Processor and processing method for lower limb arterial pulse examination operation training

Abstract

The invention relates to a processor and a processing method for lower limb arterial pulse examination operation training. The processor receives first gesture image information and first overall image information, which are acquired by the first image acquisition unit, of the arterial pulse gesture examined by the operator, and second gesture image information and second overall image information, which are acquired by the second image acquisition unit from the image acquisition angle, of the arterial pulse gesture examined by the operator; the processor performs merging processing based on the first gesture image information and the second gesture image information to obtain data of operation action skeleton points; on the basis of combination processing of the first overall image information and the second overall image information, an arm force application angle of arm force application in the pulsation measurement process is obtained, and the arm force application angle is compared with a standard arm force application angle range to determine whether the action is correct or not; and comparing a skeleton line formed based on the skeleton points and the skeleton point connecting line with a pre-stored standard gesture so as to output a result whether the gesture is correct or not.

Description

[0001] The original basis of this divisional application is the patent application with application number 202410062596.1, application date January 16, 2024, and invention name “A teaching and assessment system for lower limb arterial pulsation examination operation”. Technical Field

[0002] The present invention relates to the field of medical examination and nursing teaching, and in particular to a processor and a processing method for lower limb arterial pulse examination operation training. Background Art

[0003] For teaching aids for lower limb examinations, existing technology has designed molds that can simulate the human lower limbs. These molds can simulate the appearance and structure of the human lower limbs (simulating the skin, fat layer, muscle layer, bone, etc.). Some more advanced versions can also simulate the physiological activities of the human lower limbs, such as blood vessels and blood flow. There are multiple teaching objectives for lower limb examinations, one of which is to teach the examination of lower limb pulse conditions. This examination requires the operator to press their hand to a specific location on the lower limb to feel the pulse of the lower limb. The operation nodes lie in the pressing position, gesture, pressing force, pressing time, etc. In some cases, it is supplemented by feeling or observing the temperature and color of specific locations on the lower limb skin. For inexperienced novice medical staff, misjudgment is often made, resulting in the need for re-angiography or re-surgery. The reason for the lack of proficiency in measuring lower limb pulses is that medical staff are not familiar with the operation nodes of the measurement operation. Using lower limb examination teaching aids can better train and assess medical staff's mastery of the operation nodes.

[0004] Existing lower limb examination teaching aids use sensors (such as pressure sensors and strain sensors) placed at corresponding locations on the teaching aid model to determine whether the medical staff is pressing the correct position. However, this solution believes that this is inaccurate. The pressing gesture is also a key factor. Otherwise, it will not accurately target the specific detection location and will also affect the pulse perception. Due to the certain differences in the appearance of medical staff's hands (such as weight, joint length, etc.), how to accurately determine whether the medical staff's hand movements meet the standard is a problem that needs to be solved.

[0005] In addition, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making the present invention, but due to space limitations, not all details and contents are listed in detail. However, this does not mean that the present invention does not have the characteristics of these prior arts. On the contrary, the present invention already has all the characteristics of the prior art, and the applicant reserves the right to add relevant prior art to the background technology. Summary of the Invention

[0006] To mitigate individual differences, existing technologies have developed solutions for accurately identifying individual movements using comprehensive angle data from each joint. For example, patent publication CN108498102A discloses a rehabilitation training method that uses laser radar scan data from each patient, combined with a skeletal model, to obtain first angle data for each joint. Second angle data for each joint is calculated based on displacement data from an inertial sensor. The first and second angle data are then fused to obtain comprehensive angle data for each joint. The comprehensive angle data for each joint is then combined with an artificial neural network to obtain an error compensation matrix for each joint. During rehabilitation training, the patient's movements are identified based on the scan data and a training sample set. This solution uses scan data and a skeletal model to obtain the angles of each joint. However, the patient movement recognition process in this solution is limited to single-image data fusion, making it impossible to evaluate the specific effects of specific movements, particularly whether the examination performed by medical personnel can accurately obtain patient pulse parameter information. Based on this, this technical solution cannot provide corresponding technical inspiration for motion recognition during the lower limb arterial pulsation examination operation, especially it cannot realize the teaching and assessment of medical staff performing corresponding inspection operation processes under different assessment modes.

[0007] Based on the deficiencies of the existing technology, the present invention provides a teaching and assessment system for lower limb arterial pulsation examination operations, comprising: a foot model, which can simulate a human foot and simulate lower limb arterial pulsation, an image acquisition unit, which can acquire image information of medical personnel performing lower limb arterial pulsation examination operations, and a processor, which is communicatively connected to the image acquisition unit to receive the acquired image information. At least two image acquisition units are configured to acquire image information of medical personnel performing lower limb arterial pulsation examination operations at different imaging angles, respectively. The processor merges at least two sets of image information in an image fusion manner, and extracts skeleton points related to the operation gesture based on the merged information. The processor compares the skeleton points and the skeleton lines formed by connecting the skeleton points with pre-stored standard gestures to output a result of whether the gesture is correct.

[0008] Compared with the above-mentioned prior art, the processor of the present invention can extract the skeletal points related to the medical staff's execution of the corresponding inspection operation gestures based on the image information of the medical staff performing the lower limb arterial pulsation inspection operation. Based on the above-mentioned distinguishing technical features, the problems to be solved by the present invention may include: how to evaluate the accuracy of the inspection operation when the medical staff performs the lower limb arterial pulsation inspection operation. Specifically, the introduction of skeletal points to judge the measurement gesture is not affected by the differences in individual hands, and the judgment result is more accurate. The assessment is carried out based on the foot model data, and a comprehensive assessment can be made from three aspects: gesture, pressing pressure and pressing position. The assessment information is more comprehensive, and students can be trained in all aspects of the measurement operation behavior of the lower limb arterial pulsation.

