A real-time pain detection analysis system based on facial expressions
By designing a real-time pain detection and analysis system based on facial expressions, and utilizing the collection and comparative analysis of multiple pain features, the system solves the problem of inaccurate pain assessment caused by single-angle monitoring in existing technologies, and achieves efficient and accurate pain detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NORTH CHINA UNIVERSITY OF SCIENCE & TECHNOLOGY AFFILIATED HOSPITAL
- Filing Date
- 2026-05-02
- Publication Date
- 2026-07-21
AI Technical Summary
Existing vision-based pain detection methods typically only monitor from a single angle, resulting in insufficient accuracy of the monitoring results and making it difficult to conduct accurate pain assessment by combining multiple angles.
A real-time pain detection and analysis system based on facial expressions was designed, including a camera-facing-face adjustment module, a facial information acquisition module, an audio-visual comparison module, and a pain level output module. By collecting and comparing various pain features, combined with the pain level output module, a multi-angle pain assessment can be achieved.
It improves the accuracy and convenience of pain detection, expands the detection range, and ensures the accuracy and comprehensiveness of the test results.
Smart Images
Figure CN122432745A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of medical testing technology, specifically a real-time pain detection and analysis system based on facial expressions. Background Technology
[0002] Pain is an unpleasant subjective experience caused by actual or potential tissue damage, integrating complex neurological, psychological, and social factors. It not only causes physical and mental suffering for patients but can also affect disease recovery, increase the risk of postoperative complications, and, in the long term, become a global public health problem, placing a heavy burden on healthcare systems. Accurate and timely pain assessment is a prerequisite for effective pain management. Its core objective is to objectively reflect the patient's pain status, providing a reliable basis for clinical diagnosis, treatment plan adjustment, and efficacy evaluation. Especially for special populations who cannot express pain independently, scientific pain detection is crucial for ensuring medical safety. Among behavioral pain assessment methods, facial expression analysis has become a research hotspot in the field of objective pain assessment due to its non-invasiveness, intuitiveness, and ease of access. Research shows that pain, as a strong negative emotion, triggers characteristic contractions of facial muscles, forming regular painful facial expressions. These expressions can be decomposed into specific action units through a facial motion coding system, such as frowning, eyelid tightening, nasal wing contraction, and upper lip lifting—nine core action units in total. Their combination and activation intensity can effectively reflect the presence and severity of pain, but the following limitations still exist: Existing vision-based pain detection methods can generally only monitor from a single angle, such as the degree of facial distortion or blushing, making it difficult to combine multiple angles for monitoring, resulting in insufficient accuracy of the monitoring results. Summary of the Invention
[0003] In order to overcome the shortcomings of the prior art, the present invention provides a real-time pain detection and analysis system based on facial expressions, which effectively solves the problems mentioned in the background art.
[0004] To achieve the above objectives, the present invention provides the following technical solution: a real-time pain detection and analysis system based on facial expressions, comprising a camera-facing-face adjustment module, a face information acquisition module, an audio-visual comparison module, and a pain level output module, wherein each module is sequentially connected to form a closed-loop detection and analysis system; The camera-facing-face adjustment module includes a face acquisition camera and a rotation adjustment mechanism for positioning the face acquisition camera directly opposite the detected face. The facial information acquisition module is used to collect and classify different features of the acquired audio and video content; The audio-visual comparison module is used to compare and analyze the different features collected with features in the database; The pain level output module obtains and outputs the pain level after comparing the data.
[0005] Preferably, the facial information acquisition module includes acquisition of the patient's degree of blushing, amount of facial sweating, degree of facial muscle contortion, frequency of head shaking, and frequency of groaning sounds. By acquiring different pain characteristics, the accuracy of pain detection and analysis is improved.
[0006] Preferably, the audio-visual comparison module includes a sound capture and analysis module, an image motion capture and analysis calculation module, and an image comparison and analysis module. The sound capture and analysis module is used to determine the patient's pain level 1 by comparing the loudness and frequency of groaning sounds. The image motion capture and analysis calculation module is used to determine the patient's pain level 2 by comparing the frequency of head shaking. The image comparison and analysis module is used to determine the patient's pain level 3 by comparing the degree of redness of the patient's face, the amount of sweat on the face, and the degree of facial muscle distortion.
