed corresponding to different inspection times and with a comprehensive evaluation value greater than the preset evaluation value is determined. The center point of the delivery bed corresponding to the target delivery bed is used as the coverage center point to cover all the compression shapes of the delivery beds, thus obtaining a common coverage area. The shape of the corresponding public coverage area is determined as the reference pressure shape, and each delivery bed array cell located in the area contour of the corresponding public coverage area is determined as a contour array cell; Determine the center vectors from the coverage center point to the array center point of each corresponding contour array cell, and determine the array distribution curvature of the region contour corresponding to different contour array cells based on the center vectors. When the curvature of any array distribution along the region contour satisfies the curvature abrupt change condition, and / or the vector length of any corresponding center vector satisfies the extreme value condition, the corresponding contour array grid is determined as the identifier array grid.

[0011] Optionally, in the method according to the invention, the obtained view of the delivery bed is identified to determine the limb feature regions corresponding to different target limb features, indicating the target pregnant woman, including: Determine the image pixels that make up the view of the delivery bed, and identify the image pixels with different pixel values ​​from the reference pixels of the retrieved target delivery bed as human body pixels; Connect adjacent human body pixels and obtain the corresponding human body center point that indicates the human body recognition area of ​​the target pregnant woman; A bounding rectangle is formed that circumscribes the human body recognition area, and a longitudinal baseline is generated that passes through the center point of the human body and is parallel to the long axis extension direction of the corresponding bounding rectangle. Starting from each human pixel that makes up the human recognition contour of the corresponding human recognition area, generate morphological representation lines that intersect the vertical baseline perpendicularly, and obtain the morphological representation length of each morphological representation line. The difference between the morphological representation lines of adjacent human pixels is calculated based on the morphological representation length. Based on the obtained morphological representation differences, the limb feature regions corresponding to different target limb features are determined, including abdominal features and back features.

[0012] Optionally, in the method according to the invention, determining the limb feature regions corresponding to different target limb features, based on the obtained morphological representation differences, includes: All morphological representation differences are sorted along the human body recognition contour, and based on the obtained morphological representation sequence, human body pixels whose morphological representation differences jump from positive to negative values ​​and from negative to positive values ​​are respectively determined as the first inflection point and the second inflection point. The first inflection point with the largest corresponding morphological representation length is determined as the first extreme point, and the distance between each second inflection point and each longitudinal point on the corresponding longitudinal baseline of the first extreme point is determined. All second inflection points whose corresponding longitudinal point distance is less than the preset physiological interval are identified as candidate inflection points, and the candidate inflection point with the smallest corresponding morphological representation length is identified as the second extreme point. Based on the first extreme point, the limb feature regions corresponding to the abdominal features of the target pregnant woman are determined, and based on the second extreme point, the limb feature regions corresponding to the waist and back features of the target pregnant woman are determined.

[0013] Optionally, in the method according to the invention, determining the limb feature region corresponding to the abdominal features of the target pregnant woman based on a first extreme point, and determining the limb feature region corresponding to the waist and back features of the target pregnant woman based on a second extreme point, includes: Starting from the first extreme point, the system traverses along the human body recognition contour and determines the first inflection point, located on both sides of the first extreme point, as the abdominal boundary point. Based on human body contour recognition, the human body contour segment composed of abdominal boundary points is determined as the abdominal arc segment, and different abdominal boundary points are connected to form the limb feature area corresponding to the abdominal features of the target pregnant woman based on the obtained boundary connection line and abdominal arc segment. Starting from the second extreme point, the system traverses along the human body recognition contour and determines the first inflection point, which is located on both sides of the second extreme point, as the waist and back boundary point. Based on human body contour recognition, the human body contour segment composed of the waist and back boundary points is determined as the waist and back curvature segment. Different waist and back boundary points are connected to form the limb feature area corresponding to the waist and back features of the target pregnant woman based on the obtained boundary connection line and waist and back curvature segment.

[0014] Optionally, in the method according to the present invention, when a target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, the target limb feature is defined as an auxiliary limb feature, and the birthing bed array grids included in the target birthing bed that have an auxiliary association with the auxiliary limb feature are defined as auxiliary array grids based on the view of the birthing bed, including: When it is determined that any delivery bed array grid overlaps with any limb feature area, the delivery bed array grid is designated as the bearing array grid, and the pressure acquisition values ​​output by all pressure sensing units located under each bearing array grid are obtained. Based on all pressure acquisition values ​​corresponding to the same limb feature area, determine whether the target limb feature is in a preset auxiliary posture, and in response to any target limb feature being in a preset auxiliary posture, determine the target limb feature as an auxiliary limb feature; The response assistive limb features are waist and back features, and the load-bearing array grids whose corresponding pressure acquisition values ​​are less than the preset support threshold are determined as suspended array grids; Along the direction away from the corresponding boundary connection line of the lumbar and back curvature segment, all the birthing bed array cells that are continuously adjacent to each birthing bed array cell located in the lumbar and back curvature segment are identified as lumbar edge cells, and all suspended array cells and lumbar edge cells are identified as auxiliary array cells that have an auxiliary relationship with the auxiliary limb features. The response auxiliary limb feature is the abdominal feature. Along the direction away from the corresponding boundary connection line of the abdominal arc segment, the birthing bed array cell adjacent to each birthing bed array cell located in the abdominal arc segment is determined as the abdominal edge cell. The pressure acquisition value output by the pressure sensing unit located below each abdominal edge grid is obtained, and the abdominal edge grid with the corresponding pressure acquisition value less than the preset support threshold is determined as the auxiliary array grid with an auxiliary limb feature.

[0015] Optionally, in the method according to the present invention, determining whether the target limb feature is in a preset auxiliary posture based on all pressure acquisition values ​​corresponding to the same limb feature region includes: The response assist limb feature is the lumbar and back feature. The average value of all pressure collection values ​​corresponding to the same limb feature area is calculated to obtain the average force on the lumbar and back. When the average force on the lower back is less than the preset suspension threshold, the target limb is determined to be in a preset auxiliary posture. The response is that the auxiliary limb feature is the abdominal feature. The vertical baseline of the corresponding human body recognition area is retrieved, and based on the vertical baseline, each carrier array grid corresponding to the same limb feature area is divided into the first lateral array grid and the second lateral array grid. The average pressure values ​​of all corresponding first lateral array cells are averaged to obtain the first average force value, and the average pressure values ​​of all corresponding second lateral array cells are averaged to obtain the second average force value. If the force deviation between the first average force value and the second average force value is greater than a preset deflection threshold, it is determined that the target limb feature is in a preset auxiliary posture.

[0016] Optionally, in the method according to the invention, determining the bed board partition where the auxiliary array grid is located based on the target delivery bed includes: The target delivery bed includes adjustable bed boards arranged in a direction perpendicular to the longitudinal centerline, wherein each adjustable bed board includes a vertically arranged array of bed boards. Based on the characteristics of each auxiliary limb, the adjustment bed board where the auxiliary array grid furthest from the longitudinal center line is located is determined as the bed board partition.

[0017] Optionally, in the method according to the present invention, determining the angle adjustment value of the corresponding bed board partition, the auxiliary array grid, and the inflation value based on the characteristics of the assistive limb includes: The response assist limb features are the waist and back features. The difference between the retrieved baseline force value and the average force value of the waist and back is calculated. Based on the obtained force gap value, the retrieved preset angle mapping table is traversed to determine the angle adjustment value of the corresponding bed board zone and the force compensation amount of the corresponding angle adjustment value based on the traversal result. The difference between the stress gap value and the stress compensation amount is calculated, and the basic inflation value of the corresponding suspended array grid is determined based on the obtained remaining gap value. The lateral inflation value of the corresponding waist edge grid is determined based on the average force on the waist and back, and the expansion inflation value of the corresponding waist and back features is constructed based on the base inflation value and the lateral inflation value. The response assist limb features are abdominal features. Based on the force deviation value, the preset angle mapping table is traversed to determine the angle adjustment value of the corresponding bed board partition and the force cancellation amount of the corresponding angle adjustment value based on the traversal result. The difference between the force deviation value and the force offset is calculated, and the inflation value of the corresponding abdominal edge grid is determined based on the obtained residual imbalance value.

[0018] According to another aspect of the present invention, a personalized multifunctional integrated maternity bed management system is provided, comprising: The limb recognition module is configured to, when it is determined that the target pregnant woman is placed on the target delivery bed, control the delivery bed acquisition unit to acquire data from the target delivery bed, and identify the obtained view of the delivery bed to determine the limb feature regions that indicate the target pregnant woman and correspond to different target limb features. The auxiliary determination module is configured to determine the target limb feature as an auxiliary limb feature when the target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, and to determine the birthing bed array grids included in the target birthing bed that have an auxiliary relationship with the auxiliary limb feature as auxiliary array grids based on the view of the birthing bed. The numerical determination module is configured to determine the bed board partition where the auxiliary array grid is located based on the target delivery bed, and to determine the corresponding bed board partition, the angle adjustment value of the auxiliary array grid, and the inflation value based on the characteristics of the auxiliary limbs; The adjustment execution module is configured to control the bed board partitions to perform angle adjustments based on the angle adjustment value, and to control the auxiliary array grids to perform inflation adjustments based on the expansion inflation value.

[0019] According to the solution of the present invention, the present invention has the following beneficial effects: First, this invention acquires a view of the delivery bed by controlling the delivery bed acquisition unit and performs image recognition. Combined with real-time monitoring data from the pressure sensing unit, it determines the target limb characteristics and whether they are in a preset auxiliary posture. This achieves automatic triggering of delivery bed support adjustment and precise positioning of limb characteristic areas. There is no need for medical staff to manually judge and adjust parameters, which reduces the operational burden of clinical nursing and improves the automation level and adaptation efficiency of delivery bed management. Secondly, this invention adapts differentiated image extreme value screening and force state analysis logic to the abdominal and back features of pregnant women respectively. Based on the force deviation value or the average force value, it accurately locks the auxiliary array grid and bed board partition that have an auxiliary correlation with the auxiliary limb features. It can accurately identify the state of force imbalance and insufficient support in complex and ever-changing lying postures, avoiding the omissions and misjudgments caused by uniform judgment rules, and improving the accuracy and individualization of posture judgment. Finally, when performing support adjustment, this invention calculates the angle adjustment value of the corresponding bed board partition and the expansion inflation value of the auxiliary array grid. It achieves large-scale posture correction through overall angle adjustment of the bed board partition and achieves fine support filling through partial inflation of the auxiliary array grid. This dual-layer linkage compensation method ensures that the support strength and position are completely matched to the actual body shape of the pregnant woman. It avoids local support gaps caused by single angle adjustment, effectively corrects force imbalance, and significantly increases the safety and comfort of pregnant women during labor and examination. Attached Figure Description

[0020] Figure 1 A flowchart of a method for managing pregnant women using an individualized multifunctional integrated delivery bed according to an embodiment of the present invention is shown; Figure 2 A schematic diagram of the corresponding target delivery bed in this embodiment is shown; Figure 3 A structural block diagram of an individualized multifunctional integrated delivery bed pregnant woman management system according to another embodiment of the present invention is shown. Detailed Implementation

[0021] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0022] To address the problems existing in the prior art, the inventors proposed the solution of this invention. One embodiment of this invention provides a method for managing pregnant women using a personalized, multifunctional integrated delivery bed. This method can be executed on a computing device, which can be understood as a terminal with data processing capabilities, such as a computer.