[0009] Preferably, the image acquisition unit has two cameras to respectively obtain first gesture image information and second gesture image information related to the medical staff's operating gesture, as well as first overall image information and second overall image information related to the force exerted by the operating arm. In order to achieve accurate estimation of the human joint structure, the prior art has already developed a technical solution for detecting the human skeletal state through multi-view image information. For example, patent document No. US2013195330A1 discloses an apparatus and method for estimating the human joint structure. The technical solution uses multiple cameras arranged around the human body to obtain multi-view images for estimating the relevant skeletal structure information of the human body in any posture in a specific space. The technical solution is based on the solid skeletal structure modeling technology of the skeletal system, and estimates the joint position, skeletal structure and posture of the actual skeletal system based on the deformation information of the solid surface shape to reflect the joint movement. However, the main purpose of setting up multiple cameras in this technical solution is to achieve reverse generation of three-dimensional animation based on skeletal structure information, and it is impossible to teach and evaluate different types of lower limb artery pulsation examination operations performed by medical staff based on the detected multi-view image data.

[0010] Compared with the above-mentioned prior art, the image acquisition unit of the present invention can obtain image information of different parts according to the operation process of the medical staff performing the lower limb arterial pulsation examination. Based on the above-mentioned distinguishing technical features, the problems to be solved by the present invention may include: how to make an overall evaluation of the inspection operation posture of the medical staff when performing the lower limb arterial pulsation examination operation. Specifically, there is a difference between gesture image information and overall image information. Gesture image information is only an image of the operator's hand. The operation gesture referred to by the hand image information can only be used to evaluate the accuracy of the operator's palpation, and cannot fully reflect the accuracy of the overall inspection operation process. Overall image information is image information that can include the arm of the operator performing the action. Preferably, it can include the entire arm; more preferably, it includes all relevant parts involved in the hand force (such as shoulders, spine, etc.); if processing power permits, it can also include an image of the entire operator's body. The purpose of collecting overall image information is to be able to determine whether the action is correct by simultaneously collecting the operator's arm force angle. During lower limb artery pulse examinations, medical professionals perform different techniques for patients of different body types. For example, the femoral artery is difficult to palpate in strong or obese patients. During examination, the hip joint should be externally rotated, and palpation should be performed two fingertips above the pubic ramus of the ilium and lateral to the pubic tubercle. Palpation of the popliteal artery is also difficult. The correct technique is to have the patient lie supine with the knee slightly bent and relaxed against the examiner's hand. This allows the examiner's interphalangeal joints to hook onto the medial and lateral tendons of the knee, allowing the fingertips to penetrate deep into the popliteal fossa and palpate the popliteal artery. The posterior tibial artery is located behind the medial malleolus, and the dorsalis pedis artery is located between the first and second metatarsal bones on the dorsum of the foot. In other words, different types of lower limb artery pulse examinations require different overall postures, not just hand gestures. Combining local gesture image information with overall image information can accurately reflect and evaluate the accuracy of medical professionals' lower limb artery pulse examinations.

[0011] Preferably, the processor obtains the arm force angle regarding the arm force during the pulse measurement based on the fused image of the first overall image information and the second overall image information, and compares the arm force angle with a pre-stored standard.

[0012] Preferably, when extracting skeleton points, the processor first identifies visual singularities based on the fused image of the first gesture image information and the second gesture image information, uses the visual singularities as judgment anchor points, and matches the remaining skeleton points based on the pre-input hand structure conditions as identification conditions. Compared with the above-mentioned prior art, the processor of the present invention can determine the anchor points that guide the hand skeleton points to match based on the fused image of different gesture image information. Based on the above-mentioned distinguishing technical features, the problems to be solved by the present invention may include: how to improve the gesture matching accuracy of hand image information. Specifically, due to the defects of visual acquisition such as unclear images and easy misidentification of visual features, it is actually difficult to accurately capture user gestures because there are a large number of misidentified points, which makes it impossible to form accurate gesture skeleton points. However, based on the inherent characteristics of the hand structure, this solution proposes to quickly screen skeleton points with high availability and eliminate misidentified skeleton points, which can greatly improve the accuracy of the results while increasing the data processing speed.

[0013] Preferably, a pressure sensor is provided on the foot model, and the pressure sensor is communicatively connected to the processor. The processor compares the data from the pressure sensor with a standard value to output whether the pressure is correct.

[0014] Preferably, the image accuracy of the camera used to obtain the first gesture image information and the second gesture image information is greater than the image accuracy of the camera used to obtain the first overall image information and the second overall image information.

[0015] Preferably, the system also includes a display terminal, allowing the examiner to view the examinee's operating process in real time. Based on the received foot model data, the terminal triggers a pop-up window for changing parameters or disease patterns, allowing the examinee to select parameters to insert or change. After the examinee selects a parameter, the terminal sends the selection information to the processor to modify the disease pattern or parameter information of the foot model. The examiner can modify the assessment information in real time based on the assessment results, facilitating focused assessment based on the examinee's shortcomings, avoiding mechanical assessments, and facilitating accurate assessment results.

[0016] Preferably, the system automatically changes the foot model parameters or disease model based on the subject's previous scoring failures to assess mastery of the operation. If the subject makes an error, the assessment information is changed, and the subject is repeatedly trained or assessed until the operation is correct. Alternatively, if the subject's measurement of a disease model is completely correct, the foot model parameters or disease model are changed to ensure a realistic and effective assessment.

