Imaging type non-contact gating signal acquisition method for magnetic resonance system

Abstract

The invention relates to an imaging type non-contact gating signal obtaining method for a magnetic resonance system, and belongs to the field of physiological signal detection.The method comprises the steps that pulse and respiration signals of a magnetic resonance detection patient are collected through a camera, and face information of the patient is obtained through a reflector in a magnetic resonance instrument cavity; a living body skin detection technology is utilized to efficiently select a region of interest of the face of a patient, and an IPPG technology is combined to process a video image to obtain a curve of image gray along a time sequence; a motion detection algorithm is used for tracking micro motion of the chest, and stable and accurate breathing signals are extracted by combining frequency selection. By analyzing and quantifying pulse wave signals and delay of respiration signals and electrocardiosignals, the purpose of replacing electrocardiosignal R-peak gating triggering with pulse wave and respiration signal feature triggering is achieved. The invention provides an imaging type non-contact gating technology to solve the problems that traditional ECG and other contact gating technologies are easily affected by a strong static magnetic field, an alternating gradient magnetic field and a radio-frequency pulse electromagnetic field in a magnetic resonance imaging system, and motion artifacts can be generated on motion of fluid such as human blood, breathing and cardiac contraction. Therefore, the quality of a magnetic resonance imaging image is improved.

Description

Technical Field

[0001] The present invention relates to an imaging non-contact gating signal acquisition method for a magnetic resonance system, and in particular to a magnetic resonance imaging gating signal acquisition method and hardware device based on imaging photoplethysmography (IPPG) and optical flow, and belongs to the field of physiological signal detection. Background Art

[0002] Cardiac MRI has become the main means for clinicians to diagnose and treat heart diseases due to its advantages of being non-invasive and suitable for soft tissues. Since cardiac motion can affect MRI, resulting in image blur and image quality, traditional contact gating technologies such as ECG and pulse are susceptible to strong static magnetic fields, alternating gradient magnetic fields, and radio frequency pulse electromagnetic fields in MRI systems. These factors can cause gating signal distortion, synchronization destruction, MRI image artifacts, low scanning efficiency, and complex operation. Therefore, the development of non-contact gating technology has become the key to improving imaging quality and efficiency in order to more accurately reflect the structure and function of the heart. Imaging photoplethysmography (IPPG) uses an imaging sensor to obtain a video of skin color changes caused by the intensity of reflected light after absorption by human blood and tissues, and then extracts the pulse wave signal from a single frame of the video through image processing technology. During magnetic resonance imaging, it is usually necessary to hold your breath to eliminate artifacts caused by respiratory movement. Therefore, IPPG technology is combined with the optical flow method to obtain the human respiratory signal after filtering and denoising. The gating feature areas of the pulse signal and the respiratory signal are located according to the characteristics of the ECG gating R peak, and a composite gating signal is constructed by combining multi-signal fusion. Since IPPG technology obtains the curve of the image grayscale along the time series through video image processing, that is, the blood volume change caused by the contraction of the heart is reflected in the periodic change of the image light intensity, and the ECG signal used in conventional ECG gating reflects the contraction and relaxation of the heart, the use of IPPG technology to replace the electrocardiogram (ECG) for gated triggering can ensure that the same triggering characteristics reflecting the diastole of the heart are obtained during magnetic resonance imaging, thereby ensuring that high-quality imaging results are maintained under various physiological conditions and improving the overall robustness and accuracy of the imaging system. Summary of the invention

[0003] The purpose of the present invention is to solve the problem of motion artifacts and electromagnetic interference generated during the current traditional magnetic resonance gated imaging, which affects the patient's imaging quality. The present invention proposes an imaging non-contact gated signal acquisition method for a magnetic resonance system. When the patient undergoes magnetic resonance detection imaging, the device uses a contactless camera to collect physiological signals, and suppresses artifacts caused by the body's own heart and respiratory movements, further improving the quality of magnetic resonance imaging. The present invention is suitable for the control of gated imaging when magnetic resonance equipment detects human lesions, but is not limited to magnetic resonance imaging technology. Other imaging technologies can also use the gating technology of the present invention to achieve the purpose of artifact suppression when acquiring the patient's physiological information. The present invention avoids problems such as image blurring due to motion artifacts when the patient undergoes magnetic resonance imaging of the lesion site, so as to better assist doctors in analyzing and diagnosing the patient's cause of disease and subsequent treatment plans.