[0023] Figure 1 A flowchart of a method for managing pregnant women using an individualized multifunctional integrated delivery bed according to an embodiment of the present invention is shown, such as... Figure 1 As shown, this individualized, multifunctional integrated delivery bed pregnant woman management method begins with step S1, which includes the following: Once the target pregnant woman is identified and placed on the target delivery bed, the delivery bed acquisition unit is controlled to acquire data from the target delivery bed and to identify the view of the delivery bed to determine the limb feature regions that indicate the target pregnant woman and correspond to different target limb features.

[0024] For example, in this embodiment, the overall management process can be executed based on a server that is paired with the target delivery bed. It can be explained that the target delivery bed is an integrated maternity bed with zoned angle adjustment and point-inflatable support functions. It can perform individualized support adjustments according to the pregnant woman's body shape and lying posture. The target pregnant woman is the person who is about to give birth or undergoing obstetric examinations corresponding to this delivery bed management service. When the server determines that the target pregnant woman is fully lying down and supported on the bed surface of the target delivery bed, it automatically triggers the subsequent data collection and identification process without the need for manual operation.

[0025] Furthermore, the delivery bed acquisition unit is an image acquisition module that is pre-installed directly above the target delivery bed with the acquisition view perpendicular to the bed surface. Its acquisition range can completely cover the entire bed surface area of ​​the target delivery bed. The server can obtain a complete view of the delivery bed surface by controlling the delivery bed acquisition unit to start the acquisition operation.

[0026] After obtaining the view of the delivery bed, the server performs image recognition processing on the view, extracts the pregnant woman's body area from the bed background, and further divides it into independent areas corresponding to different target limb features based on human morphological characteristics, namely, each limb feature area. The target limb features are the key body parts that need to be supported and assisted during the labor process.

[0027] It can be noted that, in this embodiment, through the automatic collection and intelligent recognition in this step, the area corresponding to different key limb features on the pregnant woman can be quickly and accurately located, and the spatial correspondence between the pregnant woman's body parts and the support structure of the delivery bed can be established. There is no need for medical staff to manually mark and adjust the benchmark, which provides an accurate positional basis for the subsequent individualized adaptive adjustment of the delivery bed, effectively improving the automation level and adaptation efficiency of delivery bed management, and avoiding the operational errors and time losses caused by manual positioning.

[0028] Furthermore, in this embodiment, the aforementioned "when it is determined that the target pregnant woman is placed on the target delivery bed, control the delivery bed acquisition unit to acquire data from the target delivery bed" may also include the following steps: When it is determined that the pressure acquisition value output by any pressure sensing unit located under each birthing bed array cell in the target birthing bed is greater than the preset pressure threshold, the birthing bed array cell corresponding to the pressure acquisition value is determined to have pressure-bearing properties. The birthing bed compression shape, which is composed of all birthing bed array grids with compression properties, is determined, and the birthing bed compression shape is compared with the retrieved reference compression shape to determine a first evaluation value based on the obtained shape similarity. Each pre-marked array cell corresponding to the reference pressure shape is determined, and in response to any delivery bed array cell with pressure attribute whose corresponding pressure position is the same as the marking position of the corresponding mark array cell, the initial number of marks is accumulated to determine a second evaluation value based on the obtained current number of marks. The first evaluation value and the second evaluation value are weighted and summed. If the resulting comprehensive evaluation value is greater than the preset evaluation value, the target pregnant woman is determined to be placed in the target delivery bed, so as to control the delivery bed acquisition unit to acquire data from the target delivery bed.

[0029] For example, in this embodiment, the determination of whether the target pregnant woman is placed on the delivery bed can be specifically implemented based on the following method steps: First, the server collects the output data of the pressure sensing units under each grid of the target delivery bed in real time. When the pressure value collected by any pressure sensing unit is greater than the preset pressure threshold, the grid is marked as having pressure attributes. It can be noted that each delivery bed grid can be pre-deployed with an independent pressure sensing unit to sense the pressure state of the corresponding bed surface. The preset pressure threshold is a pre-set pressure threshold used to distinguish between no load and effective pressure, which can eliminate false triggering caused by items such as sheets and lightweight medical devices. Here, through real-time monitoring of the pressure sensing units, the pressure change of the bed surface can be sensed immediately while the pregnant woman is lying down, without the need to continuously start the image acquisition unit for image recognition, effectively reducing the server's operating power consumption and improving the response speed of load detection. The preset pressure threshold can be set based on actual needs, and this embodiment does not impose specific limitations on it. Next, the server can aggregate all the delivery bed array grids with pressure attributes, splice them to form a complete delivery bed pressure shape, and then compare the real-time pressure shape with the baseline pressure shape stored on the server. Based on the shape similarity between the two, the first evaluation value is calculated. It can be noted that the baseline pressure shape corresponds to the standard pressure contour formed when the pregnant woman is lying normally on the delivery bed. The higher the shape similarity, the closer the current pressure form is to the characteristics of a human lying down. Here, by comparing the contour of the overall shape, it can be preliminarily determined from a global perspective whether the current pressure on the bed surface is caused by the pregnant woman lying down, effectively eliminating the interference of non-human lying scenarios such as the placement of heavy objects or medical staff leaning against the edge of the bed, and ensuring the accuracy of the load-bearing determination from an overall perspective. Then, the server can retrieve each marker array cell corresponding to the baseline pressure shape and check one by one whether the position of the array cell with the current pressure attribute coincides with the position of the marker array cell. For each matching position, the number of markers is accumulated. Finally, the second evaluation value is calculated based on the accumulated number of markers. It can be noted that the marker array cell is a key point in the baseline pressure shape with significant human morphological features. It corresponds to the characteristic inflection point, extreme point and other landmark position of the human contour. It belongs to the typical characteristic point of the human body when lying down under pressure. Here, through the secondary matching and verification of key marker points, the morphological features of the pressure object can be further verified from the local feature dimension, making up for the misjudgment that may occur by simple contour comparison, further improving the accuracy of load determination, and avoiding detection deviation caused by items similar to the human body contour. Finally, the server can calculate a comprehensive evaluation value by weighting the first evaluation value and the second evaluation value according to preset weights. When the comprehensive evaluation value is greater than the preset evaluation value, it is determined that the target pregnant woman has been placed in the target delivery bed, and the delivery bed acquisition unit is immediately triggered to start the image acquisition operation. It can be noted that the preset evaluation value is a judgment threshold value calibrated with clinical data. The weighted summation method can take into account the weight of overall contour features and local identification features, and realize a comprehensive judgment in two dimensions. Here, through this dual-verification judgment mechanism, the accuracy of the placement detection is guaranteed, the waste of computing power caused by invalid image acquisition is avoided, and the subsequent recognition process is automatically triggered after the pregnant woman is in place, without the need for medical staff to manually start the acquisition operation, realizing the fully automatic triggering of delivery bed management, effectively improving the convenience of clinical use and the server operating efficiency. The preset evaluation value can be set based on actual needs, and this embodiment does not make specific limitations on it.

[0030] For example, in this embodiment, the determination of the evaluation value can be based specifically on the following: First, regarding the calculation of the first evaluation value, when it is determined that the pressure acquisition value output by any pressure sensing unit located under each birthing bed array grid in the target birthing bed is greater than the preset pressure threshold, the server determines the birthing bed array grid corresponding to the pressure acquisition value as having pressure-bearing attributes. Then, the server determines the birthing bed pressure-bearing shape composed of all birthing bed array grids with pressure-bearing attributes, and compares the birthing bed pressure-bearing shape with the retrieved reference pressure-bearing shape. Assuming that the total area of ​​the reference pressure-bearing shape covers 100 array grids, and the area where the currently extracted birthing bed pressure-bearing shape overlaps with the reference pressure-bearing shape is 85 array grids, then the first evaluation value is determined to be 0.85 based on the obtained shape similarity. It can be explained that the first evaluation value quantifies the degree of fit between the current actual pressure-bearing contour and the standard human lying contour from a macroscopic perspective. Secondly, regarding the calculation of the second evaluation value, the server further determines each pre-marked array grid corresponding to the baseline compression shape, such as 10 marked array grids corresponding to the extreme value points of features like the human shoulder and hip. During the verification process, if the compression position corresponding to any delivery bed array grid with compression attributes is the same as the marking position of the corresponding marked array grid, the server accumulates the initial number of markings. For example, if the initial value is 0, and assuming that 9 of the 10 marked array grids coincide with the compression position of the current delivery bed array grid with compression attributes, the server determines the second evaluation value to be 0.9 (i.e., 9 / 10) based on the current number of markings (i.e., 9) after accumulation. It can be said that the second evaluation value verifies the accuracy of the alignment of the compression bone points from the perspective of microscopic core feature points. Finally, the server can perform a weighted summation of the first evaluation value (0.85) and the second evaluation value (0.9). If the weight of the first evaluation value is set to 0.4 and the weight of the second evaluation value is set to 0.6, the resulting comprehensive evaluation value is 0.88 (0.85×0.4+0.9×0.6=0.88). Furthermore, if the comprehensive evaluation value (0.88) obtained in the response is greater than the preset evaluation value (e.g., set to 0.80), the server can determine that the target pregnant woman is on the target delivery bed and control the delivery bed acquisition unit to acquire data from the target delivery bed.

[0031] Furthermore, in this embodiment, the aforementioned "determining each pre-marked array of markers corresponding to the reference compression shape" may further include the following steps: When it is determined that the target pregnant woman has historical examination data based on her identity information, the historical examination data is determined as reference examination data; otherwise, the retrieved medical record database is traversed based on the physical data determined by the identity information in order to determine the historical examination data of other pregnant women that match the physical data as reference examination data. Based on the reference inspection data, the compression shape of each delivery bed corresponding to different inspection times and with a comprehensive evaluation value greater than the preset evaluation value is determined. The center point of the delivery bed corresponding to the target delivery bed is used as the coverage center point to cover all the compression shapes of the delivery beds, thus obtaining a common coverage area. The shape of the corresponding public coverage area is determined as the reference pressure shape, and each delivery bed array cell located in the area contour of the corresponding public coverage area is determined as a contour array cell; Determine the center vectors from the coverage center point to the array center point corresponding to each contour array cell, and determine the array distribution curvature of the region contour corresponding to different contour array cells based on the center vectors. When the curvature of any array distribution along the region contour satisfies the curvature abrupt change condition, and / or the vector length of any corresponding center vector satisfies the extreme value condition, the corresponding contour array grid is determined as the identifier array grid.