[0017] Preferably, based on the degree of proximity between the operating gesture, arm force angle and / or pressing pressure of the same subject during operation and the preset optimal standard value, the system updates the step-by-step model parameters or disease pattern of the subject's next assessment operation based on the preset association rules of the operating gesture, arm force angle and / or pressing pressure, so that the subject can gradually approach the optimal standard value based on multiple assessments.

[0018] Preferably, the surface skin of the foot model can be provided with a conductive fabric to replace the sensor. Conductive fabric refers to a conductive fabric material with good electronic transmission properties. By adding conductive fibers to the fabric, the fabric can change its resistance when subjected to pressure, thereby achieving pressure measurement. The advantage of conductive fabric is that when the conductive fabric is subjected to pressure, the contact state between the conductive fibers will change, resulting in a change in the resistance value. The change in resistance value can reflect the force situation. The greater the pressure, the more obvious the change in resistance value. Therefore, the pressure sensitivity performance of the conductive fabric is relatively good. The change in resistance value of the conductive fabric has a certain stability. During long-term use, the conductive performance of the fabric is not affected by external factors. The stability of the conductive fabric is relatively good. The above method can simultaneously determine the pressing position and pressing pressure value based on the change in current, reduce the circuit setting of the sensor, and have a simpler structure. The conductive fabric is arranged in different areas along the artery, covering the surface of the artery to collect pressure data of the pressed artery. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is the system topology diagram of the present invention;

[0020] Figure 2 This is a schematic diagram of hand skeleton points collected by the present invention;

[0021] Figure 3 This is the system logic diagram of the present invention;

[0022] Figure 4 Schematic diagram of another embodiment of the present invention in which a conductive fabric is used instead of a sensor;

[0023] Figure 5 This is a schematic diagram of the present invention collecting gesture skeleton points when an operator detects the lower limb pulse.

[0024] Reference Signs List

[0025] 100: Foot model; 110: Blood vessel structure; 120: Accommodation cavity; 130: Pump; 140: Valve; 150: Circulation path; 160: Pressure sensor; 170: Conductive fabric; 200: Image acquisition unit; 300: Processor; 400: Display terminal; 500: Database; 600: Operator; 610: Skeleton point; 620: Skeleton line. DETAILED DESCRIPTION

[0026] The following is a detailed description with reference to the accompanying drawings.

[0027] Lower limb examination is a relatively basic and main item in clinical nursing. The lower limbs of the human body contain a large number of blood vessels and nerves, especially the arteries running through the legs, which are important observation objects for judging the patient's postoperative condition. One of the items of lower limb examination is to check the patient's lower limb blood circulation status, such as detecting the dorsalis pedis artery. At present, lower limb examination is mainly performed manually. When medical staff visit patients, one of the examination operations is to press specific positions on the patient's lower limbs with their fingers to feel the arterial pulsation of the patient's lower limbs. However, it is difficult for novice medical staff to master the correct operation method of checking the patient's arterial pulsation, which may cause assessment errors and lead to some unnecessary troubles later.

[0028] The present invention cites a case: After the patient's surgery, the artery has been successfully opened. When returning to the ward, the patient's blood circulation (dorsi pedis artery pulsation, skin temperature, and color) needs to be assessed every 15 minutes. Sometimes, inexperienced novice medical staff make mistakes in the assessment, which may lead to the patient having to undergo re-angiography or even re-surgery. After re-evaluation by a senior physician, the dorsalis pedis artery pulsation is palpated, indicating good blood circulation, and the patient avoids a second surgery. It can be seen that if novice medical staff are not proficient in manually measuring the dorsalis pedis artery pulsation, misdiagnosis will occur, and they will not be able to accurately judge the arterial pulsation or determine whether the interventional surgery is successful. Therefore, institutions with training functions such as hospitals need to provide certain professional training and assessments for novice medical staff. Currently, a commonly used teaching and assessment method is to use human models, such as mannequins, or more specific limb models, such as the lower limb models involved in this solution. Lower limb models can better simulate the appearance, touch, physiological structure, etc. of the human lower limbs. Some better models can also simulate structures such as blood vessels. Using a simulated lower limb model for teaching allows medical staff to actually perform the exercises. By simulating the feel of finger contact and providing feedback based on the operation (for example, simulated flesh can dent when pressed to simulate the tactile feedback of real human flesh), it can deepen the medical staff's impression. However, detecting the lower limb, especially the arterial pulse of the lower limb, is relatively complex, requiring medical staff to touch specific locations on the lower limb in the correct manner to feel the pulse. The correct method includes the correct gesture and the correct pressure. Because each medical staff's hand shape is not exactly the same, such as hand fatness and joint range of motion, it is difficult to determine the gesture of each medical staff. Especially during the assessment process, the complexity of the operation makes it difficult to automatically determine whether the medical staff's operation is correct. Conventional methods may only focus on the contact point, and place contact sensors such as pressure sensors 160 at specific locations on the lower limb model (i.e., the correct location, the standard answer for the assessment), hoping to determine whether the medical staff's operation is correct by detecting whether the specific location is touched or pressed. It is true that the position of the contact point is important and affects whether the pulse pulsation can be felt. However, this program believes that the operator's high-precision gestures are also a focus that needs to be paid attention to and taught and tested. For the teaching and testing mode of manual inspection, it is important to establish the correct muscle memory of medical staff. It is not just about whether the contact point is correct. The degree of joint bending and stretching of the hand during the inspection will affect the force and nerve perception. If the correct muscle memory is not formed, even if the contact point is correct, it may lead to differences in the sense of pulsation, resulting in misjudgment. The existing technology lacks high-precision gesture collection for medical staff to manually inspect the pulse pulsation of the lower limb model, which makes it difficult to more accurately assess the medical staff's operational actions for measuring the lower limb pulse pulsation.

