Scoliosis probability prediction method and system

Abstract

The invention provides a scoliosis probability prediction method and system, and belongs to the technical field of image recognition, and the method comprises the steps: obtaining a body back image of a to-be-detected person; generating a back contour according to the body back image, and extracting a plurality of reference points and a plurality of groups of feature points from the back contour; at least one detection area is generated, and each detection area is enclosed by at most one reference point and at least one group of feature points; generating a reference line corresponding to each detection area; for any detection region, dividing the detection region into a first sub-region and a second sub-region through a corresponding reference line, and calculating a first area ratio of the first sub-region to the second sub-region; according to the first area ratio of all the detection areas, the scoliosis probability is obtained. According to the method, the key reference points and the feature points of the back contour are extracted, and the scoliosis probability is accurately predicted in combination with detection region division, reference line generation and area ratio calculation.

Description

Technical Field

[0001] The present invention relates to the field of image recognition technology, and in particular to a method and system for predicting the probability of scoliosis. Background Art

[0002] In the field of body posture analysis, accurate assessment of spinal morphology and its symmetry is an important part of studying human movement function and morphological balance. Since scoliosis is difficult to detect in the early stages and develops rapidly, it can easily lead to more serious consequences. Therefore, it is critical to accurately determine whether scoliosis occurs in the spine in the early stages.

[0003] Under the current technical background, the common analysis method is to manually evaluate whether the spine has scoliosis by observing the spinal morphology. However, these methods have significant limitations in practical applications. Since manual measurement methods rely on the operator's experience and judgment, it is difficult to avoid the interference of subjective factors, resulting in poor accuracy of the prediction results.

[0004] Therefore, how to improve the accuracy of scoliosis probability prediction has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The present invention provides a scoliosis probability prediction method, system, electronic device and storage medium, which are used to solve the defects in the prior art and improve the accuracy of scoliosis probability prediction.

[0006] The present invention provides a method for predicting the probability of scoliosis, comprising the following steps: Acquire a back image of the body of the person to be tested; Generate a back contour according to the back image of the body, and extract multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to characterize a reference point of the spinal posture, and each group of feature points is used to characterize a relevant point for scoliosis detection; generating at least one detection area, wherein each of the detection areas is surrounded by at most one of the reference points and at least one group of the feature points; Generate a reference line corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; For any one of the detection areas, the detection area is divided into a first sub-area and a second sub-area by a corresponding reference line, and a first area ratio of the first sub-area to the second sub-area is calculated; The probability of scoliosis is obtained according to the first area ratio of all the detection areas.

[0007] According to a scoliosis probability prediction method provided by the present invention, the reference points include an upper reference point and a lower reference point, and each group of feature points includes a first bilateral position point, a second bilateral position point, a third bilateral position point, and a fourth bilateral position point; Among them, in the vertical direction of the back contour, from top to bottom are the upper end reference point, the first bilateral position point, the second bilateral position point, the third bilateral position point, the fourth bilateral position point and the lower end reference point.

[0008] According to a scoliosis probability prediction method provided by the present invention, the detection area includes an upper area, a middle area and a lower area; The upper section area is surrounded by the upper end reference point, the first bilateral position point, the second bilateral position point and the third bilateral position point; The middle section area is surrounded by the third bilateral position point and the fourth bilateral position point; The lower section area is surrounded by the fourth bilateral position point and a first horizontal line of the back contour, wherein the first horizontal line passes through the lower end reference point.

[0009] According to a scoliosis probability prediction method provided by the present invention, the method generates a reference line corresponding to each detection area, specifically comprising: According to the upper reference point, a vertical line is drawn downward to obtain a first reference line; Based on the lower reference point, draw a vertical line upward to obtain a second reference line; The upper area corresponds to the first reference line, and the middle area and the lower area both correspond to the second reference line.

[0010] According to a method for predicting the probability of scoliosis provided by the present invention, the probability of scoliosis is obtained according to the first area ratio of all the detection areas, specifically comprising: Taking a weighted average of all the first area ratios to obtain a weighted average value; The corresponding scoliosis probability is determined according to the preset range interval that the weighted average value falls into, wherein each of the preset range intervals corresponds to a preset scoliosis probability interval.

[0011] According to a method for predicting the probability of scoliosis provided by the present invention, the method further comprises: When the fourth bilateral position point cannot be extracted from the back contour, obtaining the foot shape information and the plantar pressure information of the person to be tested; Determine the area of ​​the front area of ​​the sole and the area of ​​the rear area of ​​the sole according to the foot shape information; Determine the front plantar pressure value and the contralateral rear plantar pressure value according to the plantar pressure information; The probability of scoliosis is obtained according to a second area ratio of the front area of ​​the plantar to the rear area of ​​the plantar, and a pressure ratio of the front pressure value of the plantar to the pressure value of the contralateral rear foot.

[0012] The present invention also provides a scoliosis probability prediction system, comprising the following modules: An acquisition module, used for acquiring a back image of the person to be tested; A first processing module is used to generate a back contour according to the back image of the body, and extract multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to represent a reference point of the spinal posture, and each group of feature points is used to represent a relevant point for scoliosis detection; A second processing module, configured to generate at least one detection area, wherein each detection area is surrounded by at most one reference point and at least one group of feature points; The second processing module is further used to generate a reference line corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; A third processing module, configured to divide any one of the detection areas into a first sub-area and a second sub-area through a corresponding reference line, and calculate a first area ratio between the first sub-area and the second sub-area; The fourth processing module is used to obtain the probability of scoliosis according to the first area ratio of all the detection areas.

