CN113554609B - Neck dystonia identification system based on vision - Google Patents

Neck dystonia identification system based on vision Download PDF

Info

Publication number
CN113554609B
CN113554609B CN202110812666.7A CN202110812666A CN113554609B CN 113554609 B CN113554609 B CN 113554609B CN 202110812666 A CN202110812666 A CN 202110812666A CN 113554609 B CN113554609 B CN 113554609B
Authority
CN
China
Prior art keywords
head
neck
patient
key point
detection module
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202110812666.7A
Other languages
Chinese (zh)
Other versions
CN113554609A (en
Inventor
叶晨
靳令经
肖潏灏
李若愚
王鑫宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tongji University
Original Assignee
Tongji University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tongji University filed Critical Tongji University
Priority to CN202110812666.7A priority Critical patent/CN113554609B/en
Publication of CN113554609A publication Critical patent/CN113554609A/en
Application granted granted Critical
Publication of CN113554609B publication Critical patent/CN113554609B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20068—Projection on vertical or horizontal image axis

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Radiology & Medical Imaging (AREA)
  • General Health & Medical Sciences (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Epidemiology (AREA)
  • Primary Health Care (AREA)
  • Public Health (AREA)
  • Quality & Reliability (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

The invention belongs to the technical field of medical instruments, and particularly relates to a neck dystonia identification system based on vision. The system is characterized by comprising four main modules, namely a patient target detection module, a 2D key point detection module, a 3D key point detection module and an abnormal motion mode discrimination module. The method comprises the steps that video image data of a patient are collected through portable equipment, the position of the patient in an image is detected through a patient target detection module, the 2D key point detection module and the 3D key point detection module further detect the 2D and 3D key point coordinates of the patient respectively, the abnormal motion mode distinguishing module calculates the head and neck operation angle of the patient through the human body key point coordinates, and the abnormal motion mode of the patient is judged.