[0032] For example, in this embodiment, the construction of the reference compression shape and the determination of the identification array grid can be specifically implemented based on the following method steps: First, the server can retrieve the identity information of the target pregnant woman to verify whether she has any historical examination data from previous examinations or pregnancies in this delivery bed. If historical data exists, it can be used as reference data. If no historical data exists, it will search the pre-stored medical record database for physical data such as height, weight, and gestational age associated with the identity information, and filter out the historical examination data of other pregnant women with similar physical data as reference data. It can be noted that prioritizing the use of the pregnant woman's own historical data as a reference ensures that the baseline shape is more in line with the individual physical characteristics of the pregnant woman, ensuring the accuracy of subsequent load assessment. For pregnant women using the delivery bed for the first time, a reference baseline is generated by matching group data with similar physical characteristics, which ensures that there is still a reliable basis for judgment in the absence of historical data. This balances individualized adaptation and universality, avoiding the judgment bias caused by a single general baseline. Next, all valid birthing bed compression shapes with comprehensive evaluation values ​​greater than preset evaluation values ​​can be screened from the reference inspection data. These are the compression contour data that were determined to be in an effective bearing state in previous tests. Then, using the center point of the target birthing bed as the unified coverage center point, all the screened birthing bed compression shapes are superimposed and covered. The overlapping area covered by all shapes is extracted to obtain the common coverage area. It can be noted that the common coverage area is the common compression range that appears in multiple effective bearing scenarios. It represents the core area that will inevitably be compressed when the pregnant woman of the corresponding body type lies down. By superimposing and intersecting multiple sets of historical data, the random errors caused by lying position deviation and posture fluctuation in a single test can be eliminated, making the benchmark shape more representative and stable, and avoiding deviations in subsequent bearing judgments due to the randomness of single data. Then, the server can determine the shape of the area corresponding to the public coverage area as the final reference pressure shape, and at the same time extract all the corresponding birthing bed array grids on the area contour of the public coverage area and mark them as each contour array grid. It can be noted that the outer boundary points of the contour array grid corresponding to the reference pressure shape are the basic carriers for subsequent extraction of morphological features and screening of identification points. By locking the array grid units on the contour, the subsequent feature calculation can be limited to the contour boundary range, without having to perform full calculation on all array grids of the entire pressure area, which greatly reduces the amount of computation for feature extraction and improves the execution efficiency of reference shape construction and identification point screening. Then, the server generates center vectors pointing to the center point of each contour array grid, starting from the center point of the delivery bed. Based on the changes in direction and length of each center vector, the array distribution curvature of the region contour at different contour array grid positions is calculated. It can be seen that the center vector directly reflects the orientation and distance of the contour point relative to the center of the delivery bed, while the array distribution curvature reflects the degree of bending of the contour line at the corresponding point. Through vectorization and curvature calculation, the original contour shape can be transformed into a quantifiable numerical feature, providing a clear numerical basis for the subsequent selection of marker points, without relying on manual annotation of feature points, thus realizing automatic and intelligent selection of marker array grids. Finally, the server can verify the curvature of the array distribution and the length of the center vector point by point along the region contour. When the curvature corresponding to a certain contour array cell meets the preset curvature change condition, that is, the contour line shows obvious bending and significant shape change at that point, and / or the length of the center vector corresponding to that point meets the extreme value condition, that is, the point is the extreme position farthest or closest to the center on the contour, the contour array cell is determined as the marker array cell. It can be explained that the curvature change point and the length extreme point are the most recognizable morphological feature points on the human body pressure contour, corresponding to the contour inflection points of the human body such as the shoulder, waist, and hip, and have a very strong morphological representativeness. By selecting marker points through the dual conditions of curvature and vector length, it can be ensured that the selected marker array cells have sufficient distinguishability, which can effectively improve the recognition accuracy in the subsequent secondary verification of load determination, and at the same time avoid too many invalid points participating in the calculation, ensuring the efficiency of the determination process.

[0033] The curvature abrupt change condition and extreme value condition are explained below in conjunction with practical application scenarios: Regarding curvature abrupt change conditions, for example, the server calculates the curvature distribution of the array along the contour array grid of the public coverage area. The server can preset the curvature abrupt change threshold as a curvature change amplitude > 0.3. Among them, at two contour array grid positions corresponding to the two sides of the pregnant woman's hip, the contour line bends rapidly inward from a gentle arc, and the curvature difference between adjacent points reaches 0.42, which exceeds the preset abrupt change threshold. At this time, it can be determined that the curvature abrupt change condition is met, which means that these two points correspond to the outward expansion inflection point of the human hip bone, which are feature points with extremely high recognizability of the human body's pressure contour, and can be directly included in the identification array grid. On the other hand, the contour array grid located in the gentle transition area of ​​the thigh has a curvature difference between adjacent points of only 0.08, and the curve has no obvious bend, which does not meet the abrupt change condition, and can be removed. Regarding extreme value conditions, for example, the center vector is the line segment pointing from the center point of the delivery bed to the center of the contour array grid, and the vector length represents the straight-line distance from the point to the center of the bed. The server determines that both the maximum and minimum values ​​of the vector length satisfy the extreme value condition. In this contour, the corresponding length of the center vector of the contour array grid corresponding to the left and right acromions of the pregnant woman can reach 2.1m, which is the maximum length of the entire contour. The corresponding length of the center vector of the array grid at the lumbosacral depression is only 0.75m, which is the minimum length of the entire contour. Both points satisfy the extreme value condition and can be marked as marker array grids. In addition, since the vector length of the points on both sides of the abdomen that are gradually transitioning is between 0.75m and 2.1m and has no extreme value characteristics, they are not included in the selection of marker points.

[0034] Furthermore, in this embodiment, the aforementioned "identifying the obtained view of the delivery bed and determining the limb feature regions corresponding to different target limb features that indicate the target pregnant woman" may further include the following steps: Determine the image pixels that make up the view of the delivery bed, and identify the image pixels with different pixel values ​​from the reference pixels of the retrieved target delivery bed as human body pixels; Connect adjacent human body pixels and obtain the corresponding human body center point that indicates the human body recognition area of ​​the target pregnant woman; A bounding rectangle is formed that circumscribes the human body recognition area, and a longitudinal baseline is generated that passes through the center point of the human body and is parallel to the long axis extension direction of the corresponding bounding rectangle. Starting from each human pixel that makes up the human recognition contour of the corresponding human recognition area, generate morphological representation lines that intersect the vertical baseline perpendicularly, and obtain the morphological representation length of each morphological representation line. The difference between the morphological representation lines of adjacent human pixels is calculated based on the morphological representation length. Based on the obtained morphological representation differences, the limb feature regions corresponding to different target limb features are determined, including abdominal features and back features.

[0035] For example, in this embodiment, the identification and segmentation of limb feature regions can be specifically implemented based on the following method steps: First, the server extracts all image pixels contained in the view of the delivery bed and retrieves the pre-stored baseline pixels corresponding to the empty state of the target delivery bed. The pixel values ​​of the two types of pixels are compared one by one, and the image pixels with different pixel values ​​are identified as human pixels. It can be explained that the baseline pixels correspond to the standard pixel features of each position on the bed surface when the delivery bed is empty. The pixel features of the bed surface background are stable and uniform. However, when the pregnant woman lies down, she will cover the corresponding area of ​​the bed surface, which will cause the pixel values ​​to change. Here, human pixels are extracted by pixel difference. There is no need to deploy a complex human semantic segmentation model. The preliminary extraction of the human body region can be completed by basic pixel comparison. The calculation logic is simple and efficient, and the separation of the human body from the bed surface background can be completed quickly, effectively reducing the computing power consumption of image recognition and improving the recognition response speed. Next, the server connects and stitches together all spatially adjacent human body pixels to obtain a complete human body recognition area corresponding to the target pregnant woman. At the same time, it calculates the center point of the human body in this recognition area. It can be noted that the center point of the human body is the geometric center of the human body recognition area and serves as the core reference point for all subsequent morphological feature calculations. Here, by locking the complete human body recognition area first and then determining the center point, the subsequent feature calculations can be limited to the human body area without having to perform a full-screen analysis of the entire delivery bed view. This further narrows the effective range of recognition calculations, improves the overall recognition efficiency, and also provides a unified positioning anchor point for the subsequent generation of baselines, avoiding server-side deviations in subsequent feature positioning caused by the pregnant woman's lying position shift. Then, the server can generate an outer rectangular outline that completely encompasses the human body recognition area. Starting from the center point of the human body, a longitudinal baseline parallel to the long axis of the outer rectangle is generated. It can be noted that the long axis of the outer rectangle is basically consistent with the extension direction of the pregnant woman's torso when she is lying down. The longitudinal baseline is laid out along the longitudinal extension direction of the human torso, serving as a unified reference axis for subsequent morphological and dimensional measurements. Here, by locking the extension direction of the torso with the outer rectangle and then generating the longitudinal baseline, it can automatically adapt to different lying angles of the pregnant woman. Regardless of whether the pregnant woman is lying along the central axis of the delivery bed, a baseline matching the torso direction can be generated, ensuring that the reference for subsequent morphological measurements is always aligned with the human torso, avoiding feature recognition errors caused by lying angle deviations, and improving the scene adaptability and recognition accuracy of the recognition scheme. Then, the server can generate morphological representation lines perpendicular to the vertical baseline along the outer contour of the human body recognition area, starting from each human pixel point on the contour. Simultaneously, the length of each morphological representation line is measured, which is the morphological representation length. It can be explained that the morphological representation length essentially corresponds to the horizontal half-width dimension of the human body contour at that vertical position, intuitively reflecting the thickness of the corresponding torso position. Here, by generating morphological representation lines point by point and measuring the length, the original two-dimensional contour shape can be transformed into a one-dimensional length sequence distributed along the vertical direction. The complex contour shape recognition is transformed into a simple length value change analysis, which greatly reduces the computational complexity of subsequent limb region segmentation and makes the trend of contour feature change more intuitive and quantifiable. Finally, the server calculates the difference in morphological representation lengths corresponding to adjacent human body pixels, obtaining the morphological representation differences distributed along the human body contour. Based on the changing patterns of these length differences, it then divides the limb feature regions corresponding to different target limb features. These target limb features include at least abdominal and back features. It can be noted that the positive and negative changes in morphological representation differences directly reflect the increasing or decreasing trend of the human body contour width. When the difference changes from positive to negative, it represents a change in the contour width from continuous widening to continuous narrowing, corresponding to the dividing inflection point of the limb parts. Here, by analyzing the changing patterns of the differences, the dividing positions of different limb parts can be accurately located, and independent limb feature regions can be divided. This does not rely on a pre-trained human keypoint recognition model; the region division can be completed solely through the morphological change patterns of the contour itself. It is suitable for pregnant women of different gestational weeks and body types, with strong recognition versatility. At the same time, the calculation process is simple and efficient, and it can quickly output accurate limb region positioning results, providing a reliable location basis for subsequent orientation adjustment of the delivery bed.

