WO2024008210A1 - 锥形束ct装置的控制方法、锥形束ct装置和介质 - Google Patents

锥形束ct装置的控制方法、锥形束ct装置和介质 Download PDF

Info

Publication number
WO2024008210A1
WO2024008210A1 PCT/CN2023/114612 CN2023114612W WO2024008210A1 WO 2024008210 A1 WO2024008210 A1 WO 2024008210A1 CN 2023114612 W CN2023114612 W CN 2023114612W WO 2024008210 A1 WO2024008210 A1 WO 2024008210A1
Authority
WO
WIPO (PCT)
Prior art keywords
face
control method
cone beam
angle
keypt
Prior art date
Application number
PCT/CN2023/114612
Other languages
English (en)
French (fr)
Inventor
程阵
吉倩
王康
Original Assignee
合肥美亚光电技术股份有限公司
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 合肥美亚光电技术股份有限公司 filed Critical 合肥美亚光电技术股份有限公司
Publication of WO2024008210A1 publication Critical patent/WO2024008210A1/zh

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/54—Control of apparatus or devices for radiation diagnosis
    • A61B6/545—Control of apparatus or devices for radiation diagnosis involving automatic set-up of acquisition parameters
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03—Computed tomography [CT]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03—Computed tomography [CT]
    • A61B6/032—Transmission computed tomography [CT]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/50—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications
    • A61B6/501—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications for diagnosis of the head, e.g. neuroimaging or craniography
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/58—Testing, adjusting or calibrating thereof
    • A61B6/582—Calibration
    • A61B6/585—Calibration of detector units
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/20—Movements or behaviour, e.g. gesture recognition