Description

Neck dystonia identification system based on vision
Technical Field
The invention belongs to the technical field of medical instruments, and particularly relates to a cervical dystonia identification system based on vision.
Background
Cervical dystonia is the most common focal dystonia clinically, and is a focal dystonia which causes abnormal posture and involuntary movement of the head and neck due to clonic or rigid over-contraction of the neck muscles. Common procedures include medication, botulinum toxin injection and surgical procedures. Wherein the botulinum toxin is effective to reduce muscle contraction and pain by local injection therapy.
The target muscle for botulinum toxin therapy is generally selected clinically by abnormal movement patterns of the patient's neck muscles. Neck dystonia can be generally divided into two types, one being abnormal head movement with respect to the neck and one being abnormal neck movement with respect to the trunk, each of which can be further subdivided into flexion, roll, torsion types. For complex cervical dystonia patterns, it should be first defined which basic abnormal motor patterns are to be composed, and then all possible responsible muscles are explored.
The traditional methods commonly used for clinically assessing abnormal movement patterns of neck muscles of patients are mainly three methods:
the first method is manual measurement, i.e., measurement using a head angle measuring instrument for cervical dystonia typing. The patient needs to put the head and neck into the instrument, and the doctor manually measures the movement angle of the head and neck of the patient in different states by means of the protractor. This method is cumbersome to operate and requires a great deal of time and effort.
The second method is sensorometry, using a multi-axis sensor assembly for accurate assessment of cervical dystonia. The doctor attaches sensors to the neck, head and shoulders of the patient to measure the rotation and displacement data of the respective sites. Compared with a manual measurement method, the method has certain improvement on precision. However, the physical contact between the sensor and the body of the patient inevitably interferes with the movement of the patient, and the position where the sensor is worn each time is also deviated, and these factors can cause errors in finding out the data acquired by the sensor.
The third method adopts X-ray shooting multi-angle shooting of the skull and the spine, compares the positions of different centrums and the cross-sectional areas of corresponding muscle abdomens of the neck, and can accurately judge the abnormal movement mode of the dystonia. This method is more accurate than manual and sensor measurement methods, but increases the radiation exposure of the patient and is more complex to operate. Since the conventional CT requires the patient to be in a horizontal position, and the determination of an abnormal posture is affected, the detection of an abnormal pattern of cervical dystonia needs to be performed in a special upright CT.
The patent "head angle measuring instrument for cervical dystonia typing" (publication number: CN207412167U) is a cervical dystonia abnormal movement pattern detection device, which measures the head movement angle of a patient through a multi-plane protractor on a head-mounted device, thereby quantitatively evaluating the abnormal movement pattern of cervical dystonia. The specific operation steps are that the helmet is fixed on the intersection point of the connecting line of the ears of a person to be tested and the top of the plane of the connecting line of the nasal root and the macropore of the occiput by the elastic rope. During measurement, the head of a person to be measured inclines by a certain angle, the multi-plane protractor inclines along with the head of the person to be measured, the pointer deflects by a certain angle in the plane of the protractor under the action of gravity and machinery, and the inclination angles of the head of the person to be measured in the coronal position and the sagittal position can be obtained through measurement of the reading, so that the inclination angles are quantized.
The patent 'a multiaxial sensor composite set for accurately assessing cervical dystonia' (publication number: CN208741005U) is a multiaxial sensor composite set for accurately assessing cervical dystonia, and the measuring device comprises a head hoop, a head multiaxial sensor, a neck connecting belt, a shoulder connecting belt, a sucker base, a second cervical multiaxial sensor, a shoulder multiaxial sensor and a seventh cervical multiaxial sensor. The device can realize accurate measurement of abnormal movement patterns of cervical dystonia and perform 3D reconstruction.
The use of the above-described conventional abnormal pattern assessment method requires a professional measuring device and a high level of doctors for operation, and may cause discomfort or even pain to the patient.
Disclosure of Invention