[0036] Furthermore, in this embodiment, the aforementioned "determining the limb feature regions corresponding to different target limb features included in the target pregnant woman based on the obtained morphological representation differences" may further include the following steps: All morphological representation differences are sorted along the human body recognition contour, and based on the obtained morphological representation sequence, human body pixels whose morphological representation differences jump from positive to negative values ​​and from negative to positive values ​​are respectively determined as the first inflection point and the second inflection point. The first inflection point with the largest corresponding morphological representation length is determined as the first extreme point, and the distance between each second inflection point and each longitudinal point on the corresponding longitudinal baseline of the first extreme point is determined. All second inflection points whose corresponding longitudinal point distance is less than the preset physiological interval are identified as candidate inflection points, and the candidate inflection point with the smallest corresponding morphological representation length is identified as the second extreme point. Based on the first extreme point, the limb feature regions corresponding to the abdominal features of the target pregnant woman are determined, and based on the second extreme point, the limb feature regions corresponding to the waist and back features of the target pregnant woman are determined.

[0037] For example, in this embodiment, the process of accurately locating core feature points based on morphological representation differences and then dividing limb feature regions can be implemented based on the following method steps: First, the server sequentially arranges the morphological representation differences corresponding to all points along the longitudinal extension direction of the human body recognition contour, obtaining a continuous morphological representation sequence. Then, it verifies the changes in the differences in the sequence point by point. Specifically, when the morphological representation difference jumps from a positive value to a negative value, the corresponding human pixel at the jump position can be understood as the turning point where the contour width changes from continuously widening to continuously narrowing. Therefore, this can be identified as the first inflection point. Similarly, when the morphological representation difference jumps from a negative value to a positive value, the corresponding human pixel at the jump position can be understood as the turning point where the contour width changes from continuously narrowing to continuously widening. Therefore, it can be identified as the second inflection point. It can be explained that the positive and negative values ​​of the morphological representation difference directly correspond to the changing trend of the horizontal width of the human body contour. A positive difference represents the width continuously increasing along the vertical direction, and a negative difference represents the width continuously decreasing along the vertical direction. The position where the difference jumps to positive or negative is the turning point of the contour shape, which corresponds to the boundary position of different torso segments of the human body. Here, by sorting along the contour and identifying the difference jump point, the turning position of all contour shapes can be automatically extracted without manual annotation of feature points or reliance on preset human body templates. The calculation logic is simple and intuitive, and can quickly lock the morphological feature boundary of the waist and abdomen area. Next, the server selects the inflection point with the largest corresponding morphological representation length from all the first inflection points and marks it as the first extreme point. At the same time, using the vertical baseline as the position reference, it calculates the vertical distance between each second inflection point and the first extreme point on the vertical baseline. It can be explained that the first inflection point with the largest morphological representation length corresponds to the widest position of the human torso. For pregnant women, this is the peak position of the most prominent abdominal circumference, which is the morphological marker point of the abdominal region. By using this first extreme point as the vertical baseline anchor point, a core reference for the division of the waist and abdomen region can be established, which conforms to the physiological structural distribution law of the human torso. Here, by locking the abdominal extreme point first and then calculating the vertical distance, the subsequent feature point selection can be limited to the vertical range corresponding to the waist and abdomen, avoiding interference from inflection points of other parts such as the shoulders, hips, and lower limbs, effectively narrowing the selection range of effective feature points and improving the accuracy of positioning. Then, the server compares the longitudinal distance of all second inflection points with the preset physiological distance, selecting second inflection points whose longitudinal distance is less than the preset physiological distance as candidate inflection points. From all candidate inflection points, it selects the inflection point with the smallest corresponding morphological representation length and determines it as the second extreme point. It can be noted that the preset physiological distance is a longitudinal span threshold for the core region of the human waist and abdomen, set based on clinical ergonomic data, used to limit the effective distribution range of waist and abdomen feature points. The candidate inflection point with the smallest morphological representation length corresponds to the position where the trunk contour width is narrowest, i.e. The lumbar concave area of ​​the human lower back is the most core morphological landmark in the lower back region. Here, by using a preset physiological distance for initial screening, invalid inflection points of other segments of the torso can be eliminated, ensuring that the selected candidate inflection points are all within the physiological range corresponding to the lower back. Then, by using minimum length screening to lock the feature points of the lower back, the core position of the lower back can be accurately located, effectively avoiding misjudgment of feature points due to differences in the body shape and lying posture of pregnant women, and greatly improving the adaptability and accuracy of feature point positioning. The preset physiological distance can be set based on actual needs, and this embodiment does not impose specific limitations on it. Finally, the server uses the first extreme point as the core anchor point to delineate the limb feature region corresponding to the abdominal features, and the second extreme point as the core anchor point to delineate the limb feature region corresponding to the waist and back features. It can be explained that the two extreme points correspond to the widest part of the abdomen and the narrowest part of the waist and back, respectively, which are the most recognizable morphological features of the waist and abdomen. Using these as a benchmark to divide the limb feature regions can ensure that the region division accurately corresponds to the actual physiological structure of the pregnant woman. Here, the region division method through pure morphological numerical analysis does not rely on a pre-trained human body part recognition model, nor is it affected by the pregnant woman's gestational age, weight, or lying angle. It has strong adaptability and high computational efficiency, and can quickly output accurate limb region positioning results, providing a reliable position benchmark for subsequent directional support adjustment of the delivery bed.

[0038] Furthermore, in this embodiment, the aforementioned "determining the limb feature region corresponding to the abdominal features of the target pregnant woman based on the first extreme point, and determining the limb feature region corresponding to the waist and back features of the target pregnant woman based on the second extreme point" may further include the following steps: Starting from the first extreme point, the system traverses along the human body recognition contour and determines the first inflection point, located on both sides of the first extreme point, as the abdominal boundary point. Based on human body contour recognition, the human body contour segment composed of abdominal boundary points is determined as the abdominal arc segment, and different abdominal boundary points are connected to form the limb feature area corresponding to the abdominal features of the target pregnant woman based on the obtained boundary connection line and abdominal arc segment. Starting from the second extreme point, the system traverses along the human body recognition contour and determines the first inflection point, which is located on both sides of the second extreme point, as the waist and back boundary point. Based on human body contour recognition, the human body contour segment composed of the waist and back boundary points is determined as the waist and back curvature segment. Different waist and back boundary points are connected to form the limb feature area corresponding to the waist and back features of the target pregnant woman based on the obtained boundary connection line and waist and back curvature segment.

[0039] For example, in this embodiment, the process of accurately delineating the corresponding limb feature regions of the abdomen and lower back using extreme points as anchor points can be implemented based on the following method steps: First, the server uses the already located first extreme point as the starting point and simultaneously traverses along the human body recognition contour in two directions: head-side and foot-side. The second inflection point, located on both sides of the first extreme point, is determined as the abdominal boundary point. It can be explained that the first extreme point corresponds to the most prominent peak of the abdominal circumference, which is the widest part of the abdominal region. The second inflection point represents the turning point where the contour width changes from narrowing to widening. The first turning point encountered when extending from the peak of the abdominal width to both sides corresponds to the natural transition boundary between the protruding abdominal area and other segments of the torso. Here, in this embodiment, the extreme point is used as the center to traverse to both sides and lock the first second inflection point as the boundary. This can accurately fit the actual contour range of the pregnant woman's protruding abdomen, without including too much of the area of ​​other parts of the torso, and without omitting the effective support range of the protruding abdomen, ensuring the accuracy of the abdominal region delineation. At the same time, the traversal logic is simple and direct, without the need for complex region growth calculations, and can quickly lock the boundary point, improving the execution efficiency of region division. Next, the server extracts the contour line segment between two abdominal boundary points on the human body recognition contour, marks it as the abdominal arc segment, and then connects the two abdominal boundary points with a straight line to form a boundary connection line. The closed area enclosed by the abdominal arc segment and the boundary connection line is the limb feature area corresponding to the abdominal features. It can be explained that the abdominal arc segment corresponds to the outer contour line of the abdominal bulge, completely restoring the outward convex shape of the abdomen, while the boundary connection line corresponds to the dividing chord line of the abdominal area near the inner side of the torso. The closed area enclosed by the two completely covers the entire bulging range of the pregnant woman's abdomen. Here, the region is constructed by using the closure method of the contour arc segment and the dividing line, which not only ensures that the region range completely matches the actual shape of the abdomen and accurately matches the range of subsequent abdominal support of the delivery bed, but also avoids full-pixel-level region segmentation calculation, which greatly reduces the computing power consumption of region delineation. At the same time, the output region boundary is clear and can be mapped and matched with the spatial position of the delivery bed array grid, providing an accurate regional benchmark for subsequent support point positioning. Then, the server uses the second extreme point after positioning as the starting point and simultaneously traverses along the human body recognition contour in both the head and foot directions. The first inflection point encountered during the first traversal, located on both sides of the second extreme point, is determined as the lumbar and back boundary point. It can be explained that the second extreme point corresponds to the narrowest position of the lumbar concavity in the lumbar and back, which is the narrowest point in the lateral direction of the lumbar and back region. The first inflection point represents the turning point where the contour width changes from widening to narrowing. The first turning point encountered when extending from the narrowest point of the lumbar concavity to both sides corresponds to the natural transition boundary between the lumbar and back, and the lumbosacral region. Here, using this as the boundary of the lumbar and back region can accurately lock the concavity range of the lumbar and back, ensuring that subsequent support adjustments can be accurately applied to the core area of ​​the lumbar region that is not in the air. This avoids unnecessary bed adjustments due to an excessively large support range, and also prevents the range from being too small to completely cover the lumbar support gap, effectively ensuring the targeting and effectiveness of subsequent lumbar and back support adjustments. Finally, the server extracts the contour line segment between two waist and back boundary points on the human body recognition contour, marking it as the waist and back curvature segment. Then, the two waist and back boundary points are connected by a straight line to form a boundary connection line. The closed area enclosed by the waist and back curvature segment and the boundary connection line is the limb feature area corresponding to the waist and back features. It can be explained that the waist and back curvature segment corresponds to the inner concave contour line of the waist and back, completely presenting the concave shape of the waist. The boundary connection line corresponds to the dividing chord line of the waist and back area near the outer side of the torso. The closed area enclosed by the two completely covers the core concave range of the pregnant woman's waist and back. Here, the waist and back area is delineated by the same enclosing logic as the abdominal area, ensuring the consistency of the division rules of the two limb feature areas. This facilitates the subsequent unified mapping to the delivery bed array grid for support adjustment. At the same time, the area range completely matches the physiological shape of the pregnant woman's waist and back, and can adapt to the differences in waist concavity of pregnant women with different body shapes and different gestational weeks, providing a precise positional basis for the subsequent individualized waist support adjustment of the delivery bed.

[0040] Step S2 includes the following: When a target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, the target limb feature is identified as an auxiliary limb feature, and the birthing bed array grids included in the target birthing bed that have an auxiliary relationship with the auxiliary limb feature are identified as auxiliary array grids based on the view of the birthing bed.