[0029] Based on the above, if Figure 1 and Figure 3As shown, this solution proposes a teaching and assessment system for lower limb arterial pulse examination procedures. The system includes a foot model 100, a processor 300, an image acquisition unit 200, and a display terminal 400. Among these components, the processor 300 serves as the system's core, capable of simultaneously communicating with the foot model 100, the image acquisition unit 200, and the display terminal 400. The processor 300 can be an integrated chip with a built-in control program related to the present invention, a central processing unit, or other logical computing component. It can be integrated with the foot model 100, the image acquisition unit 200, or the display terminal 400. Alternatively, through virtualization or containerization technology, an independent computing and data processing workload area can be allocated within existing servers, cloud platforms, or other mature cloud computing infrastructure to meet the specific computing and data processing requirements of this solution. For example, this could be a cloud-based data server storing various hospital data, including electronic medical records and medical images, allowing the processor 300 to quickly and easily access and obtain the required data. Specifically, the foot model 100 can serve as a data input source, connecting its simulated foot information to the processor 300 via a high-speed data bus or network interface. This allows the foot model 100 to transmit simulated foot-related data to the processor 300 in real time for processing. The image acquisition unit 200 can also be connected to the processor 300 via a corresponding image acquisition interface. It is responsible for capturing images or video data of the medical staff (assessed personnel) performing the operation and transmitting this image or video data to the processor 300 for further processing and analysis. The display terminal 400 can communicate with the processor 300 via a user interface or network connection. It is responsible for receiving the processing and analysis results from the processor 300 and displaying them to the user in the form of graphics, text, or video. Assessors can view the medical staff's operating process in real time through the display terminal 400, thereby conveniently understanding the medical staff's skill mastery and adjusting subsequent assessment items based on their skill mastery.

[0030] This communication connection ensures efficient, accurate, and real-time data transmission between the foot model 100, image acquisition unit 200, display terminal 400, and processor 300, providing a better user experience and accurate results. This connection also makes the entire system highly scalable and flexible, allowing for adjustments and optimization based on actual needs.

[0031] Preferably, the foot model 100 is configured to simulate the external structure, skin texture, and pulse of a human foot. Specifically, the foot model 100 can be configured as a multi-layer structure in a manner that simulates the multi-layer physiological structure of the human body, and can include a skin layer, a fat layer, a muscle layer, etc. Preferably, as Figure 4As shown, the foot model 100 also includes a vascular structure 110, which is located at least beneath the cortex to simulate human blood vessels. The vascular structure 110 can be made of a material that simulates real human blood vessels, such as silicone gel, which can simulate the elastic properties of blood vessels, allowing the diameter of the vascular structure 110 to vary. The foot model 100 also includes a fluid-holding cavity 120. The cavity 120 and the vascular structure 110 are connected by a circulation passage 150 to supply fluid to the blood vessels. The fluid is a material that simulates blood, such as simulated blood, whose composition and chemical and physical properties are the same or similar to those of real blood, so as to realistically simulate the flow of blood in blood vessels. A pump 130 is provided on the circulation passage 150 between the cavity 120 and the vascular structure 110. The pump 130 is used to pump the fluid in the cavity 120 and the blood vessels. The pump 130 simulates the pumping action of the heart. Preferably, the pump 130 is communicatively connected to the processor 300 so that the working parameters of the pump 130, such as the pumping flow rate and the pumping pressure, can be controlled. Also included is a valve 140 for simulating pulse pulsation. The valve 140 is communicatively connected to the processor 300 so that the processor 300 can change the blood flow by controlling the opening and closing of the valve 140. The valve 140 is disposed in the passage of the vascular structure 110, preferably between the accommodating cavity 120 and the circulation passage 150 of the vascular structure 110. The other end of the vascular structure 110 can be connected back to the accommodating cavity 120 so as to form a closed loop with the passage of the accommodating cavity 120, so that the simulated blood fluid can be circulated. Figure 4 As shown, the valve 140 and the pump 130 are preferably disposed on a side of the foot model 100 relatively far from the vascular structure 110 to reduce interference caused by vibrations generated by the valve 140 and the pump 130 on the simulation of the pulse pulsation of the vascular structure 110 .

[0032] By controlling the frequency of valve 140 opening and closing and setting the on-off time, processor 300 can create a regular flow of fluid through vascular structure 110, simulating the heart's pumping of blood. The regularly flowing fluid through vascular structure 110 causes it to contract and expand rhythmically, allowing the shape of vascular structure 110 to be felt by touching it with the hand, thereby enabling palpation of the pulse. Furthermore, the on-off state of valve 140 can be controlled. By varying the on-off gap of valve 140 (i.e., the time interval between valve on-off switching), the flow of fluid can be altered, thereby changing the simulated pulse state. For example, shortening the on-off gap of valve 140 increases the pulse frequency; increasing the on-off gap of valve 140 decreases the pulse frequency. By varying the on-off state of valve 140, different pulse wave frequency states can be simulated. In other words, different waveform widths can be simulated, as observed from a pulse waveform diagram. Preferably, the operating parameters of pump 130 can also be controlled by processor 300 to vary, thereby simulating different pulse wave intensities, i.e., the height of the pulse waveform. By adjusting the output power of pump 130, the fluid flow intensity is varied, thereby simulating different pulsation intensities. For example, increasing the output power of pump 130 increases the simulated pulsation intensity; decreasing the output power of pump 130 decreases the simulated pulsation intensity. By separately controlling pump 130 and valve 140, a variety of simulated pulse waveforms can be combined, enabling the system to meet a wide range of educational objectives.

