Scoliosis recognition system and method based on active millimeter wave imaging technology

Abstract

The present invention relates to scoliosis recognition, and specifically to a scoliosis recognition system and method based on active millimeter wave imaging technology, an active millimeter wave scanning component, which transmits millimeter wave signals of a specific frequency band to the back of a human body, receives echo signals from the back of the human body, and sends the received echo signals to a terminal control device, thereby realizing non-invasive scanning of the back of the human body; an active millimeter wave flat-panel device, which drives the active millimeter wave scanning component to scan in a top-to-bottom order under the control of the terminal control device, thereby obtaining three-dimensional morphological data of the back of the human body; a pressure sensing device, which measures pressure distribution data of a human body when standing, and sends the collected pressure distribution data to the terminal control device; the technical solution provided by the present invention can effectively overcome the defects of the prior art in terms of low accuracy and efficiency of scoliosis recognition when facing high-throughput and rapid periodic screening.

Description

Technical Field

[0001] The present invention relates to scoliosis recognition, and in particular to a scoliosis recognition system and method based on active millimeter wave imaging technology. Background Art

[0002] In the medical field, idiopathic scoliosis is a common disease among adolescents, and its severity cannot be ignored. The most notable feature of this disease is that the spine is abnormally curved in the front view, which not only affects the patient's body appearance, but may also cause back pain, breathing difficulties and other serious complications. Therefore, timely detection and treatment of scoliosis is crucial to protecting the patient's health and improving their quality of life.

[0003] Although significant progress has been made in the diagnosis and treatment of scoliosis, in clinical practice, the symptoms of mild to moderate scoliosis are not easy to detect and are often ignored by patients and their parents. In addition, schools and medical staff lack knowledge of this disease, and misdiagnosis occurs frequently, which undoubtedly increases the risk of worsening of the disease.

[0004] Currently, medical imaging tests, such as X-rays and CT scans, play a key role in the diagnosis of scoliosis. However, these detection methods will inevitably cause a certain degree of ionizing radiation damage to patients, especially for children and adolescents in the growth and development stage. In addition, due to its cost and operational complexity, medical imaging technology is not suitable as a routine monitoring tool.

[0005] Although B-ultrasound has certain application value in scoliosis detection, it is rarely used in actual daily monitoring, mainly because there is a lack of physicians with professional knowledge to perform the operation and interpret the test results, which limits the application of B-ultrasound technology in large-scale screening and daily monitoring.

[0006] As a widely recommended screening method, the horizontal ruler flexion measurement method is favored because of its convenience. However, the accuracy of this method is low and it is easily affected by factors such as human body shaking, resulting in errors in the test results. In addition, this method relies on manual operation and is time-consuming.

[0007] As a non-invasive detection method, binocular vision detection can reduce the dependence on manpower to a certain extent, but it also has some limitations, such as being easily affected by external factors such as changes in ambient light and differences in detection angles. Similar to the horizontal ruler flexion measurement method, binocular vision detection is also easily affected by human body jitter and interference, which limits its application in daily monitoring.

[0008] In summary, the currently used scoliosis detection methods all have certain limitations, especially when faced with the need for high-throughput and rapid periodic screening, the shortcomings of these methods become particularly obvious. Therefore, the development of new scoliosis detection technologies to improve the accuracy and efficiency of scoliosis identification has become an urgent problem that needs to be solved. Summary of the invention

[0009] 1. Technical issues to be solved

[0010] In view of the above-mentioned shortcomings of the prior art, the present invention provides a scoliosis identification system and method based on active millimeter wave imaging technology, which can effectively overcome the defects of the prior art in terms of low accuracy and efficiency of scoliosis identification when facing high-throughput and rapid periodic screening.

[0011] (II) Technical solution

[0012] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0013] A scoliosis recognition system based on active millimeter wave imaging technology, including an active millimeter wave scanning component, an active millimeter wave flat panel device, a pressure sensing device, a plantar indication module, a terminal display device, and a terminal control device;

[0014] The active millimeter wave scanning component transmits millimeter wave signals of a specific frequency band to the back of the human body, receives echo signals from the back of the human body, and sends the received echo signals to the terminal control device to realize non-invasive scanning of the back of the human body;

[0015] The active millimeter wave flat panel device drives the active millimeter wave scanning component to scan from top to bottom under the control of the terminal control device to obtain three-dimensional morphological data of the back of the human body;

[0016] A pressure sensing device measures the pressure distribution data of a human body when standing, and sends the collected pressure distribution data to a terminal control device;

[0017] The terminal control device evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module, and detects whether the human spine is scoliotic based on the echo signal, and generates a corresponding diagnosis report according to the scoliosis recognition result, and sends the scoliosis recognition result and the diagnosis report to the terminal display device together;

[0018] The terminal display device shows the scanning progress, device status, and scoliosis identification results and diagnosis reports in real time.

