WO2017193461A1 - 一种ct图像扫描床去除方法及装置 - Google Patents

一种ct图像扫描床去除方法及装置 Download PDF

Info

Publication number
WO2017193461A1
WO2017193461A1 PCT/CN2016/087435 CN2016087435W WO2017193461A1 WO 2017193461 A1 WO2017193461 A1 WO 2017193461A1 CN 2016087435 W CN2016087435 W CN 2016087435W WO 2017193461 A1 WO2017193461 A1 WO 2017193461A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
dimensional
information
bed
dimensional scanned
Prior art date
Application number
PCT/CN2016/087435
Other languages
English (en)
French (fr)
Inventor
余绍德
陈璐明
姬治华
江帆
伍世宾
谢耀钦
王磊
Original Assignee
中国科学院深圳先进技术研究院
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 中国科学院深圳先进技术研究院 filed Critical 中国科学院深圳先进技术研究院
Publication of WO2017193461A1 publication Critical patent/WO2017193461A1/zh
Priority to US16/183,758 priority Critical patent/US20190073752A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/77—Retouching; Inpainting; Scratch removal
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/20—Image enhancement or restoration using local operators
    • G06T5/30—Erosion or dilatation, e.g. thinning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10081—Computed x-ray tomography [CT]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20036—Morphological image processing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/03—Recognition of patterns in medical or anatomical images