[0007] Preferably, the pain level output module combines three levels of pain intensity—pain level one, pain level two, and pain level three—to output the corresponding pain level, thereby achieving real-time pain detection and analysis based on facial expressions.
[0008] Preferably, the rotation adjustment mechanism includes a fixing plate mounted on a fixing frame above a patient chair or bed and a mounting plate for mounting a facial acquisition camera, wherein the mounting plate is located on the fixing plate near the patient's head, and a face adjustment component is provided between the fixing plate and the mounting plate.
[0009] Preferably, the face adjustment assembly includes a fixed cylinder fixedly installed on the side of the fixed plate near the mounting plate. A rotation limiting groove is provided on the inner wall of the fixed cylinder. A rotating plate is provided on the inner side of the fixed cylinder. The rotating plate is rotatably connected to the rotation limiting groove. An angle adjustment component is installed between the rotating plate and the mounting plate. A gear ring is coaxially installed on the side of the rotating plate away from the mounting plate. A first gear is meshed on the inner side of the gear ring. The first gear is fixedly connected to the output shaft of a first motor. The first motor is installed on the inner wall of the fixed cylinder.
[0010] Preferably, the angle adjustment component includes two fixed rods symmetrically installed on the side of the rotating plate near the mounting plate. One end of each fixed rod is connected to a first hinge, and the other end of the first hinge is fixedly connected to the mounting plate. The line connecting the two first hinges passes through the axis of the mounting plate.
[0011] Preferably, the rotating plate has an eccentrically formed guide plate groove, and a guide plate body is slidably installed on the inner side of the guide plate groove. A second hinge is installed at one end of the guide plate body near the mounting plate, and a slider is fixedly installed at the other end of the second hinge. The slider is slidably installed inside the slide, wherein the slide is fixedly installed on the mounting plate, and one end of the slide points to the center of the mounting plate. The line connecting the slide and the two first hinges is perpendicular to each other, and the rotation limiting direction of the second hinge is the same as the rotation limiting direction of the first hinge.
[0012] Preferably, the guide plate has equidistant toothed grooves along its length, a second gear is provided on one side of the guide plate, the second gear meshes with the toothed grooves, the second gear is fixedly connected to the output shaft of the second motor, and the second motor is mounted on the rotating plate.
[0013] Compared with the prior art, the beneficial effects of the present invention are: This invention, through a facial information acquisition module, can distinguish and collect various pain characteristics of patients, such as the degree of facial redness, the amount of facial sweating, the degree of facial muscle contortion, the frequency of head shaking, and the volume and frequency of groans. By combining and comparing multiple pain characteristics, the accuracy of real-time pain detection and analysis is improved. This invention, through an audio-visual comparison module composed of a sound capture and analysis module, an image motion capture and analysis calculation module, and an image comparison and analysis module, can compare and analyze the patient's degree of blushing, amount of facial sweating, degree of facial muscle contortion, frequency of head shaking, and frequency of groaning sounds, so that different pain features can be compared independently, avoiding mutual interference between different pain features and improving the convenience of detection. In this invention, after the guide plate is moved by the second motor, the mounting plate can be rotated to adjust the angle of the facial acquisition camera and expand the detection range. At the same time, the first motor can drive the mounting plate to rotate and adjust the orientation of the facial acquisition camera to further expand the detection range, so that the facial acquisition camera can be facing the patient's face and improve the accuracy of the detection results. Attached Figure Description
[0014] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.