[0041] For example, in this embodiment, after completing the identification and positioning of all limb feature regions, the server can perform posture determination processing for each identified limb feature region to determine whether the lying posture of the corresponding target limb feature on the delivery bed belongs to a preset auxiliary posture that requires directional auxiliary support from the delivery bed.

[0042] It can be explained that the preset auxiliary posture is a lying posture that is set in advance based on obstetric clinical nursing requirements and ergonomic standards. This posture may lead to insufficient support for the corresponding parts of the pregnant woman, imbalance of force, or discomfort and safety hazards. When the corresponding posture is detected, the delivery bed needs to perform corresponding adaptive auxiliary adjustments to improve the pregnant woman's lying support status.

[0043] Furthermore, when any target limb feature is determined to be in a preset auxiliary posture, the server marks the target limb feature as the auxiliary limb feature that needs to be supported this time, clarifying the core target and adjustment goal of the subsequent delivery bed adjustment, thereby locking the adaptation direction of this delivery bed adjustment. On this basis, the server can combine the previously collected delivery bed view to select the array grid unit with the auxiliary limb feature in all the delivery bed array grids pre-divided in the target delivery bed, and determine it as the auxiliary array grid.

[0044] It can be explained that the birthing bed array grid is the smallest functional unit that can be independently adjusted for inflation support, which is divided into the target birthing bed surface according to preset specifications. The auxiliary correlation relationship means that the corresponding array grid is located within the orthographic projection support range of the auxiliary limb feature, or within the effective area of ​​support compensation. Adjusting this part of the array grid can directly support, support or correct the posture of the auxiliary limb feature.

[0045] In this embodiment, by automatically determining the posture and adjusting the array grid for precise positioning, the limbs on the pregnant woman's body that require auxiliary support can be automatically identified, and the smallest unit on the delivery bed that needs to be adjusted can be accurately locked. This eliminates the need for medical staff to manually determine the support points and adjustment range, ensuring the individualized adaptation accuracy of the support assistance, reducing manual operation steps, improving the response speed and execution efficiency of the delivery bed's auxiliary adjustment, and avoiding the problem of inconsistent support effects caused by differences in human experience.

[0046] Furthermore, in this embodiment, the aforementioned "when a target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, the target limb feature is identified as an auxiliary limb feature, and the birthing bed array grids included in the target birthing bed that have an auxiliary association with the auxiliary limb feature are identified as auxiliary array grids based on the birthing bed view" may further include the following steps: When it is determined that any delivery bed array grid overlaps with any limb feature area, the delivery bed array grid is designated as the bearing array grid, and the pressure acquisition values ​​output by all pressure sensing units located under each bearing array grid are obtained. Based on all pressure acquisition values ​​corresponding to the same limb feature area, determine whether the target limb feature is in a preset auxiliary posture, and in response to any target limb feature being in a preset auxiliary posture, determine the target limb feature as an auxiliary limb feature; The response assistive limb features are waist and back features, and the load-bearing array grids whose corresponding pressure acquisition values ​​are less than the preset support threshold are determined as suspended array grids; Along the direction away from the corresponding boundary connection line of the lumbar and back curvature segment, all the birthing bed array cells that are continuously adjacent to each birthing bed array cell located in the lumbar and back curvature segment are identified as lumbar edge cells, and all suspended array cells and lumbar edge cells are identified as auxiliary array cells that have an auxiliary relationship with the auxiliary limb features. The response auxiliary limb feature is the abdominal feature. Along the direction away from the corresponding boundary connection line of the abdominal arc segment, the birthing bed array cell adjacent to each birthing bed array cell located in the abdominal arc segment is determined as the abdominal edge cell. The pressure acquisition value output by the pressure sensing unit located below each abdominal edge grid is obtained, and the abdominal edge grid with the corresponding pressure acquisition value less than the preset support threshold is determined as the auxiliary array grid with an auxiliary limb feature.

[0047] For example, in this embodiment, the determination of auxiliary limb features and the precise selection of auxiliary array grids can be achieved based on the following method steps: First, the server can spatially match the pre-divided birthing bed array grid with the identified limb feature regions, verifying one by one whether each birthing bed array grid overlaps with the limb feature regions. When any birthing bed array grid is detected to overlap with a limb feature region, the array grid is marked as a support array grid. At the same time, the pressure acquisition value output in real time by the pressure sensing unit below the corresponding support array grid is retrieved. It can be noted that the support array grid corresponds to the bed surface unit directly covered and directly subjected to force by the human limbs. It is the core range of action for subsequent posture determination and support adjustment. Here, by locking the support array grid first through regional overlap matching, the subsequent calculation and analysis range can be narrowed from all array grids of the entire bed to the units directly covered by the limbs, which greatly reduces the amount of invalid calculations and improves the response speed of subsequent determination and adjustment. At the same time, it also ensures the spatial correspondence between the subsequent support points and the limb regions. Next, the server can comprehensively analyze the pressure collection values ​​of all bearing array grids corresponding to the same limb feature area to determine whether the target limb feature is currently in a preset auxiliary posture. When any target limb feature is determined to meet the judgment conditions of the preset auxiliary posture, the limb feature is marked as the auxiliary limb feature that needs to be supported and adjusted. It can be said that the pressure collection value directly reflects the contact force state between the limb and the bed surface. Compared with simple image shape recognition, it can more accurately perceive actual problems such as insufficient support and force imbalance. Here, using pressure data as the core basis to determine whether auxiliary support is needed can ensure the necessity and targeting of adjustment actions, avoid performing ineffective adjustments on parts with good support, reduce unnecessary bed movements, and improve the stability and comfort of pregnant women lying down. Subsequently, in one scenario, when the assistive limb feature is determined to be a lumbar and back feature, the server verifies the pressure acquisition values ​​of the corresponding support array grids one by one. The support array grids with pressure acquisition values ​​less than the preset support threshold are identified as suspended array grids. It can be explained that the preset support threshold is the minimum pressure critical value under normal fit and support state, which is set in combination with clinical ergonomics. A pressure value lower than this threshold indicates that there is a gap between the lumbar and back at the corresponding position and the bed surface, and no effective support is obtained. This is the core area of ​​lumbar suspension. Here, by screening the suspended array grids through the pressure threshold, the specific location of the lumbar support gap can be accurately located, providing a precise target position for subsequent inflatable support, ensuring that the support adjustment directly acts on the area with the most obvious suspension. The preset support threshold can be set based on actual needs, and this embodiment does not specifically limit it. Based on this, the server can filter out all array cells that are continuously adjacent to each birthing bed array cell on the lumbar and back curvature segment, in the direction away from the boundary connection line, that is, in the direction towards the outside of the torso, and mark them as lumbar edge cells. Finally, all the suspended array cells and the lumbar edge cells are determined together as auxiliary array cells that have an auxiliary relationship with the lumbar and back features. It can be explained that lumbar support cannot only act on the suspended core points, but also needs to form a smooth transition support in the edge area to avoid local pressure and discomfort caused by sudden changes in support height. Here, including the continuously adjacent edge cells in the adjustment range allows the final support surface to form a support shape that smoothly transitions from the center to the edge, which is more in line with the physiological curve of the human lumbar and back and improves support comfort. At the same time, the continuously adjacent screening logic also ensures the integrity and continuity of the support range and avoids the problems of point support and support breakage. In another scenario, when the auxiliary limb feature is determined to be an abdominal feature, the server can filter out the array grids adjacent to each delivery bed array grid on the abdominal arc segment, in the direction away from the boundary connection line, that is, in the direction towards the outside of the torso, and mark them as abdominal edge grids. It can be explained that the auxiliary support in the abdominal area is mainly for scenarios where the abdomen is tilted to the side and the force is uneven on both sides when the pregnant woman is lying down. The lateral edge position of the protruding abdomen is a key area for support transition and force balance. Here, by first defining the edge grid range along the abdominal arc, the lateral area that may have insufficient support or force offset can be quickly identified, and a clear range can be defined for subsequent screening of effective adjustment points. Subsequently, the server retrieves the pressure values ​​from the pressure sensing units below each abdominal edge grid. Abdominal edge grids with pressure values ​​lower than a preset support threshold are identified as auxiliary array grids that have an auxiliary correlation with abdominal features. In other words, by performing a secondary screening of the abdominal edge grids through pressure values, only edge grids with insufficient actual force and inadequate support are included in the final adjustment range. This can accurately supplement the lateral support gaps in the abdomen, correct the posture of force imbalance, and avoid ineffective inflation of areas with good fit, preventing excessive support from causing unnecessary pressure on the abdomen. While ensuring the support adjustment effect, it fully protects the safety and comfort of the pregnant woman's abdomen.

[0048] It can be explained that, in this embodiment, through layered screening and part-specific adaptation logic, the core bearing range is first locked by image spatial matching, and then the posture judgment and precise screening of adjustment points are completed by combining pressure data. At the same time, different array grid screening rules are adapted for the different physiological characteristics and support needs of the waist, back and abdomen. This ensures the accuracy and targeting of the auxiliary support points, while also taking into account the smoothness and comfort of the support transition. This allows the final adjustment range to fully match the actual support needs of the pregnant woman, while effectively reducing the scope of calculation and adjustment, and improving the server's response efficiency and adjustment accuracy.

[0049] Furthermore, in this embodiment, the aforementioned "determining whether the target limb feature is in a preset auxiliary posture based on all pressure acquisition values ​​corresponding to the same limb feature region" may also include the following steps: The response assist limb feature is the lumbar and back feature. The average value of all pressure collection values ​​corresponding to the same limb feature area is calculated to obtain the average force on the lumbar and back. When the average force on the lower back is less than the preset suspension threshold, the target limb is determined to be in a preset auxiliary posture. The response is that the auxiliary limb feature is the abdominal feature. The vertical baseline of the corresponding human body recognition area is retrieved, and based on the vertical baseline, each carrier array grid corresponding to the same limb feature area is divided into the first lateral array grid and the second lateral array grid. The average pressure values ​​of all corresponding first lateral array cells are averaged to obtain the first average force value, and the average pressure values ​​of all corresponding second lateral array cells are averaged to obtain the second average force value. If the force deviation between the first average force value and the second average force value is greater than a preset deflection threshold, it is determined that the target limb feature is in a preset auxiliary posture.