The conventional cervical dystonia abnormal pattern evaluation method requires professional measuring equipment and a high-level doctor to operate, contact type protractor measurement and sensor measurement can cause discomfort and even pain to a patient, and a non-contact type X-ray camera law can increase the radiation exposure of the patient. The scheme of the invention realizes a non-contact evaluation method by using a computer vision technology, and evaluates the abnormal motion mode of the patient by using the image, video and other data of the patient, the method only needs to use a common camera device to collect data, such as a smart phone and the like, and does not need to be in direct contact with the patient, thereby being convenient and quick while maintaining the evaluation accuracy, not bringing discomfort to the patient, and being easy to be applied in remote auxiliary diagnosis.
In order to achieve the above object, the present invention proposes the following technical solutions:
the utility model provides a neck dystonia identification system based on vision, its characterized in that includes four main modules, is patient target detection module, 2D key point detection module, 3D key point detection module, unusual motion mode discrimination module respectively, wherein:
the patient target detection module is used for detecting the specific position of the patient in the image;
the 2D key point detection module is responsible for evaluating the specific 2D coordinates of the key points of the patient in the image; providing the abnormal motion pattern to an abnormal motion pattern judging module;
the 3D key point detection module is responsible for mapping the 2D key point coordinates of the patient to a 3D space to obtain 3D key point coordinates; providing the abnormal motion mode to an abnormal motion mode judging module;
the abnormal motion mode judging module is used for calculating the motion angle of the head and neck of the patient through the 2D and 3D key points of the patient and judging whether an abnormal motion mode exists or not;
the method comprises the steps that video image data of a patient are collected through portable equipment, the position of the patient in an image is detected through a patient target detection module, the 2D key point detection module and the 3D key point detection module further detect the 2D and 3D key point coordinates of the patient respectively, the abnormal motion mode distinguishing module calculates the head and neck operation angle of the patient through the human body key point coordinates, and the abnormal motion mode of the patient is judged.
In the patient target detection module, the specific position of the patient in the video image is detected to obtain an image with a bounding box.
In the 2D key point detection module, 2D human key points of a patient are detected, and the detected 2D human key points are detected.
In the 3D human key point detection module, the 2D human key points are promoted to a 3D space to obtain the 3D human key points.
In an abnormal motion mode judging module, evaluating the abnormal motion mode of the neck muscles of the patient by using the previously detected 2D and 3D human body key point information; the "head-neck" concept divides cervical dystonia into two types, according to the functional anatomical features of head and neck movement:
(1) head and neck complexes are mainly involved, abnormal head movements relative to the neck; beginning or ending in the skull or C1Abnormal contraction of the muscles of the vertebral body, pulling the head, causing head side inclination, head torsion and head flexion and extension;
(2) major involvement of C2-7A vertebral body abnormally moving with respect to a neck of a trunk; start and end at C2The muscles below the vertebral body contract, pulling the neck, causing a cervical roll, a cervical twist and a cervical flexion and extension;
there were 6 abnormal movement patterns evaluated, head roll, neck roll, head torsion, neck torsion, and head flexion and extension; when the head is tilted, the included angle between the longitudinal axis of the head and the cervical vertebra is reduced, and the included angle between the cervical vertebra and the thoracic vertebra is reduced; the cervical inclination is opposite, the included angle between the cervical vertebra and the thoracic vertebra is reduced, and the included angle between the head longitudinal axis and the cervical vertebra is not changed; when the patient is observed from the side, the included angle between the longitudinal axis of the head and the cervical vertebra is reduced when the head bends and stretches, and the included angle between the cervical vertebra and the thoracic vertebra is reduced when the neck bends and stretches; the laryngeal prominence lies substantially in the midline of the body when the head is twisted, and tends to deviate from the midline when the neck is twisted.
In the abnormal motion mode determination module, the head roll is determined by calculating an angle θ between a straight line passing through both eyes (Reye and Leye) and a horizontal line, and the angle θ is close to 0 degree in a normal state.