[0033] Furthermore, the processor 300 has or can be connected to the database 500. The database 500 pre-stores a pulse simulation scheme, and according to the pulse simulation scheme, the processor 300 can adjust the working parameters of the pump 130 and the valve 140 respectively to simulate the corresponding pulse wave. The pulse simulation scheme may contain target working parameters of the pump 130 and the valve 140, and the processor 300 can make adjustments according to the target working parameters. The pulse simulation scheme data can be uploaded to the database 500 in advance, and the data can be manually pre-written. Preferably, the database 500 also provides an on-site input interface, and a customized pulse simulation scheme can be input on-site into the database 500 through the input device, so as to facilitate teaching and examination personnel to customize teaching and examination schemes. Preferably, the pulse simulation scheme can be obtained by actually collecting data from several real patients. Specifically, an existing pulse acquisition tool is used, or experienced medical staff manually examines the pulse wave of a real patient, processes the acquired pulse wave into target operating parameters for adjusting the pump 130 and valve 140, and combines the parameters with the physiological condition of the real patient and saves them as a pulse simulation scheme. Physiological conditions may include parameters such as age, weight, and blood routine. By acquiring the pulses of multiple real patients with different physiological conditions, a set of pulse simulation schemes with a relatively large amount of data can be constructed in the database 500. These schemes can correspond to the pulse conditions of most patients with physiological conditions, so that the data can be more in line with the real situation of the person. When the system executes pulse simulation, the processor 300 can automatically match the corresponding pulse simulation scheme according to the matching instructions input by the user. The matching method can be to match using the above-mentioned physiological conditions as keywords.

[0034] Furthermore, the image acquisition unit 200 is configured to acquire operation images of medical personnel and transmit the operation images to the processor 300. Figure 2As shown, the processor 300 extracts action skeleton points 610 based on the image, thereby obtaining accurate data on the medical practitioner's gesture. Specifically, at least two image acquisition units 200 are provided. Each of the two image acquisition units 200 captures images of the medical practitioner's gestures at different imaging angles and simultaneously transmits the captured images to the processor 300. The imaging angles of the two image acquisition units 200 intersect at an angle of at least greater than 10°, preferably between 45° and 80°. For ease of description, the two image acquisition units 200 are referred to as the first image acquisition unit and the second image acquisition unit, respectively. The first image acquisition unit and the second image acquisition unit capture images using a first coordinate system and a second coordinate system, respectively. The first and second coordinate systems can be predetermined when the two image acquisition units 200 are set up. The first image acquisition unit captures first gesture image information of the operator's gesture for checking the arterial pulse, and simultaneously captures first overall image information of the operator's large limb area from its imaging angle. Similarly, the second image acquisition unit captures second gesture image information and second overall image information of the operator's gesture for checking the arterial pulse from its imaging angle. There's a difference between gesture image information and overall image information. Gesture image information is an image of only the operator's hand. Overall image information includes image information of the operator's arm performing the action. Preferably, it includes the entire arm; more preferably, it includes all relevant parts involved in exerting force with the hand (e.g., shoulder, spine, etc.); if processing power allows, it may also include an image of the operator's entire body. Preferably, the two image acquisition units 200 can be configured with multiple cameras of different focal lengths. The cameras used to capture gesture image information can be telephoto cameras with higher capture resolution, while the cameras used to capture overall image information can be wide-angle cameras with relatively lower capture resolution. This configuration allows for the simultaneous capture of relatively detailed gesture image information and relatively coarse overall image information. The reason for simultaneous capture is that the processor 300 relies on two highly corresponding images over time to perform joint processing to obtain accurate data on the skeletal points 610 of the action and the arm force angle associated with the former.