[0013] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a method for predicting the probability of scoliosis as described above is implemented.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for predicting the probability of scoliosis as described in any one of the above is implemented.

[0015] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for predicting the probability of scoliosis as described above is implemented.

[0016] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: By acquiring the back image of the person to be tested, complete and clear back morphology data can be provided, laying the foundation for the subsequent quantitative analysis of spinal posture; by generating at least one detection area, each detection area is surrounded by at most one reference point and at least one group of feature points, and local segmentation can be performed according to the characteristics of different areas of the back, thereby refining the detection area and improving the pertinence and accuracy of the analysis. For any detection area, by dividing the detection area into a first sub-area and a second sub-area through the corresponding reference line, and calculating the first area ratio of the first sub-area to the second sub-area, the symmetry characteristics within the detection area can be quantified, thereby achieving accurate analysis of local morphological changes. Overall, this technical solution constructs a standardized and precise spinal posture analysis process through back image data acquisition, reference point and feature point extraction, detection area division, reference line generation, and area ratio calculation. It not only improves the accuracy and efficiency of detection, but also has strong adaptability and robustness, and is suitable for automated detection needs in various scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is one of the flow charts of the method for predicting the probability of scoliosis provided by the present invention.

[0019] Figure 2 It is a schematic diagram of back contour reference points and feature points provided by the present invention.

[0020] Figure 3 It is a schematic diagram of the upper area of ​​the back contour provided by the present invention.

[0021] Figure 4 It is a schematic diagram of the middle area of ​​the back contour provided by the present invention.

[0022] Figure 5 It is a schematic diagram of the lower area of ​​the back contour provided by the present invention.

[0023] Figure 6 It is a schematic diagram of the reference line corresponding to each detection area provided by the present invention.

[0024] Figure 7 This is the second flow chart of the scoliosis probability prediction method provided by the present invention.

[0025] Figure 8This is the third flow chart of the scoliosis probability prediction method provided by the present invention.

[0026] Fig. 9 It is a structural schematic diagram of the scoliosis probability prediction system provided by the present invention.

[0027] Fig.10 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] It should be noted that, in the description of the present invention, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "include one..." do not exclude the existence of other identical elements in the process, method, article or device including the elements. The orientation or positional relationship indicated by the terms "upper", "lower", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the system or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances.

[0030] The terms "first", "second", etc. in the present invention are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present invention can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally of the same type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" means at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0031] Combine the following Figure 1-Figure 10 The present invention describes a scoliosis probability prediction method, system, electronic device and storage medium.

[0032] Figure 1 FIG. 1 is one of the flow charts of the method for predicting the probability of scoliosis provided by the present invention, such as Figure 1 As shown, including but not limited to the following steps: Step 101: Acquire a back image of the body of the person to be tested.

[0033] In step 101, by obtaining the back image of the person to be tested, the foundation is laid for the subsequent spinal posture analysis and related tests. The core of this step is to provide clear and complete back surface information, which is convenient for extracting the back contour and key feature points, thereby ensuring the accuracy and consistency of subsequent calculations. Specifically, the person to be tested needs to maintain an upright posture, facing forward, standing with both feet parallel, and the soles of the feet fully touching the ground to ensure that the posture is natural and meets the detection requirements. At the same time, in order to avoid interference factors, the person to be tested is required to take off his shoes and socks and expose his back.

[0034] The image acquisition device is placed in a horizontal position directly behind the back of the person to be tested, perpendicular to the ground, and about 1 meter away from the person to be tested. The device uses depth camera technology to capture the three-dimensional morphological information of the back and ensure high-precision resolution to ensure the accuracy of feature point extraction. During the acquisition process, the device automatically calibrates to maintain the optical axis aligned with the midline of the back, and combines algorithms to optimize image quality in real time to reduce the impact of external light and noise on the results. In addition, to further improve the reliability of the back morphological data, the multiple frames of the acquired images can be superimposed and averaged to filter out errors caused by short-term motion or jitter.

[0035] Through this step, high-precision and high-completeness back image data can be obtained, providing basic support for generating back contours and extracting reference points and feature points. This method effectively avoids errors caused by manual measurement or low-precision imaging in traditional inspections, thus laying a technical foundation for the standardization and automation of subsequent analysis and ensuring the stability and applicability of the inspection process.

[0036] Step 102: Generate a back contour based on the back image of the body, and extract multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to represent a reference point of the spinal posture, and each group of feature points is used to represent a relevant point for scoliosis detection.

[0037] In a possible implementation, the reference point includes an upper reference point and a lower reference point, and each group of feature points includes a first bilateral position point, a second bilateral position point, a third bilateral position point, and a fourth bilateral position point; Among them, in the vertical direction of the back contour, from top to bottom are the upper reference point, the first bilateral position point, the second bilateral position point, the third bilateral position point, the fourth bilateral position point and the lower reference point.