In the abnormal motion mode judging module, for the neck roll, the included angle theta between a straight line between the middle point of the shoulders (Lshoulder and Rshoulder) and the middle point of the ears (real and Lear) and a horizontal line is calculated to judge, the theta is close to 90 degrees in a normal state, and alpha is reduced when the roll phenomenon occurs. If α is smaller than a predetermined threshold value, it can be discriminated as the neck roll.
In the abnormal movement pattern discrimination module, the head torsion is judged by calculating the ratio eta of the Euclidean distance between ears (real and bear) and a nose (nose), and the eta evaluation formula is as follows:
Figure BDA0003168799770000031
the distances between ears and a nose are equal in a normal state, namely the value of eta is close to 1; when the head twist phenomenon occurs, the value of η becomes small, and if η is smaller than a predetermined threshold value, it can be determined that the head twist has occurred.
In the abnormal motion mode judging module, for neck torsion, judging by calculating a projection included angle beta of a thorax-spine vector and a thorax-tack vector on an x-y plane, wherein the included angle beta is close to 0 in a normal state, and the theta can be increased when a torsion phenomenon occurs; if β is greater than a specified threshold, neck torsion may be determined.
In the abnormal motion mode judging module, for head flexion and extension, the included angle gamma of the neck-thorax vector and the neck-head vector on the y-z plane is calculated for judgment, when the head flexion and extension phenomenon occurs, the included angle gamma is reduced, and if the gamma is smaller than a specified threshold value, the head flexion and extension can be judged.
In the abnormal motion mode judging module, neck flexion and extension are judged by calculating an included angle omega of a thorax-spine vector and a thorax-neck vector on a y-z plane, when the neck flexion and extension phenomenon occurs, the included angle omega is reduced, and if the included angle omega is smaller than a specified threshold value, neck flexion and extension can be judged.
Advantageous effects
1. According to the scheme, the abnormal movement mode of the cervical dystonia can be judged by using the video image data of the patient, and the evaluation can be completed only by using common terminal equipment, such as a smart phone, without professional medical equipment and professional operation skills.
2. The scheme is a non-contact evaluation method, which does not interfere with the autonomous movement of the patient and does not bring discomfort to the patient.
In conclusion, the scheme has great beneficial effect on the promotion of the abnormal cervical dystonia mode evaluation field.
Drawings
FIG. 1 flow chart of the main module of the evaluation model
FIG. 2 is a flow chart of the operation of the inventive arrangements
FIG. 32D human Key points
FIG. 43D human Key points
FIG. 5 Algorithm for abnormal motion pattern discrimination
FIG. 6 head roll
FIG. 7 cervical roll
Fig. 8 head twist
FIG. 9 neck twist
FIG. 10 shows the head flexed and extended
FIG. 11 shows the flexion and extension of the neck
Detailed Description
The evaluation model provided by the scheme of the invention mainly comprises four main modules, namely a patient target detection module, a 2D key point detection module, a 3D key point detection module and an abnormal motion mode discrimination module. The flow chart is shown in fig. 1. Wherein:
the patient target detection module is responsible for detecting the specific position of the patient in the image;
the 2D key point detection module is responsible for evaluating the specific 2D coordinates of the key points of the patient in the image; providing the abnormal motion pattern to an abnormal motion pattern judging module;
the 3D key point detection module is responsible for mapping the 2D key point coordinates of the patient to a 3D space to obtain 3D key point coordinates; providing the abnormal motion pattern to an abnormal motion pattern judging module;
the abnormal motion mode judging module is used for calculating the motion angle of the head and the neck of the patient through the 2D and 3D key points of the patient and judging whether an abnormal motion mode exists.
The neck muscle tension abnormal movement mode evaluation method based on the computer vision technology only needs one common terminal device for collecting data, such as a smart phone and the like, does not need to be in direct contact with a patient, and does not need to be instructed and operated by a professional doctor.
The operational flow diagram of the inventive arrangement is shown in fig. 2. Firstly, the portable equipment is used for collecting video image data of a patient, then the position of the patient in the image is detected, the 2D and 3D key point coordinates of the patient are further detected, the head and neck operation angle of the patient is calculated by utilizing the human body key point coordinates, and finally the abnormal motion mode of the patient is judged.