[0019] Preferably, the terminal control device comprises a pressure distribution data preprocessing module, a pressure distribution data analysis module, a key position detection module and a standing posture and position assessment module;

[0020] The pressure distribution data preprocessing module digitizes the received pressure distribution data, filters out noise, and converts it into a data format suitable for algorithm analysis through normalization;

[0021] The pressure distribution data analysis module performs pressure distribution analysis on both feet based on the pre-processed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution;

[0022] The key position detection module detects the heel, arch and forefoot positions of the human body based on the pre-processed pressure distribution data, and compares and analyzes the position with the foot position mark on the pressure sensing device;

[0023] The standing posture assessment module evaluates whether the standing posture of the human body is standard by combining the pressure distribution symmetry and the standing position comparison analysis results. If the standing posture is not standard, it generates corresponding feedback information and sends the feedback information to the sole indication module, which guides the user to adjust the standing posture.

[0024] Among them, the plantar indication module guides the user to adjust his or her standing posture through voice or visual prompts.

[0025] Preferably, the pressure distribution data analysis module performs pressure distribution analysis on both feet based on the preprocessed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution, including:

[0026] The pressure distribution of the feet is calculated using the following formula:

[0027]

[0028]

[0029] Among them, C L (t) is the pressure value distribution of the left foot at time t, P i (t) is the pressure value detected by the i-th pressure sensor under the left foot at time t, x i is the coordinate of the i-th pressure sensor under the left foot on the plane, and m is the number of pressure sensors under the left foot;

[0030] C R (t) is the right foot pressure value distribution at time t, P j (t) is the pressure value detected by the jth pressure sensor under the right foot at time t, y j is the coordinate of the jth pressure sensor under the right foot on the plane, and n is the number of pressure sensors under the right foot;

[0031] The weight of the center of gravity of the human body is calculated based on the pressure distribution of the two feet using the following formula:

[0032]

[0033]

[0034] Among them, W L (t), W R (t) are the weights of the left foot and right foot at time t, respectively, L , S R They are the left foot contact area and the right foot contact area respectively;

[0035] The pressure distribution symmetry score is calculated based on the center of gravity weight of the human body using the following formula:

[0036]

[0037] Among them, S(t) is the pressure distribution symmetry score at time t. When the pressure distribution symmetry score is higher than 0.85, it means that the standing posture of the human body is symmetrical and meets the posture symmetry requirement. Otherwise, the standing posture of the human body is asymmetrical and does not meet the posture symmetry requirement.

[0038] Preferably, the terminal control device includes an echo signal preprocessing module, a back key point recognition module, a spine fitting analysis module and a diagnosis report generation module;

[0039] The echo signal preprocessing module filters and removes noise from the received echo signal and converts it into three-dimensional morphological data of the human back;

[0040] The back key point recognition module uses the deep learning algorithm BlazePose to identify the back key points from the 3D morphological data of the human back, and then segments the 3D data of the human back;

[0041] The spine fitting analysis module uses an improved active contour algorithm to identify the initial curve of the spine contour based on the three-dimensional data of the human back, and then performs cubic B-spline curve smoothing to obtain the fitting curve of the spine contour;

[0042] A diagnostic report generation module detects whether the human spine is scoliotic based on the calculated Cobb angle of the spine, generates a corresponding diagnostic report based on the scoliosis recognition result, and sends the scoliosis recognition result and the diagnostic report together to a terminal display device;

[0043] Among them, the key points of the back include the bilateral acromion, scapula, lumbar dimples, and iliac crests. The diagnostic report includes the Cobb angle of the spine, the severity range of scoliosis, and corresponding medical recommendations.