Definitions

  • the present disclosure relates to the technical field of head positioning, and in particular, to a control method of a cone beam CT device, a cone beam CT device and a medium.
  • the head adjustment method of CBCT (Cone Beam CT) is for the doctor to manually adjust the motor and observe the position of the laser beam to determine whether the head position is appropriate.
  • the method of human observation has great disadvantages. Different doctors have different understandings of the standards, so the positioning position may not be appropriate, thus affecting the shooting quality of the CBCT machine, affecting the image diagnosis, and in severe cases, causing bad films. Continuously adjusting the CBCT machine will also increase the load on the radiation source and reduce the service life of the radiation source.
  • the present disclosure proposes a control method for a cone beam CT device.
  • the cone beam CT device includes a column, a camera, and a laser beam.
  • the control method includes: acquiring a face image through the camera, and Perform key point detection on the face image to obtain the key points of the face; determine whether the face posture is normal based on the key points of the face; if normal, turn on the laser beam and extract it from the face image Laser beam position; drive the column to move in the first direction according to the laser beam position and the key points of the human face; when the laser beam position reaches the first designated position, control the cone beam CT device Take a CT scan;
  • the control method also includes: determining whether the face image contains a complete face; if not, driving the column to move in a set direction; if yes, performing key point detection on the face image A step of;
  • Determining whether the facial posture is normal based on the facial key points includes: calculating a first pitch angle, a first yaw angle and a first roll angle based on the facial key points; based on the first pitch angle, the first yaw angle and the first The roll angle determines whether the facial posture is normal;
  • the first pitch angle is calculated by the following formula:
  • KeyPt 51 represents the key point in the middle of the nose and the highest position among the nose key points
  • KeyPt 16 represents the key point in the middle of the face and the lowest position among the face outline key points
  • KeyPt 51.y represents the key point KeyPt 51 .
  • the coordinates in the first direction KeyPt 51.x represents the coordinates of the key point KeyPt 51 in the second direction
  • Pitch represents the first pitch angle
  • k 1 is the first preset constant
  • KeyPt 97 represents the key point in the middle of the eye among the left eye key points
  • Yaw represents the first yaw angle
  • k 2 is the second preset constant
  • RollTan (KeyPt 97 .x-KeyPt 96 .x)/(KeyPt 97 .y-KeyPt 96 .y)
  • KeyPt 96 represents the key point in the middle of the eye among the right eye key points
  • Roll represents the first roll angle
  • k 3 is the third preset constant
  • Determining whether the human face posture is normal based on the first pitch angle, the first yaw angle and the first roll angle includes: determining the first pitch angle, the first yaw angle and the first roll angle. Whether the absolute values of the first roll angles are all less than or equal to the first preset threshold; if so, it is determined that the facial posture is normal, otherwise it is determined that the facial posture is abnormal;
  • the control method further includes: prompting in at least one of the following ways: when the absolute value of the first pitch angle is greater than the first preset threshold, sending a first prompt message to prompt the patient to perform a pitching operation; When the absolute value of the first yaw angle is greater than the first preset threshold, a second prompt message is issued to prompt the patient to perform a shaking operation; when the absolute value of the first roll angle is greater than the first preset threshold, When the threshold is set, a third prompt message is issued to prompt the patient to proceed Perform head turning operation;
  • control method further includes: determining whether the straight line connecting the key point KeyPt 51 and the key point KeyPt 16 coincides with the second designated position; if not, issuing a fourth prompt information to prompt the patient to move; if they overlap, perform the step of turning on the laser beam;
  • Extracting the laser beam position from the face image includes: determining all laser beam rows in the face image, and recording the row coordinates of each laser beam row, wherein the row coordinates are The coordinates of the laser beam row in the first direction; calculate the mean value of all row coordinates, and use the mean value as the laser beam position; wherein all pixels of the laser beam row satisfy the R pixel value and The difference between the G pixel values is greater than the second preset threshold, and the difference between the R pixel values and the B pixel values is greater than the third preset threshold;
  • the position of the laser beam reaches the first designated position, including: the position of the laser beam coincides with the coordinates of the key point KeyPt 31 in the first direction, where KeyPt 31 represents the first key point of the human face outline. The second highest key point in the direction;
  • the control method also includes: obtaining a second pitch angle, a second yaw angle and a second roll angle based on the current face image; and determining the second pitch angle, second yaw angle and second roll angle. Whether the absolute values of the two pitch angles, the second yaw angle and the second roll angle are greater than the first preset threshold; when the second pitch angle, the second yaw angle and the second roll angle If any of the angles is greater than the first preset threshold, control the cone beam CT device to stop CT imaging;
  • the control method also includes: calculating a first difference between the third pitch angles of two frames of face images separated by a first preset time period, the second difference between the third yaw angle and the third difference between the third roll angle; in the absolute value of the first difference, the absolute value of the second difference, the third difference When any of the absolute values of the differences is greater than the fourth preset threshold, the cone beam CT device is controlled to stop CT imaging.
  • the control method of the cone beam CT device in the embodiment of the present disclosure roughly determines the facial posture in advance, that is, obtains the facial image through the camera and detects key points of the facial image, and determines the facial posture based on the key points of the human face. Is it normal? When the face posture is normal, adjust the shooting position of the device, that is, turn on the laser, the device automatically extracts the position of the laser beam, and drives the column to move according to the position of the laser beam until the laser beam moves to the designated position and then performs CT shooting.
  • the relative position of the beam on the human face can indicate the position of the cone beam CT shooting area relative to the patient's head. This can automatically adjust the shooting position and shooting posture for different patients, improve the quality of the captured images, reduce the bad film rate, and reduce radiation to the patient and improve the service life of the machine.
  • the present disclosure proposes a cone beam CT device, which device includes: a column, a camera, and a controller, wherein the controller includes a memory, a processor, and a computer program stored on the memory, said computer When the program is executed by the processor, the above control method is implemented.
  • the cone beam CT device in the embodiment of the present disclosure automatically adjusts the shooting position and shooting posture for different patients through the above-mentioned control method of the cone beam CT device, improves the quality of the captured images, reduces the bad film rate, and at the same time reduces the radiation to the patient. Increase the service life of the machine.
  • the present disclosure proposes a computer-readable storage medium on which a computer program is stored.
  • the computer program is executed by a processor, the above-mentioned control method of a cone beam CT device is implemented.
  • Figure 1 is a flow chart of a control method of a cone beam CT device according to an embodiment of the present disclosure
  • Figure 2 is a flow chart for determining a face image according to an embodiment of the present disclosure
  • Figure 3 is a flow chart for determining whether the facial posture is normal based on facial key points according to an embodiment of the present disclosure
  • Figure 4 is a flow chart of step S22 according to an embodiment of the present disclosure.
  • Figure 5 is a flow chart of whether the facial key points satisfy the second specified position according to an embodiment of the present disclosure
  • Figure 6 is a flow chart for extracting the position of a laser beam from a face image according to one embodiment of the present disclosure
  • Figure 7 is a flow chart of the CT imaging process of the cone beam CT device according to an embodiment of the present disclosure.
  • Figure 8 is a flow chart for judging any two frames of face images according to an embodiment of the present disclosure.
  • Figure 9 is a schematic diagram of human face key points according to an embodiment of the present disclosure.
  • Figure 10 is a schematic diagram of pitch angle, yaw angle, and roll angle according to an embodiment of the present disclosure
  • Figure 11 is a schematic structural diagram of a cone beam CT device according to an embodiment of the present disclosure.
  • Figure 12 is a front view of a cone beam CT device according to an embodiment of the present disclosure.
  • the cone beam CT device of this embodiment can be applied to special CBCT in various directions such as the oral cavity, ears, nose, and other parts of the head, for example, oral CBCT.
  • a cone beam CT device includes a column, a camera and a laser, and both the camera and the laser are installed on the column.
  • the column is movable and can move up and down according to control instructions; the camera is used to capture facial images.
  • a cone beam CT device may include a base and a column.
  • the column includes a movable part, a fixed part and a positioning mechanism.
  • the bottom end of the movable part is fixed on the base.
  • the movable part is slidably located on the fixed part.
  • the movable part is relative to the fixed part. After the position is changed, the positioning mechanism can be used for positioning so that the movable part and the fixed part are relatively fixed at this position.
  • the aforementioned column is movable and corresponds to the movable part in this example.
  • FIG. 1 is a flowchart of a control method of a cone beam CT apparatus according to an embodiment of the present disclosure.
  • control method of the cone beam CT device includes:
  • S1 obtain the face image through the camera, perform key point detection on the face image, and obtain the key points of the face.
  • the camera can be installed on the cone beam CT, or a stand for the camera can be installed separately.
  • the position of the camera can be adjusted up and down.
  • the height of the camera is adjusted according to the height of the current patient's head when shooting, so that the patient's face is at least partially located in the shooting area of the camera. For example, for standing cone beam CT, it can be adjusted according to the height of the patient; for seated cone beam CT, it can be adjusted according to the height of the patient's head after sitting down.
  • the face image when controlling the cone beam CT device, can be collected in real time through the camera, and the collection frequency of the camera can be 30-60fps (frames per second). If the complete face is not collected in the face image and key point detection is performed directly, the face key point detection will be incomplete, which will affect the later judgment of the face key points. Therefore, after collecting the face image, you can first make a preliminary judgment on the collected face image (such as judging whether the collected face image contains the patient's complete face), and then perform key point detection.
  • control method of the cone beam CT device also includes: determining whether the face image contains a complete face; if not, driving the column to move in the set direction; if so , then perform the step of detecting key points on the face image.
  • a face image contains a complete face
  • the set direction can be the same as the direction of the incomplete part of the face image relative to the entire face image. For example, if the face image is collected The features on the face image obtained do not include the chin feature. The face image lacks the chin part.
  • the direction of the chin part relative to the entire face image is downward, so the driving column moves downward so that the missing chin part can enter the camera.
  • the shooting area if the features on the collected face image do not include forehead features, the face image lacks the forehead part, and the direction of the forehead part relative to the entire face image is upward, then the driving column moves upward to make the missing forehead part You can enter the camera's shooting area.
  • the camera still collects face images in real time until the collected face image contains a complete face and stops moving the column. Take this time The collected face images of complete faces are used to detect facial key points. If the face image initially collected by the camera already contains a complete face, there is no need to drive the column and the key points of the face can be detected directly.
  • the face area can be intercepted from the collected face image containing the complete face, only the key points of the face area can be detected, and the intercepted face area can be sent to the human face area.
  • the face key points shown in Figure 9 are obtained.
  • the face key point algorithm can detect 98 key points.
  • the specific distribution of the 98 key points is shown in Figure 9.
  • the facial key point algorithm can also be 68 key point detection, 106 key point detection, etc., which are not limited here.
  • the key points can be identified through the neural network algorithm.
  • the identified key points at least include the key points used to judge the face posture.
  • the neural network framework is built in advance, and then the training samples marked with these key points are used. By training the built neural network, you can obtain a neural network model that can be used to identify these key points.
  • the key points to be identified are the key points used to calculate the pitch angle, yaw angle, and roll angle later.
  • Facial posture can be represented by facial key points, such as the relative positions between key points, including angle, orientation, distance, etc.
  • the face pose standard is set based on this representation, so that it can be judged whether the face pose meets the preset face pose standard. If it satisfies the preset face pose standard, the face pose is considered normal. If it does not, the face pose is considered normal.
  • the facial posture is abnormal.
  • the face posture standard is set based on whether the shooting requirements are met. For example, the shooting requirements may include facing the side of the camera and facing directly forward.
  • pitch angle represents the degree to which the head raises or lowers its head
  • yaw angle represents the degree to which the head tilts left or right
  • roll represents the degree to which the head can be turned left or right.
  • the standard face pose is represented by the rotation angle in the three-dimensional direction of space calculated based on key points, such as pitch angle, yaw angle and roll angle.
  • the standard face pose can be on the side of the face facing the camera and facing forward. Based on the front position, a rotation test in two directions is performed at each rotation angle, so that the rotation amplitude limit in each direction, which is the face posture standard, can be determined, thereby eliminating the need for precise calibration of the camera installation, such as , after lowering the head to a certain angle, the cone beam CT can shoot normally. This lowering angle is smaller than the lowering angle at which the cone beam CT cannot shoot normally. At this time, the pitch angle corresponding to this position is the standard face posture in one direction.
  • the pitch angle corresponding to the raised head corresponds to the five directions respectively.
  • Face pose standard can be equal. The specific direction is determined based on the positive and negative values of the pitch angle, yaw angle, and roll angle, which will be described in detail later.
  • determining whether the facial posture is normal based on the key points of the human face may include:
  • the first pitch angle is calculated by the following formula:
  • KeyPt 51 represents the key point in the middle of the nose and the highest position among the nose key points
  • KeyPt 16 represents the key point in the middle of the face and the lowest position among the facial contour key points
  • KeyPt 51.y represents the key point KeyPt 51 in the The coordinates in one direction
  • KeyPt 51.x represents the coordinates of the key point KeyPt 51 in the second direction
  • Pitch represents the first pitch angle.
  • k 1 is the first preset constant
  • arcsin is the inverse sine function.
  • the first direction is the vertical direction in Figure 9
  • the second direction is the left-right direction in Figure 9.
  • KeyPt 97 represents the key point in the middle of the eye among the left eye key points
  • Yaw represents the first yaw angle
  • k 2 is the second preset constant
  • arcsin is the inverse sine function.
  • KeyPt 96 represents the key point in the middle of the eye among the right eye key points
  • Roll represents the first roll angle
  • k 3 is the third preset constant
  • arctan is the arctangent function.
  • k 1 , k 2 , and k 3 can be adjusted according to actual needs, and can take the same value or different values.
  • each angle can be normalized to radian units, then the three constants can be preset as
  • the sign of each angle is adjusted to change the positive and negative values indicating the patient's head deflection state; of course, the three values can also be 1 or -1.
  • the first direction is the Y-axis direction
  • the second direction is the X-axis direction.
  • the angle formula calculates the first roll angle. After the first roll angle is obtained, it is rounded.
  • KeyPt 51 represents the key point numbered 51
  • KeyPt 16 represents the key point numbered 16
  • KeyPt 97 represents the key point numbered 97
  • KeyPt 96 represents the key point numbered 97. 96 key points.
  • the key points used can refer to the positions of the key points in Figure 9, and select alternative key points, for example, using key points with similar positions. Substitutions are made so that the corresponding angles can be calculated in a variety of ways.
  • S22 Determine whether the human face posture is normal based on the first pitch angle, the first yaw angle and the first roll angle.
  • determining whether the human face posture is normal based on the first pitch angle, the first yaw angle, and the first roll angle includes:
  • the first pitch angle, the first yaw angle and the first roll angle are used to represent the human face posture
  • the first preset threshold is the preset human face posture standard.
  • the first preset threshold can take a value in the range of 3-7 degrees, for example, the value can be 5 degrees. If the absolute values of k 1 , k 2 , and k 3 are not 1, the values can also be taken within the range of 3-7 degrees and then multiplied by the corresponding absolute values of k 1 , k 2 , or k 3 .
  • the first preset threshold is 5 degrees
  • the absolute values of the first pitch angle, the first yaw angle, and the first roll angle are all less than or equal to 5 degrees
  • the first pitch angle When the value of either of the first yaw angle and the first roll angle is greater than 5 degrees, it is determined that the facial posture is abnormal.
  • the facial posture When the facial posture is abnormal, the patient can be prompted to adjust the posture.
  • the first preset thresholds set for the first pitch angle, the first yaw angle and the first roll angle may be the same or different, and may be specifically set according to the CT imaging effect. , you can set a smaller threshold in the direction that has a relatively greater impact on the shooting effect, and conversely, set a larger threshold in the direction that has a relatively smaller impact on the shooting effect.
  • the first preset threshold can be set according to post-processing requirements after cone beam CT imaging, a panoramic view can be generated according to the position of the middle incisor in the CT image, or the middle incisor can be identified according to the CT image. If the left and right rotation angle is too large, it may cause the middle incisor to be in an offset position of the overall teeth, eventually leading to inaccurate or unrecognizable recognition, and it may not be possible to generate a panorama. In this case, the yaw angle corresponding to the third The first preset threshold value can be set slightly smaller.
  • the first preset threshold value corresponding to the roll angle can also be set slightly smaller, while the first preset threshold value corresponding to the pitch angle can be set slightly larger, so as to ensure that the requirements are met.
  • the imaging requirements will not make adjustments too frequent and increase the patient's anxiety.
  • the first preset threshold corresponding to the first roll angle is 4.5 degrees
  • the first preset threshold corresponding to the first yaw angle is 4.5 degrees
  • the first preset threshold corresponding to the first pitch angle is 5.5 degrees.
  • the first prompt message is issued to prompt the patient to perform a pitching action
  • a second prompt message is issued to prompt the patient to perform a head-turning movement
  • a third prompt message is issued to prompt the patient to perform a head-turning movement.
  • the patient can adjust the head posture according to the display animation, voice prompts or the doctor's guidance. If the face posture is judged to be abnormal based on the first pitch angle, the patient will be prompted to hold his head up based on the positive or negative of the first pitch angle. Or lower the head; if the face posture is judged to be abnormal based on the first yaw angle, prompt the patient to shake his head left or right according to the positive or negative of the first yaw angle; if the face posture is judged to be abnormal based on the first roll angle If it is normal, the patient will be prompted to tilt to the left or right according to the positive or negative value of the first roll angle. That is to say, according to the positive and negative of each angle, it can be determined that the patient's head is currently in a certain deflection state in a certain direction, which can prompt adjustment in the opposite direction.
  • the first pitch angle is positive and the value is greater than the first preset threshold, it means that the patient is in a head-up state, and the patient should be prompted to lower his head slightly.