[0038] In step 102, a back contour is generated from the back image of the body obtained in step 101, and multiple reference points and multiple groups of feature points are extracted therefrom to achieve a standardized representation of the back structure. The purpose of this step is to accurately extract key points from the back image to provide necessary data support for the subsequent generation of detection areas and calculation of area ratios. The generation of the back contour can not only reflect the overall characteristics of the back surface morphology, but also provide a clear contour benchmark for further analysis.

[0039] In the specific implementation, the back image is converted into a three-dimensional data representation through the point cloud processing technology of the depth image. First, an algorithm is used to identify the outer contour area of ​​the back, and the non-back area such as the arms or clothing occlusion is removed to ensure the purity of the contour. Subsequently, multiple reference points and feature points are automatically located in the contour based on the known human anatomical structure model. Among them, the reference points include the upper reference point and the lower reference point, which are used to calibrate the reference position of the spinal posture; the feature points include the first bilateral position point, the second bilateral position point, the third bilateral position point, and the fourth bilateral position point, which are used to calibrate other key structural positions of the back.

[0040] The specific upper reference point, lower reference point, first bilateral position point, second bilateral position point, third bilateral position point and fourth bilateral position point can be referred to Figure 2 , Figure 2 It is a schematic diagram of back contour reference points and feature points provided by the present invention.

[0041] In this solution, the upper reference point is preferably the seventh cervical vertebra position point (i.e., C7 point), and the lower reference point is preferably the first sacral position point (i.e., S1 point). The first bilateral position point is preferably the bilateral inner shoulder position point, the second bilateral position point is preferably the bilateral outer shoulder position point, the third bilateral position point is preferably the bilateral axillary position point, and the fourth bilateral position point is preferably the bilateral lumbar concave position point.

[0042] In order to improve the accuracy of extraction, a constraint algorithm based on the bone structure is used, combined with the density characteristics of the point cloud data, to automatically match and identify the positions of these points. During the recognition process, the system identifies the upper reference point, each group of feature points, and the lower reference point from top to bottom based on the vertical relationship of the back contour to ensure the logical order and rationality of the spatial distribution between the points. In addition, if some feature points are not visible under certain conditions, the system will fit and complete them according to the local morphology of the back contour to ensure the integrity of the final data.

[0043] Through this step, accurate reference points and feature points can be efficiently extracted from the back image. This not only provides a standardized benchmark for the subsequent generation of detection areas, but also reflects the morphological characteristics of the back through the relative relationship between the points, thereby significantly improving the accuracy and consistency of detection.

[0044] Step 103: Generate at least one detection area, wherein each detection area is surrounded by at most one reference point and at least one set of feature points.

[0045] In a possible implementation, the detection area includes an upper area, a middle area, and a lower area; The upper section area is surrounded by the upper reference point, the first bilateral position point, the second bilateral position point and the third bilateral position point; The middle section area is surrounded by the third bilateral position point and the fourth bilateral position point; The lower section area is surrounded by the fourth bilateral position point and the first horizontal line of the back contour, wherein the first horizontal line passes through the lower end reference point.

[0046] In step 103, at least one detection area is generated and surrounded by at most one reference point and at least one set of feature points to provide a basis for local division and quantitative analysis of the back morphological features. The core of this step is to decompose the complex contour information of the whole back into multiple areas that can be evaluated separately through reasonable division of the detection area, thereby ensuring the pertinence and accuracy of subsequent calculations.

[0047] In a specific implementation, according to the reference points and feature points extracted in step 102, the spinal posture is used as the core reference point, and the vertical relationship of the back morphology is combined to perform regional division. In a possible implementation, the detection area is divided into an upper area, a middle area, and a lower area. The upper area is preferably an upper thoracic area, which is surrounded by the upper reference point, the first bilateral position point, the second bilateral position point, and the third bilateral position point, and is used to reflect the symmetry and morphological characteristics of the upper thoracic section of the back; the middle area is preferably a thoracolumbar area, which is surrounded by the third bilateral position point and the fourth bilateral position point, and is used to evaluate the morphological changes of the thoracolumbar section of the back; the lower area is preferably a lumbar area, which is surrounded by the fourth bilateral position point and the first horizontal line of the back contour, wherein the first horizontal line passes through the lower reference point, and is used to analyze the overall morphology of the lumbar section of the back.

[0048] For specific upper area division, please refer to Figure 3 , Figure 3 It is a schematic diagram of the upper area of ​​the back contour provided by the present invention.

[0049] The specific division of the middle section can be referred to Figure 4 , Figure 4 It is a schematic diagram of the middle area of ​​the back contour provided by the present invention.

[0050] For specific division of the lower area, please refer to Figure 5 , Figure 5 It is a schematic diagram of the lower area of ​​the back contour provided by the present invention.

[0051] In the process of region generation, the reference points and feature points are connected by geometric algorithms in combination with the point cloud data of the depth image to construct the boundary of the detection area. At the same time, the boundary is mapped to a two-dimensional plane using three-dimensional space projection technology to ensure the stability and accuracy of region generation. For complex back contours, the region division algorithm will dynamically adjust the point connection order to adapt to individual differences and maintain the closure of the region.