In the patient target detection module, a deep learning target detection method, such as fast R-CNN, YOLO, etc., is used to detect the specific position of the patient in the video image, and an image with a bounding box is obtained. In the 2D keypoint detection module, a deep learning human keypoint detection method, such as an HRNet model, a Hourglass model, etc., is used to detect 2D human keypoints of a patient, and the detected 2D human keypoints are as shown in fig. 3.
In the 3D human key point detection module, a deep learning 3D human key point detection method, such as simplebaeline 3D and videodose 3D, is used to promote the 2D human key points into the 3D space, and the obtained 3D human key points are as shown in fig. 4.
In the abnormal motion pattern discrimination module, the abnormal motion pattern of the neck muscles of the patient is evaluated by using the previously detected 2D and 3D human body key point information. The "head-neck" concept divides cervical dystonia into two types, according to the functional anatomical features of head and neck movement:
(1) the head and neck complex is mainly involved, and abnormal movement of the head relative to the neck is involved. Initiation or termination with skull or C1Abnormal contraction of the muscles of the vertebral body, pulling on the head, can lead to head roll, head torsion, and head flexion and extension.
(2) Major involvement of C2-7Vertebral bodies, abnormally moving relative to the neck of the torso. Start and end at C2Contraction of the muscles below the vertebral body, stretching the neck, can result in neck roll, neck torsion, and neck flexion and extension.
In summary, there are 6 abnormal motion patterns to be evaluated, head roll, neck roll, head torsion, neck torsion, and head flexion and neck flexion. When the head is tilted, the included angle between the longitudinal axis of the head and the cervical vertebra is reduced, and the included angle between the cervical vertebra and the thoracic vertebra is reduced; the neck side inclination is opposite, the included angle between the cervical vertebra and the thoracic vertebra becomes smaller, and the included angle between the head longitudinal axis and the cervical vertebra does not change. When the patient is observed from the side, the included angle between the longitudinal axis of the head and the cervical vertebra is reduced when the head bends and stretches, and the included angle between the cervical vertebra and the thoracic vertebra is reduced when the neck bends and stretches. The laryngeal prominence lies substantially in the midline of the body when the head is twisted, and tends to deviate from the midline when the neck is twisted.
An algorithm flow chart for abnormal motion pattern discrimination is shown in fig. 5.
For the head roll, the judgment is made by calculating an angle θ between a straight line through both eyes (Reye and Leye) and a horizontal line, as shown in fig. 6. In the normal state, θ approaches 0 degrees, and becomes large when the head roll phenomenon occurs, and if θ is larger than a predetermined threshold value, it can be determined that the head roll occurs.
For the cervical roll, the determination is made by calculating the angle θ between the horizontal line and the straight line passing through the center point of the shoulders (Lshoulder and Rshoulder) and the center point of the ears (real and Lear), as shown in fig. 7. In the normal state, θ approaches 90 degrees, and when the roll phenomenon occurs, α becomes small. If α is smaller than a predetermined threshold value, it can be discriminated as the neck roll.
For head torsion, the judgment is carried out by calculating the ratio eta of Euclidean distances between ears (real and bear) and nose (nose), and the eta evaluation formula is as follows:
Figure BDA0003168799770000051
the distances between ears and nose are equal in normal state, i.e. the value of η is close to 1. When the head twist phenomenon occurs, the value of η becomes small as shown in fig. 8. If η is smaller than a predetermined threshold, it can be determined as head twist.
For neck torsion, judgment is carried out by calculating the projection included angle beta of the thorax-spine vector and the thorax-neck vector on the x-y plane, as shown in FIG. 9. In a normal state, the included angle β is close to 0, and when the twisting phenomenon occurs, θ increases. If β is greater than a specified threshold, neck torsion may be determined.
For head flexion and extension, the determination is made by calculating the angle γ between the neck-thorax vector and the neck-head vector in the y-z plane, as shown in FIG. 10. When the head bends and stretches, the included angle gamma is reduced, and if gamma is smaller than a specified threshold value, the head bends and stretches can be judged.
The neck flexion and extension is judged by calculating the included angle omega of the thorax-spine vector and the thorax-neck vector on the y-z plane, as shown in FIG. 11. When the neck flexion and extension phenomenon occurs, the included angle ω is reduced, and if ω is smaller than a specified threshold, the neck flexion and extension can be determined.