  • the first pitch angle is negative and the value is greater than the first preset threshold, when the first yaw angle is positive and greater than the first preset threshold, it means that the patient is shaking his head to the right, and the patient should be prompted to shake his head slightly to the left.
  • first yaw angle is negative and the value is greater than the first preset threshold, it means that the patient is shaking his head to the left, and the patient should be prompted to shake his head slightly to the right; if the first roll angle is positive and the value is greater than the first preset threshold , indicating that the patient is tilting to the left, and the patient should be prompted to tilt to the right. If the first roll angle is negative and the value is greater than the first preset threshold, it means that the patient is tilting to the right, and the patient should be prompted to tilt to the left. .
  • the first preset threshold is 5 degrees.
  • a voice prompt "Please raise the patient's head slightly” is issued.
  • a voice prompt is issued.
  • the voice prompt "Please shake the patient's head slightly to the left” is given.
  • the voice prompt "Please tilt the patient's head slightly to the right” is given; when the first roll angle is positive and the value is greater than 5 degrees, the voice prompt "Please tilt the patient slightly to the right” is given.
  • a voice prompt is issued, "Please tilt the patient slightly to the left."
  • the patient can also adjust the head posture according to the image guidance prompts on the display screen. This allows the patient to know how to adjust his posture to quickly achieve shooting conditions.
  • the voice prompt when issued based on the calculation results of the first pitch angle, the first yaw angle and the first roll angle, it may also include instructing the patient to raise his head or shake his head or turn his head at a specific angle.
  • the camera collects the face image in real time and calculates the adjusted pitch angle, yaw angle and roll angle until the adjusted pitch angle , the yaw angle and the roll angle are both less than or equal to the first preset threshold. If both are less than or equal to 5 degrees, the face posture at this time is normal.
  • turning on the laser can be achieved by turning on the laser to emit a laser beam.
  • the second designated position is The setting is determined based on the relative positional relationship between the camera shooting area and the CT shooting area. In the left and right directions of the face, when the face is in the required first position of the CT shooting area, the face is in the second position of the image collected by the camera. Then the second position is the second designated position. If the patient's head is in the central area of the ray source and detector of the cone beam CT in the left and right directions of the face, that is, the first position, it is conducive to improving the CT shooting effect.
  • the patient's face is in the first position in the left and right directions of the face.
  • the middle area of the image collected by the camera is the second position, then the second designated position can be that the face represented in the image collected by the camera is in the middle of the shooting area, specifically it can refer to the center of the image in the left and right directions of the face
  • the line position can also be the center area position that is symmetrical to the center line position.
  • the control method of the cone beam CT device further includes: determining whether the straight line connecting the key point KeyPt 51 and the key point KeyPt 16 is consistent with The second designated positions coincide with each other; if they do not coincide with each other, a fourth prompt message is sent to prompt the patient to move; if they coincide with each other, the steps of turning on the laser are performed.
  • the second designated position here may be, for example, the center line position of the image collected by the camera in the left and right direction of the human face, or a central area of a certain width defined to the left and right of the center line position.
  • the straight line connecting key point KeyPt 51 and key point KeyPt 16 is between the key point in the middle of the nose and the highest position among the nose key points and the key point in the middle of the face and the lowest position among the face contour key points.
  • the straight line connecting key point KeyPt 51 and key point KeyPt 16 coincides with the second specified position, it means that the position of the patient's head has been adjusted from the left and right directions of the face.
  • a fourth prompt message is sent to prompt the patient to move.
  • the fourth prompt message is in the same form as the above prompt message and can be a message on the display screen. Image guidance information and voice prompt information.
  • a voice prompt "Please move the patient to the left" is issued.
  • a line representing the second designated position is displayed on the display screen, and the patient can move according to the display screen so that the tip of his nose falls on the line. Turn on the laser until the patient has moved to the second designated position.
  • the aforementioned coincidence is in a broad sense, and may mean that the connection line strictly overlaps the center line, or is within a preset distance range, or the connection line is in a set central area.
  • the position of the patient's head in the left and right directions of the face can be adjusted, thereby achieving more accurate CT shooting.
  • the first designated position is determined based on the relative positional relationship between the camera shooting area and the CT shooting area during shooting.
  • the position of the laser beam relative to the CT shooting area is fixed.
  • the fourth position can be the first designated position.
  • a certain feature on the human face can be selected as the alignment position, such as the inner corner point of the eye.
  • the laser beam When the sample patient is at the required position in the CT shooting area, the laser beam is irradiated to the inner corner point of the eye, and the face is in the camera image collection area. In this way, during the specific adjustment process, it can be judged based on the face image collected by the camera whether the laser beam reaches the focus in the eye.
  • the laser beam can be adjusted together with the shooting module of the cone beam CT.
  • the shooting area of the cone beam CT can also cover the patient's head, that is, the position of the patient's head meets the shooting conditions of the cone beam CT.
  • first designated position and the second designated position limit the patient's head position from the first direction and the second direction respectively, achieving a more accurate prompt for the patient to adjust the head position to meet the shooting requirements as soon as possible. condition.
  • extracting the laser beam position from the face image includes:
  • S31 Determine all laser beam rows in the face image, and record the row coordinates of each laser beam row, where the row coordinates are the coordinates of the laser beam row in the first direction.
  • the laser beam row is the corresponding pixel row in the image after the laser beam irradiates the human face and then is imaged.
  • the laser beam irradiates the human face, causing the color or brightness of the area to change. Therefore, the laser beam row can be identified based on the color difference. According to the color characteristics of the laser beam lines in the image, or the number of pixels in each line that meet this characteristic, the laser beam lines can be accurately identified. Limiting the number of pixels in each row that meet this characteristic can avoid the problem of inaccurate recognition caused by the intermittent state of the laser beam.
  • the color of the laser beam used is red
  • the R pixel value also called the red value
  • the R pixel value and the B pixel value also called the blue value
  • the original image of the face with the laser beam is collected, and the width and height of the original image of the face are extracted.
  • retrieve the original face image by row if the difference between the R pixel value and the G pixel value on a certain row is greater than the second preset threshold, and the difference between the R pixel value and the B pixel value is greater than the third preset threshold , determine that this line is a laser beam line, and record the line coordinates of the laser beam line.
  • the line coordinates are the coordinates of the line on the Y axis.
  • the conditions for identifying the laser beam row can also add restrictions: the difference between the R pixel value and the G pixel value is greater than the second preset threshold, and the difference between the R pixel value and the B pixel value is both greater than the third preset threshold.
  • the proportion of pixels in the entire row exceeds a certain threshold, such as 50%.
  • the second preset threshold and the third preset threshold can be set as needed.
  • the second preset threshold can be set to 50 to 200
  • the third preset threshold can be set to 50 to 200. The specific needs are adjusted according to the on-site environment. .
  • N laser beam rows that meet the above pixel requirements can be obtained, and the average row coordinate of the N laser beam rows is taken as the position of the laser beam.
  • the number of pixel points corresponding to the laser beam row that meet the above color difference conditions can be obtained, and the average value of the row coordinates of the pixel points is used as the position of the laser beam row.
  • the laser beam position can be calculated by:
  • H is the laser beam position
  • N is the number of identified laser beam rows.
  • the difference between the R pixel value and the G pixel value is greater than the second preset threshold, and the difference between the R pixel value and the B pixel value is satisfied.
  • the number of rows whose difference is greater than the third preset threshold, h i is the row coordinate of the identified laser beam row in the first direction, if the difference between the R pixel value and the G pixel value is greater than the second preset threshold A threshold is set, and the difference between the R pixel value and the B pixel value is greater than the row coordinate of the pixel in the first direction in the third preset threshold.
  • the laser beam position can be calculated by:
  • H is the laser beam position
  • M is the number of pixels in the identified laser beam row that meet the color difference condition.
  • the number of pixels that meet the color difference condition exceeds the preset number of pixels in the laser beam row.
  • the color difference condition is that the difference between the R pixel value and the G pixel value is greater than the second preset threshold, and the difference between the R pixel value and the B pixel value is greater than the third preset threshold, and the color difference is satisfied.
  • the number of pixels in the value condition accounts for more than 50% of the pixel rows in the laser beam row.
  • Pi is the row coordinate of the pixel that meets the color difference condition, such as the difference between the R pixel value and the G pixel value is greater than the second preset threshold, and the difference between the R pixel value and the B pixel value Row coordinates in the first direction of pixels greater than the third preset threshold.
  • the first direction may be perpendicular to the row direction, and the second direction may be perpendicular to the first direction.
  • the first direction may be a vertical direction, and the second direction may be horizontal direction.
  • the laser beam row at the location irradiated by the laser can be identified based on the color.
  • the position of the laser beam reaches the first designated position, including: the position of the laser beam coincides with the coordinates of the key point KeyPt 31 in the first direction, where KeyPt 31 represents the key point of the face outline at the first
  • the first direction can be the Y-axis direction.
  • the adjustment of S4 is a fine adjustment and will not cause the face to exceed the shooting area of the camera.
  • control method of the cone beam CT device in the process of controlling the cone beam CT device to perform CT imaging, further includes:
  • the second pitch angle, the second yaw angle, and the second roll angle are consistent with the calculation formulas of the first pitch angle, the first yaw angle, and the first roll angle, and will not be described again here.
  • S52 Determine whether the absolute values of the second pitch angle, the second yaw angle and the second roll angle are greater than the first preset threshold.
  • the cone beam CT device needs to be controlled to stop taking pictures. .
  • the camera acquires a face image in real time, and uses the above formula to calculate the second pitch angle, the second yaw angle, and the second roll angle based on the current face image. If any of the absolute values of the second pitch angle, the second yaw angle and the second roll angle is greater than the first preset threshold, it indicates that the patient is shaking significantly in one or several directions and dimensions.
  • the tapered column CT device was controlled to stop CT shooting. If the absolute values of the second pitch angle, the second yaw angle and the second roll angle are all less than or equal to the first preset threshold, it indicates that the patient is slightly shaken, and the slight shake has little impact on the shooting effect, and the tapered column CT device is controlled to continue CT. Shoot.
  • the first preset threshold can take a value in the range of 3-7 degrees, for example, the first preset threshold can take a value of 5 degrees.
  • the first preset thresholds set for the first pitch angle, the first yaw angle and the first roll angle may be the same or different. Specifically, they may be set according to the CT imaging effect. Set a smaller threshold in the direction that has a relatively large impact, and conversely set a larger threshold in the direction that has a relatively small impact on the shooting effect. For example, the first preset threshold corresponding to the first roll angle is 4.5 degrees, the first preset threshold corresponding to the first yaw angle is 4.5 degrees, and the first preset threshold corresponding to the first pitch angle is 5.5 degrees.
  • the control method of the cone beam CT device in the process of controlling the cone beam CT device to perform CT imaging, in order to prevent the patient from moving and affecting the quality of the imaging, the control method of the cone beam CT device also includes:
  • S55 Calculate the first difference between the third pitch angles of the two frames of face images separated by the first preset time period, the second difference between the third yaw angles, and the third difference between the third roll angles. Three differences.
  • the monitoring data of the two frames of the camera separated by the first preset time are real-time monitored, and the difference between the pitch angle, yaw angle and roll angle of the two frames separated by the first preset time is calculated. If the absolute value of the difference between the pitch angle, yaw angle or roll angle of the two frames at a preset time is within the fourth preset threshold, the CT shooting is continued.
  • the fourth preset threshold can take a value in the range of 3-7 degrees, such as a value of 5 degrees, when the absolute value of the difference between the pitch angle, yaw angle or roll angle at any interval of the first preset time
  • the cone beam CT device is controlled to stop CT imaging.
  • the main reason for this situation is that the face moves and swings back and forth.
  • the changes in the face are divided into two types: fast changes and slow changes.
  • the degree of deviation should be determined according to the degree of deviation. , terminate the shooting in time and in advance, reduce the rate of bad films and reduce the radiation dose to the patient.
  • the first preset time period can be set according to shooting needs, and can be adjusted specifically with reference to the shooting effect.
  • the first preset time can be set according to the camera shooting frequency. The faster the camera shooting frequency, the shorter the first preset time can be set.
  • the doctor can remotely observe the patient's status in real time by touching the image projected on the display screen outside the lead room.
  • the doctor can remind the patient to maintain the status in real time through the voice module, and can terminate the shooting in time if an abnormality is discovered.
  • the control method of the cone beam CT device captures a face image through a camera, detects key points on the face image, adjusts the face posture according to the key points of the face, and turns on the laser after the face posture meets the requirements. , and extract the laser beam position from the face image, and drive the column to move so that the laser beam position reaches the designated position, then control the cone beam CT device to perform CT shooting.
  • this cone beam CT device it can be automatically adjusted for different patients to achieve the best shooting position and shooting posture, so that the quality of the captured images is optimal, the bad film rate is reduced, and the radiation to the patient is reduced. It can also increase the service life of the machine.
  • the cone beam CT shooting stage is entered, and the background real-time detection state is entered at this time.
  • the control method of the cone beam CT device specifically includes:
  • step 3 Determine whether there is a patient in the image, such as whether there is a complete face of the patient. If no patient is detected, it means that the detector or radiation source of the cone beam CT is rotated to the front of the person and blocks the camera. Repeat steps 1) and 2. ) continue to wait, if the patient image detects a complete face, perform step 3);
  • facial key points use artificial intelligence methods to achieve key point detection, using 98 key point detection; the specific steps can refer to the aforementioned key point detection steps, and will not be described in detail here.
  • control method of the cone beam CT device may also include:
  • the first frame pitch angle PitchB, yaw angle YawB and roll angle RollB
  • the second frame pitch angle PitchC, Yaw angle YawC and roll angle RollC
  • abs(PitchC-PitchB) ⁇ th1 and abs(YawC-YawB) ⁇ th1 and abs(RollC-RollB) ⁇ th1 prompt the patient to maintain the state, if th1 is 0-3 degrees
  • abs(PitchC-PitchB)>th1 or abs(YawC-YawB)>th1 or abs(RollC-RollB)>th1 it will be considered abnormal and the shooting will be terminated.
  • the first preset time period may be a time interval between two adjacent frames, or may be a time interval between multiple frames.
  • the first preset time period is separated by multiple frames
  • the first time period is 600ms, which is an interval of 20 frames
  • the second preset time period is 300ms, which is an interval of 10 frames.
  • th1 is 3 degrees
  • th2 is 2.5 degrees
  • th3 is 2 degrees. .
  • the present disclosure also proposes a cone beam CT device.
  • the conical column CT device 100 includes: a column 10 , a camera 20 and a controller 30 , where the controller 30 includes a memory, a processor and a computer stored on the memory 30 Program, when the computer program is executed by the processor, the above-mentioned control method of the tapered column CT device is implemented.
  • Figure 12 is a front view of a cone beam CT device according to an embodiment of the present disclosure.
  • the cone beam CT device 100 includes a base 1, a fixed column 2, a sliding column 3, a laser 4, a touch screen 5, a detector 6, a cantilever 7, a radiation source 8, and a side camera module 9.
  • the laser 4, the touch screen 5 and the C-shaped arm 10 are all fixed on the sliding column 3, and can move up and down together with the sliding column 3 relative to the fixed column 2.
  • the detector 6 and the ray source 8 can be disposed on the C-shaped arm 10 oppositely.
  • the laser 4 can emit laser light in a horizontal direction and irradiate a horizontal laser line on the patient's face.
  • the first designated position will change as the position of the laser 4 changes.
  • the first designated position is the tip of the patient's nose
  • the first designated position is the tip of the patient's nose
  • the first designated position may be the inner corner of the eye.
  • the touch screen 5 is used to set shooting parameters, such as selecting a predetermined shooting model by inputting adult or child settings.
  • shooting parameters such as selecting a predetermined shooting model by inputting adult or child settings.
  • each component of the cone beam CT device 100 that sets these parameters It is relatively fixed and will not affect the doctor's parameter setting by touching the display screen 5 .
  • some components of the cone beam CT apparatus 100 that set these parameters are movable, but through design, the location where the doctor sets the parameters can be avoided.
  • the touch display screen 5 is also used to display the position of the patient's head in real time when positioning the patient, so as to prompt the patient for accurate positioning in real time.
  • the touch display screen 5 is a movable screen (similar to a tablet) and uses wireless communication. The doctor can hold it in his hand to operate when setting parameters. After the operation is completed, it can be placed opposite the patient to guide the patient. And adjusting the parameters of the device and guiding the patient to perform head positioning are realized through the same touch screen 5, which can save equipment costs.
  • the camera 11 is embedded in the middle of the upper frame of the touch display screen 5 and can be integrated with the touch display screen 5 .
  • the present disclosure also proposes a computer-readable storage medium.
  • a computer program is stored on the computer-readable storage medium.
  • the computer program is executed by the processor, the above-mentioned control method of the cone beam CT device is implemented.
  • the control method of the cone beam CT device, the cone beam CT device and the medium of the embodiments of the present disclosure capture a face image through a camera, detect the key points of the face image, and adjust the face posture according to the key points of the face. After the face posture meets the requirements, the laser is turned on, the laser beam position is extracted from the face image, and the column is driven to move so that the laser beam position reaches the designated position, and then the cone beam CT device is controlled to perform CT shooting.
  • the control method of this cone beam CT device it can be automatically adjusted for different patients to achieve the best shooting position and shooting posture, so that the quality of the captured images is optimal, the bad film rate is reduced, and the radiation to the patient is reduced. It can also increase the service life of the machine.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Medical Informatics (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Optics & Photonics (AREA)
  • Animal Behavior & Ethology (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Biophysics (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Psychiatry (AREA)
  • Social Psychology (AREA)
  • Pulmonology (AREA)
  • Neurology (AREA)
  • Neurosurgery (AREA)
  • Dentistry (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