[0044] The method for identifying scoliosis based on active millimeter wave imaging technology comprises the following steps:

[0045] S1. Start the active millimeter wave flat panel device and the active millimeter wave scanning component, the system enters the scanning state, and automatically performs the inspection and spatial calibration procedures to ensure the scanning accuracy;

[0046] S2, the user stands at the designated position according to the foot position mark on the pressure sensing device;

[0047] S3, the pressure sensing device measures the pressure distribution data of the human body when standing, and sends the collected pressure distribution data to the terminal control device;

[0048] S4. The terminal control device evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module;

[0049] S5. Repeat S3 to S4 until the standing posture meets the standard requirements;

[0050] S6. The active millimeter wave flat panel device and the active millimeter wave scanning component start to perform non-invasive scanning of the back of the human body. The active millimeter wave scanning component transmits a millimeter wave signal of a specific frequency band to the back of the human body, and sends the received echo signal of the back of the human body to the terminal control device;

[0051] S7, the terminal control device detects whether the human spine is scoliotic based on the echo signal, generates a corresponding diagnosis report according to the scoliosis identification result, and sends the scoliosis identification result and the diagnosis report to the terminal display device together;

[0052] S8. The terminal display device displays the scoliosis identification results and diagnosis reports in real time, and can be electronically transmitted or printed through the terminal control device for medical personnel to further analyze and formulate treatment plans.

[0053] Preferably, in S4, the terminal control device evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module, including:

[0054] S41, digitally processing the received pressure distribution data, filtering out noise, and converting it into a data format suitable for algorithm analysis through normalization processing;

[0055] S42, performing pressure distribution analysis on both feet based on the preprocessed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution;

[0056] S43, detecting the positions of the heel, arch and forefoot of the human body based on the preprocessed pressure distribution data, and performing position comparison analysis with the foot position mark on the pressure sensing device;

[0057] S44, evaluating whether the standing posture of the human body is standard based on the pressure distribution symmetry and the standing position comparison analysis results, generating corresponding feedback information when the standing posture is not standard, and sending the feedback information to the sole indication module, so that the sole indication module guides the user to adjust the standing posture;

[0058] Among them, the plantar indication module guides the user to adjust his or her standing posture through voice or visual prompts.

[0059] Preferably, in S7, the terminal control device detects whether the human spine is scoliotic based on the echo signal, generates a corresponding diagnosis report according to the scoliosis identification result, and sends the scoliosis identification result and the diagnosis report to the terminal display device together, including:

[0060] S71, filtering and denoising the received echo signal, and converting it into three-dimensional morphological data of the back of the human body;

[0061] S72, using the deep learning algorithm BlazePose to identify back key points from the three-dimensional morphological data of the human back, and then segmenting the three-dimensional data of the human back;

[0062] S73, based on the three-dimensional data of the human back, using an improved active contour algorithm to identify a preliminary curve of the spine contour, and then performing cubic B-spline curve smoothing to obtain a fitting curve of the spine contour;

[0063] S74, detecting whether the human spine is scoliotic based on the calculated Cobb angle of the spine, generating a corresponding diagnosis report according to the scoliosis recognition result, and sending the scoliosis recognition result and the diagnosis report together to the terminal display device;

[0064] Among them, the key points of the back include the bilateral acromion, scapula, lumbar dimples, and iliac crests. The diagnostic report includes the Cobb angle of the spine, the severity range of scoliosis, and corresponding medical recommendations.

[0065] Preferably, in S73, based on the three-dimensional data of the human back, an improved active contour algorithm is used to identify a preliminary curve of the spine contour, and then cubic B-spline curve smoothing is performed to obtain a fitting curve of the spine contour, including:

[0066] S731, initializing a curve at the central axis position of the three-dimensional data of the back of the human body as an initial curve approximating the actual spine contour;

[0067] S732, the active contour algorithm constructs an energy function including an internal energy term and an external energy term, wherein the internal energy maintains the smoothness and structure of the curve, and the external energy is based on the image gradient information and utilizes the gradient size to attract the initial curve to the edge of the spine;

[0068] S733, updating the position of the initial curve according to the gradient descent direction of the energy function, and minimizing the energy function by iteration so that the initial curve gradually approaches the true spine contour;

[0069] S734, setting the stop condition as the energy function reduction amplitude being lower than a threshold or the number of iterations reaching an upper limit, after the iteration is completed, performing outlier removal to obtain a preliminary curve of the spine contour;

[0070] S735. Use a cubic B-spline curve to smoothly obtain a fitting curve of the spine contour.

[0071] (III) Beneficial effects

[0072] Compared with the prior art, the scoliosis identification system and method based on active millimeter wave imaging technology provided by the present invention, by introducing active millimeter wave imaging technology, provides a radiation-free, high-accuracy, high-throughput, fast-periodic and user-friendly scoliosis identification system and method. Based on the principle of millimeter wave radar, the three-dimensional morphological data of the human back is obtained by transmitting and receiving millimeter wave signals, and then the deep learning algorithm and the surface fitting algorithm are used to respectively identify the key points of the back and analyze the three-dimensional morphology of the spine, and finally the Cobb angle of the spine is calculated and the corresponding diagnosis report is generated. This method not only improves the accuracy and efficiency of scoliosis identification, but also helps to reduce the risk of missed diagnosis and misdiagnosis, while protecting the privacy of users, and laying a solid foundation for the early treatment of scoliosis patients and the improvement of their quality of life.