[0052] By generating detection regions, the back morphology can be decomposed into several standardized sub-regions, which facilitates independent quantitative calculations of different regional characteristics in subsequent analysis. This division method not only effectively reduces the complexity of the overall analysis, but also ensures clear boundaries between regions, thereby improving the accuracy and consistency of the detection results. Compared with traditional analysis methods based on overall features, this step significantly enhances the resolution and adaptability of the detection through regional division, providing a solid foundation for standardized analysis of spinal posture.

[0053] Step 104: Generate a reference line corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point.

[0054] In a possible implementation, the method generates a reference line corresponding to each detection area, specifically including: Based on the upper reference point, draw a vertical line downward to obtain the first reference line; Based on the lower reference point, draw a vertical line upward to obtain the second reference line; The upper area corresponds to the first reference line, and the middle area and the lower area both correspond to the second reference line.

[0055] In step 104, a reference line corresponding to each detection area is generated to provide a clear baseline for subsequent area division and area ratio calculation. The core purpose of this step is to provide a unified and standardized reference benchmark for the detection area to ensure consistency and comparability of quantitative analysis between different detection areas. The generation of the reference line is based on the vertical projection of the reference point, so as to reflect the local characteristics of the spinal posture in each area.

[0056] In the specific implementation, the upper reference point and the lower reference point extracted in step 102 are used to generate corresponding reference lines in combination with the characteristics of each detection area. The upper reference point (C7) is used as the reference position of the upper section area, and a first reference line is generated by drawing a vertical line downward, and the first reference line runs through the core of the upper section area; the lower reference point (S1) is used as the reference position of the middle section area and the lower section area, and a second reference line is generated by drawing a vertical line upward, and the second reference line covers the middle section area and the lower section area. The generation of the reference line strictly follows the geometric rules of the vertical direction of the back to ensure a high degree of fit with the shape of the back contour.

[0057] The reference lines corresponding to each detection area can be found at Figure 6 , Figure 6 It is a schematic diagram of the reference line corresponding to each detection area provided by the present invention.

[0058] In the process of generating reference lines, a method combining deep image point cloud data with geometric algorithms is used to accurately calculate the direction of the vertical line using the three-dimensional coordinates of the points. At the same time, the system uses an automatic correction algorithm to optimize the projection position of the vertical line in the back contour to avoid reference line deviation caused by asymmetric back morphology or data collection errors. In addition, for reference point deviations caused by abnormal postures in specific individuals, the algorithm can dynamically adjust the starting point of the vertical line generation to ensure that the reference line always accurately reflects the core benchmark of each area.

[0059] By generating reference lines corresponding to each detection area, a clear and unified baseline can be provided for subsequent detection area division and area ratio calculation. This not only improves the accuracy and stability of area division, but also ensures the comparability of inter-regional analysis through standardized baseline design. Compared with traditional analysis methods that rely on manual calibration or a single baseline, this solution has significant advantages in terms of automation and accuracy, providing efficient and reliable technical support for spinal posture analysis.

[0060] Step 105: For any detection area, divide the detection area into a first sub-area and a second sub-area through a corresponding reference line, and calculate a first area ratio of the first sub-area to the second sub-area.

[0061] In step 105, by dividing any detection area into a first sub-area and a second sub-area, and calculating a first area ratio between the first sub-area and the second sub-area, the morphological feature difference within the detection area is quantified, providing accurate data support for subsequent spinal posture analysis. The core of this process is to divide the detection area into two symmetrical left and right parts with the help of a reference line, and use the area ratio as a measurement indicator to achieve a standardized evaluation of local symmetry or asymmetry.

[0062] In a specific implementation, the reference line generated in step 104 is used as a basis for division, and each detection area is divided into a first sub-area on the left side of the reference line and a second sub-area on the right side of the reference line. For example, for the upper section area, the reference line is the first reference line, which vertically passes through the upper reference point; the middle section area and the lower section area are divided vertically upward using the second reference line with the lower reference point as the reference. In the division process, the intersection of the reference line and the detection area boundary is calculated by a geometric algorithm in combination with the point cloud data of the back contour, and two independent closed sub-areas are further generated.

[0063] After the division is completed, the areas of the first sub-area and the second sub-area are calculated respectively. The area calculation is based on the plane projection of the point cloud data, and the geometric area of ​​each sub-area is accurately obtained through the integral algorithm. Subsequently, the area of ​​the first sub-area is calculated to be a ratio of the area of ​​the second sub-area to obtain the first area ratio of the detection area.

[0064] It should be noted that the calculation method of the first area ratio is preferably: ; Where A is the first area ratio, is the area of ​​the first sub-region, is the area of ​​the second sub-region, To take the smaller value of the area of ​​the first sub-region and the area of ​​the second sub-region, The first area ratio is the larger value of the area of ​​the first sub-region and the area of ​​the second sub-region. If the area of ​​the first sub-region is equal to the area of ​​the second sub-region, the first area ratio is 1.

[0065] To ensure the stability and consistency of the data, the system also introduces a redundant calculation mechanism to calculate the area ratio of each region multiple times and take the average value to reduce possible measurement errors.

[0066] Through this step, the morphological feature differences within the detection area can be quantified as a standardized area ratio index. This analysis method based on area ratio can intuitively reflect the changes in left-right symmetry within the detection area, especially when the local morphology is abnormal, the deviation of the area ratio will be clearly reflected.