[0004] The purpose of the present invention is achieved through the following technical solutions.

[0005] An imaging non-contact gating signal acquisition method for a magnetic resonance system, characterized in that it comprises the following steps:

[0006] Step 1: The subject lies in the cavity of the magnetic resonance device, and a plastic robotic arm is used to control a convex mirror so that the subject's face is reflected by the convex mirror, and a camera is used to collect a video of the subject's facial skin tissue;

[0007] Step 2: Perform image processing operations such as filtering on the video collected in step 1 to obtain an IPPG signal.

[0008] Step 2-1, using IPPG superpixel segmentation to perform live skin detection on the subject's video, obtain skin pixels of the video and determine the subject's facial region of interest (ROI), where the size of the region of interest is smaller than the image size of the CMOS camera;

[0009] Step 2-2, taking the method of step 2-1 for each frame of the video to obtain a region of interest;

[0010] Step 2-3, calculating the pixel mean of the three channels of R, G, and B in the region of interest of each frame of the image, and obtaining the original IPPG signal by the change of the pixel mean of multiple frames of images;

[0011] Step 3, using amplitude threshold method and window sliding method to determine the pulse gating feature position and area based on the ECG R peak feature for the original IPPG signal;

[0012] Step 4: Use a CMOS camera to record the subject's body posture;

[0013] Step 4-1, calculate the grayscale of adjacent frames of the video and combine it with the optical flow algorithm to obtain the optical flow signal;

[0014] Step 4-2, filtering and denoising the signal extracted from the optical flow field of the ROI area to obtain the breathing signal of the subject;

[0015] Step 5, performing periodic characteristic analysis on the respiratory signal obtained in step 4-1, and locating the respiratory gating characteristic area using frequency domain analysis method and amplitude change method;

[0016] Step 6: Combine the physiological gating signals obtained in steps 3 and 5 to perform feature fusion to construct a composite signal, and trigger the magnetic resonance imaging based on the gating feature area of ​​the composite signal, so as to achieve the purpose of removing respiratory and motion artifacts.

[0017] Beneficial Effects

[0018] 1. The present invention provides an imaging non-contact gated signal acquisition method for a magnetic resonance system, which is suitable for extracting robust pulse waves and respiratory signals of the human body and can effectively improve the signal-to-noise ratio.

[0019] 2. The present invention provides an imaging non-contact gated signal acquisition method for a magnetic resonance system, which is suitable for the fusion of multiple signal features during gated imaging, eliminates the influence of respiratory movement on single pulse signals, and improves the quality and accuracy of gated imaging.

[0020] 3. The present invention provides an imaging non-contact gated signal acquisition method for a magnetic resonance system, which utilizes IPPG technology and optical flow method to extract physiological information, thereby realizing a contactless, low-interference, and highly robust human magnetic resonance imaging process, and providing doctors with more accurate image information for their diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A video acquisition schematic diagram of a method for acquiring imaging non-contact gated signals for a magnetic resonance system provided in an embodiment;

[0022] Figure 2 The present invention provides an overall flow chart of a method for acquiring imaging non-contact gating signals for a magnetic resonance system provided in an embodiment. DETAILED DESCRIPTION

[0023] In order to make the purpose, advantages and features of the present invention clearer, the following is a further detailed description of an imaging non-contact gated signal acquisition method for a magnetic resonance system proposed by the present invention in combination with the accompanying drawings and specific embodiments. It should be noted that the accompanying drawings are all in a very simplified form and use non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention. The structure shown in the accompanying drawings is part of the actual structure; the gating technology included in the present invention is not limited to magnetic resonance imaging, and other imaging technologies that need to overcome cardiac motion and respiratory artifacts are also applicable. The present invention is also not limited to the IPPG signal acquisition site on the human face, and other sites that can be detected by living skin are also applicable. It is only convenient for magnetic resonance imaging to select the face as the IPPG signal extraction site.