一种锥形束CT装置（100）的控制方法、锥形束CT装置（100）和介质，锥形束CT装置（100）包括立柱（10）、摄像头（20,11）和激光器（4）,控制方法包括：通过摄像头（11）获取人脸图像，并对人脸图像进行关键点检测，得到人脸关键点（S1）;根据人脸关键点判断人脸姿态是否正常（S2）;如果正常，则开启激光器（4）,并从人脸图像中提取激光束位置（S3）;根据激光束位置和人脸关键点，驱动立柱（10）在第一方向上移动（S4）;待激光束位置达到第一指定位置时，控制锥形束CT装置（100）进行CT拍摄（S5）。该锥形束CT装置（100）的控制方法可以针对不同患者自动调整拍摄位置和拍摄姿态，进而可提高拍摄影像质量,降低坏片率，同时减少对患者的辐射，也可以提高机器的使用寿命。

Description

锥形束CT装置的控制方法、锥形束CT装置和介质
相关申请的交叉引用
本公开要求于2022年07月04日提交的申请号为202210787730.5、名称为“锥形束CT装置的控制方法、锥形束CT装置和介质”的中国专利申请的优先权，其全部内容通过引用结合在本公开中。
技术领域
本公开涉及头部定位技术领域，尤其涉及一种锥形束CT装置的控制方法、锥形束CT装置和介质。
背景技术
相关技术中，CBCT(Cone Beam CT，锥形束CT)头部调整方法是通过医生手动调整电机，观察激光束位置来判断头部位置是否合适。但是人眼观察的方法，存在很大的弊端。不同医生对标准的理解不一样导致定位位置不一定合适，从而影响CBCT机器的拍摄质量，影响影像诊断，严重时会导致坏片。不断调整CBCT机器也会增加射源负荷，降低射源使用寿命。
发明内容
一种锥形束CT装置的控制方法、锥形束CT装置和介质，以针对不同患者自动调整拍摄位置和拍摄姿态，提高拍摄影像质量，降低坏片率，同时减少对患者的辐射，提高机器的使用寿命。
第一方面，本公开提出了一种锥形束CT装置的控制方法，所述锥形束CT装置包括立柱、摄像头和激光束，所述控制方法包括：通过所述摄像头获取人脸图像，并对所述人脸图像进行关键点检测，得到人脸关键点；根据所述人脸关键点判断人脸姿态是否正常；如果正常，则开启所述激光束，并从所述人脸图像中提取激光束位置；根据所述激光束位置和所述人脸关键点，驱动所述立柱在第一方向上移动；待所述激光束位置达到第一指定位置时，控制所述锥形束CT装置进行CT拍摄；
所述控制方法还包括：判断所述人脸图像是否包含完整人脸；如果否，则按照设定方向驱动所述立柱移动；如果是，则执行所述对所述人脸图像进行关键点检测的步骤；
所述根据所述人脸关键点判断人脸姿态是否正常，包括：根据所述人脸关键点计算第一俯仰角、第一偏航角和第一翻滚角；根据所述第一俯仰角、所述第一偏航角和所述第一 翻滚角判断人脸姿态是否正常；
通过下式计算所述第一俯仰角：

其中，KeyPt51表示鼻子关键点中处于鼻子中间且位置最高的关键点，KeyPt16表示人脸轮廓关键点中处于人脸中间且位置最低的关键点，KeyPt51.y表示关键点KeyPt51在所述第一方向上的坐标，KeyPt51.x表示关键点KeyPt51在第二方向上的坐标，Pitch表示所述第一俯仰角，k1为第一预设常数；
和/或，通过下式计算所述第一偏航角：