[0050] For example, in this embodiment, the server can use differentiated judgment logic to identify preset auxiliary postures for different types of assistive limb features, ensuring that the judgment results conform to the physiological characteristics and support failure modes of different parts, thereby improving the accuracy of posture judgment and scene adaptability. Specifically: In one scenario, when the assisted limb feature is the lower back, the server first calculates the average pressure values ​​of all load-bearing array grids within the limb feature area to obtain the average force on the lower back. It can be noted that insufficient support in the lower back usually manifests as overall suspension and poor bed surface fit, rather than abnormal force at a single point. Here, the calculation method of the overall regional average can filter out interference caused by occasional factors such as single-point pressure fluctuations and local clothing wrinkles, reflecting the fit and support status of the lower back and bed surface from an overall perspective, and avoiding misjudgment due to abnormal values ​​at individual points. Subsequently, the server can compare the average force on the lower back with the preset suspension threshold. When the average force is less than the threshold, it is determined that the overall support of the lower back is insufficient and it is in a preset auxiliary posture of suspension and lack of support. It can be noted that the preset suspension threshold is the minimum force threshold set by combining obstetric clinical data and ergonomic standards. It is the minimum force threshold under normal lower back support. If it is lower than the threshold, it means that the physiological curvature of the lower back is not effectively supported, and there is a risk of lower back strain and decreased comfort. It is necessary to activate the auxiliary support adjustment. The specific value can be set based on actual needs. This embodiment does not limit this. Here, by using the overall mean to determine the overall unsupported state of the lower back, the judgment logic is simple and efficient, and can quickly respond to support needs while ensuring the stability and reliability of the judgment results. In another scenario, when the auxiliary limb feature is an abdominal feature, the server can first retrieve the previously generated longitudinal baseline corresponding to the human body recognition area. Using this baseline as the dividing axis, all the bearing array grids within the same limb feature area are divided into the first lateral array grid and the second lateral array grid distributed on both sides of the baseline, which are the bearing units corresponding to the left and right sides of the pregnant woman's torso. It can be explained that the preset auxiliary posture of the abdominal area is mainly manifested as lying side-biased and unbalanced force on both sides of the abdomen. Long-term uneven force can easily increase the burden on the pregnant woman's body and the risk of fetal position deviation. Therefore, the judgment logic of comparing the force on both sides is more in line with the abnormal posture characteristics of this part. Subsequently, the server can calculate the average value of the pressure collected by the array grids on both sides to obtain the first average value of the force and the second average value of the force on both sides. Then, the difference between the two sets of average values ​​is calculated to obtain the force deviation value. Finally, the force deviation value is compared with the preset deflection threshold. When the force deviation value is greater than the preset deflection threshold, it is determined that the force on both sides of the abdomen is seriously uneven and the pregnant woman is in a preset assisted posture of lateral deviation. It can be noted that the preset deflection threshold is the maximum allowable force difference between the two sides, which is set in combination with clinical safety standards. Exceeding this threshold means that the degree of lateral deviation of the body exceeds the safe and comfortable range, and it is necessary to activate the auxiliary adjustment to correct the force balance. The specific value can be set based on actual needs, and this embodiment does not limit it.

[0051] In this embodiment, different posture judgment logics are adapted for the lower back and abdomen respectively. The server can accurately match the abnormal posture characteristics of different parts and avoid missed judgments and misjudgments caused by a single judgment rule. At the same time, all judgments are completed based on real-time pressure acquisition data, without the need for additional manual evaluation and auxiliary equipment. The judgment process runs automatically in real time and can identify the state of insufficient support and force imbalance in the first time. This provides a reliable trigger basis for the adaptive adjustment of the delivery bed. While ensuring the accuracy of the judgment, the server's response efficiency and operational stability are also taken into account.

[0052] Step S3 includes the following: Based on the target delivery bed, determine the bed board partition where the auxiliary array grid is located, and based on the characteristics of the auxiliary limbs, determine the corresponding bed board partition, the angle adjustment value of the auxiliary array grid, and the inflation value.

[0053] For example, in this embodiment, after the auxiliary array grid is selected and located, the server further performs the matching and calculation of adjustment parameters to provide a precise execution basis for the subsequent adjustment of the delivery bed. It can be explained that the main body of the target delivery bed is composed of multiple bed board sections that can be independently adjusted in angle. Each bed board section can change the tilt angle of the corresponding area through angle adjustment, thereby optimizing the support posture of the corresponding limb parts from an overall perspective. The delivery bed array grid can be embedded in the bed surface of each bed board section and achieve refined support compensation of local points through independent inflation and expansion, filling the support gaps that still exist after angle adjustment.

[0054] For example, Figure 2 A schematic diagram of the corresponding target delivery bed in this embodiment is shown, as follows: Figure 2 As shown, the target delivery bed's surface is physically divided into multiple independently controlled bed board sections. Each bed board section is strip-shaped and arranged side-by-side along a direction perpendicular to the longitudinal centerline of the target delivery bed (i.e., the transverse / width direction of the delivery bed), forming an independent strip-shaped bearing surface. Furthermore, within each strip-shaped bed board section, multiple vertically arranged delivery bed arrays (such as...) are supported. Figure 2 As shown in the dashed box, this structural layout allows the target delivery bed to form a matrix-like physical topology with horizontal strip partitions and vertical array units. During actual adjustment, each independent bed board partition can independently perform macroscopic angle deflection adjustments without interfering with each other, while each delivery bed array grid distributed within it can independently perform microscopic point inflation and expansion adjustments. This perfectly supports the aforementioned individualized adjustment method of dual-layer linkage compensation at the underlying hardware structure.

[0055] Furthermore, the server first matches and determines the bed board partition to which all auxiliary array grids belong based on the spatial distribution of all auxiliary array grids on the delivery bed surface, and identifies the bed board units that need to be adjusted in angle. This ensures that the effective range of angle adjustment precisely corresponds to the coverage area of ​​the auxiliary limb features, avoids irrelevant bed boards from performing invalid adjustment actions, and ensures the targeting of the adjustment process. On this basis, the server can calculate the angle adjustment value of the corresponding bed board partition and the inflation value of each auxiliary array grid by combining the specific type of auxiliary limb features and the current degree of posture deviation.

[0056] Among them, the angle adjustment value corresponds to the angle value that the bed board section needs to be deflected, which is used to realize the overall posture correction and support angle optimization of the corresponding limb parts. The expansion inflation value corresponds to the amount of gas that needs to be filled into the auxiliary array grid, which is used to realize the precise support filling of local points and make up for the support gap that still exists after the angle adjustment.

[0057] It can be explained that, in this embodiment, by matching the partitioning and accurately calculating the adjustment parameters, the support requirements of the limb features can be broken down into two layers of execution logic: overall bed board angle adjustment and array grid partial inflation adjustment. This achieves large-scale posture correction through bed board partitioning and individualized detailed support compensation through array grid inflation. While ensuring the support and assistance effect, all adjustment parameters are fully matched to the pregnant woman's individual body shape and current lying posture. There is no need for medical staff to manually adjust the bed board angle and support inflation volume, which effectively improves the accuracy and individualized adaptation of the delivery bed adjustment, while also reducing the burden of manual operation and the time spent on parameter adjustment.

[0058] Furthermore, in this embodiment, the aforementioned "determining the bed board partition where the auxiliary array grid is located based on the target delivery bed" may further include the following steps: The target delivery bed includes adjustable bed boards arranged in a direction perpendicular to the longitudinal centerline, wherein each adjustable bed board includes a vertically arranged array of bed boards. Based on the characteristics of each auxiliary limb, the adjustment bed board where the auxiliary array grid furthest from the longitudinal center line is located is determined as the bed board partition.

[0059] For example, in this embodiment, the precise locking of the bed board partitions can be achieved based on the following method steps: First, the server clarifies the structural composition of the target delivery bed, identifying all independently adjustable bed panels. All adjustable bed panels are arranged sequentially along the direction perpendicular to the longitudinal centerline of the delivery bed, i.e., side-by-side along the width of the delivery bed. Each adjustable bed panel contains multiple vertically arranged arrays along the length of the delivery bed. It can be explained that the longitudinal centerline corresponds to the central axis of the delivery bed along its length, aligning with the direction of the pregnant woman's torso. Here, the adjustable bed panels, arranged in sections along the width, allow for independent angle adjustments to different lateral areas of the torso. This precisely adapts to different support needs, such as insufficient support on one side or imbalance of force. Compared to adjusting the entire bed as a whole, this method is more targeted, avoiding unrelated areas from following and adjusting in tandem. This reduces ineffective bed movements and improves the stability and comfort of the pregnant woman while lying down. Subsequently, for each auxiliary limb feature, the server calculates the vertical distance from all corresponding auxiliary array grids to the longitudinal center line of the delivery bed, selects the auxiliary array grid with the largest distance value, and determines the adjustment bed board to which the furthest array grid belongs as the bed board partition for the corresponding auxiliary limb feature. It can be explained that the auxiliary array grid furthest from the longitudinal center line corresponds to the outermost boundary point of the auxiliary support area. Here, using the adjustment bed board where the outermost point is located as the adjustment partition can ensure that the effective range of angle adjustment completely covers all auxiliary array grids, avoiding the situation where the outermost support gap cannot be effectively compensated by angle adjustment due to insufficient adjustment partition range, and ensuring that the support effect of angle adjustment fully covers the entire required area.

[0060] It can be noted that in this embodiment, the method of locking the partition by taking the outermost point is adopted. There is no need to perform partitioning statistics and merging calculations for all auxiliary array grids one by one. The target bed board partition can be quickly locked by extreme value screening. The calculation logic is simple and efficient, which can effectively shorten the time of partitioning determination and improve the server response speed. At the same time, this method naturally has sufficient coverage redundancy, which can ensure that the angle adjustment range completely covers all points that require auxiliary support. While ensuring the targeting of adjustment and avoiding irrelevant bed boards from performing invalid actions, it also eliminates the problem of support failure caused by insufficient adjustment range, and takes into account both execution efficiency and the reliability of adjustment effect.

[0061] Furthermore, in this embodiment, the aforementioned "determining the angle adjustment value of the corresponding bed board partition, the auxiliary array grid, and the inflation value based on the characteristics of the auxiliary limbs" may also include the following steps: The response assist limb features are the waist and back features. The difference between the retrieved baseline force value and the average force value of the waist and back is calculated. Based on the obtained force gap value, the retrieved preset angle mapping table is traversed to determine the angle adjustment value of the corresponding bed board zone and the force compensation amount of the corresponding angle adjustment value based on the traversal result. The difference between the stress gap value and the stress compensation amount is calculated, and the basic inflation value of the corresponding suspended array grid is determined based on the obtained remaining gap value. The lateral inflation value of the corresponding waist edge grid is determined based on the average force on the waist and back, and the expansion inflation value of the corresponding waist and back features is constructed based on the base inflation value and the lateral inflation value. The response assist limb features are abdominal features. Based on the force deviation value, the preset angle mapping table is traversed to determine the angle adjustment value of the corresponding bed board partition and the force cancellation amount of the corresponding angle adjustment value based on the traversal result. The difference between the force deviation value and the force offset is calculated, and the inflation value of the corresponding abdominal edge grid is determined based on the obtained residual imbalance value.