Definitions

  • the present application relates to the field of image segmentation technology, and in particular, to a CT image scan bed removal method and device.
  • CT photography technology has matured, the scanned images have become clearer, and the amount of data generated has also increased.
  • the spatial resolution of the more commonly used CT images today is 512*512, and there are more than 260,000 pixels in a single scan image. If the gray value is stored in 8 bytes, the amount of data reaches 2 megabytes. Section, and for a CT volume data, the number of slices scanned is generally greater than 100, that is, the amount of data of a commonly used three-dimensional CT scan image exceeds 200 megabytes, the huge amount of data of the image to be processed and the existing medicine
  • the limitations of image segmentation algorithms seriously affect the efficiency of clinical treatment, and the acceleration of image segmentation is the basis of real-time clinical diagnosis.
  • the acceleration methods commonly used in image segmentation mainly include hardware acceleration and software acceleration.
  • Hardware acceleration is to increase the speed of image segmentation by using the large memory, large capacity, and multiple CPUs of high-configuration devices, but its drawback is that (1) hardware needs to be designed according to actual applications, equipment costs are increased, and maintenance costs are high. It is difficult; (2) the influence of the limitations of the existing image segmentation algorithm, the acceleration effect is not obvious.
  • Software acceleration is the simplification or innovation of the algorithm itself by deeply studying the principle of image segmentation algorithm, such as reducing the inner loop or downsampling preprocessed image, etc., but its drawback is that (1) need to study the essence of the algorithm, from the basic principle of the algorithm It is extremely difficult to implement and time-consuming to rewrite the code by the complexity and diversity of the algorithm. (2) The speed is limited, such as image preprocessing, gray scale statistics or multi-layer loop in the image segmentation process. This method is suitable for researchers to carry out image segmentation algorithm innovation, and has a limited limitation.
  • the CT scanning bed is used to complete the omni-directional scanning with the scanning device.
  • the scanning bed has the functions of moving up and down, etc., and the scanning bed is adjusted according to different scanning positions in use to achieve a reasonable position.
  • the CT image taken usually contains the image of the scanning bed, and the image of the scanning bed will interfere with the CT image, which affects the accuracy of clinical diagnosis. Therefore, removing the CT scanning bed is the basis of CT image processing.
  • CT scan bed removal methods are implemented by CT devices with built-in go-to-bed algorithms.
  • the CT device with built-in bed-out algorithm is based on the model characteristics of the scanning bed in the device, and adds the bed-lifting program. The result of the shooting is directly the CT image after going to bed.
  • CT equipment with built-in bed-out algorithm has no universality due to the built-in bed-out algorithm.
  • the device algorithms between different manufacturers are different and cannot be used universally.
  • the bed-out algorithm is not visible, and researchers and doctors cannot modify the algorithm according to actual needs.
  • CT devices with built-in bed-out algorithms usually use hardware acceleration or software acceleration, and the acceleration effect is not obvious. It takes a long time to remove the scanning bed image, and the bed-out effect is not good.
  • the present invention provides a CT image scanning bed removal method and device, which solves the technical problem that the built-in bed-out algorithm in the prior art has no versatility, takes a long time to remove the scanning bed image, and has poor bed-cutting effect.
  • a CT image scanning bed removal method including:
  • Step a reading a three-dimensional CT image through the main thread of the image processing device, counting the number of CT device cores, and initializing the sub-algorithm;
  • Step b The main thread of the image processing device extracts the two-dimensional scanned image from the input three-dimensional CT image, and through the shared memory, the image processing device automatically allocates the two-dimensional scanned image to the kernel, and realizes multi-thread parallel processing to perform the two-dimensional scanned image.
  • Step c The image processing device ends the parallel operation and outputs the three-dimensional CT image after the bed is removed.
  • step b specifically includes:
  • Step b1 extracting a two-dimensional scanned image from the input three-dimensional CT image, reading the two-dimensional scanned image, and dividing the read two-dimensional scanned image;
  • Step b2 extracting target area image information in the two-dimensional scanned image
  • Step b3 performing morphological opening operation on the extracted image information of the target area
  • Step b4 acquiring grayscale information of the target area image in the two-dimensional scanned image
  • Step b5 Combine the grayscale information of the target area image in the two-dimensional scanned image acquired by each thread, and go to the bed to scan the bed information.
  • the technical solution adopted by the embodiment of the present invention further includes: in the step b1, segmenting the read two-dimensional scanned image by using Otsu threshold segmentation.
  • the technical solution adopted by the embodiment of the present invention further includes: in the step b2, extracting the target area image information in the two-dimensional scanned image comprises extracting the body part information in the two-dimensional scanned image.
  • the technical solution adopted by the embodiment of the present invention further includes: in the step b3, acquiring gray information of the target area image in the two-dimensional scanned image: acquiring gray information of the body part in the two-dimensional scanned image, and removing the CT image scanning Bed information.
  • a CT image scanning bed removing device comprising an image reading module, an image processing module and an image output module, wherein the image reading module reads a three-dimensional CT image, and the statistical CT device
  • the number of cores is initialized, and the image processing module extracts a two-dimensional scanned image from the input three-dimensional CT image, and automatically distributes the two-dimensional scanned image to the kernel through the shared memory, thereby implementing multi-thread parallel processing on the two-dimensional scanned image.
  • a bed-out operation is performed, and the image output module is configured to end the parallel operation and output the three-dimensional CT image after the bed is removed.