[0035] The processor 300 obtains the data collected by the first image acquisition unit and the second image acquisition unit and performs the following calculation and processing operations: based on the merging of the first gesture image information and the second gesture image information, obtains the data of the operation action skeleton point 610; based on the merging of the first overall image information and the second overall image information, obtains the arm force angle information. When the processor 300 obtains the data of the skeleton point 610, it first fuses the first gesture image information and the second gesture image information to form a three-dimensional gesture image. The three-dimensional gesture image can contain gesture image information of two angles. When fusion is performed, it refers to the predetermined first coordinate system and second coordinate system of the first image acquisition unit and the second image acquisition unit and the imaging angle. The fusion image algorithm uses a certain coordinate system as the standard coordinate system (for example, the first coordinate system is used as the standard) to establish a three-dimensional space, and first constructs the standard coordinate system into the three-dimensional space as the origin. Based on the imaging angle and the second coordinate system, the second image is converted to the corresponding coordinates, and then image fusion is performed. Preferably, the human hand structure can be used as a boundary condition to assist convergence during fusion. Figure 2 、 Figure 5As shown, after fusing the images, the processor 300 extracts data on the skeleton points 610 of the operator's hand based on the three-dimensional fused image. The extraction process is an artificial intelligence discrimination algorithm based on image recognition. First, the visual unique points of the operator's hand image 600 are identified, such as the hand joints, the base of the thumb, and the nails. The unique points have visual characteristics, such as shape and color. After identifying the unique points, the remaining skeleton points 610 are matched based on the pre-entered hand structure conditions combined with visual recognition. Skeletal points 610 are key nodes that constitute gestures. The joints of the operator's hand 600 and the contact points (finger endpoints) of the fingers can be selected as skeleton points 610. The skeleton points 610 are connected according to the hand structure of the operator's hand 600 to form skeleton lines 620. These lines can represent the direction of the skeleton of the operator's hand 600. Preferably, a large number of possible skeletal points 610 are initially matched. The skeletal points 610 with higher availability are then used as judgment anchor points, combined with preset hand structure conditions, to determine the availability of other skeletal points 610 in the initially matched set of skeletal points 610. Points with availability below the judgment conditions are then deleted. Availability refers to whether the skeletal point 610 is close to a real hand joint. Skeletal points 610 with distinct visual features, such as interphalangeal joints, can be selected as judgment anchor points. Hand structure conditions predefine the range of motion of a hand joint. Conditional encoding formats can be selected. For example, when the interphalangeal joint is in the x position, the metacarpophalangeal joint can only move within the yz range. This condition can be derived based on theoretical hand kinematics. Using hand structure conditions, a large number of skeletal points 610 that do not meet the conditions can be quickly eliminated. However, due to visual acquisition limitations such as unclear images and easily misidentified visual features, accurately capturing user gestures is difficult due to the large number of misidentified points, making it impossible to accurately identify gesture skeletal points 610. Based on the inherent characteristics of the hand structure, this solution proposes a method for rapidly screening highly available skeletal points 610 and eliminating misidentified skeletal points 610, significantly improving result accuracy while also increasing data processing speed. The skeletal points 610 that can be collected include interphalangeal joints, metacarpophalangeal joints, carpometacarpal joints, distal interphalangeal joints, and proximal interphalangeal joints. Based on the matched skeletal points 610 and the skeletal lines 620 connecting them, the processor 300 can form a precise gesture model and compare this gesture model to standard gestures in the database 500. Standard gestures in the database 500 also have skeletal points 610 and skeletal lines 620. During the comparison, each skeletal point 610 and skeletal line 620 is compared in a one-to-one manner. Each skeletal point 610 or skeletal line 620 has a comparison redundancy interval, and this redundancy interval is set differently depending on the position of each point or line. Preferably, the redundancy interval for the skeletal points 610 of the fingertip joints is relatively small, while the redundancy interval for the carpometacarpal joints is relatively large.The above solution enables regional enhanced matching and teaching guidance of medical staff's gestures, can distinguish the focus of actions, and when there may be errors in both collection and personnel gestures, focus on matching whether the finger part meets the standards to maximize the determination of whether the gesture is correct.

[0036] Furthermore, the processor 300 also determines the arm force angle based on the first overall image information and the second overall image information. Specifically, the processor 300 fuses the first overall image information and the second overall image information, matches the image information of the arm based on the fused image, and forms an abstract model of the arm. Preferably, the skeleton point 610 matching method can also be adopted in this step, but the skeleton point 610 matching accuracy of this step is lower than the above-mentioned step of forming precise operation gestures. On the one hand, this is because the accuracy of the acquired image is not high, and on the other hand, this is because this step does not require accurate acquisition of the precise posture of the arm, but only needs to obtain a few key parameters. After forming the abstract model of the arm, the arm force angle is extracted based on the angular relationship between the arm and the palm. The arm force angle is compared with the standard arm force angle range.

[0037] Preferably, the model is also provided with a pressure sensor 160 to collect pressure information. The processor 300 obtains the pressure information and compares it with a pre-stored standard.

[0038] If the pressure, gesture, and arm force angle are all within the standard range, a qualified operation message is output. If one or all of the pressure, gesture, and arm force angle are outside the standard range, a failed operation message is output. This solution focuses on the relationship between arm force angle and precise gesture, and simultaneously assesses both. Only when both are met is the operation considered qualified.

[0039] During the assessment, the subject's gesture image information is collected, and the skeleton points 610 and their connecting lines are extracted based on the gesture image information, and the connecting line angle is calculated. The pressing pressure and pressing position are confirmed based on the massage position and massage pressure data collected by the pressure sensor 160 provided on the foot model 100. The gesture, pressing pressure, and pressing position are aligned based on the time information to form the assessment data of the assessment subject. The assessment subject's gesture is compared with the lower limb artery pulsation measurement model to confirm whether the gesture is qualified. The assessment subject's pressing pressure is compared with the pressure threshold, and the pressing position is compared with the preset pressing area to comprehensively determine whether the assessment subject has passed the assessment.

[0040] Preferably, a display terminal 400 is also included. The examiner can view the examinee's operation process in real time on the display terminal 400. Based on the received data (pressing position or pressure value) from the foot model 100, the display terminal 400 triggers a pop-up window for changing parameters or disease modes, allowing the examinee to select parameters to insert or change. After the examinee makes a selection, the display terminal 400 transmits the selection information to the processor 300, which then changes the disease mode or parameter information on the foot model 100.

[0041] If the examiner does not make a selection, the display terminal 400 does not automatically insert or change the disease pattern or parameter information of the foot model 100 .

[0042] The above scheme enables the assessor to change the assessment information in real time according to the assessment situation, which is conducive to conducting key assessments based on the shortcomings of the assessment object, avoiding mechanical assessments, and facilitating obtaining true assessment results.

[0043] According to a preset change mode, the processor 300 automatically changes the disease mode or parameter information based on the received data of the foot model 100 (pressing position or pressing pressure value).

[0044] For example, if the subject makes an error in their operation, the assessment information can be changed, and the subject can be repeatedly trained or assessed until the subject performs the operation correctly. Alternatively, if the subject's measurement operation for a disease pattern is completely correct, the parameters of the foot model 100 or the disease pattern can be changed to achieve a true and effective assessment.