[0062] For example, in this embodiment, for two different assistive limb features—the lower back and abdomen—the server can use a layered and progressive parameter calculation logic. First, it completes large-scale force compensation by adjusting the bed board angle, and then it completes precise filling of local gaps by inflating the array grid, achieving an individualized adjustment effect. The specific process is as follows: In one scenario, when the assistive limb feature is the lower back, the server first retrieves a preset baseline force value and calculates the difference between it and the previously calculated average lower back force value to obtain the current lower back force gap value. It can be noted that the baseline force value is a standard force value set based on obstetric clinical ergonomic data, representing the ideal support level for the lower back under normal support. It can be set based on actual needs, and this embodiment does not impose any limitations on this. Furthermore, the force gap value directly reflects the difference between the actual force on the lower back and the ideal support state. The larger the gap value, the more severe the degree of lumbar unsupport. Here, quantifying the support gap through difference calculation transforms the originally vague support requirements into calculable numerical indicators, providing a clear quantitative basis for subsequent adjustment parameter matching and avoiding blind adjustments. The baseline force value can be set based on actual needs, and this embodiment does not impose any specific limitations on this. Subsequently, the server can retrieve a pre-calibrated preset angle mapping table and traverse and match the mapping table using the stress gap value as the search condition to determine the angle adjustment value of the corresponding bed board partition and the stress compensation amount corresponding to the angle adjustment value. It can be noted that the preset angle mapping table is a mapping relationship library pre-built based on the structural parameters of the delivery bed and clinical test data. It can record the changes in the bed board support stress corresponding to different angle adjustment amounts. Here, by looking up the table and matching, the angle adjustment value that can compensate for most of the stress gap can be quickly obtained without performing complex mechanical simulation calculations in real time, which greatly improves the response speed of parameter calculation, while ensuring the accuracy of the correspondence between the angle adjustment amount and the stress compensation effect, and avoiding excessive or insufficient angle adjustment. Based on this, the server can further calculate the difference between the force gap value and the force compensation amount corresponding to the angle adjustment to obtain the remaining gap value that is not filled after the angle adjustment. Then, based on the remaining gap value, the basic inflation value corresponding to the suspended array grid is calculated. It can be noted that the angle adjustment of the bed board is a large-scale overall support adjustment, which generally cannot completely fit the local concavity of the physiological curvature of the waist. Therefore, there will be a certain remaining support gap. Based on this, the basic inflation amount of the core suspended area is calculated based on the remaining gap value. This allows the inflation support to accurately fill the support gap left after the angle adjustment, and achieve precise fitting support of the core suspended area. This avoids the insufficient support accuracy caused by simply relying on the angle adjustment, and also prevents the support redundancy and resource waste caused by full-range inflation. This ensures that the support strength and support position are completely matched with the actual degree of waist suspension. Next, the server can calculate the lateral inflation value corresponding to the waist edge grid based on the average force on the waist and back. Finally, the basic inflation value of the core area and the lateral inflation value of the edge area together constitute the complete expansion inflation value corresponding to the waist and back features. It can be explained that the transition boundary of the support area corresponding to the waist edge grid adopts a lower lateral inflation value than the core area. This allows the support surface to form a smooth height transition from the central suspended area to the normal edge area, avoiding local pressure and discomfort caused by sudden changes in support height. This makes the inflatable support surface more in line with the natural physiological curve of the human waist and back. While filling the support gap, it fully ensures the comfort and fit of the pregnant woman when lying down. The lateral inflation value can be calculated by multiplying the average force on the waist and the preset conversion coefficient. The preset conversion coefficient can be set according to the needs. This embodiment does not make specific limitations on this. In another scenario, when the auxiliary limb feature is the abdominal feature, the server uses the previously calculated force deviation value as the search condition to traverse and match the preset angle mapping table, thereby determining the angle adjustment value of the corresponding bed board section and the force cancellation amount corresponding to that angle adjustment value. It can be explained that the focus of the abdominal area assistance is to correct the bilateral force imbalance and improve the body's lateral tilt state. Therefore, by directly matching the angle adjustment amount based on the bilateral force deviation value, the angle deflection of the bed board section can accurately cancel the force deviation, correcting the posture of the trunk's lateral tilt from an overall perspective. In addition, the preset angle mapping table pre-marks the bilateral force cancellation effect corresponding to different angle deflection amounts. The matching angle adjustment value can be quickly obtained by looking up the table, ensuring that the angle adjustment can specifically improve the force imbalance problem. The adjustment logic is directly targeted and has high execution efficiency. Subsequently, the server can calculate the difference between the force deviation value and the force offset corresponding to the angle adjustment to obtain the remaining imbalance value after angle adjustment. Then, based on the remaining imbalance value, the expansion inflation value corresponding to the abdominal edge grid is calculated. It can be explained that angle adjustment corrects most of the force deviation at the overall level, but the local shape difference of the protruding abdomen may still leave a slight imbalance problem. Here, the abdominal edge grid is directionally inflated based on the remaining imbalance value, which can form a moderate lateral support on the side with less force, further make up for the remaining force imbalance, and avoid pressure on the abdominal area. Under the premise of correcting the force balance, the safety and comfort of the pregnant woman's abdomen are fully guaranteed.

[0063] It can be explained that, in this embodiment, by using a part-specific and layer-specific parameter calculation method, the server adapts differentiated calculation logic to the different support needs of the lower back and abdomen. This ensures both the overall adjustment effect and the fine-grained adaptation of local points. All parameters are calculated based on real-time collected force data, perfectly matching the actual condition of each pregnant woman. No manual parameter adjustment is required. While ensuring the accuracy and individualized adaptation of the adjustment, it also effectively improves the server's operating efficiency and adjustment response speed.

[0064] Step S4 includes the following: The control bed board partition performs angle adjustment based on the angle adjustment value, and the control auxiliary array grid performs inflation adjustment based on the expansion inflation value.

[0065] For example, in this embodiment, after the positioning of the bed board partitions and the calculation of all adjustment parameters are completed, the server can enter the adjustment execution stage and send control commands to the corresponding execution unit of the target delivery bed to drive the delivery bed to complete the corresponding adaptive adjustment action. It can be noted that the entire adjustment execution process in this step is automatically triggered and controlled by the server, without the need for medical staff to manually operate the adjustment controls. This ensures that the calculated adjustment parameters can be accurately implemented and avoids parameter deviations and response delays caused by manual operation.

[0066] Furthermore, the server first sends control commands to the angle drive mechanism of the corresponding bed board section, controlling the bed board section to perform a smooth angle deflection according to the calculated angle adjustment value. This adjusts the support angle of the corresponding limbs at the overall level, improves the stress state of the pregnant woman's corresponding parts, and achieves a wide range of posture correction and support optimization. Based on the completion of the angle adjustment of the bed board section, the server can further send commands to the inflation control unit corresponding to each auxiliary array grid, controlling each auxiliary array grid to perform inflation according to the corresponding inflation value. Based on the angle adjustment, this completes the fine support filling of local points, fills the support gap between the limbs and the bed surface, and allows the support surface to completely conform to the individual body shape of the pregnant woman, achieving point-level individualized precise support.

[0067] It can be explained that, in this embodiment, the adjustment plan calculated in the early stage can be smoothly implemented through the dual-layer linkage adjustment. The overall angle adjustment of the bed board is used to achieve posture correction, and the array pattern partial inflation is used to achieve precise fit. The two work together to form a complete individualized support system, which can effectively improve the problems of force imbalance and insufficient support during the pregnant woman's lying process, improve the comfort and safety during labor and examination, and realize the fully automated operation of the delivery bed adjustment process. There is no need for medical staff to manually adjust the parameters one by one, which greatly reduces the operational burden of clinical nursing. At the same time, it ensures that each adjustment is completely matched with the real-time body posture and support needs of the pregnant woman, which significantly improves the adaptability and execution efficiency of delivery bed management.

[0068] Another embodiment of the present invention provides a personalized, multifunctional, integrated maternity bed management system for pregnant women. Figure 3 Its corresponding system block diagram includes: The limb recognition module is configured to, when it is determined that the target pregnant woman is placed on the target delivery bed, control the delivery bed acquisition unit to acquire data from the target delivery bed, and identify the obtained view of the delivery bed to determine the limb feature regions that indicate the target pregnant woman and correspond to different target limb features. The auxiliary determination module is configured to determine the target limb feature as an auxiliary limb feature when the target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, and to determine the birthing bed array grids included in the target birthing bed that have an auxiliary relationship with the auxiliary limb feature as auxiliary array grids based on the view of the birthing bed. The numerical determination module is configured to determine the bed board partition where the auxiliary array grid is located based on the target delivery bed, and to determine the corresponding bed board partition, the angle adjustment value of the auxiliary array grid, and the inflation value based on the characteristics of the auxiliary limbs; The adjustment execution module is configured to control the bed board partitions to perform angle adjustments based on the angle adjustment value, and to control the auxiliary array grids to perform inflation adjustments based on the expansion inflation value.

[0069] In the specification provided herein, the algorithms and displays are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used with the examples of this invention. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing preferred embodiments of the invention.

[0070] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0071] Similarly, it should be understood that, in order to streamline this disclosure and aid in understanding one or more of the various aspects of the invention, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof.

[0072] Those skilled in the art will understand that modules, units, or components of the devices disclosed in the examples herein can be arranged in the devices described in this embodiment, or alternatively, can be located in one or more devices different from the devices in this example. The modules in the foregoing examples can be combined into a single module or, in addition, can be divided into multiple sub-modules.

[0073] Those skilled in the art will understand that the modules in the device of the embodiment can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiment can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components.

[0074] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention and form different embodiments.

[0075] Furthermore, some of the embodiments described herein are methods or combinations of method elements that can be implemented by a processor of a computer system or by other means of performing the functions. Therefore, a processor having the necessary instructions for implementing the methods or method elements forms means for implementing the methods or method elements. Furthermore, the elements described herein in the apparatus embodiments are examples of means for implementing the functions performed by elements for the purposes of carrying out the invention.

[0076] As used herein, unless otherwise specified, the use of ordinal numbers such as “first,” “second,” “third,” etc., to describe ordinary objects merely indicates different instances of similar objects and is not intended to imply that the objects being described must have a given order in time, space, ordering, or any other manner.

[0077] Although the invention has been described with respect to a limited number of embodiments, those skilled in the art will understand from the foregoing description that other embodiments are conceivable within the scope of the invention described herein. Furthermore, it should be noted that the language used in this specification has been chosen primarily for readability and edibility purposes, and not for the purpose of explaining or limiting the subject matter of the invention.

Claims

1. A method for managing pregnant women using an individualized, multifunctional integrated delivery bed, characterized in that: Includes the following steps: Once the target pregnant woman is identified as being placed on the target delivery bed, the delivery bed acquisition unit is controlled to acquire data from the target delivery bed and to identify the obtained view of the delivery bed, thereby determining the limb feature regions that indicate the target pregnant woman and correspond to different target limb features. When the target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, the target limb feature is determined as an auxiliary limb feature, and the birthing bed array grids included in the target birthing bed that have an auxiliary relationship with the auxiliary limb feature are determined as auxiliary array grids based on the view of the birthing bed. Based on the target delivery bed, determine the bed board partition where the auxiliary array grid is located, and based on the characteristics of the auxiliary limbs, determine the corresponding bed board partition, the angle adjustment value of the auxiliary array grid, and the inflation value; The control bed board partition performs angle adjustment based on the angle adjustment value, and the control auxiliary array grid performs inflation adjustment based on the expansion inflation value.