Claims (4)

1. The utility model provides a neck dystonia identification system based on vision, its characterized in that includes four main modules, is patient target detection module, 2D key point detection module, 3D key point detection module, unusual motion mode discrimination module respectively, wherein:
the patient target detection module is responsible for detecting the specific position of the patient in the image;
the 2D key point detection module is responsible for evaluating the specific 2D coordinates of the key points of the patient in the image; providing the abnormal motion pattern to an abnormal motion pattern judging module;
the 3D key point detection module is responsible for mapping the 2D key point coordinates of the patient to a 3D space to obtain 3D key point coordinates; providing the abnormal motion pattern to an abnormal motion pattern judging module;
the abnormal motion mode judging module is used for calculating the motion angle of the head and neck of the patient through the 2D and 3D key points of the patient and judging whether an abnormal motion mode exists or not;
the method comprises the steps that video image data of a patient are collected by using portable equipment, a patient target detection module detects the position of the patient in an image, a 2D key point detection module and a 3D key point detection module respectively detect the 2D and 3D key point coordinates of the patient, an abnormal motion mode distinguishing module calculates the head and neck operation angle of the patient by using the human body key point coordinates, and the abnormal motion mode of the patient is judged;
the method is characterized in that in an abnormal motion mode discrimination module, the abnormal motion mode of neck muscles of a patient is evaluated by using the 2D and 3D human body key point information detected before; the "head-neck" concept divides cervical dystonia into two types according to the functional anatomic features of head and neck movement:
(1) head and neck complexes are mainly involved, abnormal head movements relative to the neck; beginning or ending in the skull or C1Abnormal contraction of the muscles of the vertebral body, pulling the head, causing head roll, head torsion and head flexion and extension;
(2) major involvement of C2-7A vertebral body abnormally moving with respect to a neck of a trunk; start and end at C2The muscles below the vertebral body contract, pulling the neck, causing a cervical roll, a cervical twist and a cervical flexion and extension;
there were 6 abnormal movement patterns evaluated, head roll, neck roll, head torsion, neck torsion, and head flexion and extension; when the head is tilted, the included angle between the longitudinal axis of the head and the cervical vertebra is reduced, and the included angle between the cervical vertebra and the thoracic vertebra is reduced; the cervical inclination is opposite, the included angle between the cervical vertebra and the thoracic vertebra is reduced, and the included angle between the head longitudinal axis and the cervical vertebra is not changed; when the patient is observed from the side, the included angle between the longitudinal axis of the head and the cervical vertebra is reduced when the head bends and stretches, and the included angle between the cervical vertebra and the thoracic vertebra is reduced when the neck bends and stretches; the laryngeal prominence lies substantially in the midline of the body when the head is twisted, and tends to deviate from the midline when the neck is twisted;
in the abnormal motion mode judging module, for head roll, judging by calculating an included angle theta between a straight line of eyes (eye and eye) and a horizontal line, wherein theta is close to 0 degree in a normal state, and can be increased when a head roll phenomenon occurs, and if theta is larger than a specified threshold value, the head roll can be judged;
in the abnormal motion mode judging module, for neck roll, judging by calculating an included angle theta between a straight line passing through the middle points of shoulders (Lshoulder and Rshoulder) and ears (real and Lear) and a horizontal line, wherein theta is close to 90 degrees in a normal state, and alpha is reduced when a roll phenomenon occurs; if alpha is smaller than a specified threshold value, the neck is judged to be inclined;
in the abnormal movement pattern discrimination module, the head torsion is judged by calculating the ratio eta of the Euclidean distance between ears (real and bear) and a nose (nose), and the eta evaluation formula is as follows:
Figure FDA0003566874930000011
the distances between ears and a nose are equal in a normal state, namely the value of eta is close to 1; when the head torsion phenomenon occurs, the value of eta is reduced, and if eta is smaller than a specified threshold value, the head torsion can be judged;
in the abnormal motion mode judging module, for neck torsion, judging by calculating a projection included angle beta of a thorax-spine vector and a thorax-tack vector on an x-y plane, wherein the included angle beta is close to 0 in a normal state, and the theta can be increased when a torsion phenomenon occurs; if beta is larger than a specified threshold value, neck torsion can be judged;
in the abnormal motion mode judging module, judging the head flexion and extension by calculating an included angle gamma between a neck-thorax vector and a neck-head vector on a y-z plane, wherein the included angle gamma is reduced when the head flexion and extension phenomenon occurs, and if the gamma is smaller than a specified threshold value, the head flexion and extension can be judged;
in the abnormal motion mode judging module, neck flexion and extension are judged by calculating an included angle omega of a thorax-spine vector and a thorax-neck vector on a y-z plane, when the neck flexion and extension phenomenon occurs, the included angle omega is reduced, and if the included angle omega is smaller than a specified threshold value, neck flexion and extension can be judged.
2. A vision-based cervical dystonia identification system as claimed in claim 1 wherein in the patient target detection module, the specific position of the patient in the video image is detected to obtain a bounding box image.
3. The vision-based cervical dystonia identification system of claim 1, wherein 2D human key points of the patient are detected in the 2D key point detection module, the detected 2D human key points.
4. The vision-based cervical dystonia identification system of claim 1, wherein in the 3D human keypoint detection module, 2D human keypoints are promoted into 3D space, resulting in 3D human keypoints.
CN202110812666.7A 2021-07-19 2021-07-19 Neck dystonia identification system based on vision Active CN113554609B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110812666.7A CN113554609B (en) 2021-07-19 2021-07-19 Neck dystonia identification system based on vision