[0024] An imaging non-contact gated signal acquisition method for a magnetic resonance system, the facial video acquisition schematic diagram is as follows Figure 1 The overall flow chart is as shown in Figure 2 shown.

[0025] Step 1: Use a plastic robotic arm to remotely control the convex lens into the magnetic resonance cavity and move it to a position where facial information can be recorded;

[0026] Step 1-1, using a visible light source and a near-infrared light source to illuminate facial skin tissue, and using a camera to capture a video of the corresponding illuminated skin tissue area;

[0027] Step 1-2, start the light source and camera:

[0028] The subject lies in the cavity of the magnetic resonance instrument, and the mirror facial skin tissue is evenly illuminated using a visible light source and a near-infrared light source, while the CMOS camera is turned on. A polarizer is placed in front of the visible light source, and an analyzer is placed in front of the CMOS camera to remove the influence of mirror reflected light. The human skin tissue area is not limited to the facial area, and is applicable to any part of the human body from which pulse waves can be extracted, such as arms and fingers. The implementation of the design of the magnetic resonance cavity is described using a human facial video as the part for extracting IPPG signals.

[0029] Steps 1-3, video acquisition:

[0030] The imaging video containing pulse information of the corresponding mirror-reflected facial area is collected by a CMOS camera. The frame rate of the CMOS camera for collecting facial videos is above 30fps, and the acquisition band is 4 bands. The subject remains still during the shooting.

[0031] Step 2: Perform live skin detection on the video collected in step 1 to segment the area of ​​interest containing the pulse wave, and obtain the IPPG signal after image processing:

[0032] Step 2-1, for the collected facial video of the subject, perform real-time live skin detection on the face by using the trained superpixel segmentation SPASD algorithm, obtain the skin pixels of the video and determine the region of interest (ROI) of the subject's face, where the size of the region of interest is smaller than the image size of the CMOS camera;

[0033] Step 2-2, taking the method of step 2-1 for each frame in the video to obtain a region of interest;

[0034] Step 2-3, calculate the pixel mean of the region of interest of each frame of the image, and obtain the original IPPG signal by the change of the pixel mean of multiple frames:

[0035] Each pixel value in the matrix [a1, b1] area of ​​the IPPG imaging device can be calculated by formula (1):

[0036] C(x,y)=I×(ρ s (t)+ρ d (t))+V n (1)

[0037] Where C(x,y) represents the light intensity value corresponding to the pixel with coordinates (x,y); I represents the light intensity of the light source; ρ s (t) and ρ d (t) represent the specular reflection coefficient and diffuse reflection coefficient respectively; V n Represents the quantization noise of the image sensor.

[0038] The quantization noise V of the image sensor is removed by formula (2) n , that is, all pixels are averaged in each frame of the image.

[0039]

[0040] in, Represents the average light intensity of all pixels in a frame of image. The polarizer and analyzer remove the non-physiological parameter light intensity information related to mirror reflection, and according to formula (3), the average light intensity of all pixels in the time series t is obtained. The collection constitutes the IPPG signal.

[0041]

[0042] Step 2-4: Perform peak detection on the acquired pulse wave signal to find the peak value of the pulse wave corresponding to the R peak of the ECG monitoring, thereby determining the area or position of the diastolic gating feature of the pulse signal:

[0043] Step 2-5: Establish a delay time prediction model for the ECG R peak and calculate the prediction relationship using the least squares method.