[0015] In the attached diagram: Figure 1 This is a schematic diagram of the structure of the real-time pain detection and analysis system based on facial expressions according to the present invention; Figure 2 This is a schematic diagram of the camera-facing-face adjustment module structure of the present invention; Figure 3 This is a schematic diagram of the face adjustment component structure of the present invention; Figure 4This is a schematic diagram of the angle adjustment component of the present invention; In the diagram: 1. Fixing plate; 2. Mounting plate; 3. Facial capture camera; 4. Face adjustment assembly; 401. Fixing cylinder; 402. Rotation limit groove; 403. Rotating plate; 404. Gear ring; 405. First gear; 406. First motor; 407. Angle adjustment component; 4071. Fixing rod; 4072. First hinge; 4073. Guide plate groove; 4074. Guide plate; 4075. Second hinge; 4076. Slider; 4077. Carriage; 4078. Gear groove; 4079. Second gear; 40710. Second motor. Detailed Implementation
[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0017] Example 1, by Figure 1 The present invention relates to a real-time pain detection and analysis system based on facial expressions, including a camera facing the face adjustment module, a face information acquisition module, an audio-visual comparison module, and a pain level output module. Each module is connected in sequence to form a closed-loop detection and analysis system. The camera-facing-face adjustment module includes a face acquisition camera 3 and a rotation adjustment mechanism for positioning the face acquisition camera 3 so that it faces the detected face. The facial information acquisition module is used to collect and classify different features of the acquired audio and video content. The facial information acquisition module can distinguish and collect various patient pain features, such as the degree of redness of the patient's face, the amount of sweat on the face, the degree of facial muscle contortion, the frequency of head shaking, and the volume and frequency of groans. By combining and comparing multiple pain features, the accuracy of real-time pain detection and analysis is improved. The audio-visual comparison module is used to compare and analyze the different features collected with features in the database. The audio-visual comparison module consists of a sound capture and analysis module, an image motion capture and analysis calculation module, and an image comparison and analysis module. It can compare and analyze the patient's degree of blushing, amount of facial sweating, degree of facial muscle contortion, frequency of head shaking, and frequency of groaning sounds, so that different pain features can be compared independently, avoiding mutual interference between different pain features and improving the convenience of detection. The pain level output module obtains and outputs the pain level after comparing the data.
[0018] The facial information acquisition module includes the acquisition of the patient's facial redness level, facial sweating amount, facial muscle contortion degree, head shaking frequency, and groaning sound volume and frequency. By acquiring different pain characteristics, the accuracy of pain detection and analysis is improved.
[0019] The audio-visual comparison module includes a sound capture and analysis module, an image motion capture and analysis module, and an image comparison and analysis module. The sound capture and analysis module is used to determine the patient's pain level 1 by comparing the loudness and frequency of groans. The image motion capture and analysis module is used to determine the patient's pain level 2 by comparing the frequency of head shaking. The image comparison and analysis module is used to determine the patient's pain level 3 by comparing the degree of redness of the patient's face, the amount of sweat on the face, and the degree of facial muscle contortion.
[0020] The pain level output module combines three levels of pain intensity—Pain Level 1, Pain Level 2, and Pain Level 3—to output the corresponding pain level, enabling real-time pain detection and analysis based on facial expressions.
[0021] Depend on Figures 2-4 The rotating adjustment mechanism includes a fixed plate 1 mounted on a fixed frame above a patient chair or bed and a mounting plate 2 for mounting a facial acquisition camera 3. The mounting plate 2 is located on the fixed plate 1 near the patient's head, and a face adjustment component 4 is provided between the fixed plate 1 and the mounting plate 2. The face adjustment assembly 4 includes a fixed cylinder 401 fixedly installed on the side of the fixed plate 1 near the mounting plate 2. A rotation limiting groove 402 is provided on the inner wall of the fixed cylinder 401. A rotating plate 403 is provided on the inner side of the fixed cylinder 401. The rotating plate 403 is rotatably connected to the rotation limiting groove 402. An angle adjustment component 407 is installed between the rotating plate 403 and the mounting plate 2. A gear ring 404 is coaxially installed on the side of the rotating plate 403 away from the mounting plate 2. A first gear 405 is meshed on the inner side of the gear ring 404. The first gear 405 is fixedly connected to the output shaft of the first motor 406. The first motor 406 is installed on the inner wall of the fixed cylinder 401. The angle adjustment component 407 includes two fixed rods 4071 symmetrically mounted on the rotating plate 403 near the mounting plate 2. One end of each