[0073] The present invention specifically includes the following beneficial effects:

[0074] 1) Protect user privacy: Using active millimeter wave imaging technology to detect scoliosis does not require users to undress or take a specific posture, thus protecting user privacy;

[0075] 2) No direct contact required: Millimeter wave imaging technology allows non-invasive scanning within a certain distance, without the need for direct physical contact between medical personnel and users, which respects user privacy while reducing the workload of medical personnel;

[0076] 3) Convenient detection process: The entire scanning process is fast and automated. Users only need to stand on the scanning platform according to the instructions without any other complicated coordination, making the detection process more convenient;

[0077] 4) Reduce human errors: Due to the use of automated scanning and analysis processes, the reliance on operators is reduced, thus avoiding the occurrence of inaccurate scoliosis identification caused by human factors;

[0078] 5) Instant result feedback: The system can provide scoliosis identification results and diagnostic reports instantly, without users having to wait for a long time, and can meet the needs of high-throughput and rapid periodic screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0080] Figure 1 It is a system schematic diagram of the present invention;

[0081] Figure 2 For the present invention Figure 1 A schematic diagram of the structure of the pressure sensing device from above;

[0082] Figure 3 This is a schematic diagram of the back key point recognition result in the present invention;

[0083] Figure 4 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0084] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments 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.

[0085] Scoliosis recognition system based on active millimeter wave imaging technology, such as Figure 1 As shown, it includes an active millimeter wave scanning component 1, an active millimeter wave flat panel device 2, a pressure sensing device 3, a foot sole indication module 4, a terminal display device 5 and a terminal control device 6;

[0086] The active millimeter wave scanning component 1 transmits a millimeter wave signal of a specific frequency band (Ka band) to the back of the human body, receives an echo signal from the back of the human body, and sends the received echo signal to the terminal control device 6, thereby realizing non-invasive scanning of the back of the human body;

[0087] The active millimeter wave flat panel device 2 drives the active millimeter wave scanning component 1 to scan from top to bottom under the control of the terminal control device 6 to obtain three-dimensional morphological data of the back of the human body;

[0088] Pressure sensing device 3, measuring the pressure distribution data of the human body when standing (such as Figure 2 As shown, two groups of pressure sensor arrays are used to collect pressure distribution data of two feet respectively), and the collected pressure distribution data is sent to the terminal control device 6;

[0089] The terminal control device 6 evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module 4, and detects whether the human spine is scoliotic based on the echo signal, and generates a corresponding diagnosis report according to the scoliosis recognition result, and sends the scoliosis recognition result and the diagnosis report to the terminal display device 5;

[0090] The terminal display device 5 displays the scanning progress, device status, and scoliosis identification results and diagnosis reports in real time.

[0091] ① The terminal control device 6 includes a pressure distribution data preprocessing module, a pressure distribution data analysis module, a key position detection module and a standing posture assessment module;

[0092] The pressure distribution data preprocessing module digitizes the received pressure distribution data, filters out noise, and converts it into a data format suitable for algorithm analysis through normalization processing (to eliminate the influence of different user weights);

[0093] The pressure distribution data analysis module performs pressure distribution analysis on both feet based on the pre-processed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution;

[0094] The key position detection module detects the heel, arch and forefoot positions of the human body based on the pre-processed pressure distribution data, and compares them with the foot position mark (such as the foot position mark) on the pressure sensing device 3. Figure 2 The station comparison analysis was performed as shown in Figure 1.

[0095] The standing posture assessment module assesses whether the standing posture of the human body is standard by combining the pressure distribution symmetry and the standing position comparison analysis results, generates corresponding feedback information when the standing posture is not standard, and sends the feedback information to the sole indication module 4, which guides the user to adjust the standing posture;

[0096] The sole indication module 4 guides the user to adjust his / her standing posture through voice or visual prompts (such as "please move your left foot forward a little" or "please straighten your waist", etc.).

[0097] Specifically, the pressure distribution data analysis module performs pressure distribution analysis on both feet based on the preprocessed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution, including:

[0098] The pressure distribution of the feet is calculated using the following formula:

[0099]

[0100]

[0101] Among them, C L (t) is the pressure value distribution of the left foot at time t, P i (t) is the pressure value detected by the i-th pressure sensor under the left foot at time t, x i is the coordinate of the i-th pressure sensor under the left foot on the plane, and m is the number of pressure sensors under the left foot;

[0102] C R (t) is the right foot pressure value distribution at time t, P j (t) is the pressure value detected by the jth pressure sensor under the right foot at time t, y j is the coordinate of the jth pressure sensor under the right foot on the plane, and n is the number of pressure sensors under the right foot;

[0103] The weight of the center of gravity of the human body is calculated based on the pressure distribution of the two feet using the following formula:

[0104]

[0105]

[0106] Among them, W L (t), W R (t) are the weights of the left foot and right foot at time t, respectively, L , S R They are the left foot contact area and the right foot contact area respectively;

[0107] The pressure distribution symmetry score is calculated based on the center of gravity weight of the human body using the following formula:

[0108]

[0109] Among them, S(t) is the pressure distribution symmetry score at time t. When the pressure distribution symmetry score is higher than 0.85, it means that the standing posture of the human body is symmetrical and meets the posture symmetry requirement. Otherwise, the standing posture of the human body is asymmetrical and does not meet the posture symmetry requirement.

[0110] ② The terminal control device 6 includes an echo signal preprocessing module, a back key point recognition module, a spine fitting analysis module and a diagnosis report generation module;

[0111] The echo signal preprocessing module filters and removes noise from the received echo signal and converts it into three-dimensional morphological data of the human back;

[0112] The back key point recognition module uses the deep learning algorithm BlazePose to identify the back key points (such as Figure 3 As shown), the three-dimensional data of the back of the human body is then segmented;

[0113] The spine fitting analysis module uses an improved active contour algorithm to identify the initial curve of the spine contour based on the three-dimensional data of the human back, and then performs cubic B-spline curve smoothing to obtain the fitting curve of the spine contour;

[0114] A diagnostic report generation module detects whether the human spine is scoliotic based on the calculated Cobb angle of the spine (an important parameter for measuring the degree of scoliosis), generates a corresponding diagnostic report based on the scoliosis identification result, and sends the scoliosis identification result and the diagnostic report together to the terminal display device 5;

[0115] Among them, the key points of the back include the bilateral acromion, scapula, lumbar dimples, and iliac crests. The diagnostic report includes the Cobb angle of the spine, the severity range of scoliosis, and corresponding medical recommendations.

[0116] The technical solution of this application also provides a scoliosis identification method based on active millimeter wave imaging technology, such as Figure 4 As shown, the following steps are included:

[0117] S1, start the active millimeter wave flat panel device 2 and the active millimeter wave scanning component 1, the system enters the scanning state, and automatically performs the inspection and spatial calibration procedures to ensure the scanning accuracy;

[0118] S2, the user stands at a designated position according to the foot position mark on the pressure sensing device 3;

[0119] S3, the pressure sensing device 3 measures the pressure distribution data of the human body when standing, and sends the collected pressure distribution data to the terminal control device 6;

[0120] S4, the terminal control device 6 evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module 4;

[0121] S5. Repeat S3 to S4 until the standing posture meets the standard requirements;

[0122] S6, the active millimeter wave flat panel device 2 and the active millimeter wave scanning component 1 start to perform non-invasive scanning of the back of the human body, the active millimeter wave scanning component 1 transmits a millimeter wave signal of a specific frequency band to the back of the human body, and sends the received echo signal of the back of the human body to the terminal control device 6;

[0123] S7, the terminal control device 6 detects whether the human spine is scoliotic based on the echo signal, generates a corresponding diagnosis report according to the scoliosis identification result, and sends the scoliosis identification result and the diagnosis report to the terminal display device 5;

[0124] S8. The terminal display device 5 displays the scoliosis identification result and diagnosis report in real time, and can be electronically transmitted or printed through the terminal control device 6 for medical personnel to further analyze and formulate a treatment plan.

[0125] ① In S4, the terminal control device 6 evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module 4, including:

[0126] S41, digitally processing the received pressure distribution data, filtering out noise, and converting it into a data format suitable for algorithm analysis through normalization processing;

[0127] S42, performing pressure distribution analysis on both feet based on the preprocessed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution;

[0128] S43, based on the pre-processed pressure distribution data, the heel, arch and forefoot positions of the human body are detected, and compared with the foot position mark (such as the foot type position mark) on the pressure sensing device 3. Figure 2 The station comparison analysis was performed as shown in Figure 1.

[0129] S44, evaluating whether the standing posture of the human body is standard in combination with the pressure distribution symmetry and the standing position comparison analysis results, generating corresponding feedback information when the standing posture is not standard, and sending the feedback information to the sole indication module 4, so that the sole indication module 4 guides the user to adjust the standing posture;

[0130] The sole indication module 4 guides the user to adjust his / her standing posture through voice or visual prompts.