[0067] Step 106: Obtain the probability of scoliosis according to the first area ratio of all detection areas.

[0068] In step 106, the probability of scoliosis is obtained according to the first area ratio of all the detection areas, so as to achieve an overall evaluation and quantitative analysis of the back morphology. The core of this step is to integrate the local features of different detection areas and unify multiple area ratios into a comprehensive index to reflect the overall symmetry and change trend of the back morphology, thereby providing a standardized evaluation result.

[0069] In one possible implementation, refer to Figure 7 , Figure 7 FIG. 2 is a flow chart of the method for predicting the probability of scoliosis provided by the present invention. Figure 7 As shown, step 106 specifically includes steps 201-202: Step 201: Perform weighted averaging on all first area ratios to obtain a weighted average value.

[0070] In step 201, a comprehensive weighted average value is obtained by weighted averaging all first area ratios to uniformly quantify the feature differences of different detection areas. The core of this step is to comprehensively consider the importance and contribution weight of each detection area, integrate multiple local features into an overall indicator, and thus provide a standardized basis for the evaluation of scoliosis probability.

[0071] In a specific implementation, firstly, the first area ratios of all detection areas calculated in step 105 are collected, including the upper area, the middle area, and the lower area. Since the roles of different detection areas in the overall analysis may be different, it is necessary to assign different weights to the area ratios of each area. For example, the upper area may have a more significant effect on the symmetry of the upper back shape, while the lower area more reflects the overall structural balance. These weights are pre-set based on actual needs and are usually obtained through experimental data and statistical analysis.

[0072] When performing weighted processing, the system uses the following formula: ; in, represents the area ratio of the i-th detection area, represents the weight corresponding to the area, and W is the weighted average. The weights of all areas satisfy In the specific calculation process, the system automatically reads the area ratio of each detection area in combination with the algorithm, and calculates according to the preset weight to ensure the efficiency and accuracy of the results.

[0073] A comprehensive weighted average value is generated by weighted averaging the area ratios of each detection area. This value can fully reflect the contribution of multiple local areas in the overall morphological analysis, avoiding the situation where the characteristics of a single area excessively affect the overall evaluation. Compared with simple average calculation, weighted average takes into account the differences in regional importance, making the evaluation results more scientific and practical. Finally, the weighted average serves as the core input for determining the probability of scoliosis in the subsequent step 202, laying a technical foundation for the accuracy and consistency of the results.

[0074] Step 202: Determine the corresponding scoliosis probability according to the preset range interval that the weighted average falls into, wherein each preset range interval corresponds to a preset scoliosis probability interval.

[0075] In step 202, the corresponding scoliosis probability is determined by comparing the weighted average obtained in step 201 with the preset range. The core of this step is to convert the calculated comprehensive index into an intuitive probability result to achieve a quantitative assessment of the back morphological characteristics and clarify the risk level, thereby providing an operational decision-making basis for subsequent analysis.

[0076] In the specific implementation, first, the weighted average value is classified into the corresponding interval range according to the preset range interval. These preset range intervals are pre-determined based on a large amount of experimental data and statistical analysis, and each interval is associated with a clear probability or evaluation result. For example, the system divides the weighted average value into several intervals, such as [0.8, 1] represents a low probability interval, [0.5, 0.8) represents a medium probability interval, and [0, 0.5) represents a high probability interval. The closer the weighted average value is to 1, the higher the similarity between the first sub-area and the second sub-area divided by the corresponding reference line for any detection area, that is, the lower the probability of scoliosis. The system uses a threshold comparison algorithm to compare the calculated value with these intervals one by one, and quickly locates the corresponding interval.

[0077] To improve the resolution and flexibility of the assessment, the system uses interpolation calculations near the interval boundaries. For example, when the weighted average is close to the boundary of two intervals, the assessment results are further refined through linear interpolation or other interpolation algorithms. This dynamic adjustment mechanism can effectively reduce the uncertainty caused by measurement errors or boundary effects.

[0078] In one possible implementation, refer to Figure 8 , Figure 8 FIG. 3 is a flow chart of the method for predicting the probability of scoliosis provided by the present invention. Figure 8 As shown, the method further includes steps 301-304: Step 301: when the fourth bilateral position point cannot be extracted from the back contour, the foot shape information and the plantar pressure information of the person to be tested are obtained.

[0079] In step 301, by obtaining the foot shape information and the plantar pressure information of the person to be tested, alternative input data is provided when the fourth bilateral position point cannot be extracted from the back contour. The core of this step is to use the correlation between the plantar data and the overall structure of the body to make up for the lack of data caused by the lack of local features of the back, and ensure the integrity and accuracy of the scoliosis probability assessment.

[0080] In the specific implementation, the person to be tested is required to take off his shoes and socks and stand naturally on the plantar pressure plate with both feet to ensure the authenticity and reliability of the measurement. The system first collects the plantar pressure information of the person to be tested through the plantar pressure plate, including the pressure distribution in the front and back areas of the plantar. At the same time, the foot type information is obtained using a foot scanning device, and the detailed features of the plantar contour, including the area of ​​the front and back areas of the plantar, are captured through three-dimensional point cloud technology.