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110812666.7A CN113554609B (en) 2021-07-19 2021-07-19 Neck dystonia identification system based on vision

Publications (2)

Publication Number Publication Date
CN113554609A CN113554609A (en) 2021-10-26
CN113554609B true CN113554609B (en) 2022-07-08

Family

ID=78132010

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110812666.7A Active CN113554609B (en) 2021-07-19 2021-07-19 Neck dystonia identification system based on vision

Country Status (1)

Country Link
CN (1) CN113554609B (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110210446A (en) * 2019-06-12 2019-09-06 广东工业大学 A kind of sitting posture condition detection method, device, equipment and the medium of target object

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9550029B2 (en) * 2012-10-30 2017-01-24 Elwha Llc Systems and methods for guiding injections
CN108577871B (en) * 2018-05-17 2024-09-17 上海市同济医院 Multi-axis sensor combination device for accurately evaluating cervical dystonia
CN109460702B (en) * 2018-09-14 2022-02-15 华南理工大学 Passenger abnormal behavior identification method based on human body skeleton sequence
CN111414780B (en) * 2019-01-04 2023-08-01 卓望数码技术（深圳）有限公司 Real-time intelligent sitting posture distinguishing method, system, equipment and storage medium
CN110188728A (en) * 2019-06-06 2019-08-30 四川长虹电器股份有限公司 A kind of method and system of head pose estimation
CN112084967A (en) * 2020-09-12 2020-12-15 周美跃 Limb rehabilitation training detection method and system based on artificial intelligence and control equipment

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110210446A (en) * 2019-06-12 2019-09-06 广东工业大学 A kind of sitting posture condition detection method, device, equipment and the medium of target object

Also Published As

Publication number Publication date
CN113554609A (en) 2021-10-26

Similar Documents

Publication Publication Date Title
CN101901485B (en) 3D free head moving type gaze tracking system
Yang et al. Automatic 3-D imaging and measurement of human spines with a robotic ultrasound system
Dunne et al. Wearable monitoring of seated spinal posture
CN110772255A (en) Method for measuring human body scoliosis angle based on posture and position sensor
CN110059670B (en) Non-contact measuring method and equipment for head and face, limb movement angle and body posture of human body
CN113674257A (en) Method, device and equipment for measuring scoliosis angle and storage medium
Bumacod et al. Image-processing-based digital goniometer using OpenCV
CN114360043B (en) Model parameter calibration method, sight tracking method, device, medium and equipment
CN112446162A (en) Intervertebral disc stress measuring device and method based on attitude recognition
CN114983396A (en) Automatic detection system for scoliosis
CN113554609B (en) Neck dystonia identification system based on vision
CN106618580B (en) Strabismus and nystagmus head position detection method, device and system
Weidenbacher et al. Detection of head pose and gaze direction for human-computer interaction
TWI681755B (en) System and method for measuring scoliosis
CN101879057B (en) Three-dimensional head spatial repositioning offset measuring system and operation method thereof
CN116580359A (en) Construction personnel severe fatigue detection system and method based on space-time characteristics
CN213874121U (en) Facial proportion and symmetry measuring tool
JP2023549865A (en) Method and system for measuring binocular distance for children
CN201505136U (en) Three-dimensional skull space recovery deflection measuring system
Johnston et al. A Fully Automated System for Sizing Nasal PAP Masks Using Facial Photographs
CN109567816B (en) Monocular vision-based cervical vertebra mobility measuring system and method
Guihard et al. Three‐Dimensional Quantification of Facial Morphology and Movements Using a Wearable Helmet
Kutílek et al. Determining the position of head and shoulders in neurological practice with the use of cameras
Giorgi et al. Morphological Analysis of 3D Faces for Weight Gain Assessment.
Li et al. A photogrammetric method for the measurement of three-dimensional cervical range of motion

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant