A PET attenuation correction method and system for five-modality integrated brain imaging equipment

Abstract

The present invention provides a PET attenuation correction method and system for a five-modality integrated brain imaging device, relating to the field of brain imaging technology. The method comprises: generating a first attenuation coefficient map for a fixed-position hardware component; generating a second attenuation coefficient map for the optode and optical fiber of a non-fixed-position functional near-infrared spectrometer; superimposing the first attenuation coefficient map with the second attenuation coefficient map to generate a third attenuation coefficient map of the PET image under scanning conditions; and performing registration and attenuation correction on the third attenuation coefficient map to obtain an attenuation-corrected PET image. The present invention can more accurately obtain an attenuation template under each actual acquisition environment, and maximizes the correction of the effects of hardware components in the imaging field of view on the attenuation and scattering of gamma photons in PET imaging.

Description

Technical Field

[0001] The present invention relates to the field of brain imaging technology, and in particular to a PET attenuation correction method and system for a five-modality integrated brain imaging device. Background Art

[0002] Existing integrated PET / MR systems have overcome the challenge of attenuation correction for MRI hardware and other imaging accessories within the PET field of view. In early integrated PET / MR systems, the impact of hardware components such as the MRI head coil, spine coil, and bedplate on PET imaging and quantification was significantly reduced by implementing attenuation template-based correction. Specifically, the bedplate and other hardware components were pre-CT scanned to obtain CT images of each component. The CT images of the hardware components were then bilinearly transformed and converted into an attenuation coefficient map (μmap) for 511keV gamma photons, or the attenuation template. The attenuation template was then input into the PET reconstruction pipeline. During each reconstruction, the attenuation template was aligned with the PET image based on the coordinate information obtained at the time of acquisition, accurately achieving PET attenuation and scatter correction.

[0003] Previous studies have evaluated the impact of EEG electrodes, eye tracker mirrors, and fNIRS optodes and fibers on PET imaging. The results indicate that only the fNIRS optodes and fibers significantly attenuate gamma photons, leading to a significant decrease in SUV values. However, in actual scanning, the position and configuration of the fNIRS optodes and fibers are not fixed from scan to scan, depending on factors such as the subject's head shape and the placement and configuration of the fiber. Therefore, existing methods based on fixed templates for CT scans cannot be used to correct for attenuation. Summary of the Invention

[0004] In order to overcome the shortcomings of the existing technology, the purpose of the present invention is to provide a PET attenuation correction method and system for a five-modality integrated brain imaging device, which can more accurately obtain the attenuation template under each real acquisition environment and correct the influence of hardware components in the imaging field of view on the attenuation and scattering of gamma photons in PET imaging to the greatest extent.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A PET attenuation correction method for a five-modality integrated brain imaging device, comprising:

[0007] generating a first attenuation coefficient map for a hardware component at a fixed position;

[0008] generating a second attenuation coefficient map for an optode and an optical fiber of a functional near-infrared spectrometer having a non-fixed position;

[0009] superimposing the first attenuation coefficient map and the second attenuation coefficient map to generate a true third attenuation coefficient map of the PET image under scanning conditions;

[0010] The third attenuation coefficient map is registered and attenuation corrected to obtain an attenuation-corrected PET image.

[0011] Preferably, generating a first attenuation coefficient map for a hardware component at a fixed position includes:

[0012] Performing CT scanning on the hardware components at fixed positions to obtain an initial CT image of each of the hardware components;

[0013] interpolating the initial CT image to a resolution the same as that of the PET image voxels to obtain an interpolated CT image;

[0014] The unit of the grayscale value corresponding to each voxel in the interpolated CT image is converted from HU to the linear attenuation coefficient of the PET image to obtain the first attenuation coefficient map.

[0015] Preferably, the conversion formula for converting the unit of the grayscale value corresponding to each voxel in the interpolated CT image from HU to the linear attenuation coefficient of the PET image is:

[0016]

[0017] in, is the PET linear attenuation coefficient after conversion of the CT value of any voxel; is the linear attenuation coefficient of 511keV gamma photons at the effective CT energy of water; is the linear attenuation coefficient of 511keV gamma photons for bone at the effective CT energy; CT is the grayscale value of any voxel in the interpolated CT image, is the linear attenuation coefficient of water under CT energy; is the linear attenuation coefficient of bone under CT energy.

[0018] Preferably, generating a second attenuation coefficient map for an optode and an optical fiber of a functional near-infrared spectrometer with a non-fixed position comprises:

[0019] Removing the head region from the initial UTE sequence to obtain a first UTE image without the head region;

[0020] removing a head coil region from the first UTE image to obtain a second UTE image where the head coil region is removed;

[0021] The optode and the optical fiber are extracted from the second UTE image, and the second attenuation coefficient map of the optode and the optical fiber of the functional near-infrared spectrometer is prepared.

[0022] Preferably, removing the head region in the initial UTE sequence to obtain a first UTE image with the head region removed includes:

[0023] Automatically determine a first optimal threshold for separating the head and background by using the Otsu method-maximum inter-class variance method, and convert the water-fat separation sequence image into a first binary image according to the first optimal threshold;

[0024] Applying morphological opening and closing operations to remove noise in the first binary image and fill small internal holes to obtain a binary mask of the head area;

[0025] Interchanging the 0 value and the 1 value in the binary mask of the head region to obtain a first conversion mask;

[0026] Perform a dot product between the first conversion mask and the corresponding voxel of the UTE image to obtain the first UTE image.

[0027] Preferably, removing the head coil region from the first UTE image to obtain a second UTE image from which the head coil region is removed includes:

[0028] performing a top-hat filtering on the first UTE image to highlight the high signal area of the head coil while suppressing background noise and residual head tissue signals, thereby obtaining a top-hat filtered image;

[0029] automatically determining a second optimal threshold for separating the head coil and the background by using the Otsu method-maximum inter-class variance method, and binarizing the top-hat filtered image according to the second optimal threshold to segment the head coil region to obtain a second binary image;

[0030] performing an iterative dilation operation and an iterative closing operation on the second binary image to obtain a binary mask of the head coil;

[0031] swapping the 0 value and the 1 value in the binary mask of the head coil to obtain a second conversion mask;

[0032] The second conversion mask is multiplied by the corresponding voxels of the UTE image to obtain a second UTE image with the head coil area removed.

[0033] Preferably, extracting the optode and the optical fiber from the second UTE image and producing the second attenuation coefficient map of the optode and the optical fiber of the functional near-infrared spectrometer comprises:

[0034] automatically determining a third optimal threshold for separating the optical electrode and the optical fiber from the background by using the Otsu method-maximum inter-class variance method, and binarizing the second UTE image according to the third optimal threshold to obtain a third binary image;

[0035] determining masks of the optical electrode and the optical fiber according to the third binary image;

[0036] The voxel value size of the voxel position of the mask of the optical electrode and the optical fiber is changed to the pre-measured linear attenuation coefficient μ value to obtain the second attenuation coefficient map.

[0037] A PET attenuation correction system for a five-modality integrated brain imaging device, comprising:

[0038] a first generating unit, configured to generate a first attenuation coefficient map for a hardware component at a fixed position;

[0039] a second generating unit, configured to generate a second attenuation coefficient map for an optode and an optical fiber of a functional near-infrared spectrometer whose positions are not fixed;

[0040] a third generating unit, configured to superimpose the first attenuation coefficient map and the second attenuation coefficient map to generate a true third attenuation coefficient map of the PET image under scanning conditions;

[0041] The attenuation correction unit is used to perform registration and attenuation correction on the third attenuation coefficient map to obtain an attenuation-corrected PET image.

[0042] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0043] The present invention provides a PET attenuation correction method and system for a five-modality integrated brain imaging device. The method comprises: generating a first attenuation coefficient map for fixed-position hardware components; generating a second attenuation coefficient map for the optode and optical fiber of a non-fixed functional near-infrared spectrometer; superimposing the first and second attenuation coefficient maps to generate a third attenuation coefficient map representing the true PET image under scanning conditions; and performing registration and attenuation correction on the third attenuation coefficient map to obtain an attenuation-corrected PET image. The present invention more accurately obtains an attenuation template for each actual acquisition environment, maximally correcting the effects of hardware components in the imaging field of view on gamma photon attenuation and scattering in PET imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0046] Figure 2A schematic diagram of a five-modality integrated brain imaging device provided by an embodiment of the present invention;

[0047] Figure 3 Schematic diagram of the PET attenuation correction process for the five-modality integrated brain imaging device provided by an embodiment of the present invention.

[0048] Description of reference numerals:

[0049] 1. EEG / NIR integrated head cap; 2. Head coil; 3. Eye tracker mirror; 4. MR; 5. PET. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0051] The purpose of the present invention is to provide a PET attenuation correction method and system for a five-modality integrated brain imaging device, which can more accurately obtain the attenuation template under each real acquisition environment and correct the influence of hardware components in the imaging field of view on the attenuation and scattering of gamma photons in PET imaging to the greatest extent.

[0052] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0053] Figure 1 A flow chart of the method provided in the embodiment of the present invention is shown in FIG. Figure 1 As shown, the present invention provides a PET attenuation correction method for a five-modality integrated brain imaging device, comprising:

[0054] Step 100: Generate a first attenuation coefficient map for a hardware component at a fixed position;

[0055] Step 200: generating a second attenuation coefficient map for the optode and optical fiber of the functional near-infrared spectrometer with a non-fixed position;

[0056] Step 300: superimposing the first attenuation coefficient map and the second attenuation coefficient map to generate a true third attenuation coefficient map of the PET image under scanning conditions;

[0057] Step 400: performing registration and attenuation correction on the third attenuation coefficient map to obtain an attenuation-corrected PET image.

[0058] like Figure 2As shown, the five-modality integrated brain imaging device of this embodiment includes: an EEG / near-infrared integrated head cap 1, a head coil 2, an eye tracker mirror 3, MR4, and PET5. Among them, the EEG / near-infrared integrated head cap 1 is used for the synchronous acquisition of scalp EEG and near-infrared signals; the head coil 2 is used to transmit radio frequency pulses during magnetic resonance signal acquisition and receive magnetic resonance signals generated by the head; the eye tracker mirror 3 is used to reflect the eye image into the field of view of the eye tracker camera; MR4 is a general term for hardware devices used to acquire MR signals and reconstruct MR images; PET5 is a general term for devices that acquire positron annihilation events generated by radioactive nuclides decay and reconstruct PET images based on them.

[0059] The EEG / NIR integrated head cap 1 is put on the subject's head, close to the scalp, and connected to the EEG amplifier and near-infrared acquisition equipment; the head coil 2 is set at the specified position of the MRI bed and can only be used after being recognized by the MRI system; the eye tracker mirror 3 is placed on the upper part of the head coil, and the angle of the mirror is adjusted until the eye image is located in the center of the field of view of the eye tracker camera; the components of PET5 and MR4 are organically combined, and the detector ring of PET5 is installed on the side wall of the aperture of MR4; MR4, PET5 and near-infrared, EEG, and eye trackers are connected together through the main control computer.

[0060] like Figure 3 As shown, the specific technical route of this embodiment is as follows:

[0061] Step 1: Generate an attenuation template for fixed-position components.

[0062] For fixed components such as the MRI bed plate and coil, a CT scan is performed in advance to obtain the CT images of each component. The CT images are then subjected to bilinear transformation to calculate the attenuation template μ-map of 511keV γ photons. For example, the MRI bed plate belongs to Figure 2 Identifies a part of an MRI system. The coil specifically refers to the head coil noted above. A μ-map is a three-dimensional spatial distribution diagram used to describe the linear attenuation coefficient (μ) of a scanned object (such as a human body or fixed equipment) for 511keV gamma photons.

[0063] The CT image is bilinearly transformed to calculate the attenuation template μ-map of 511keV γ photons, which includes:

[0064] Step 11: Interpolate the CT image to the same resolution as the PET image voxels to obtain the interpolated CT image.

[0065] Step 12: Convert the CT value from Hounsfield units (HU) to the PET linear attenuation coefficient in cm -1, the conversion formula is:

[0066]

[0067] in, is the PET linear attenuation coefficient after conversion of the CT value of any voxel; is the linear attenuation coefficient of 511keV gamma photons at the effective CT energy of water; is the linear attenuation coefficient of 511keV gamma photons for bone at the effective CT energy; CT is the grayscale value of any voxel in the interpolated CT image, is the linear attenuation coefficient of water under CT energy; is the linear attenuation coefficient of bone under CT energy. According to the above formula, the attenuation template μmap is obtained from the CT image.

[0068] Step 2: Generate attenuation templates for fNIRS optodes and optical fibers

[0069] For fNIRS optodes and optical fibers with fluidic positions, their positions are identified in ultra-short echo time series (UTE) images (hereinafter referred to as UTE images). Pre-measured μ values are assigned to corresponding locations to generate μ-maps of the fNIRS optodes and optical fibers. Ultra-short echo time series can be used to image NIRS fiber optodes.

[0070] The locations of the fNIRS optodes and optical fibers are identified in the UTE image, specifically:

[0071] Step 21: Remove the head area to obtain a UTE image with the head area removed.