[0045] Preferably, the system automatically changes the parameters of the foot model 100 or the disease mode based on the test subject's previous scoring failure to assess mastery of the operation. The processor 300 changes the window information displayed on the display terminal 400 based on the test subject's proficiency in the operation (which can be assessed based on time), allowing the examinee to quickly select appropriate insertion information or change the disease mode to promote rapid mastery of the operation. A scoring failure characteristic can mean that a test subject's operation does not meet the system's preset standards, such as an inaccurate arm force angle. In the next test for this student, the foot model 100 or disease parameters will be modified to prioritize the arm force angle, for example, adjusting the arterial pulse to a weak state, as pulse detection is difficult without the correct arm force angle. This approach can strengthen the test subject's impression and help them quickly master the correct operation technique.

[0046] Furthermore, based on the proximity of the same subject's operating gestures, arm force angles and / or pressing pressures to the preset optimal standard values during the operation, the system updates the step-by-step model parameters or disease patterns of the subject's next assessment operation based on the preset association rules of the operating gestures, arm force angles and / or pressing pressures, so that the subject can gradually approach the optimal standard values for the operating gestures, arm force angles and / or pressing pressures based on multiple assessments. This solution is concerned that in teaching or assessment tasks, some personnel's operations have met the standards in each item, but there are still cases of misjudgment after taking up the job. The reason is that compared to the training goal of satisfying a relatively broad standard, it is more difficult to maintain accuracy in each examination. Medical staff need to correctly learn or feel the relationship between the operating gestures, arm force angles or pressing pressures, rather than achieving the three conditions in isolation. Therefore, this solution is based on the relationship between the three. When one of the subject's operations is close to the preset optimal standard value, the best-performing item is used to train the subject's operating experience for the remaining items, so that the subject can achieve the best overall performance. The association rules of operation gestures, arm force angles and pressing pressures can be pre-set. The relationship between gestures, arm force angles and pressing pressures is determined based on human kinesiology, and then the association rules are written based on the pulse examination requirements.

[0047] Preferably, if Figure 4 As shown, the surface skin of the foot model 100 can be covered with a conductive fabric 170, replacing the sensor. Conductive fabric 170 is a conductive fabric material with excellent electron transfer properties. By adding conductive fibers to the fabric, the fabric changes its resistance when subjected to pressure, thereby enabling pressure measurement. The advantage of conductive fabric 170 is that when pressure is applied to the fabric, the contact between the conductive fibers changes, resulting in a change in resistance. This change in resistance reflects the force applied; greater pressure results in a more pronounced change in resistance. Therefore, conductive fabric 170 exhibits excellent pressure sensitivity. The resistance change of conductive fabric 170 is relatively stable, and its conductivity is unaffected by external factors during long-term use. Conductive fabric 170 exhibits excellent stability. This approach allows for simultaneous determination of the pressure location and pressure value based on current changes, reducing the sensor circuitry and simplifying the structure. Conductive fabric 170 is arranged in zones along the artery, covering the arterial surface to collect pressure data.

[0048] The process of checking a lower limb arterial pulse involves a healthcare professional using their hands to feel the pulsation of a specific location on the patient's lower limb. This process can be considered a form of exercise: the healthcare professional extends their hand and presses on a specific part of the patient's lower limb, then slightly adjusts their arm to facilitate the feeling of the pulse. Because each healthcare professional has different movement habits, even relatively simple movements like the one described above can exhibit significant differences in detail when broken down. Existing techniques have designed standard movements for teaching purposes. When the healthcare professional's movements match the standard movements, the procedure is considered correct; otherwise, it is incorrect. However, this approach has found that standard movements do not cover all correct operations. Some movements can accurately detect a pulse, but may be considered incorrect because some of the sub-movements (e.g., the arm force angle) do not conform to the standard movements. Furthermore, existing techniques use the final test results to determine whether the operation is correct, without providing a scientific and accurate analysis of the causes of incorrect operations. This is detrimental to teaching, as students do not understand how to improve their remaining incorrect sub-movements even if some of their sub-movements are correct. Therefore, when it is detected that at least one skeleton point 610 meets the standard, the processor 300 obtains the predicted position or predicted trajectory of the remaining skeleton points 610 and / or the arm force angle based on the correct skeleton point 610 through reverse kinematic calculation. Reverse kinematic calculation belongs to the theory of inverse kinematics, which is a calculation method that reverses the motion process based on the result. By inputting the boundary conditions of the hand movement into the processor 300 in advance, all the motion modes that may cause the skeleton point 610 to appear in this position are reversed based on the determined correct skeleton point 610 as input. The motion must meet the boundary conditions of the hand movement. The boundary conditions of hand movement refer to the movements that can be performed based on the physiological conditions of the hand. For example, the reverse bending of the fingers during the movement is an impossible movement. Based on this boundary condition, the predicted movement will be deleted. One problem with reverse kinematic calculation is that it is difficult to find a unified and available standard coordinate. Therefore, this solution further proposes the following solution: based on the processor 300 processing the first gesture image information and the second gesture image information, it searches for a judgment anchor point with unique visual features. The processor 300 uses the judgment anchor point as the origin or reference point of the initial coordinate system for the reverse motion calculation process, thereby realizing the reverse motion calculation of the operation. In addition, based on the reverse motion calculation results, the preset standard operation data in the program is updated, thereby updating and changing the evaluation criteria for the medical staff's operation. This solution uses the judgment anchor point to construct the hand skeleton point 610 and the initial coordinate system for reverse motion calculation. This solves the problems of low hand posture detection accuracy in traditional technologies and the difficulty in finding a highly reliable standard coordinate system for performing reverse motion calculations.By incorporating inverse kinematics calculations, medical staff can now identify all possible motions based on a specific decomposed motion (such as a hand gesture) as a reference, allowing them to easily compare correct and incorrect methods. Furthermore, based on updated evaluation criteria, some decomposed motions can be incorporated into the evaluation criteria, allowing medical staff to perform examinations using movements that most closely match their own habits, significantly improving teaching efficiency, accuracy, and adaptability.