其中，KeyPt97表示左眼关键点中处于眼睛中间位置的关键点，Yaw表示所述第一偏航角，k2为第二预设常数；
和/或，通过下式计算所述第一翻滚角：
RollTan＝(KeyPt97.x-KeyPt96.x)/(KeyPt97.y-KeyPt96.y)
其中，KeyPt96表示右眼关键点中处于眼睛中间位置的关键点，Roll表示所述第一翻滚角；k3为第三预设常数；
所述根据所述第一俯仰角、所述第一偏航角和所述第一翻滚角判断人脸姿态是否正常，包括：判断所述第一俯仰角、所述第一偏航角和所述第一翻滚角的绝对值是否均小于或等于第一预设阈值；如果是，则判定人脸姿态正常，否则判定人脸姿态不正常；
所述控制方法还包括：按照以下方式的至少一者进行提示：在所述第一俯仰角的绝对值大于所述第一预设阈值时，发出第一提示信息，以提示患者进行俯仰操作；在所述第一偏航角的绝对值大于所述第一预设阈值时，发出第二提示信息，以提示患者进行摇头操作；在所述第一翻滚角的绝对值大于所述第一预设阈值时，发出第三提示信息，以提示患者进 行转头操作；
在判定人脸姿态正常之后，所述控制方法还包括：判断所述关键点KeyPt51和所述关键点KeyPt16连成的直线是否与第二指定位置重合；如果不重合，则发出第四提示信息，以提示患者移动；如果重合，则执行所述开启所述激光束的步骤；
所述从所述人脸图像中提取所述激光束位置，包括：确定所述人脸图像中的所有激光束行，并记录各所述激光束行的行坐标，其中，所述行坐标为所述激光束行在所述第一方向上的坐标；计算所有行坐标的均值，并将所述均值作为所述激光束位置；其中，所述激光束行的所有像素均满足R像素值与G像素值之间的差值大于第二预设阈值，且R像素值与B像素值之间的差值大于第三预设阈值；
所述激光束位置达到第一指定位置，包括：所述激光束位置与关键点KeyPt31在所述第一方向上的坐标重合，其中，KeyPt31表示人脸轮廓关键点中在所述第一方向上位置次高的关键点；
在控制所述锥形束CT装置进行CT拍摄的过程中，所述控制方法还包括：根据当前人脸图像，得到第二俯仰角、第二偏航角和第二翻滚角；判断所述第二俯仰角、所述第二偏航角和所述第二翻滚角的绝对值是否大于第一预设阈值；在所述第二俯仰角、所述第二偏航角和所述第二翻滚角中任一角度大于所述第一预设阈值的情况下，控制所述锥形束CT装置停止CT拍摄；
在控制所述锥形束CT装置进行CT拍摄的过程中，所述控制方法还包括：计算间隔第一预设时间段的两帧人脸图像的第三俯仰角之间的第一差值、第三偏航角之间的第二差值和第三翻滚角之间的第三差值；在所述第一差值的绝对值、所述第二差值的绝对值、所述第三差值的绝对值中的任一者大于第四预设阈值时，控制所述锥形束CT装置停止CT拍摄。
本公开实施例的锥形束CT装置的控制方法，预先对人脸姿态进行大致的判断，即通过摄像头获取人脸图像并对人脸图像进行关键点检测，根据人脸关键点判断人脸姿态是否正常。在人脸姿态正常的情况下，再对设备进行拍摄位置的调整，即开启激光器，装置自动提取激光束位置，根据激光束位置驱动立柱移动，直到激光束移动到指定位置后进行CT拍摄，激光束的位置在人脸上的相对位置可以表示锥形束CT拍摄区域相对患者头部的位置，这样可以针对不同患者自动调整拍摄位置和拍摄姿态，提高拍摄影像质量，降低坏片率，同时减少对患者的辐射，提高机器的使用寿命。
第二方面，本公开提出了一种锥形束CT装置，所述装置包括：立柱、摄像头和控制器，其中，所述控制器包括存储器、处理器和存储在所述存储器上的计算机程序，所述计算机 程序被所述处理器执行时，实现如上述的控制方法。
本公开实施例的锥形束CT装置，通过上述的锥形束CT装置的控制方法针对不同患者自动调整拍摄位置和拍摄姿态，提高拍摄影像质量，降低坏片率，同时减少对患者的辐射，提高机器的使用寿命。
第三方面，本公开提出了一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时，实现如上述的锥形束CT装置的的控制方法。
本公开附加的方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本公开的实践了解到。
附图说明
本公开的上述和/或附加的方面和优点从结合下面附图对实施例的描述中将变得明显和容易理解，其中：
图1是本公开一个实施例的锥形束CT装置的控制方法的流程图；
图2是本公开一个实施例的判断人脸图像的流程图；
图3是本公开一个实施例的根据人脸关键点判断人脸姿态是否正常的流程图；
图4是本公开一个实施例的步骤S22的流程图；
图5是本公开一个实施例的人脸关键点是否满足第二指定位置的流程图；
图6是本公开一个实施例的从人脸图像中提取激光束位置的流程图；
图7是本公开一个实施例的锥形束CT装置进行CT拍摄过程的流程图；
图8是本公开一个实施例的判断任意两帧人脸图像的流程图；
图9是本公开一个实施例的人脸关键点的示意图；
图10是本公开一个实施例的俯仰角、偏航角、翻滚角的示意图；
图11是本公开一个实施例的锥形束CT装置的结构示意图；
图12是本公开一个实施例的锥形束CT装置的正视图。
具体实施方式
下面详细描述本公开的实施例，所述实施例的示例在附图中示出。下面通过参考附图描述的实施例是示例性的，旨在用于解释本公开，而不能理解为对本公开的限制。
下面将结合说明书附图1-12以及具体的实施方式对本公开实施例的锥形束CT装置的控制方法、锥形束CT装置和介质进行详细地说明。
本实施例的锥形束CT装置可以应用于口腔、耳、鼻，以及头上其他部位等各方向的专用CBCT，例如，口腔CBCT。
在本公开的实施例中，锥形束CT装置包括立柱、摄像头和激光器，摄像头和激光器均安装在立柱上。立柱是可以活动的，并且立柱可根据控制指令上下移动；摄像头用于拍摄人脸图像。例如，锥形束CT装置可以包括底座和立柱，立柱包括活动部分、固定部分和定位机构，活动部分底端固定于底座上，活动部分可滑动的设于固定部分上，活动部分相对于固定部分位置改变后，可以通过定位机构进行定位，使得活动部分和固定部分在该位置处相对固定，前述立柱是可以活动的，对应的就是本示例中的活动部分。
图1是本公开一个实施例的锥形束CT装置的控制方法的流程图。
在本公开的实施例中，如图1所示，锥形束CT装置的控制方法包括：
S1，通过摄像头获取人脸图像，并对人脸图像进行关键点检测，得到人脸关键点。
摄像头可以设置在锥形束CT上，也可以单独设置安装摄像头的支架，摄像头的位置是可以上下调整的。使用时，根据当前患者拍摄时头部的高度来调整摄像头的高度，使得患者脸部至少部分位于摄像头的拍摄区域。例如，对于站式的锥形束CT，可以根据患者的身高进行调整；对于坐式锥形束CT，可以根据患者落座后的头部所处的高度进行调整。
具体的，在对锥形束CT装置进行控制时，可通过摄像头实时采集人脸图像，摄像头的采集频率可以是30-60fps(每秒的帧数)。如果人脸图像中没有采集到完整人脸而直接进行关键点检测，则人脸关键点检测将检测不完全，从而影响后期对人脸关键点的判断。因此，采集到人脸图像之后，可先对采集到的人脸图像做初步的判断(如判断采集到的人脸图像是否包含患者完整的人脸)，之后再进行关键点检测。
在本公开的一个实施例中，如图2所示，锥形束CT装置的控制方法还包括：判断人脸图像是否包含完整人脸；如果否，则按照设定方向驱动立柱移动；如果是，则执行对人脸图像进行关键点检测的步骤。
具体地，判断人脸图像是否包含完整人脸可通过提取人脸图像上的特征，判断所提取的特征是否包含一张完整人脸的所有特征。若人脸图像上的特征包含一张完整人脸的所有特征，则摄像头采集到的人脸图像包含完整人脸。若人脸图像上的特征不包含一张完整人脸的所有特征，则摄像头采集到的人脸图像不包含完整人脸。若摄像头采集到的人脸图像不包含完整人脸，则按照设定方向驱动立柱移动，设定的方向可以与人脸图像不完整部分相对于整张人脸图像的方向相同，例如，若采集到的人脸图像上的特征不包含下巴特征，人脸图像缺少下巴部位，下巴部位相对于整张人脸图像的方向为下，则驱动立柱向下移动，使得缺少的下巴部位可以进入摄像头的拍摄区域，若采集到的人脸图像上的特征不包含额头特征，人脸图像缺少额头部位，额头部位相对于整张人脸图像的方向为上，则驱动立柱向上移动，使得缺少的额头部位可以进入摄像头的拍摄区域。在立柱移动的过程中，摄像头仍实时采集人脸图像，直到采集到的人脸图像包含完整人脸，停止移动立柱。对此时采 集到的完整人脸的人脸图像进行人脸关键点检测。若摄像头最开始采集到的人脸图像已经包含完整人脸，则无需驱动立柱，直接进行人脸关键点检测。
进一步具体地，为提高关键点检测效率，可对采集到的包含完整人脸的人脸图像，截取人脸区域，只对人脸区域进行关键点检测，将截取到的人脸区域送入人脸关键点算法中，获得如图9所示的人脸关键点。
可选地，人脸关键点算法可为98点关键点检测，98个关键点的具体分布如图9所示。人脸关键点算法也可为68点关键点检测、106点关键点检测等，在此不做限定。
需要说明的是，识别关键点可以通过神经网络的算法进行识别，识别到的关键点至少包括用于判断人脸姿态的关键点，预先搭建神经网络框架，然后利用标记了这些关键点的训练样本对搭建的神经网络进行训练，即可获得可以用于识别这些关键点的神经网络模型。例如，所要识别的关键点为后文中用于计算俯仰角、偏航角、翻滚角的关键点。
S2，根据人脸关键点判断人脸姿态是否正常。
人脸姿态满足预设的拍摄需求则是正常的，否则为不正常。人脸姿态可以用人脸关键点来表示，如关键点间的相对位置，包括角度、方位、距离等等。
用关键点对人脸姿态进行表示，并且基于该表示来设置人脸姿态标准，从而可以判断人脸姿态是否满足预设的人脸姿态标准，满足则认为人脸姿态正常，不满足则认为人脸姿态不正常。人脸姿态标准是根据是否满足拍摄需求来设置的，例如，拍摄需求可以包括面向摄像头的一侧，且朝向正前方。
具体地，判断人脸姿态是否正常可通过三个维度进行判断，三个维度分别是俯仰角、偏航角和翻滚角。俯仰角、偏航角和翻滚角这三个维度的具体形式如图10所示，俯仰间的大小代表头部抬头或低头的程度，偏航角的大小代表头部左倾或右倾的程度，翻滚角的大小代表头部左转或右转的程度。
设置人脸姿态用基于关键点计算的空间三维方向的旋转角来表示，如俯仰角、偏航角和翻滚角，而人脸姿态标准，则是可以在脸面向摄像头的一侧，且朝向正前方位置的基础上，在各个旋转角旋转方向上进行两个方向的旋转测试，从而可以确定各方向的转动幅度限制即人脸姿态标准，从而也可以省去对摄像头安装时的精确校准，例如，低头一定角度后，锥形束CT可以正常拍摄，这个低头角度是小于锥形束CT刚好不能正常拍摄的低头幅度的，此时，该位置对应的俯仰角即为一个方向的人脸姿态标准，如果低头超过这个幅度则认为人脸姿态不正常，否则为正常，同理，仰头对应的俯仰角、左右旋转方向的偏航角和左右倾斜方向的翻滚角等分别对应的五个方向的人脸姿态标准。当然，在对称方向上，人脸姿态标准对应的阈值可以是相等的，具体方向是根据俯仰角、偏航角和翻滚角的正负来分辨的，具体后续会详细陈述。
当然，具体判断依据距离也是可以实现的，例如，识别图9中点64和点68，计算它们之间高低相差的距离，设置距离标准如5个像素，点64比点68高超过5个像素，说明人脸是倾斜，也就是不正常的，应该向点64一侧倾斜来调整。
需要说明的是，本公开实施例中对于关键点位置次序说明时，表示该点在图9所示的关键点中的位置，对应的是相应位置人脸特征的点，并不代表人脸关键点检测时必须把图9中的关键点要全部检出，当然，全部检出也是可以的。如果仅检测出本实施例参与计算位置、姿态等的关键点，则可以根据神经网络模型进行，标记样本时参照图9中的位置标记相应点即可。
在本公开的一个实施例中，如图3所示，根据人脸关键点判断人脸姿态是否正常，可包括：
S21，根据人脸关键点计算第一俯仰角、第一偏航角和第一翻滚角。
具体地，通过下式计算第一俯仰角：

其中，KeyPt51表示鼻子关键点中处于鼻子中间且位置最高的关键点，KeyPt16表示人脸轮廓关键点中处于人脸中间且位置最低的关键点，KeyPt51.y表示关键点KeyPt51在第一方向上的坐标，KeyPt51.x表示关键点KeyPt51在第二方向上的坐标，Pitch表示第一俯仰角。k1为第一预设常数，arcsin为反正弦函数。
其中，第一方向为图9中竖直方向，第二方向为图9中左右方向，实际应用时，可以参照图9中第一方向和第二方向与人脸的相对位置来做适应性调整。
通过下式计算第一偏航角：