[0081] During the acquisition process, the system uses a multi-touch pressure sensor to monitor the plantar pressure distribution in real time, and combines the foot shape information to correct the accuracy of the plantar area division. For example, the plantar data is divided into the front area and the back area, and the geometric boundaries and pressure ranges of the two are clarified. To improve data quality, the system filters the collected pressure information and foot shape contour data to remove possible noise or abnormal data to ensure measurement accuracy.

[0082] By obtaining foot shape information and plantar pressure information, it is possible to provide effective alternative feature input when the back contour data is insufficient. Plantar data can not only reflect the overall balance of human posture, but also further support the analysis of spinal features by comparing regional area and pressure distribution.

[0083] Step 302: Determine the area of ​​the front region of the sole and the area of ​​the rear region of the sole according to the foot shape information.

[0084] In step 302, the front area and the rear area of ​​the sole are determined according to the foot shape information to quantify the structural characteristics of the sole, providing key input data for the subsequent analysis of the probability of scoliosis. The core of this step is to use the geometric distribution of the foot shape data to analyze the proportional relationship between the front and rear areas, thereby indirectly reflecting the distribution of the center of gravity and posture characteristics of the body.

[0085] In a specific implementation, the foot shape information obtained in step 301 is used to divide the sole area into the front area and the back area. The division is based on the geometric features of the foot shape, for example, the line connecting the front edge point and the heel point of the sole is used as the vertical axis reference line, and combined with the foot length ratio (such as the front and back areas are divided according to 1 / 3 of the foot length or other ratios) to determine the boundaries of the front area and the back area. The system extracts the boundary curve of each area through the dense distribution of point cloud data, and closes the boundary curve to generate a polygonal contour.

[0086] After obtaining the contours of the front and back areas, the polygon area calculation formula is used to accurately calculate the front and back area through numerical integration or grid division. In order to improve the accuracy of the calculation, the system introduces redundant sampling and multi-frame overlay processing to smooth the local irregular points of the foot data to ensure the stability of the area calculation results.

[0087] By calculating the area of ​​the front and back regions of the foot, the geometric characteristics of the weight-bearing area of ​​the sole can be quantified, providing important parameters for subsequent analysis. The proportional relationship between the area of ​​the front region and the area of ​​the back region of the sole can indirectly reflect the anterior-posterior distribution characteristics of the body's center of gravity, especially when the posture is asymmetric or the center of gravity is offset, the ratio will change significantly. Compared with the traditional single analysis method that relies on back data, this step introduces the geometric characteristics of the foot as a supplement, which not only improves the adaptability of the analysis, but also enhances the robustness and integrity of the evaluation model, providing reliable support for spinal analysis in complex data scenarios.

[0088] Step 303: Determine the front plantar pressure value and the contralateral rear plantar pressure value according to the plantar pressure information.

[0089] In step 303, the plantar pressure value of the front and the contralateral rear pressure values ​​are determined according to the plantar pressure information to quantify the plantar pressure distribution characteristics, providing key support for further analysis of the probability of scoliosis. The core of this step is to use the regional distribution of plantar pressure to indirectly reflect the balance of body posture and the deviation of the center of gravity, thereby making up for the lack of analysis caused by the missing back feature points.

[0090] In a specific implementation, the plantar pressure information obtained in step 301 is used to divide the data collected by the plantar pressure plate into a front pressure area and a rear pressure area. The division basis is consistent with the front and rear areas of the foot type information to ensure that the pressure area matches the geometric area. During the data processing process, the system extracts the pressure value of the corresponding grid through the distribution grid of the pressure sensor, and accumulates all the pressure values ​​in the front area and the rear area to obtain the front pressure value and the rear pressure value of the plantar, respectively.

[0091] To further evaluate the left-right balance, the system further divides the rear pressure area into left and right areas, and calculates the pressure values ​​of the left and right rear areas respectively. Combined with the symmetry analysis of the back shape, the system calculates the difference between the pressure of the left rear area and the pressure of the right rear area to obtain the pressure value of the contralateral rear foot. During the entire calculation process, the system filters abnormal pressure points (such as abnormal values ​​that significantly deviate from the surrounding data) to ensure the accuracy and stability of the calculation results.

[0092] By calculating the pressure values ​​of the front of the plantar and the pressure values ​​of the contralateral rear foot, the distribution characteristics of the body's center of gravity on the plantar can be intuitively reflected. In the case of asymmetric posture or scoliosis, the plantar pressure distribution usually shows an imbalance in the proportion of the front and back regions or an abnormal pressure difference between the left and right regions. This step introduces the pressure distribution characteristics into the analysis model, effectively supplements the deficiencies of the back data, and provides an important basis for the multi-dimensional feature quantification of spinal analysis. Compared with the traditional analysis method that only relies on a single morphological feature, this step significantly improves the system's sensitivity to complex posture changes and analysis accuracy by introducing pressure distribution data, laying a solid foundation for subsequent probability evaluation.

[0093] Step 304: Obtain the probability of scoliosis according to the second area ratio of the front area of ​​the plantar to the rear area of ​​the plantar, and the pressure ratio of the front pressure value of the plantar to the pressure value of the contralateral rear foot.