[0131] ② In S7, the terminal control device 6 detects whether the human spine is scoliotic based on the echo signal, generates a corresponding diagnosis report according to the scoliosis identification result, and sends the scoliosis identification result and the diagnosis report to the terminal display device 5, including:

[0132] S71, filtering and denoising the received echo signal, and converting it into three-dimensional morphological data of the back of the human body;

[0133] S72, using the deep learning algorithm BlazePose to identify back key points from the three-dimensional morphological data of the human back, and then segmenting the three-dimensional data of the human back;

[0134] S73, based on the three-dimensional data of the human back, using an improved active contour algorithm to identify a preliminary curve of the spine contour, and then performing cubic B-spline curve smoothing to obtain a fitting curve of the spine contour;

[0135] S74, detecting whether the human spine is scoliotic based on the calculated Cobb angle of the spine, generating a corresponding diagnosis report according to the scoliosis identification result, and sending the scoliosis identification result and the diagnosis report together to the terminal display device 5;

[0136] Among them, the key points of the back include the bilateral acromion, scapula, lumbar dimples, and iliac crests. The diagnostic report includes the Cobb angle of the spine, the severity range of scoliosis, and corresponding medical recommendations.

[0137] Specifically, in S73, based on the three-dimensional data of the human back, an improved active contour algorithm is used to identify the preliminary curve of the spine contour, and then cubic B-spline curve smoothing is performed to obtain a fitting curve of the spine contour, including:

[0138] S731, initializing a curve at the central axis position of the three-dimensional data of the back of the human body as an initial curve approximating the actual spine contour;

[0139] S732, the active contour algorithm constructs an energy function including an internal energy term and an external energy term, wherein the internal energy maintains the smoothness and structure of the curve, and the external energy is based on the image gradient information and utilizes the gradient size to attract the initial curve to the edge of the spine;

[0140] S733, updating the position of the initial curve according to the gradient descent direction of the energy function, and minimizing the energy function by iteration so that the initial curve gradually approaches the true spine contour;

[0141] S734, setting the stop condition as the energy function reduction amplitude being lower than a threshold or the number of iterations reaching an upper limit, after the iteration is completed, performing outlier removal to obtain a preliminary curve of the spine contour;

[0142] S735. Use a cubic B-spline curve to smoothly obtain a fitting curve of the spine contour.

[0143] The technical solution of the present application, by introducing active millimeter wave imaging technology, provides a radiation-free, high-accuracy, high-throughput, fast-periodic and user-friendly scoliosis identification system and method. Based on the principle of millimeter wave radar, the three-dimensional morphological data of the human back is obtained by transmitting and receiving millimeter wave signals, and then the deep learning algorithm and the surface fitting algorithm are used to respectively identify the key points of the back and analyze the three-dimensional morphology of the spine. Finally, the Cobb angle of the spine is calculated and the corresponding diagnosis report is generated. This method not only improves the accuracy and efficiency of scoliosis identification, but also helps to reduce the risk of missed diagnosis and misdiagnosis, while protecting the privacy of users, laying a solid foundation for the early treatment of scoliosis patients and improving their quality of life.