[0072] Acquire MR images from the Water-Fat-Imaging (WFI) sequence (the following technical solution refers to WFI images). First, use the WFI image to obtain a binary mask of the head region, obtain the location information of the head region, and then remove the head region from the UTE image. This includes:

[0073] Step 211: The optimal threshold for separating the head and background is automatically determined by Otsu's method, and the WFI sequence images are converted into binary images. Voxels above the threshold are identified as the head area and assigned a value of 1, and voxels below the threshold are identified as the background area and assigned a value of 0.

[0074] For example, Otsu's method automatically determines the optimal threshold by maximizing the inter-class variance between foreground (head) and background in the image histogram. The specific steps are as follows:

[0075] (1) Calculate the image grayscale histogram and count the number of voxels at each grayscale level.

[0076] (2) Traverse all possible thresholds (0≤T≤N), where N is the maximum voxel value in the image, and calculate the inter-class variance corresponding to each threshold. T is the preset threshold of the above-mentioned image binary method.

[0077] (3) Select the threshold that maximizes the inter-class variance as the optimal threshold.

[0078] Step 212: Apply morphological opening (erosion followed by dilation) and closing (dilation followed by erosion) operations to remove noise and fill small internal holes in the binary image obtained in step 211. Finally, a binary mask of the head region is obtained.

[0079] The erosion operation convolves the image with the structuring element to remove the edges of objects in the image. The shape of the structuring element is chosen to be circular, and its radius is determined according to the actual situation. For the image area covered by the structuring element, the corresponding point in the resulting image is set to 1 only when the structuring element is completely contained within the object. Mathematically, the erosion operation is expressed as: , where A is the input image; B is the structural element; z is the position in the image; and (B)z is the position of the structural element B after z translation.

[0080] The dilation operation is the inverse of erosion. It expands the edges of objects in the image by convolving the structuring element with the image. For the image area covered by the structuring element, as long as there is any overlap between the structuring element and the object, the corresponding points in the resulting image will be set to 1. Mathematically, the dilation operation can be expressed as: ,in, is the complement of the structuring element B; z is the position in the image.

[0081] Step 213: swap the 0 and 1 values in the head region mask to obtain a mask, that is, assigning the voxels in the head region to 0 and the voxels in the background region to 1, which is used for removing the head region in the UTE image in the next step.

[0082] Step 214: perform dot multiplication of the mask obtained in step 213 and the corresponding voxel of the UTE image to obtain the UTE image with the head area removed.

[0083] Step 22: Remove the head coil and obtain a UTE image of the area without the head coil. This includes:

[0084] Step 221: Top-hat filtering is performed on the UTE image after removing the head region to highlight the high-signal areas of the head coil while suppressing background noise and residual head tissue signals. The shape of the structuring element used in this operation is circular. The optimal number of times and size of the structuring element are currently unknown and will depend on the specific results of the actual operation.

[0085] Among them, the top hat filter is the difference between the original image and the result of the opening operation.

[0086] Step 222: Binarization and morphological processing to obtain a binary mask of the head coil. First, use Otsu's method to binarize the UTE image after top-hat filtering in step 221 to segment the head coil region. Voxels above the Otsu threshold are considered head coil voxels and assigned a value of 1, while voxels below the threshold are considered background voxels and assigned a value of 0. Next, a dilation operation is performed on the binary image, iterating multiple times until the coil is continuous and connecting broken coil segments. Subsequently, 1-2 closing operation iterations are performed to fill holes in the coil and smooth the edges. After completing the above morphological operations, only connected domains with an area within an empirically preset range are retained to eliminate noise and small fragments.

[0087] Step 223: swap the 0 and 1 values in the head coil mask obtained in step 222 to obtain a new mask, that is, assigning the head coil voxels to 0 and the background area voxels to 1, which is used for removing the head coil area in the UTE image in the next step.

[0088] Step 224: Finally, the mask obtained in step 223 is multiplied by the corresponding voxel of the UTE image to obtain the UTE image with the head coil area removed.

[0089] Step 23: Extract the optodes and optical fibers from the UTE image and create μ-maps of the fNIRS optodes and optical fibers. This includes:

[0090] Step 231: Because optodes and optical fibers have high voxel values in the image, Otsu's method is used to automatically determine the segmentation threshold. The UTE image, which removes the head and head coil regions from step 224, is binarized. Voxels above the threshold are labeled 1, indicating that the location is an optical fiber or optode. Voxels below the threshold are labeled 0, indicating that the location is background. This results in a mask for the optode and optical fiber.

[0091] Step 232: The voxel value of the voxel position of the optode and the optical fiber is changed to the pre-measured linear attenuation coefficient μ value, and finally the μ-map of the fNIRS optode and the optical fiber is obtained.

[0092] Step 3: Generate a real attenuation template to obtain the μ-map of PET under real scanning conditions.

[0093] The μ-map generated by the fixed-position components in step 1 was superimposed with the μ-map of the fNIRS optode and optical fiber in step 2, whose positions were not fixed. That is, the corresponding voxel values were directly added together to obtain the μ-map of PET under real scanning conditions.

[0094] Step 4: Attenuation Correction

[0095] The μ-map of the PET scan under actual scanning conditions obtained in step 3 is registered and attenuation corrected to obtain an attenuation-corrected PET image. For example, the u-map obtained under actual scanning conditions is a linear attenuation coefficient map of the combined fixed component and fiber optode.

[0096] Specifically, using a PET reconstruction system, uploading the real attenuation template generated in step 3 to the PET reconstruction system can automatically complete the registration and attenuation correction.

[0097] Corresponding to the above method, this embodiment further provides a PET attenuation correction system for a five-modality integrated brain imaging device, comprising:

[0098] a first generating unit, configured to generate a first attenuation coefficient map for a hardware component at a fixed position;

[0099] a second generating unit, configured to generate a second attenuation coefficient map for an optode and an optical fiber of a functional near-infrared spectrometer whose positions are not fixed;

[0100] a third generating unit, configured to superimpose the first attenuation coefficient map and the second attenuation coefficient map to generate a true third attenuation coefficient map of the PET image under scanning conditions;

[0101] The attenuation correction unit is used to perform registration and attenuation correction on the third attenuation coefficient map to obtain an attenuation-corrected PET image.

[0102] The beneficial effects of the present invention are as follows:

[0103] This paper proposes a PET attenuation correction method for five-modality integrated simultaneous imaging, designed to accurately correct the attenuation of hardware components within the PET imaging field of view. By using a UTE sequence to image and automatically segment hardware components (fNIRS optodes and optical fibers) with variable position and morphology, and assigning μ values, this method more accurately generates an attenuation template for each real-world acquisition, maximizing the correction of the effects of hardware components within the imaging field of view on gamma photon attenuation and scattering during PET imaging. This method is also applicable to attenuation correction of other components with variable position within the scanning field of view.

[0104] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0105] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A PET attenuation correction method for a five-modality integrated brain imaging device, characterized in that: The device includes: an EEG / near-infrared integrated head cap, a head coil, an eye tracker reflector, MR, and PET; the method includes: Generating a first attenuation coefficient map for a fixed-position hardware component; the fixed-position hardware component includes an MRI bed and a coil; generating a second attenuation coefficient map for an optode and an optical fiber of a functional near-infrared spectrometer having a non-fixed position; superimposing the first attenuation coefficient map and the second attenuation coefficient map to generate a true third attenuation coefficient map of the PET image under scanning conditions; Performing registration and attenuation correction on the third attenuation coefficient map to obtain an attenuation-corrected PET image, specifically: uploading the true third attenuation coefficient map to a PET reconstruction system to automatically perform registration and attenuation correction; Generate a second attenuation coefficient map for the optode and fiber of a non-fixed functional near-infrared spectrometer, including: Removing the head region from the initial UTE sequence to obtain a first UTE image without the head region; removing a head coil region from the first UTE image to obtain a second UTE image where the head coil region is removed; extracting the optode and the optical fiber from the second UTE image, and producing the second attenuation coefficient map of the optode and the optical fiber of the functional near-infrared spectrometer; Extracting the optode and the optical fiber from the second UTE image and producing the second attenuation coefficient map of the optode and the optical fiber of the functional near-infrared spectrometer, including: automatically determining a third optimal threshold for separating the optical electrode and the optical fiber from the background by using the Otsu method-maximum inter-class variance method, and binarizing the second UTE image according to the third optimal threshold to obtain a third binary image; determining masks of the optical electrode and the optical fiber according to the third binary image; The voxel value size of the voxel position of the mask of the optical electrode and the optical fiber is changed to the pre-measured linear attenuation coefficient μ value to obtain the second attenuation coefficient map.