其中，KeyPt97表示左眼关键点中处于眼睛中间位置的关键点，Yaw表示第一偏航角。k2为第二预设常数，arcsin为反正弦函数。
通过下式计算第一翻滚角：
RollTan＝(KeyPt97.x-KeyPt96.x)/(KeyPt97.y-KeyPt96.y)
其中，KeyPt96表示右眼关键点中处于眼睛中间位置的关键点，Roll表示第一翻滚角。k3为第三预设常数，arctan为反正切函数。
需要说明的是，可以k1、k2、k3可以根据实际需求来调整，可以取相同的值也可以取不同的值。例如，可以将各个角度归一化到弧度单位衡量，则该三个常数可以预先设置为再例如，调整各个角度的符号，以改变表示患者头部偏向状态的正负值；当然，该三个值也可取1或-1。
具体地，第一方向为Y轴方向，第二方向为X轴方向。在根据公式计算第一翻滚角之前，首先判断KeyPt97.x是否等于KeyPt96.x，如果等于则执行Roll＝0，如果KeyPt97.x不等于KeyPt96.x，则根据上述的第一翻滚角公式计算第一翻滚角，得到第一翻滚角之后还对其进行取整运算，第一翻滚角取证运算公式如下：
Roll＝(int)Roll
以98点关键点检测为例，参见图9，上述的KeyPt51表示编号为51的关键点，KeyPt16表示编号为16的关键点，KeyPt97表示编号为97的关键点，KeyPt96表示编号为96的关键点。
需要说明的是，上述计算俯仰角、偏航角和翻滚角的公式中，所采用的关键点可以参照图9中的关键点的位置，选择可以替代的关键点，例如利用位置相近的关键点进行代替，从而可以利用多种方式计算相应的角度。
S22，根据第一俯仰角、第一偏航角和第一翻滚角判断人脸姿态是否正常。
在本公开的一个实施例中，如图4所示，根据第一俯仰角、第一偏航角和第一翻滚角判断人脸姿态是否正常，包括：
S221，判断第一俯仰角、第一偏航角和第一翻滚角的绝对值是否均小于或等于第一预设阈值。
S222，如果是，则判定人脸姿态正常，否则判定人脸姿态不正常。
具体地，第一俯仰角、第一偏航角和第一翻滚角用来表示人脸姿态，第一预设阈值为预设的人脸姿态标准。
例如，k1、k2、k3的绝对值为1时，第一预设阈值可在3-7度范围内取值，如取值可为5度。如果k1、k2、k3的绝对值不为1时，也可在3-7度范围内取值后乘以相应的k1、k2或k3的绝对值。例如，第一预设阈值为5度时，当第一俯仰角、第一偏航角和第一翻滚角的绝对值均小于或等于5度时，判定人脸姿态正常，第一俯仰角、第一偏航角和第一翻滚角中任意一者的值大于5度时，判定人脸姿态不正常。在人脸姿态不正常的情况下，可提示患者进行姿态调整。
另外需要说明的是，针对第一俯仰角、第一偏航角和第一翻滚角设置的第一预设阈值可以是相同的，也可以是不同的，具体可以根据CT拍摄成像效果来设定，可以在对拍摄效果影响相对较大的方向上设置较小的阈值，相反在对拍摄效果影响相对较小的方向设置较大的阈值。
一种实施例中，可以根据锥形束CT成像后的后处理需求对第一预设阈值进行设定，根据CT图像中的中间门牙的位置来生成全景图，或者根据CT图像识别中间门牙，如果左右旋转角度过大，可能导致中间门牙处于整体牙齿的偏置位置，最终导致识别不准确或无法识别，也就可能无法生成全景图，这种情况下，就可以将偏航角对应的第一预设阈值设置稍微小一些，其次，将翻滚角对应的第一预设阈值也可以设置稍微小一些，而俯仰角对应的第一预设阈值则可以设置稍微大一些，这样既可以保证满足成像需求，也不至于使得调整过于频繁，增加患者的焦虑情绪等。
例如，第一翻滚角对应的第一预设阈值为4.5度，第一偏航角对应的第一预设阈值为4.5度，而第一俯仰角对应的第一预设阈值为5.5度。
在本公开一个示例中，在第一俯仰角大于对应的第一预设阈值时，发出第一提示信息，以提示患者进行俯仰动作；在第一偏航角大于对应的第一预设阈值时，发出第二提示信息，以提示患者进行摇头动作；在第一翻滚角大于对应的第一预设阈值时，发出第三提示信息，以提示患者进行转头动作。
具体地，患者可以根据显示屏动画、语音提示或医生的指导进行头部姿态的调整，若根据第一俯仰角判断人脸姿态不正常，则根据第一俯仰角的正负，提示患者昂头或低头动作；若根据第一偏航角判断人脸姿态不正常，则根据第一偏航角的正负，提示患者向左或向右摇头动作；若根据第一翻滚角判断人脸姿态不正常，则根据第一翻滚角的正负，提示患者向左或向右倾斜动作。也就说根据各个角度的正负可以确定患者头部在当前处于某一个方向的某种偏向状态，从而可以提示向相反的方向进行调整。
需要说明的是，针对上述计算俯仰角、偏航角、翻滚角可以通过设置k1、k2、k3的值 进行调整，可以改变相应角度的符号，也就是说，一个方向角度的正负，一者表示患者头部在该方向偏向于一侧，则另一者表示患者头部在该方向偏向于另一侧，具体可以通过预设k1、k2、k3的值进行调整。
例如，若第一俯仰角为正值且值大于第一预设阈值时，表示患者处于昂头状态，应提示患者略微低头动作，若第一俯仰角为负值且值大于第一预设阈值时，表示患者处于低头状态，应提示患者略微抬头动作；若第一偏航角为正且大于第一预设阈值时，表示患者处于向右摇头状态，应提示患者略微向左摇头动作，若第一偏航角为负且值大于第一预设阈值时，表示患者处于向左摇头状态，应提示患者略微向右摇头动作；若第一翻滚角为正且值大于第一预设阈值时，表示患者处于向左倾斜状态，应提示患者向右倾斜动作，若第一翻滚角为负值且值大于第一预设阈值时，表示患者处于向右倾斜状态，应提示患者向左倾斜动作。
再例如，第一预设阈值为5度，在第一俯仰角为负值且值大于5度时，发出语音提示“请患者略微抬头”，在第一俯仰角为正值且值大于5度时，发出语音提示“请患者略微低头”；在第一偏航角为正值且值大于5度时，发出语音提示“请患者略微向左摇头”，在第一偏航角为负值且值大于5度时，发出语音提示“请患者略微向右摇头”；在第一翻滚角为正值且值大于5度时，发出语音提示“请患者略微向右倾斜”，在第一翻滚角为负值且值大于5度时，发出语音提示“请患者略微向左倾斜”，患者也可根据显示屏的图像指导提示调整头部姿态。由此，使得患者可明确如何调整姿态，以便快速达到拍摄条件。
需要说明的是，根据第一俯仰角、第一偏航角和第一翻滚角计算结果发出语音提示时，也可包含指导患者抬头或摇头或转头的具体角度。
进一步具体地，患者根据语音提示和显示屏的图像指导提示调整头部的过程中，摄像头实时采集人脸图像，并计算调整后的俯仰角、偏航角和翻滚角，直到调整后的俯仰角、偏航角和翻滚角均小于或等于第一预设阈值，如均小于或等于5度，此时的人脸姿态正常。
S3，如果正常，则开启激光器，并从人脸图像中提取激光束位置。
具体地，开启激光器可以通过开启激光器，使其发射出激光束来实现，在人脸姿态正常的情况下，开启激光器之前，还可以判断人脸是否处于第二指定位置，该第二指定位置的设置是根据摄像头拍摄区域与CT拍摄区域的相对位置关系确定的，脸部左右方向上，当人脸处于CT拍摄区域的所需的第一位置时，人脸处于摄像头采集图像的第二位置，则该第二位置即为第二指定位置。如患者头部在其脸部的左右方向上处于锥形束CT的射线源和探测器的中心区域即第一位置，利于提升CT拍摄效果，此时，人脸在脸部的左右方向上处于摄像头所采集图像的中间区域即第二位置，则第二指定位置可以是摄像头采集的图像中表示的人脸处于该拍摄区域的中部位置，具体可以指该图像在脸部左右方向的中心 线位置，也可以是以该中心线位置左右对称的中心区域位置。
在本公开的一个实施例中，如图5所示，在判定人脸姿态正常之后，锥形束CT装置的控制方法还包括：判断关键点KeyPt51和关键点KeyPt16连成的直线是否与第二指定位置重合；如果不重合，则发出第四提示信息，以提示患者移动；如果重合，则执行开启激光器的步骤。这里的第二指定位置如可以为人脸左右方向上，摄像头所采集图像的中心线位置，或者在该中心线位置左右限定的一定宽度的中心区域。
具体地，关键点KeyPt51和关键点KeyPt16连成的直线，即为鼻子关键点中处于鼻子中间且位置最高的关键点与人脸轮廓关键点中处于人脸中间且位置最低的关键点之间的直线，当关键点KeyPt51和关键点KeyPt16连成的直线与第二指定位置重合时，说明从脸部的左右方向上患者头部位置已调整好。若关键点KeyPt51和关键点KeyPt16连成的直线与第二指定位置不重合，发出第四提示信息，以提示患者移动，第四提示信息与上述提示信息形式相同，可为显示屏上的图像指导信息和语音提示信息，例如，当关键点KeyPt51和关键点KeyPt16连成的直线处于第二位置的右侧，则发出语音提示“请患者向左平移”。同时，显示屏上显示代表第二指定位置的线，患者可根据显示屏移动，使其鼻尖落在该线上。直到患者平移到第二指定位置后，开启激光器。
前述重合是广义的，可以是该连线与该中心线严格重叠，或者相距在预设距离范围内，或该连线处于设定的中心区域中。
根据判断人脸是否处于第二指定位置，可以调整脸部左右方向上患者头部的位置，从而实现更加精准的CT拍摄。
在开启激光器之后，还需提取激光束在人脸上的位置，并判断激光束位置是否达到第一指定位置。其中，第一指定位置是根据拍摄时，摄像头拍摄区域与CT拍摄区域的相对位置关系确定的，其中，激光束位置相对CT拍摄区域位置是固定不动的，竖直方向上，当人脸处于CT拍摄区域的所需的第三位置时，激光束照射到人脸的第四位置，则该第四位置可以为第一指定位置。例如，可以选择人脸上某一特征作为对齐的位置如眼睛的内角点，当样本患者处于CT拍摄区域所需的位置后，激光束照射到眼睛内角点，且人脸是处于摄像头图像采集区域的，这样就可以在具体调整过程中，根据摄像头所采集的人脸图像判断激光束是否到达眼睛内焦点，激光束可以是跟锥形束CT的拍摄模块一起调整的，当激光束到达第一指定位置后，锥形束CT的拍摄区域也能覆盖患者头部，即患者的头部的位置达到锥形束CT的拍摄条件。
需要说明的是，激光束在调整过程中，人脸在竖直方向上一直处于摄像头采集图像的区域中，不会跳出该区域。
需要说明的是，第一指定位置和第二指定位置分别从第一方向和第二方向的两个方向限制了患者头部位置，实现较精准的提示患者进行头部位置调整，以尽快满足拍摄条件。
在本公开的一个实施例中，如图6所示，从人脸图像中提取激光束位置，包括：
S31，确定人脸图像中的所有激光束行，并记录各激光束行的行坐标，其中，行坐标为激光束行在第一方向上的坐标。其中激光束行是由于激光束照射人脸再进行成像后在图像中对应的像素行。
激光束照射到人脸上，使得该区域的颜色或亮度发生变化，因此，根据颜色区别即可识别确定出激光束行。根据图像中激光束行的颜色特性，或每行满足该特性的像素多少，可以准确的识别出激光束行。限制每行满足该特性的像素多少，可以避免因为激光束断续状态造成识别不准确的问题。
S32，计算所有行坐标的均值，并将均值作为激光束位置。
一种实施方式中，采用的激光束的颜色是红色的，被激光束照射的人脸原始图像上R像素值(也称红值)为最大。其中，所述激光束行的所有像素均满足R像素值与G像素值(也称绿值)之间的差值大于第二预设阈值，且R像素值与B像素值(也称蓝值)之间的差值大于第三预设阈值。
具体地，开启激光器之后，采集具有激光束的人脸原始图像，并提取人脸原始图像的宽与高。对人脸原始图像按行进行检索，若某一行上R像素值与G像素值的差值大于第二预设阈值，以及R像素值与B像素值的差值均大于第三预设阈值时，判定这一行为激光束行，并记录该激光束行的行坐标，行坐标为该行在Y轴上的坐标。进一步的，识别激光束行的条件还可以增加限制：R像素值与G像素值的差值大于第二预设阈值，以及R像素值与B像素值的差值均大于第三预设阈值的像素点占全行比例超过一定阈值，如50％。
其中，第二预设阈值和第三预设阈值可根据需要设定，如可设为第二预设阈值为50～200,第三预设阈值为50～200,具体需要根据现场环境进行调整。
进一步具体地，对人脸原始图像上的所有行的像素值进行判断之后，可得到N个满足上述像素要求的激光束行，取N个激光束行的行坐标均值作为激光束的位置。或者可以得到激光束行对应的满足上述颜色差值条件的像素点的个数，利用像素点的行坐标的均值作为激光束行的位置。