fixed rod 4071 is connected to a first hinge 4072, and the other end of the first hinge 4072 is fixedly connected to the mounting plate 2. The line connecting the two first hinges 4072 passes through the axis of the mounting plate 2. An eccentrically formed guide plate groove 4073 is provided on the rotating plate 403. A guide plate body 4074 is slidably mounted inside the guide plate groove 4073. A second hinge 4075 is mounted on one end of the guide plate body 4074 near the mounting plate 2. A slider 4076 is fixedly mounted on the other end of the second hinge 4075. The slider 4076 is slidably mounted inside a slide 4077, which is fixedly mounted on the mounting plate 2, with one end of the slide 4077 pointing towards the center of the mounting plate 2. The slide 4077 and the two first hinges 4072 are connected... The connecting lines are perpendicularly arranged. The rotation limiting direction of the second hinge 4075 is the same as that of the first hinge 4072. The guide plate 4074 has toothed grooves 4078 equidistantly opened along its length. A second gear 4079 is provided on one side of the guide plate 4074. The second gear 4079 meshes with the toothed grooves 4078. The second gear 4079 is fixedly connected to the output shaft of the second motor 40710. The second motor 40710 is mounted on the rotating plate 403. After the second motor 40710 drives the guide plate 4074 to move, it can pull the mounting plate 2 to rotate, adjust the angle of the facial acquisition camera 3, and expand the detection range. At the same time, the first motor 406 can drive the mounting plate 2 to rotate, adjust the orientation of the facial acquisition camera 3, further expand the detection range, and enable the facial acquisition camera 3 to face the patient's face, thereby improving the accuracy of the detection results.
[0022] Working principle: In use, the patient first lies on a hospital chair or bed, and then the angle of the face adjustment module is adjusted so that the facial capture camera 3 is aimed at the patient's face. The adjustment is performed as follows: During adjustment, when angle adjustment is required, the second motor 40710 is activated, driving the second gear 4079 to rotate. The second gear 4079 meshes with the toothed groove 4078, thereby driving the guide plate 4074 to move along the guide plate groove 4073. This pulls the mounting plate 2 to rotate around the first hinge 4072 as the rotation center. During the movement of the guide plate 4074 along the guide plate groove 4073, the slider 4076 rotates with the guide plate 4074 through the second hinge 4075. At the same time, the slider 4076 slides within the carriage 4077, facilitating angle adjustment and expanding the detection range. Then, the first motor 406 is activated, driving the first gear 405 to rotate. The first gear 405 meshes with the toothed ring 404 on the rotating plate 403, thereby driving the rotating plate 403 to rotate. This, in turn, drives the mounting plate 2 to rotate, further adjusting the orientation of the face acquisition camera 3 and further expanding the detection range. During the patient's treatment, the facial capture camera 3 captures the patient's face and transmits the image to the facial information acquisition module. The facial information acquisition module analyzes the video recording and distinguishes different pain characteristics, including the patient's degree of blushing, amount of facial sweating, degree of facial muscle contortion, frequency of head shaking, and frequency of groaning sounds. The distinguished acquisition results are then input into the audio-visual comparison module. The audio-visual comparison module includes a sound capture and analysis module, an image motion capture and analysis module, and an image comparison and analysis module. The sound capture and analysis module is used to compare the volume and frequency of groans with different pain level standards to determine the patient's pain level 1. The image motion capture and analysis module is used to compare the frequency of head shaking with different pain level standards to determine the patient's pain level 2. The image comparison and analysis module is used to compare the patient's degree of blushing, amount of facial sweating, and degree of facial muscle contortion with different pain level standards to determine the patient's pain level 3. Then, based on the combination of three levels of pain—pain level 1, pain level 2, and pain level 3—the pain level output module outputs the corresponding pain level, realizing real-time pain detection and analysis based on facial expressions.
Claims
1. A real-time pain detection and analysis system based on facial expressions, characterized in that: The system includes a camera-facing-face adjustment module, a face information acquisition module, an audio-visual comparison module, and a pain level output module. These modules are sequentially connected to form a closed-loop detection and analysis system. The camera-facing-face adjustment module includes a face acquisition camera (3) and a rotation adjustment mechanism for positioning the face acquisition camera (3) facing the detected face. The facial information acquisition module is used to collect and classify different features of the acquired audio and video content; The audio-visual comparison module is used to compare and analyze the different features collected with features in the database; The pain level output module obtains and outputs the pain level after comparing the data.