[0144] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. The scoliosis recognition system based on active millimeter wave imaging technology is characterized by: It comprises an active millimeter wave scanning component (1), an active millimeter wave flat panel device (2), a pressure sensing device (3), a foot sole indication module (4), a terminal display device (5) and a terminal control device (6); An active millimeter wave scanning component (1) transmits a millimeter wave signal of a specific frequency band to the back of a human body, receives an echo signal from the back of the human body, and sends the received echo signal to a terminal control device (6), thereby realizing non-invasive scanning of the back of the human body; The active millimeter wave flat panel device (2) drives the active millimeter wave scanning component (1) to scan in a top-to-bottom order under the control of the terminal control device (6) to obtain three-dimensional morphological data of the back of the human body; A pressure sensing device (3) measures pressure distribution data of a human body when standing, and sends the collected pressure distribution data to a terminal control device (6); The terminal control device (6) evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module (4), and detects whether the human spine is scoliotic based on the echo signal, and generates a corresponding diagnosis report according to the scoliosis identification result, and sends the scoliosis identification result and the diagnosis report together to the terminal display device (5); A terminal display device (5) displays the scanning progress, device status, and scoliosis identification results and diagnosis reports in real time; The terminal control device (6) comprises an echo signal preprocessing module, a back key point recognition module, a spine fitting analysis module and a diagnosis report generation module; The echo signal preprocessing module filters and removes noise from the received echo signal and converts it into three-dimensional morphological data of the human back; The back key point recognition module uses the deep learning algorithm BlazePose to identify the back key points from the 3D morphological data of the human back, and then segments the 3D data of the human back; The spine fitting analysis module uses an improved active contour algorithm to identify the initial curve of the spine contour based on the three-dimensional data of the human back, and then performs cubic B-spline curve smoothing to obtain the fitting curve of the spine contour; A diagnostic report generating module detects whether the human spine is scoliotic based on the calculated Cobb angle of the spine, generates a corresponding diagnostic report based on the scoliosis identification result, and sends the scoliosis identification result and the diagnostic report together to a terminal display device (5); Among them, the key points of the back include the bilateral acromion, scapula, lumbar dimples, and iliac crests. The diagnostic report includes the Cobb angle of the spine, the severity range of scoliosis, and corresponding medical advice; The method uses an improved active contour algorithm to identify the preliminary curve of the spine contour based on the three-dimensional data of the human back, and then performs cubic B-spline curve smoothing to obtain a fitting curve of the spine contour, including: Initialize a curve at the center axis position of the three-dimensional data of the back of the human body as an initial curve approximating the real spine contour; The active contour algorithm constructs an energy function that includes internal energy terms and external energy terms. The internal energy maintains the smoothness and structure of the curve, and the external energy is based on the image gradient information and uses the gradient size to attract the initial curve to the edge of the spine. The initial curve position is updated according to the gradient descent direction of the energy function, and the energy function is iteratively minimized so that the initial curve gradually approaches the true spine contour; The stopping condition is set as the energy function reduction amplitude is lower than the threshold or the number of iterations reaches the upper limit. After the iteration is completed, the abnormal points are removed to obtain the preliminary curve of the spine contour; The cubic B-spline curve was used to smooth the spine contour.

2. The scoliosis identification system based on active millimeter wave imaging technology according to claim 1, characterized in that: The terminal control device (6) comprises a pressure distribution data preprocessing module, a pressure distribution data analysis module, a key position detection module and a standing posture and position evaluation module; The pressure distribution data preprocessing module digitizes the received pressure distribution data, filters out noise, and converts it into a data format suitable for algorithm analysis through normalization; The pressure distribution data analysis module performs pressure distribution analysis on both feet based on the pre-processed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution; A key position detection module detects the positions of the heel, arch and forefoot of the human body based on the pre-processed pressure distribution data, and performs position comparison analysis with the foot position mark on the pressure sensing device (3); The standing posture and position evaluation module evaluates whether the standing posture and position of the human body are standard by combining the pressure distribution symmetry and the standing position comparison analysis results, generates corresponding feedback information when the standing posture and position are not standard, and sends the feedback information to the sole indication module (4), so that the sole indication module (4) guides the user to adjust the standing posture and position; The sole indication module (4) guides the user to adjust his / her standing posture through voice or visual prompts.

3. The scoliosis identification system based on active millimeter wave imaging technology according to claim 2, characterized in that: The pressure distribution data analysis module performs pressure distribution analysis on both feet based on the preprocessed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution, including: The pressure distribution of the feet is calculated using the following formula: Among them, C L (t) is the pressure value distribution of the left foot at time t, P i (t) is the pressure value detected by the i-th pressure sensor under the left foot at time t, x i is the coordinate of the i-th pressure sensor under the left foot on the plane, and m is the number of pressure sensors under the left foot; C R (t) is the right foot pressure value distribution at time t, P j (t) is the pressure value detected by the jth pressure sensor under the right foot at time t, y j is the coordinate of the jth pressure sensor under the right foot on the plane, and n is the number of pressure sensors under the right foot; The weight of the center of gravity of the human body is calculated based on the pressure distribution of the two feet using the following formula: Among them, W L (t), W R (t) are the weights of the left foot and right foot at time t, respectively, L , S R They are the left foot contact area and the right foot contact area respectively; The pressure distribution symmetry score is calculated based on the center of gravity weight of the human body using the following formula: Among them, S(t) is the pressure distribution symmetry score at time t. When the pressure distribution symmetry score is higher than 0.85, it means that the standing posture of the human body is symmetrical and meets the posture symmetry requirement. Otherwise, the standing posture of the human body is asymmetrical and does not meet the posture symmetry requirement.