2. The method according to claim 1, characterized in that, Once the target pregnant woman is identified as being placed in the target delivery bed, the delivery bed acquisition unit is controlled to acquire data from the target delivery bed, including: When it is determined that the pressure acquisition value output by any pressure sensing unit located under each birthing bed array cell in the target birthing bed is greater than the preset pressure threshold, the birthing bed array cell corresponding to the pressure acquisition value is determined to have pressure-bearing properties. The birthing bed compression shape, which is composed of all birthing bed array grids with compression properties, is determined, and the birthing bed compression shape is compared with the retrieved reference compression shape to determine a first evaluation value based on the obtained shape similarity. Each pre-marked array cell corresponding to the reference pressure shape is determined, and in response to any delivery bed array cell with pressure attribute whose corresponding pressure position is the same as the marking position of the corresponding mark array cell, the initial number of marks is accumulated to determine a second evaluation value based on the obtained current number of marks. The first evaluation value and the second evaluation value are weighted and summed. If the resulting comprehensive evaluation value is greater than the preset evaluation value, the target pregnant woman is determined to be placed in the target delivery bed, so as to control the delivery bed acquisition unit to acquire data from the target delivery bed.

3. The method according to claim 2, characterized in that, Determine each pre-marked array of cells corresponding to the reference compression shape, including: When it is determined that the target pregnant woman has historical examination data based on her identity information, the historical examination data is determined as reference examination data; otherwise, the retrieved medical record database is traversed based on the physical data determined by the identity information in order to determine the historical examination data of other pregnant women that match the physical data as reference examination data. Based on the reference inspection data, the compression shape of each delivery bed corresponding to different inspection times and with a comprehensive evaluation value greater than the preset evaluation value is determined. The center point of the delivery bed corresponding to the target delivery bed is used as the coverage center point to cover all the compression shapes of the delivery beds, thus obtaining a common coverage area. The shape of the corresponding public coverage area is determined as the reference pressure shape, and each delivery bed array cell located in the area contour of the corresponding public coverage area is determined as a contour array cell; Determine the center vectors from the coverage center point to the array center point corresponding to each contour array cell, and determine the array distribution curvature of the region contour corresponding to different contour array cells based on the center vectors. When the curvature of any array distribution along the region contour satisfies the curvature abrupt change condition, and / or the vector length of any corresponding center vector satisfies the extreme value condition, the corresponding contour array grid is determined as the identifier array grid.

4. The method according to claim 1, characterized in that, The obtained views of the delivery bed are identified to determine the limb feature regions corresponding to different target limb features, including: Determine the image pixels that make up the view of the delivery bed, and identify the image pixels with different pixel values ​​from the reference pixels of the retrieved target delivery bed as human body pixels; Connect adjacent human body pixels and obtain the corresponding human body center point that indicates the human body recognition area of ​​the target pregnant woman; A bounding rectangle is formed that circumscribes the human body recognition area, and a longitudinal baseline is generated that passes through the center point of the human body and is parallel to the long axis extension direction of the corresponding bounding rectangle. Starting from each human pixel that makes up the human recognition contour of the corresponding human recognition area, generate morphological representation lines that intersect the vertical baseline perpendicularly, and obtain the morphological representation length of each morphological representation line. The difference between the morphological representation lines of adjacent human pixels is calculated based on the morphological representation length. Based on the obtained morphological representation differences, the limb feature regions corresponding to different target limb features are determined, including abdominal features and back features.

5. The method according to claim 4, characterized in that, Based on the obtained morphological representation differences, the limb feature regions corresponding to different target limb features of the pregnant woman are determined, including: All morphological representation differences are sorted along the human body recognition contour, and based on the obtained morphological representation sequence, human body pixels whose morphological representation differences jump from positive to negative values ​​and from negative to positive values ​​are respectively determined as the first inflection point and the second inflection point. The first inflection point with the largest corresponding morphological representation length is determined as the first extreme point, and the distance between each second inflection point and each longitudinal point on the corresponding longitudinal baseline of the first extreme point is determined. All second inflection points whose corresponding longitudinal point distance is less than the preset physiological interval are identified as candidate inflection points, and the candidate inflection point with the smallest corresponding morphological representation length is identified as the second extreme point. Based on the first extreme point, the limb feature regions corresponding to the abdominal features of the target pregnant woman are determined, and based on the second extreme point, the limb feature regions corresponding to the waist and back features of the target pregnant woman are determined.

6. The method according to claim 5, characterized in that, Based on the first extreme point, the limb feature regions corresponding to the abdominal features of the target pregnant woman are determined, and based on the second extreme point, the limb feature regions corresponding to the waist and back features of the target pregnant woman are determined, including: Starting from the first extreme point, the system traverses along the human body recognition contour and determines the first inflection point, located on both sides of the first extreme point, as the abdominal boundary point. Based on human body contour recognition, the human body contour segment composed of abdominal boundary points is determined as the abdominal arc segment, and different abdominal boundary points are connected to form the limb feature area corresponding to the abdominal features of the target pregnant woman based on the obtained boundary connection line and abdominal arc segment. Starting from the second extreme point, the system traverses along the human body recognition contour and determines the first inflection point, which is located on both sides of the second extreme point, as the waist and back boundary point. Based on human body contour recognition, the human body contour segment composed of the waist and back boundary points is determined as the waist and back curvature segment. Different waist and back boundary points are connected to form the limb feature area corresponding to the waist and back features of the target pregnant woman based on the obtained boundary connection line and waist and back curvature segment.

7. The method according to claim 6, characterized in that, When a target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, the target limb feature is identified as an auxiliary limb feature. Furthermore, based on the view of the delivery bed, the delivery bed array grids included in the target delivery bed that have an auxiliary relationship with the auxiliary limb feature are identified as auxiliary array grids, including: When it is determined that any delivery bed array grid overlaps with any limb feature area, the delivery bed array grid is designated as the bearing array grid, and the pressure acquisition values ​​output by all pressure sensing units located under each bearing array grid are obtained. Based on all pressure acquisition values ​​corresponding to the same limb feature area, determine whether the target limb feature is in a preset auxiliary posture, and in response to any target limb feature being in a preset auxiliary posture, determine the target limb feature as an auxiliary limb feature; The response assistive limb features are waist and back features, and the load-bearing array grids whose corresponding pressure acquisition values ​​are less than the preset support threshold are determined as suspended array grids; Along the direction away from the corresponding boundary connection line of the lumbar and back curvature segment, all the birthing bed array cells that are continuously adjacent to each birthing bed array cell located in the lumbar and back curvature segment are identified as lumbar edge cells, and all suspended array cells and lumbar edge cells are identified as auxiliary array cells that have an auxiliary relationship with the auxiliary limb features. The response auxiliary limb feature is the abdominal feature. Along the direction away from the corresponding boundary connection line of the abdominal arc segment, the birthing bed array cell adjacent to each birthing bed array cell located in the abdominal arc segment is determined as the abdominal edge cell. The pressure acquisition value output by the pressure sensing unit located below each abdominal edge grid is obtained, and the abdominal edge grid with the corresponding pressure acquisition value less than the preset support threshold is determined as the auxiliary array grid with an auxiliary limb feature.

8. The method according to claim 7, characterized in that, Determine whether the target limb feature is in a preset assisted posture based on all pressure data collected from the same limb feature region, including: The response assist limb feature is the lumbar and back feature. The average value of all pressure collection values ​​corresponding to the same limb feature area is calculated to obtain the average force on the lumbar and back. When the average force on the lower back is less than the preset suspension threshold, the target limb is determined to be in a preset auxiliary posture. The response is that the auxiliary limb feature is the abdominal feature. The vertical baseline of the corresponding human body recognition area is retrieved, and based on the vertical baseline, each carrier array grid corresponding to the same limb feature area is divided into the first lateral array grid and the second lateral array grid. The average pressure values ​​of all corresponding first lateral array cells are averaged to obtain the first average force value, and the average pressure values ​​of all corresponding second lateral array cells are averaged to obtain the second average force value. If the force deviation between the first average force value and the second average force value is greater than a preset deflection threshold, it is determined that the target limb feature is in a preset auxiliary posture.

9. The method according to claim 8, characterized in that, Determine the bed board partition where the auxiliary array grid is located based on the target delivery bed, including: The target delivery bed includes adjustable bed boards arranged in a direction perpendicular to the longitudinal centerline, wherein each adjustable bed board includes a vertically arranged array of bed boards. Based on the characteristics of each auxiliary limb, the adjustment bed board where the auxiliary array grid furthest from the longitudinal center line is located is determined as the bed board partition.

10. The method according to claim 9, characterized in that, Based on the characteristics of the assistive limbs, the corresponding bed board partitions, the angle adjustment values ​​of the assistive array grids, and the inflation values ​​are determined, including: The response assist limb features are the waist and back features. The difference between the retrieved baseline force value and the average force value of the waist and back is calculated. Based on the obtained force gap value, the retrieved preset angle mapping table is traversed to determine the angle adjustment value of the corresponding bed board zone and the force compensation amount of the corresponding angle adjustment value based on the traversal result. The difference between the stress gap value and the stress compensation amount is calculated, and the basic inflation value of the corresponding suspended array grid is determined based on the obtained remaining gap value. The lateral inflation value of the corresponding waist edge grid is determined based on the average force on the waist and back, and the expansion inflation value of the corresponding waist and back features is constructed based on the base inflation value and the lateral inflation value. The response assist limb features are abdominal features. Based on the force deviation value, the preset angle mapping table is traversed to determine the angle adjustment value of the corresponding bed board partition and the force cancellation amount of the corresponding angle adjustment value based on the traversal result. The difference between the force deviation value and the force offset is calculated, and the inflation value of the corresponding abdominal edge grid is determined based on the obtained residual imbalance value.

11. A personalized, multifunctional, integrated maternity bed management system for pregnant women, characterized in that, include: The limb recognition module is configured to, when it is determined that the target pregnant woman is placed on the target delivery bed, control the delivery bed acquisition unit to acquire data from the target delivery bed, and identify the obtained view of the delivery bed to determine the limb feature regions that indicate the target pregnant woman and correspond to different target limb features. The auxiliary determination module is configured to determine the target limb feature as an auxiliary limb feature when the target limb feature is determined to be in a preset auxiliary posture based on any limb feature region, and to determine the birthing bed array grids included in the target birthing bed that have an auxiliary relationship with the auxiliary limb feature as auxiliary array grids based on the view of the birthing bed. The numerical determination module is configured to determine the bed board partition where the auxiliary array grid is located based on the target delivery bed, and to determine the corresponding bed board partition, the angle adjustment value of the auxiliary array grid, and the inflation value based on the characteristics of the auxiliary limbs; The adjustment execution module is configured to control the bed board partitions to perform angle adjustments based on the angle adjustment value, and to control the auxiliary array grids to perform inflation adjustments based on the expansion inflation value.

Images