[0044] d=t Rpeak -t PPG (4)

[0045] Step 3: Use a CMOS camera to record the subject's body posture;

[0046] Step 3-1: According to the human posture detection framework, the chest and abdomen areas are efficiently and real-time located, and the optical flow algorithm is used to solve the optical flow field of the ROI area after image grayscale processing.

[0047] Step 3-2: Select a bandpass Butterworth filter with a cutoff frequency of 0.1 to 0.8 Hz to filter and obtain the subject's breathing signal.

[0048] Step 3-3, performing periodic analysis on the acquired respiratory signal, specifically dividing the respiratory signal according to phase and amplitude to locate the gating feature position of the respiratory signal;

[0049] Step 4: Combine the physiological gating signals obtained in steps 3 and 4 to perform feature fusion to construct a composite signal, obtain the composite signal by signal weighting, and set the threshold of the composite signal to trigger imaging when it reaches a certain threshold:

[0050] S f =ω1·S ippg +ω2·S breath (5)

[0051] Among them, S f is a composite signal, ω1 and ω2 are the weights of the pulse signal and the respiratory signal respectively.

[0052] The above embodiments are only used to illustrate rather than limit the technical solutions of the present invention. Although the present invention is described in detail with reference to the above embodiments, those skilled in the art should understand that the present invention can still be modified or replaced by equivalents. Any modification or partial replacement that does not depart from the spirit and scope of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for acquiring imaging gated signals for a magnetic resonance system, characterized in that: The following steps are involved: Step 1: The subject lies in the cavity of the magnetic resonance device, and a non-magnetic mechanical arm is used to control the convex mirror so that the subject's face is reflected by the convex mirror, and a camera is used to capture a video of the subject's pulse beating part; Step 2: Perform image processing operations such as filtering on the video collected in step 1 to obtain the IPPG signal; Step 3: using the amplitude threshold method and the window sliding method to determine the pulse gating feature position and area based on the ECG R peak feature on the original IPPG signal; Step 4: Use a CMOS camera to record the subject's body posture, calculate the grayscale of adjacent frames of the video, and combine it with the optical flow algorithm to obtain the optical flow signal; Step 5: Perform periodic feature analysis on the respiratory signal obtained in step 4, and use frequency domain analysis and amplitude change method to locate the respiratory gating feature area; Step 6: Combine the physiological gating signals obtained in steps 3 and 5 to perform feature fusion to construct a composite signal, and trigger magnetic resonance imaging based on the gating feature area of ​​the composite signal.

2. The method according to claim 1, wherein step 2-1 includes performing live skin detection on the subject's video using superpixel segmentation of IPPG, acquiring skin pixels of the video and determining a region of interest (ROI) on the subject's face, wherein the size of the region of interest is smaller than the image size of the CMOS camera.

3. The method according to claim 1, wherein step 2-2 comprises obtaining the region of interest by taking the method of step 2-1 for each frame of the video.

4. The method according to claim 1, wherein steps 2-3 include calculating the pixel mean of the three channels R, G, and B of the region of interest of each frame of the image, and the change of the pixel mean of multiple frames of images obtains the original IPPG signal.

5. The method according to claim 1, wherein step 4-1 comprises calculating the grayscale of adjacent frames of the video and combining the optical flow algorithm to obtain an optical flow signal.

6. The method according to claim 1, wherein step 4-2 comprises filtering and denoising the signal extracted from the optical flow field of the ROI region, thereby obtaining the breathing signal of the subject.

7. The method according to claim 1, wherein step 5 comprises performing periodic characteristic analysis on the respiratory signal acquired in step 4-1, and locating the respiratory gating characteristic area using frequency domain analysis method and amplitude change method.

8. The method according to claim 1, wherein step 6 comprises combining the physiological gating signals obtained in steps 3 and 5 to perform feature fusion to construct a composite signal, and triggering magnetic resonance imaging based on the gating feature area of ​​the composite signal, thereby achieving the purpose of removing respiratory and motion artifacts.

Images