4. A scoliosis recognition method based on active millimeter wave imaging technology, applied to a scoliosis recognition system based on active millimeter wave imaging technology as claimed in claim 1, characterized in that: The following steps are involved: S1, starting the active millimeter wave flat panel device (2) and the active millimeter wave scanning component (1), the system enters a state ready for scanning, and automatically performs a check and spatial calibration procedure to ensure scanning accuracy; S2, the user stands at a designated position according to the foot position mark on the pressure sensing device (3); S3, the pressure sensing device (3) measures the pressure distribution data of the human body when standing, and sends the collected pressure distribution data to the terminal control device (6); S4, the terminal control device (6) evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module (4); S5. Repeat S3 to S4 until the standing posture meets the standard requirements; S6, the active millimeter wave flat panel device (2) and the active millimeter wave scanning component (1) start to perform non-invasive scanning of the back of the human body, the active millimeter wave scanning component (1) transmits a millimeter wave signal of a specific frequency band to the back of the human body, and sends the received echo signal of the back of the human body to the terminal control device (6); S7, the terminal control device (6) detects whether the human spine is scoliotic based on the echo signal, generates a corresponding diagnosis report according to the scoliosis identification result, and sends the scoliosis identification result and the diagnosis report together to the terminal display device (5); S8, the terminal display device (5) displays the scoliosis identification result and diagnosis report in real time, and can be electronically transmitted or printed through the terminal control device (6) for medical personnel to further analyze and formulate a treatment plan; In S7, the terminal control device (6) detects whether the human spine is scoliotic based on the echo signal, generates a corresponding diagnosis report according to the scoliosis identification result, and sends the scoliosis identification result and the diagnosis report to the terminal display device (5), including: S71, filtering and denoising the received echo signal, and converting it into three-dimensional morphological data of the back of the human body; S72, using the deep learning algorithm BlazePose to identify back key points from the three-dimensional morphological data of the human back, and then segmenting the three-dimensional data of the human back; S73, based on the three-dimensional data of the human back, using an improved active contour algorithm to identify a preliminary curve of the spine contour, and then performing cubic B-spline curve smoothing to obtain a fitting curve of the spine contour; S74, detecting whether the human spine is scoliotic based on the calculated Cobb angle of the spine, generating a corresponding diagnosis report according to the scoliosis identification result, and sending the scoliosis identification result and the diagnosis report together to the terminal display device (5); Among them, the key points of the back include the bilateral acromion, scapula, lumbar dimples, and iliac crests. The diagnostic report includes the Cobb angle of the spine, the severity range of scoliosis, and corresponding medical advice; In S73, based on the three-dimensional data of the human back, an improved active contour algorithm is used to identify the preliminary curve of the spine contour, and then cubic B-spline curve smoothing is performed to obtain the fitting curve of the spine contour, including: S731, initializing a curve at the central axis position of the three-dimensional data of the back of the human body as an initial curve approximating the actual spine contour; S732, the active contour algorithm constructs an energy function including an internal energy term and an external energy term, wherein the internal energy maintains the smoothness and structure of the curve, and the external energy is based on the image gradient information and utilizes the gradient size to attract the initial curve to the edge of the spine; S733, updating the position of the initial curve according to the gradient descent direction of the energy function, and minimizing the energy function by iteration so that the initial curve gradually approaches the true spine contour; S734, setting the stop condition as the energy function reduction amplitude being lower than a threshold or the number of iterations reaching an upper limit, after the iteration is completed, performing outlier removal to obtain a preliminary curve of the spine contour; S735. Use a cubic B-spline curve to smoothly obtain a fitting curve of the spine contour.

5. The method for identifying scoliosis based on active millimeter wave imaging technology according to claim 4, characterized in that: The terminal control device (6) in S4 evaluates whether the standing posture of the human body is standard based on the pressure distribution data, generates corresponding feedback information when the standing posture is not standard, and guides the user to adjust the standing posture through the sole indication module (4), including: S41, digitally processing the received pressure distribution data, filtering out noise, and converting it into a data format suitable for algorithm analysis through normalization processing; S42, performing pressure distribution analysis on both feet based on the preprocessed pressure distribution data to determine the weight of the center of gravity of the human body and the symmetry of the pressure distribution; S43, detecting the positions of the heel, arch and forefoot of the human body based on the pre-processed pressure distribution data, and performing position comparison analysis with the foot position mark on the pressure sensing device (3); S44, evaluating whether the standing posture of the human body is standard by combining the pressure distribution symmetry and the standing position comparison analysis results, generating corresponding feedback information when the standing posture is not standard, and sending the feedback information to the sole indication module (4), so that the sole indication module (4) guides the user to adjust the standing posture; The sole indication module (4) guides the user to adjust his / her standing posture through voice or visual prompts.

Images