[0094] In step 304, the probability of scoliosis is finally determined based on the second area ratio of the front area of ​​the sole to the rear area, and the pressure ratio of the front pressure value of the sole to the pressure value of the opposite rear foot. The core of this process is to comprehensively utilize the foot shape information and the sole pressure information, combine the proportional characteristics of the two, and quantitatively analyze the balance and structural symmetry of the body posture, so as to make up for the evaluation of the probability of scoliosis when the back data is insufficient.

[0095] In a specific implementation, the system first combines the front area and rear area of ​​the plantar region calculated in step 302 to calculate the second area ratio using the following formula: second area ratio = front area of ​​the plantar region / rear area of ​​the plantar region.

[0096] Then, according to the anterior plantar pressure value and the contralateral hindfoot pressure value obtained in step 303, a pressure ratio is calculated: pressure ratio=anterior plantar pressure value / contralateral hindfoot pressure value.

[0097] After completing the calculation of the two ratios, the system inputs the second area ratio and the pressure ratio into the preset probability model for joint analysis. The probability model is trained with a large amount of experimental data to establish the correspondence between the ratio change and the spinal posture change. For example, the system divides the ratio into low probability area, medium probability area and high probability area according to a specific threshold range. When the ratio deviates significantly from symmetry, the system identifies it as a high probability area. For boundary cases, the system can introduce interpolation methods to further optimize the probability calculation results to ensure the continuity and stability of the results.

[0098] Reference Fig. 9 , Fig. 9 Schematic diagram of the structure of the scoliosis probability prediction system provided by the present invention, the system comprises: An acquisition module, used for acquiring a back image of the person to be tested; A first processing module is used to generate a back contour according to the back image of the body, and extract multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to represent a reference point of the spinal posture, and each group of feature points is used to represent a relevant point for scoliosis detection; A second processing module, configured to generate at least one detection area, wherein each detection area is surrounded by at most one reference point and at least one set of feature points; The second processing module is further used to generate a reference line corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; A third processing module, configured to divide any detection area into a first sub-area and a second sub-area through a corresponding reference line, and calculate a first area ratio between the first sub-area and the second sub-area; The fourth processing module is used to obtain the probability of scoliosis according to the first area ratio of all detection areas.

[0099] In a possible implementation manner, the second processing module is further configured to: Based on the upper reference point, draw a vertical line downward to obtain the first reference line; Based on the lower reference point, draw a vertical line upward to obtain the second reference line; The upper area corresponds to the first reference line, and the middle area and the lower area both correspond to the second reference line.

[0100] In a possible implementation manner, the fourth processing module is further configured to: Perform weighted averaging on all first area ratios to obtain a weighted average value; The corresponding scoliosis probability is determined according to the preset range interval that the weighted average falls into, wherein each preset range interval corresponds to a preset scoliosis probability interval.

[0101] In a possible implementation, the system further includes a fifth processing module; The acquisition module is also used to acquire the foot shape information and the plantar pressure information of the person to be tested when the fourth bilateral position point cannot be extracted from the back contour; A fifth processing module, used to determine the area of ​​the front area of ​​the sole and the area of ​​the rear area of ​​the sole according to the foot shape information; The fifth processing module is further used to determine the front plantar pressure value and the contralateral rear plantar pressure value according to the plantar pressure information; The fifth processing module is further used to obtain the probability of scoliosis according to the second area ratio of the front area of ​​the plantar to the rear area of ​​the plantar, and the pressure ratio of the front pressure value of the plantar to the pressure value of the contralateral rear foot.

[0102] It should be noted that the scoliosis probability prediction system provided by the present invention can execute the scoliosis probability prediction method of any of the above-mentioned embodiments during specific operation, which will not be elaborated in this embodiment.

[0103] Fig.10 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Fig.10 As shown, the electronic device may include: a processor 1010 (processor), a communication interface 1020 (CommunicationsInterface), a memory 1030 (memory) and a communication bus 1040 , wherein the processor 1010 , the communication interface 1020 , and the memory 1030 communicate with each other through the communication bus 1040 . The processor 1010 can call the logic instructions in the memory 1030 to execute a scoliosis probability prediction method, which includes: obtaining a back image of the body of the person to be tested; generating a back contour based on the back image of the body, and extracting multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to characterize the reference point of the spinal posture, and each group of feature points is used to characterize the relevant points of scoliosis detection; generating at least one detection area, wherein each detection area is surrounded by at most one reference point and at least one group of feature points; generating reference lines corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; for any detection area, dividing the detection area into a first sub-area and a second sub-area by the corresponding reference line, and calculating the first area ratio of the first sub-area to the second sub-area; and obtaining the scoliosis probability according to the first area ratio of all detection areas.

[0104] In addition, the logic instructions in the above-mentioned memory 1030 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0105] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the scoliosis probability prediction method provided by the above-mentioned embodiments, and the method includes: obtaining a back image of the body of the person to be tested; generating a back contour based on the back image of the body, and extracting multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to characterize a reference point of the spinal posture, and each group of feature points is used to characterize the relevant points of scoliosis detection; generating at least one detection area, wherein each detection area is surrounded by at most one reference point and at least one group of feature points; generating a reference line corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; for any detection area, the detection area is divided into a first sub-area and a second sub-area by the corresponding reference line, and the first area ratio of the first sub-area to the second sub-area is calculated; according to the first area ratio of all detection areas, the scoliosis probability is obtained.