可通过下式计算激光束位置：
其中，H为激光束位置，N为识别激光束行的个数，如，满足R像素值与G像素值之间的差值大于第二预设阈值，且R像素值与B像素值之间的差值大于第三预设阈值的行的个数，hi为识别的激光束行在第一方向上的行坐标，如满足R像素值与G像素值之间的差值大于第二预设阈值，且R像素值与B像素值之间的差值大于第三预设阈值的像素在第一方向上的行坐标。
或者可通过下式计算激光束位置：
其中，H为激光束位置，M为识别的激光束行中满足颜色差值条件的像素点个数，其中，满足颜色差值条件的像素点的数量超过预设数量的像素行为激光束行。例如，颜色差值条件为R像素值与G像素值之间的差值大于第二预设阈值，且R像素值与B像素值之间的差值大于第三预设阈值，满足该颜色差值条件的像素点个数占所在行的比例超过50％的像素行为激光束行。Pi为满足颜色差值条件的像素点的行坐标，如为满足R像素值与G像素值之间的差值大于第二预设阈值，且R像素值与B像素值之间的差值大于第三预设阈值的像素点在第一方向上的行坐标。
第一方向可以是与行的方向垂直的，第二方向是与第一方向垂直的，对于站式或者坐式锥形束CT来说，第一方向可以为竖直方向，第二方向可以为水平方向。
同理，若采用的激光束的颜色是其他颜色，可以根据颜色识别出被激光照射处的激光束行。
S4，根据激光束位置和人脸关键点，驱动立柱在第一方向上移动。
得到激光束位置之后，还需比较激光束位置和人脸关键点KeyPt31的行坐标是否重合。通常情况下，激光束位置很难一次性达到要求，此时驱动立柱进行微调，如果H-KeyPt31.y<0，则驱动向上移动abs(H-KeyPt31.y)，否则驱动向下移动abs(H-KeyPt31.y),其中，abs为取绝对值的函数，文中其它处与此相同，不再赘述。
S5，待激光束位置达到第一指定位置时，控制锥形束CT装置进行CT拍摄。
在本公开的一个实施例中，激光束位置达到第一指定位置，包括：激光束位置与关键点KeyPt31在第一方向上的坐标重合，其中，KeyPt31表示人脸轮廓关键点中在第一方向上位置次高的关键点，第一方向可为Y轴方向。
通过合理的设置激光器、摄像头、锥形束CT的拍摄模块的相对位置，可以实现相对 于在人脸检测之前调整摄像头位置的调整，S4的调整属于微调，不会导致人脸超出摄像头的拍摄区域。
在本公开的一个实施例中，如图7所示，在控制锥形束CT装置进行CT拍摄的过程中，锥形束CT装置的控制方法还包括：
S51，根据当前人脸图像，得到第二俯仰角、第二偏航角和第二翻滚角。
具体地，第二俯仰角、第二偏航角和第二翻滚角同第一俯仰角、第一偏航角和第一翻滚角的计算公式一致，在此不再赘述。
S52，判断第二俯仰角、第二偏航角和第二翻滚角的绝对值是否大于第一预设阈值。
S53，在第二俯仰角、第二偏航角和第二翻滚角的绝对值中任一角度大于第一预设阈值的情况下，控制锥形束CT装置停止拍摄。
具体地，锥形柱CT开启激光器进行拍摄的过程中，患者不可能完全保持静止不动，若患者动作幅度较大则会严重影响CT拍摄的质量，此时需控制锥形束CT装置停止拍摄。在控制锥形束CT装置进行CT拍摄的过程中，摄像头实时获取人脸图像，根据当前的人脸图像利用上述公式计算得到第二俯仰角、第二偏航角和第二翻滚角。若第二俯仰角、第二偏航角和第二翻滚角的绝对值中人任一角度大于第一预设阈值，表明患者在某个或某几个方向维度上的大幅度抖动，此时已严重影响CT拍摄质量，控制锥形柱CT装置停止CT拍摄。若第二俯仰角、第二偏航角和第二翻滚角的绝对值均小于等于第一预设阈值，表明患者轻微抖动，轻微抖动对拍摄效果影响较小，控制锥形柱CT装置继续CT拍摄。
需要说明的是，第一预设阈值可在3-7度范围内取值，如取值可为5度。针对第一俯仰角、第一偏航角和第一翻滚角设置的第一预设阈值可以是相同的，也可以是不同的，具体可以根据CT拍摄成像效果来设定，可以在对拍摄效果影响相对较大的方向上设置较小的阈值，相反在对拍摄效果影响相对较小的方向设置较大的阈值。例如，第一翻滚角对应的第一预设阈值为4.5度，第一偏航角对应的第一预设阈值为4.5度，而第一俯仰角对应的第一预设阈值为5.5度。
在本公开的一个实施例中，如图8所示，在控制锥形束CT装置进行CT拍摄的过程中，为防止患者移动而影响拍片质量，锥形束CT装置的控制方法还包括：
S55，计算间隔第一预设时间段的两帧人脸图像的第三俯仰角之间的第一差值、第三偏航角之间的第二差值和第三翻滚角之间的第三差值。
具体地，在拍摄期间实时监控摄像头间隔第一预设时间的两帧的监控数据，并计算间隔第一预设时间的两帧的俯仰角、偏航角和翻滚角的差值，若间隔第一预设时间的两帧俯仰角或偏航角或翻滚角差值的绝对值均在第四预设阈值以内，则继续CT拍摄。
S56，在第一差值的绝对值、第二差值的绝对值、第三差值的绝对值中的任一者大于第 四预设阈值时，控制锥形束CT装置停止CT拍摄。
具体地，第四预设阈值可在3-7度范围内取值，如取值为5度，当任意间隔第一预设时间的俯仰角或偏航角或翻滚角的差值的绝对值在5度以上时，控制锥形束CT装置停止CT拍摄。出现该情况的主要原因为人脸移动和来回摆动，人脸的变化分为快速变化和缓慢变化两种，偏差较大时，对图像质量影响较大，会影响图像的重建效果，应根据偏差程度，适时提前终止拍摄，减少坏片率，降低患者的辐射剂量。
第一预设时间段可以根据拍摄需求来设置，具体可以参照拍摄效果来调整。一种实施方式中，第一预设时间可根据摄像头拍摄频率设定，摄像头拍摄频率越快，第一预设时间可设置越短。
同时，医生在铅房外可通过触摸显示屏实时投射过来的图像，远程实时观察到患者的状态，医生可通过语音模块实时提醒患者保持状态，一旦发现异常可及时终止拍摄。
本公开实施例的锥形束CT装置的控制方法，通过摄像头拍摄人脸图像，并对人脸图像进行关键点检测，根据人脸关键点调整人脸姿态，在人脸姿态满足要求后开启激光器，并从人脸图像中提取激光束位置，并驱动立柱移动使得激光束位置达到指定位置之后，控制锥形束CT装置进行CT拍摄。通过该锥形束CT装置的控制方法，可以针对不同患者自动调整以实现最佳的拍摄位置和拍摄姿态，使得拍摄影像质量达到最佳，降低了坏片率，同时减少了对患者的辐射，也可以提高机器的使用寿命。
一种实施方式中，人脸姿态调整完成后，进入锥形束CT拍摄阶段，此时进入后台实时检测状态，锥形束CT装置的控制方法具体包括：
(1)实时采集患者图像；
(2)判断图像里是否有患者如是否有患者完整的人脸，如果没有检测到患者说明锥形束CT的探测器或射线源旋转到人的正面挡住了摄像头，重复步骤1)和步骤2)继续等待，如果患者图像检测到完整的人脸，则执行步骤3)；
(3)通过所述摄像头获取人脸图像，并对所述人脸图像进行关键点检测，得到人脸关键点，如截取患者人脸区域，送入人脸关键点算法进行人脸关键点检测，人脸关键点采用人工智能方法来实现关键点检测，采用98点关键点检测；具体步骤可以参照前述关键点检测步骤，在此不作赘述。
(4)通过人脸关键点计算俯仰角(Pitch)、偏航角(Yaw)和翻滚角(Roll)，计算方法参照前述部分，在此不做赘述；
(5)通过俯仰角、偏航角及翻滚角判断患者人脸姿态是否正常，具体步骤可以参照前述人脸姿态调是否正常的判断过程，在此不作赘述。
进一步的，锥形束CT装置的控制方法还可以包括：
计算间隔第一预设时间段的两帧人脸图像的俯仰角、偏航角和翻滚角，第一帧：俯仰角PitchB、偏航角YawB和翻滚角RollB；第二帧：俯仰角PitchC，偏航角YawC和翻滚角RollC；如果abs(PitchC-PitchB)<th1且abs(YawC-YawB)<th1且abs(RollC-RollB)<th1，提示患者保持状态,如th1为0-3度，如果abs(PitchC-PitchB)>th1或abs(YawC-YawB)>th1或abs(RollC-RollB)>th1，则认为出现异常，终止拍摄。
其中，第一预设时间段可以为相邻两帧的时间间隔，也可以为间隔多帧的时间间隔。
进一步的，在第一预设时间段是间隔多帧的情况下，如果第一预设时间段内abs(PitchC-PitchB)、abs(YawC-YawB)、abs(RollC-RollB)中的至少一者大于th2，th2<th1，则针对该第一预设时间段内，计算间隔第二预设时间段内的两帧相应角度对应两帧的差值，如果该差值大于th3，则认为出现异常，应终止拍摄。如abs(PitchC-PitchB)大于th2，则计算对应该两帧的俯仰角的差值，如果该差值大于th3则认为出现异常。
也就是说，一定时间段内如果角度变化较快，也认为出现异常，应终止拍摄。
例如，以相机30fps为例，第一时间段为600ms，也就是间隔20帧，第二预设时间段为300ms，也就是间隔10帧，th1为3度，th2为2.5度，th3为2度。
基于上述的锥形束CT装置的控制方法，本公开还提出了一种锥形束CT装置。
在本公开的实施例中，如图11所示，锥形柱CT装置100包括：立柱10、摄像头20和控制器30，其中，控制器30包括存储器、处理器和存储在存储器30上的计算机程序，计算机程序被处理器执行时，实现如上述的锥形柱CT装置的控制方法。
图12是本公开一个实施例的锥形束CT装置的正视图。
如图12所示，锥形束CT装置100包括底座1、固定立柱2、滑动立柱3，激光器4、触摸显示屏5、探测器6、悬臂7、射线源8、侧围拍摄模块9、C型臂10、摄像头11。
激光器4、触摸显示屏5和C型臂10均固定在滑动立柱3上，可以与滑动立柱3一起相对固定立柱2上下移动。拍摄时，患者位于拍摄区域，且面对滑动立柱3。探测器6和射线源8可相对地设置在C型臂10上。激光器4可以发射水平方向的激光，在患者脸上照射出水平的激光线。
需要说明的是，在其他模块位置不变的情况下，第一指定位置会随着激光器4的位置变动而变动，例如，激光器4在一位置时，第一指定位置为患者鼻尖，而在另一靠上的位置时，第一指定位置可能为眼睛内角点位置。
触摸显示屏5用于设置拍摄参数，例如通过输入设置成人或小孩来选择预定的拍摄模型。可选的，在设置这些参数的过程中，设置这些参数的锥形束CT装置100的各个部件 相对固定的，这样不会影响医生通过触摸显示屏5进行参数设置。作为另一种示例，设置这些参数的过程中，设置这些参数的锥形束CT装置100的部分部件是活动的，但通过设计可以避开医生设置参数所处的位置。
触摸显示屏5还用于患者定位时，实时显示患者头部位置，以便实时提示患者进行定位找准。同时触摸显示屏5是可移动屏(类似平板)，采用无线通信方式，医生设置参数时可以拿在手上进行操作，操作完成后可以放置在患者对面作为指导患者使用。并且调整设备的参数与指导患者进行头部定位通过同一触摸显示屏5实现，这样可以节省设备成本。
摄像头11嵌于触摸显示屏5上边框的中部，可与触摸显示屏5集成于一体。
基于上述的锥形束CT装置的控制方法，本公开还提出了一种计算机可读存储介质。
在本公开的实施例中，计算机可读存储介质上存储有计算机程序，计算机程序被处理器执行时，实现如上述的锥形束CT装置的控制方法。
本公开实施例的锥形束CT装置的控制方法、锥形束CT装置和介质，通过摄像头拍摄人脸图像，并对人脸图像进行关键点检测，根据人脸关键点调整人脸姿态，在人脸姿态满足要求后开启激光器，并从人脸图像中提取激光束位置，并驱动立柱移动使得激光束位置达到指定位置之后，控制锥形束CT装置进行CT拍摄。通过该锥形束CT装置的控制方法，可以针对不同患者自动调整以实现最佳的拍摄位置和拍摄姿态，使得拍摄影像质量达到最佳，降低了坏片率，同时减少了对患者的辐射，也可以提高机器的使用寿命。
在本说明书的描述中，参考术语“一个实施例”、“一些实施例”、“示意性实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本公开的至少一个实施例或示例中。在本说明书中，对上述术语的示意性表述不一定指的是相同的实施例或示例。
尽管已经示出和描述了本公开的实施例，本领域的普通技术人员可以理解：在不脱离本公开的原理和宗旨的情况下可以对这些实施例进行多种变化、修改、替换和变型，本公开的范围由权利要求及其等同物限定。