2. The PET attenuation correction method for a five-modality integrated brain imaging device according to claim 1, characterized in that: Generating a first attenuation coefficient map for a hardware component at a fixed position includes: Performing CT scanning on the hardware components at fixed positions to obtain an initial CT image of each of the hardware components; interpolating the initial CT image to a resolution the same as that of the PET image voxels to obtain an interpolated CT image; The unit of the grayscale value corresponding to each voxel in the interpolated CT image is converted from HU to the linear attenuation coefficient of the PET image to obtain the first attenuation coefficient map.

3. The PET attenuation correction method for a five-modality integrated brain imaging device according to claim 2, characterized in that: The conversion formula for converting the unit of the gray value corresponding to each voxel in the interpolated CT image from HU to the linear attenuation coefficient of the PET image is: in, is the PET linear attenuation coefficient after conversion of the CT value of any voxel; is the linear attenuation coefficient of 511keV gamma photons at the effective CT energy of water; is the linear attenuation coefficient of 511keV gamma photons for bone at the effective CT energy; CT is the grayscale value of any voxel in the interpolated CT image, is the linear attenuation coefficient of water under CT energy; is the linear attenuation coefficient of bone under CT energy.

4. The PET attenuation correction method for a five-modality integrated brain imaging device according to claim 1, characterized in that: Removing the head region from the initial UTE sequence to obtain a first UTE image with the head region removed includes: Automatically determine a first optimal threshold for separating the head and background by using the Otsu method-maximum inter-class variance method, and convert the water-fat separation sequence image into a first binary image according to the first optimal threshold; Applying morphological opening and closing operations to remove noise in the first binary image and fill small internal holes to obtain a binary mask of the head area; Interchanging the 0 value and the 1 value in the binary mask of the head region to obtain a first conversion mask; Perform a dot product between the first conversion mask and the corresponding voxel of the UTE image to obtain the first UTE image.

5. The PET attenuation correction method for a five-modality integrated brain imaging device according to claim 1, characterized in that: Removing the head coil region from the first UTE image to obtain a second UTE image from which the head coil region is removed includes: performing a top-hat filtering on the first UTE image to highlight the high signal area of the head coil while suppressing background noise and residual head tissue signals, thereby obtaining a top-hat filtered image; automatically determining a second optimal threshold for separating the head coil and the background by using the Otsu method-maximum inter-class variance method, and binarizing the top-hat filtered image according to the second optimal threshold to segment the head coil region to obtain a second binary image; performing an iterative dilation operation and an iterative closing operation on the second binary image to obtain a binary mask of the head coil; swapping the 0 value and the 1 value in the binary mask of the head coil to obtain a second conversion mask; The second conversion mask is multiplied by the corresponding voxels of the UTE image to obtain a second UTE image with the head coil area removed.

6. A PET attenuation correction system for a five-modality integrated brain imaging device, characterized in that: The equipment includes: an EEG / near-infrared integrated head cap, a head coil, an eye tracker mirror, MR, and PET; the system includes: A first generating unit is configured to generate a first attenuation coefficient map for a hardware component in a fixed position; the hardware component in a fixed position includes an MRI bed and a coil; a second generating unit, configured to generate a second attenuation coefficient map for an optode and an optical fiber of a functional near-infrared spectrometer whose positions are not fixed; a third generating unit, configured to superimpose the first attenuation coefficient map and the second attenuation coefficient map to generate a true third attenuation coefficient map of the PET image under scanning conditions; an attenuation correction unit, configured to perform registration and attenuation correction on the third attenuation coefficient map to obtain an attenuation-corrected PET image, specifically by uploading the true third attenuation coefficient map to a PET reconstruction system and automatically completing the registration and attenuation correction; Generate a second attenuation coefficient map for the optode and fiber of a non-fixed functional near-infrared spectrometer, including: Removing the head region from the initial UTE sequence to obtain a first UTE image without the head region; removing a head coil region from the first UTE image to obtain a second UTE image where the head coil region is removed; extracting the optode and the optical fiber from the second UTE image, and producing the second attenuation coefficient map of the optode and the optical fiber of the functional near-infrared spectrometer; Extracting the optode and the optical fiber from the second UTE image and producing the second attenuation coefficient map of the optode and the optical fiber of the functional near-infrared spectrometer, including: automatically determining a third optimal threshold for separating the optical electrode and the optical fiber from the background by using the Otsu method-maximum inter-class variance method, and binarizing the second UTE image according to the third optimal threshold to obtain a third binary image; determining masks of the optical electrode and the optical fiber according to the third binary image; The voxel value size of the voxel position of the mask of the optical electrode and the optical fiber is changed to the pre-measured linear attenuation coefficient μ value to obtain the second attenuation coefficient map.

Images