2. The real-time pain detection and analysis system based on facial expressions according to claim 1, characterized in that: The facial information acquisition module includes acquisition of the patient's facial redness level, facial sweating amount, facial muscle contortion degree, head shaking frequency, and groaning sound volume and frequency. By acquiring different pain characteristics, the accuracy of pain detection and analysis is improved.
3. The real-time pain detection and analysis system based on facial expressions according to claim 2, characterized in that: The audio-visual comparison module includes a sound capture and analysis module, an image motion capture and analysis module, and an image comparison and analysis module. The sound capture and analysis module is used to determine the patient's pain level 1 by comparing the loudness and frequency of groans. The image motion capture and analysis module is used to determine the patient's pain level 2 by comparing the frequency of head shaking. The image comparison and analysis module is used to determine the patient's pain level 3 by comparing the degree of redness of the patient's face, the amount of sweat on the face, and the degree of facial muscle contortion.
4. The real-time pain detection and analysis system based on facial expressions according to claim 3, characterized in that: The pain level output module combines three levels of pain intensity—pain level 1, pain level 2, and pain level 3—to output the corresponding pain level, enabling real-time pain detection and analysis based on facial expressions.
5. The real-time pain detection and analysis system based on facial expressions according to claim 1, characterized in that: The rotation adjustment mechanism includes a fixed plate (1) installed on a fixed frame above a patient chair or bed and an mounting plate (2) for installing a facial acquisition camera (3), wherein the mounting plate (2) is located on the fixed plate (1) near the patient's head, and a face adjustment component (4) is provided between the fixed plate (1) and the mounting plate (2).
6. A real-time pain detection and analysis system based on facial expressions according to claim 5, characterized in that: The face adjustment assembly (4) includes a fixed cylinder (401) fixedly installed on the side of the fixed plate (1) near the mounting plate (2). A rotation limiting groove (402) is provided on the inner wall of the fixed cylinder (401). A rotating plate (403) is provided on the inner side of the fixed cylinder (401). The rotating plate (403) is rotatably connected to the rotation limiting groove (402). An angle adjustment component (407) is installed between the rotating plate (403) and the mounting plate (2). A gear ring (404) is coaxially installed on the side of the rotating plate (403) away from the mounting plate (2). A first gear (405) is meshed on the inner side of the gear ring (404). The first gear (405) is fixedly connected to the output shaft of the first motor (406). The first motor (406) is installed on the inner wall of the fixed cylinder (401).
7. A real-time pain detection and analysis system based on facial expressions according to claim 6, characterized in that: The angle adjustment component (407) includes two fixed rods (4071) symmetrically installed on the side of the rotating plate (403) near the mounting plate (2). One end of the fixed rod (4071) is connected to a first hinge (4072), and the other end of the first hinge (4072) is fixedly connected to the mounting plate (2). The line connecting the two first hinges (4072) passes through the axis of the mounting plate (2).
8. A real-time pain detection and analysis system based on facial expressions according to claim 7, characterized in that: The rotating plate (403) is eccentrically provided with a guide plate groove (4073). A guide plate body (4074) is slidably installed on the inner side of the guide plate groove (4073). A second hinge (4075) is installed on one end of the guide plate body (4074) near the mounting plate (2). A slider (4076) is fixedly installed on the other end of the second hinge (4075). The slider (4076) is slidably installed inside the slide (4077). The slide (4077) is fixedly installed on the mounting plate (2), and one end of the slide (4077) points to the center of the mounting plate (2). The line connecting the slide (4077) and the two first hinges (4072) is perpendicular. The rotation limit direction of the second hinge (4075) is the same as the rotation limit direction of the first hinge (4072).
9. A real-time pain detection and analysis system based on facial expressions according to claim 8, characterized in that: The guide plate (4074) has toothed grooves (4078) equidistantly spaced along its length. A second gear (4079) is provided on one side of the guide plate (4074). The second gear (4079) meshes with the toothed groove (4078). The second gear (4079) is fixedly connected to the output shaft of the second motor (40710). The second motor (40710) is mounted on the rotating plate (403).