Claims (13)

  1. 一种锥形束CT装置的控制方法，其特征在于，所述锥形束CT装置包括立柱、摄像头和激光器，所述控制方法包括：
    通过所述摄像头获取人脸图像，并对所述人脸图像进行关键点检测，得到人脸关键点；
    根据所述人脸关键点判断人脸姿态是否正常；
    如果正常，则开启所述激光器，并从所述人脸图像中提取激光束位置；
    根据所述激光束位置和所述人脸关键点，驱动所述立柱在第一方向上移动；
    待所述激光束位置达到第一指定位置时，控制所述锥形束CT装置进行CT拍摄。
  2. 根据权利要求1所述的锥形束CT装置的控制方法，其特征在于，所述控制方法还包括：
    判断所述人脸图像是否包含完整人脸；
    如果否，则按照设定方向驱动所述立柱移动；
    如果是，则执行所述对所述人脸图像进行关键点检测的步骤。
  3. 根据权利要求1所述的锥形束CT装置的控制方法，其特征在于，所述根据所述人脸关键点判断人脸姿态是否正常，包括：
    根据所述人脸关键点计算第一俯仰角、第一偏航角和第一翻滚角；
    根据所述第一俯仰角、所述第一偏航角和所述第一翻滚角判断人脸姿态是否正常。
  4. 根据权利要求3所述的锥形束CT装置的控制方法，其特征在于，
    通过下式计算所述第一俯仰角：

    其中，KeyPt51表示鼻子关键点中处于鼻子中间且位置最高的关键点，KeyPt16表示人脸轮廓关键点中处于人脸中间且位置最低的关键点，KeyPt51.y表示关键点KeyPt51在所述第一方向上的坐标，KeyPt51.x表示关键点KeyPt51在第二方向上的坐标，Pitch表示所述第一俯仰角，k1为第一预设常数；
    和/或，通过下式计算所述第一偏航角：

    其中，KeyPt97表示左眼关键点中处于眼睛中间位置的关键点，Yaw表示所述第一偏航角，k2为第二预设常数；
    和/或，通过下式计算所述第一翻滚角：
    RollTan＝(KeyPt97.x-KeyPt96.x)/(KeyPt97.y-KeyPt96.y)
    其中，KeyPt96表示右眼关键点中处于眼睛中间位置的关键点，Roll表示所述第一翻滚角，k3为第三预设常数。
  5. 根据权利要求3所述的锥形束CT装置的控制方法，其特征在于，所述根据所述第一俯仰角、所述第一偏航角和所述第一翻滚角判断人脸姿态是否正常，包括：
    判断所述第一俯仰角、所述第一偏航角和所述第一翻滚角的绝对值是否均小于或等于第一预设阈值；
    如果是，则判定人脸姿态正常，否则判定人脸姿态不正常。
  6. 根据权利要求5所述的锥形束CT装置的控制方法，其特征在于，所述控制方法还包括：
    按照以下方式中的至少一者进行提示：
    在所述第一俯仰角的绝对值大于所述第一预设阈值时，发出第一提示信息，以提示患者进行俯仰操作；
    在所述第一偏航角的绝对值大于所述第一预设阈值时，发出第二提示信息，以提示患者进行摇头操作；
    在所述第一翻滚角的绝对值大于所述第一预设阈值时，发出第三提示信息，以提示患者进行转头操作。
  7. 根据权利要求4-6中任一项所述的锥形束CT装置的控制方法，其特征在于，在判定人脸姿态正常之后，所述控制方法还包括：
    判断所述关键点KeyPt51和所述关键点KeyPt16连成的直线是否与第二指定位置重合；
    如果不重合，则发出第四提示信息，以提示患者移动；
    如果重合，则执行所述开启所述激光束的步骤。
  8. 根据权利要求1所述的锥形束CT装置的控制方法，其特征在于，所述从所述人脸图像中提取所述激光束位置，包括：
    确定所述人脸图像中的所有激光束行，并记录各所述激光束行的行坐标，其中，所述行坐标为所述激光束行在所述第一方向上的坐标；
    计算所有行坐标的均值，并将所述均值作为所述激光束位置；
    其中，所述激光束行的所有像素均满足R像素值与G像素值之间的差值大于第二预设阈值，且R像素值与B像素值之间的差值大于第三预设阈值。
  9. 根据权利要求1所述的锥形束CT装置的控制方法，其特征在于，所述激光束位置达到第一指定位置，包括：
    所述激光束位置与关键点KeyPt31在所述第一方向上的坐标重合，其中，KeyPt31表示人脸轮廓关键点中所述在第一方向上位置次高的关键点。
  10. 根据权利要求1所述的锥形束CT装置的控制方法，其特征在于，在控制所述锥形束CT装置进行CT拍摄的过程中，所述控制方法还包括：
    根据当前人脸图像，得到第二俯仰角、第二偏航角和第二翻滚角；
    判断所述第二俯仰角、所述第二偏航角和所述第二翻滚角的绝对值是否大于第一预设阈值；
    在所述第二俯仰角、所述第二偏航角和所述第二翻滚角的绝对值中任一角度大于所述第一预设阈值的情况下，控制所述锥形束CT装置停止CT拍摄。
  11. 根据权利要求1所述的锥形束CT装置的控制方法，其特征在于，在控制所述锥形束CT装置进行CT拍摄的过程中，所述控制方法还包括：
    计算间隔第一预设时间段的两帧人脸图像的第三俯仰角之间的第一差值、第三偏航角之间的第二差值和第三翻滚角之间的第三差值；
    在所述第一差值的绝对值、所述第二差值的绝对值、所述第三差值的绝对值中的任一者大于第四预设阈值时，控制所述锥形束CT装置停止CT拍摄。
  12. 一种锥形束CT装置，其特征在于，所述装置包括：立柱、摄像头和控制器，其中，
    所述控制器包括存储器、处理器和存储在所述存储器上的计算机程序，所述计算机程序被所述处理器执行时，实现如权利有求1-11中任一项所述的控制方法。
  13. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时，实现如权利要求1-11中任一项所述的控制方法。
PCT/CN2023/114612 2022-07-04 2023-08-24 锥形束ct装置的控制方法、锥形束ct装置和介质 WO2024008210A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210787730.5A CN117414144A (zh) 2022-07-04 2022-07-04 锥形束ct装置的控制方法、锥形束ct装置和介质
CN202210787730.5 2022-07-04

Publications (1)

Publication Number Publication Date
WO2024008210A1 true WO2024008210A1 (zh) 2024-01-11

Family

ID=89454326

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/114612 WO2024008210A1 (zh) 2022-07-04 2023-08-24 锥形束ct装置的控制方法、锥形束ct装置和介质

Country Status (2)

Country Link
CN (1) CN117414144A (zh)
WO (1) WO2024008210A1 (zh)

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1642483A (zh) * 2002-04-04 2005-07-20 株式会社日立医药 X射线ct设备
US20080056439A1 (en) * 2005-10-17 2008-03-06 Imtec Imaging, Llc Multi-use CT scanning apparatus
CN102551772A (zh) * 2012-03-16 2012-07-11 中国科学院深圳先进技术研究院 射线成像装置
CN102626319A (zh) * 2012-04-26 2012-08-08 上海优益基医疗器械有限公司 Cbct的患者射线照射位置自动调整的系统及方法
US20120300895A1 (en) * 2010-02-02 2012-11-29 Juha Koivisto Dental imaging apparatus
CN103610470A (zh) * 2013-12-05 2014-03-05 赛诺威盛科技（北京）有限公司 一种ct设备扫描的实时监控装置和方法
CN107616808A (zh) * 2017-09-28 2018-01-23 广州柿子信息技术有限公司 自动化口腔锥形束ct机
CN110338831A (zh) * 2019-08-14 2019-10-18 晓智科技（成都）有限公司 一种调整x光成像系统中各部件物理对齐的方法
CN111914693A (zh) * 2020-07-16 2020-11-10 上海云从企业发展有限公司 一种人脸姿态调整方法、系统、装置、设备及介质

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1642483A (zh) * 2002-04-04 2005-07-20 株式会社日立医药 X射线ct设备
US20080056439A1 (en) * 2005-10-17 2008-03-06 Imtec Imaging, Llc Multi-use CT scanning apparatus
US20120300895A1 (en) * 2010-02-02 2012-11-29 Juha Koivisto Dental imaging apparatus
US20170265744A1 (en) * 2010-02-02 2017-09-21 Planmeca Oy Dental imaging apparatus
CN102551772A (zh) * 2012-03-16 2012-07-11 中国科学院深圳先进技术研究院 射线成像装置
CN102626319A (zh) * 2012-04-26 2012-08-08 上海优益基医疗器械有限公司 Cbct的患者射线照射位置自动调整的系统及方法
CN103610470A (zh) * 2013-12-05 2014-03-05 赛诺威盛科技（北京）有限公司 一种ct设备扫描的实时监控装置和方法
CN107616808A (zh) * 2017-09-28 2018-01-23 广州柿子信息技术有限公司 自动化口腔锥形束ct机
CN110338831A (zh) * 2019-08-14 2019-10-18 晓智科技（成都）有限公司 一种调整x光成像系统中各部件物理对齐的方法
CN111914693A (zh) * 2020-07-16 2020-11-10 上海云从企业发展有限公司 一种人脸姿态调整方法、系统、装置、设备及介质

Also Published As

Publication number Publication date
CN117414144A (zh) 2024-01-19

Similar Documents

Publication Publication Date Title
CN1293446C (zh) 一种非接触式目控操作系统和方法
KR100522514B1 (ko) 안면화상 촬영장치 및 안면화상 촬영방법
CN100343867C (zh) 一种判别视线方向的方法和装置
CN108294772B (zh) 一种ct扫描视觉定位方法及ct系统
WO2020062773A1 (zh) 一种用于经颅磁刺激治疗的tms定位导航方法
CN109924994B (zh) 一种x光拍摄过程中的检测位置自动校准方法及系统
CN106707674A (zh) 投影设备的自动对焦方法及投影设备
EP3776583B1 (en) Guidance method and system for teledentistry imaging
JP2023547595A (ja) 眼底カメラ及び眼底画像全自動撮影方法
JP2009520546A (ja) コンピュータに基づいた顔面麻痺の評価
CN101163189A (zh) 人脸图像校正方法
WO2021204211A1 (zh) 人脸和虹膜图像采集方法、装置、可读存储介质及设备
WO2022199545A1 (zh) 图像处理方法、装置、计算机设备及存储介质
US20190130598A1 (en) Medical apparatus
JP2006268248A (ja) 撮影装置、顔の向きの判定方法
TW201310993A (zh) 顯示裝置及該顯示裝置的轉動方法
WO2024008210A1 (zh) 锥形束ct装置的控制方法、锥形束ct装置和介质
CN113180716B (zh) 一种可以实现智能摆位的医用数字化x射线系统
CN109700423B (zh) 一种自动距离感知的智能视力检测方法和装置
CN105047239B (zh) 核燃料组件修复跟踪检测方法及装置
JP2009181130A (ja) 顔画像撮影装置および顔画像撮影方法
JP2008161322A (ja) 顔撮影装置
TWI819439B (zh) 視線追蹤方法及應用其之視線追蹤裝置
CN110536044B (zh) 一种证件照自动拍摄方法及装置
CN211460163U (zh) 一种眼球运动检测装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23834975

Country of ref document: EP

Kind code of ref document: A1