[0049] It should be noted that the above-mentioned specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also belong to the disclosure scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the present invention specification and its drawings are illustrative and do not constitute a limitation on the claims. The scope of protection of the present invention is defined by the claims and their equivalents. The present invention specification contains multiple inventive concepts, such as "preferably", "according to a preferred embodiment" or "optionally", all of which indicate that the corresponding paragraph discloses an independent concept, and the applicant reserves the right to file a divisional application based on each inventive concept. Throughout the text, the features guided by "preferably" are only an optional method and should not be understood as having to be set, so the applicant reserves the right to abandon or delete the relevant preferred features at any time.

Claims

1. A processor for lower limb arterial pulse examination operation training, wherein the processor (300) is communicatively connected with an image acquisition unit (200) to receive acquired image information, characterized in that: In the case where the two image acquisition units (200) respectively acquire operation images of the medical staff at different imaging angles, the processor (300) receives first gesture image information and first overall image information of the operator's gesture for checking arterial pulsation acquired by the first image acquisition unit, and second gesture image information and second overall image information of the operator's gesture for checking arterial pulsation acquired by the second image acquisition unit from its imaging angle; The processor (300) is configured to: Based on the combined processing of the first gesture image information and the second gesture image information, data of the operation action skeleton point (610) is obtained; obtaining an arm force angle during the pulse measurement based on the combined processing of the first overall image information and the second overall image information, and comparing the arm force angle with a standard arm force angle range to determine whether the action is correct; The skeleton points (610) and the skeleton lines (620) formed by connecting the skeleton points (610) are compared with the pre-stored standard gestures to output a result indicating whether the gesture is correct.

2. The processor according to claim 1, wherein: When extracting the skeleton points (610), the processor (300) identifies visual singular points based on a fusion image of the first gesture image information and the second gesture image information, uses the visual singular points as judgment anchor points, and matches the remaining skeleton points (610) based on a pre-input hand structure condition as an identification condition.

3. The processor according to claim 1 or 2, characterized in that The processor (300) is in communication with a pressure sensor (160) provided on the foot model (100). The processor (300) receives pressure information, compares the pressure information with a standard value, and outputs whether the pressure is correct.

4. The processor according to any one of claims 1 to 3, wherein: When acquiring the data of the skeleton point (610), the processor (300) fuses the first gesture image information and the second gesture image information to form a three-dimensional gesture image. The three-dimensional gesture image includes gesture image information from two angles. The processor (300) refers to the predetermined first coordinate system and second coordinate system of the first image acquisition unit and the second image acquisition unit, as well as the imaging angle, during fusion.

5. The processor according to any one of claims 1 to 4, characterized in that: The processor (300) extracts data of the skeleton points (610) of the operator's hand from the three-dimensional fused image based on an artificial intelligence discrimination algorithm for image recognition.

6. The processor according to any one of claims 1 to 5, characterized in that: The processor (300) matches a plurality of possible skeleton points (610) for the first time, uses the skeleton points (610) with higher availability as judgment anchor points, and determines the availability of each skeleton point (610) in the set of the first matched skeleton points (610) in combination with a preset hand structure condition, and deletes points with availability lower than the judgment condition; The availability refers to whether the skeleton point (610) is close to the joint point of the real hand.

7. The processor according to any one of claims 1 to 6, characterized in that: The processor (300) fuses the first overall image information and the second overall image information, matches the image information of the arm based on the fused image, and forms an abstract model of the arm; After the arm abstract model is formed, the arm force angle is extracted based on the angular relationship between the arm and the palm.

8. The processor according to any one of claims 1 to 7, wherein: If the pressure, gesture and arm force angle are all within the standard range, the processor (300) outputs operation qualification information.

9. The processor according to any one of claims 1 to 8, wherein: Based on the degree of proximity between the operation gesture, arm force angle and / or pressing pressure of the same subject during operation and the preset optimal standard value, the processor (300) updates the foot model (100) parameters or disease pattern of the subject's next assessment operation based on the preset association rules of the operation gesture, arm force angle and / or pressing pressure, so that the subject can gradually approach the optimal standard value for the operation gesture, arm force angle and / or pressing pressure based on multiple assessments.

10. A processing method for lower limb arterial pulse examination operation training, characterized in that: The method comprises: When two image acquisition units (200) respectively acquire operation images of medical personnel at different imaging angles, first gesture image information of the operator's gesture of checking arterial pulsation and first overall image information acquired by the first image acquisition unit and second gesture image information of the operator's gesture of checking arterial pulsation and second overall image information acquired by the second image acquisition unit from its imaging angle are received; Based on the combined processing of the first gesture image information and the second gesture image information, data of the operation action skeleton point (610) is obtained; obtaining an arm force angle during the pulse measurement based on the combined processing of the first overall image information and the second overall image information, and comparing the arm force angle with a standard arm force angle range to determine whether the action is correct; The skeleton points (610) and the skeleton lines (620) formed by connecting the skeleton points (610) are compared with the pre-stored standard gestures to output a result indicating whether the gesture is correct.

Images