[0106] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented by a processor to execute the scoliosis probability prediction method provided by the above-mentioned embodiments when the computer program is executed, the method comprising: obtaining a back image of the body of the person to be tested; generating a back contour based on the back image of the body, and extracting multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to characterize a reference point of the spinal posture, and each group of feature points is used to characterize relevant points for scoliosis detection; generating at least one detection area, wherein each detection area is surrounded by at most one reference point and at least one group of feature points; generating reference lines corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; for any detection area, dividing the detection area into a first sub-area and a second sub-area by the corresponding reference line, and calculating a first area ratio between the first sub-area and the second sub-area; and obtaining the scoliosis probability according to the first area ratio of all detection areas.

[0107] The system embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0108] Through the description of the above implementation modes, those skilled in the art can clearly understand that each implementation mode can be implemented by means of software plus a necessary general hardware platform, or of course by hardware. Based on such an understanding, the above technical solution can essentially or in other words be embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the probability of scoliosis, characterized in that: include: Acquire a back image of the body of the person to be tested; Generate a back contour according to the back image of the body, and extract multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to characterize a reference point of the spinal posture, and each group of feature points is used to characterize a relevant point for scoliosis detection; generating at least one detection area, wherein each of the detection areas is surrounded by at most one of the reference points and at least one group of the feature points; Generate a reference line corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; For any one of the detection areas, the detection area is divided into a first sub-area and a second sub-area by a corresponding reference line, and a first area ratio of the first sub-area to the second sub-area is calculated; The probability of scoliosis is obtained according to the first area ratio of all the detection areas.

2. The method for predicting scoliosis probability according to claim 1, characterized in that: The reference points include an upper reference point and a lower reference point, and each group of feature points includes a first bilateral position point, a second bilateral position point, a third bilateral position point, and a fourth bilateral position point; Among them, in the vertical direction of the back contour, from top to bottom are the upper end reference point, the first bilateral position point, the second bilateral position point, the third bilateral position point, the fourth bilateral position point and the lower end reference point.

3. The method for predicting scoliosis probability according to claim 2, characterized in that: The detection area includes an upper area, a middle area and a lower area; The upper section area is surrounded by the upper end reference point, the first bilateral position point, the second bilateral position point and the third bilateral position point; The middle section area is surrounded by the third bilateral position point and the fourth bilateral position point; The lower section area is surrounded by the fourth bilateral position point and a first horizontal line of the back contour, wherein the first horizontal line passes through the lower end reference point.

4. The method for predicting scoliosis probability according to claim 3, characterized in that: The method generates a reference line corresponding to each detection area, specifically comprising: According to the upper reference point, a vertical line is drawn downward to obtain a first reference line; Based on the lower reference point, draw a vertical line upward to obtain a second reference line; The upper area corresponds to the first reference line, and the middle area and the lower area both correspond to the second reference line.

5. The method for predicting scoliosis probability according to claim 1, characterized in that: The step of obtaining the probability of scoliosis according to the first area ratio of all the detection areas specifically includes: Taking a weighted average of all the first area ratios to obtain a weighted average value; The corresponding scoliosis probability is determined according to the preset range interval that the weighted average value falls into, wherein each of the preset range intervals corresponds to a preset scoliosis probability interval.

6. The method for predicting scoliosis probability according to claim 2, characterized in that: The method further comprises: When the fourth bilateral position point cannot be extracted from the back contour, obtaining the foot shape information and the plantar pressure information of the person to be tested; Determine the area of ​​the front area of ​​the sole and the area of ​​the rear area of ​​the sole according to the foot shape information; Determine the front plantar pressure value and the contralateral rear plantar pressure value according to the plantar pressure information; The probability of scoliosis is obtained according to a second area ratio of the front area of ​​the plantar to the rear area of ​​the plantar, and a pressure ratio of the front pressure value of the plantar to the pressure value of the contralateral rear foot.

7. A scoliosis probability prediction system, characterized in that: include: An acquisition module, used for acquiring a back image of the person to be tested; A first processing module is used to generate a back contour according to the back image of the body, and extract multiple reference points and multiple groups of feature points from the back contour, wherein each reference point is used to represent a reference point of the spinal posture, and each group of feature points is used to represent a relevant point for scoliosis detection; A second processing module, configured to generate at least one detection area, wherein each detection area is surrounded by at most one reference point and at least one group of feature points; The second processing module is further used to generate a reference line corresponding to each detection area, wherein each reference line corresponds to a reference point, and each reference line is obtained by drawing a perpendicular line along a preset direction from the corresponding reference point; A third processing module, configured to divide any one of the detection areas into a first sub-area and a second sub-area through a corresponding reference line, and calculate a first area ratio between the first sub-area and the second sub-area; The fourth processing module is used to obtain the probability of scoliosis according to the first area ratio of all the detection areas.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the scoliosis probability prediction method according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the probability of scoliosis as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting the probability of scoliosis as described in any one of claims 1 to 6 is implemented.

Images