  • the technical solution adopted by the embodiment of the present invention further includes: the image processing module includes an image segmentation module, an image extraction module, an image operation module, an information acquisition module, and an image combining module, wherein the image segmentation module reads the two-dimensional scanned image, and The read two-dimensional scanned image is subjected to threshold segmentation, and the image extracting module extracts target region image information in the two-dimensional scanned image, and the image computing module pairs extract The target area image information is subjected to a morphological opening operation, and the information acquiring module acquires the target area image gradation information in the two-dimensional scanned image, and the image combining module grays out the target area image in the two-dimensional scanned image acquired by each thread The degree information is combined to go to the bed to scan the bed information.
  • the image processing module includes an image segmentation module, an image extraction module, an image operation module, an information acquisition module, and an image combining module, wherein the image segmentation module reads the two-dimensional scanned image, and The read two-dimensional scanned
  • the technical solution adopted by the embodiment of the present invention further includes: the image segmentation module performs threshold segmentation on the read two-dimensional scanned image by using Otsu threshold segmentation.
  • the technical solution adopted by the embodiment of the present invention further includes: extracting the target area image information in the two-dimensional scanned image by the image extracting module, specifically: extracting the target area image information in the two-dimensional scanned image, including extracting the body in the two-dimensional scanned image Part of the information.
  • the technical solution adopted by the embodiment of the present invention further includes: acquiring, by the information acquiring module, the grayscale information of the target region image in the two-dimensional scanned image, specifically: acquiring gray information of the body part in the two-dimensional scanned image, and removing the CT image scanning bed. information.
  • the image segmentation algorithm used in the CT image scanning bed removal method and apparatus of the embodiment of the present invention is very effective and accurate, and the body bed information is not lost while the scanning bed is removed.
  • the CT image stripping method and apparatus of the embodiment of the present invention are large. The amplitude increases the removal speed of the scanning bed to meet the real-time requirements.
  • FIG. 1 is a flow chart of a CT image scanning bed removing method according to an embodiment of the present invention
  • FIG. 2 is a flow chart showing a method of removing a two-dimensional scanned image by a CT image scanning bed removing method according to an embodiment of the present invention
  • FIG. 3 is a flow chart of a CT image scanning bed removing method according to another embodiment of the present invention.
  • FIG. 4 is a schematic structural view of a CT image scanning bed removing apparatus according to an embodiment of the present invention.
  • FIG. 5 is a view showing experimental results of a CT image scanning bed removing method according to an embodiment of the present invention.
  • FIG. 6 is an accuracy of three-dimensional data segmentation of a CT image scanning bed removal method according to an embodiment of the present invention.
  • FIG. 1 is a flowchart of a CT image scanning bed removing method according to an embodiment of the present invention.
  • the CT image scanning bed removal method of the embodiment of the invention includes:
  • Step 10 reading a three-dimensional CT image through an image processing device (main thread), counting the number of CT device cores, and initializing the sub-algorithm;
  • the image processing device that reads the three-dimensional CT image may be disposed in the CT device, or may be disposed outside of the CT device or independent of the CT device.
  • Step 20 The main thread of the image processing device extracts the two-dimensional scanned image from the input three-dimensional CT image, and the image processing device automatically allocates the two-dimensional scanned image to the kernel through the shared memory, and realizes multi-thread parallel processing to perform the two-dimensional scanned image.
  • Bed operation
  • Step 30 The image processing device (main thread) ends the parallel operation and outputs the three-dimensional CT image after the bed is removed.
  • FIG. 2 is a flow chart showing a method for removing a two-dimensional scanned image by a CT image scanning bed removing method according to an embodiment of the present invention.
  • the CT image scanning bed removal method of the embodiment of the present invention performs the bed-out operation on the two-dimensional scanned image, and specifically includes:
  • Step 210 Extract a two-dimensional scanned image from the input three-dimensional CT image, read a two-dimensional scanned image, and perform an Otsu (OTSU) threshold segmentation on the read two-dimensional scanned image;
  • Otsu Otsu
  • the Otsu (OTSU) threshold segmentation is performed on the read two-dimensional scanned image according to the principle of the bed-out algorithm.
  • Otsu (OTSU) threshold segmentation is based on the grayscale characteristics of the image, dividing the image into backgrounds. And the target two parts. The larger the variance between the background and the target, the greater the difference between the two parts that make up the image. When part of the target is divided into the background or part of the background is divided into the target, the difference between the two parts will be smaller. Therefore, the segmentation that maximizes the variance between classes means that the probability of misclassification is minimal.
  • Step 220 Extract target area image information in the two-dimensional scanned image
  • step 220 extracting target area image information in the two-dimensional scanned image includes extracting body part information in the two-dimensional scanned image.
  • Step 230 Perform a morphological opening operation on the extracted target area image information.
  • the morphology is mainly to obtain the topological and structural information of the object, and obtain some more essential forms of the object through some operations of interaction between the object and the structural element.
  • the application in image processing is mainly to use the basic operations of morphology to observe and process images to achieve the purpose of improving image quality. Corrosion and expansion in image morphology can well denoise binary images.
  • the specific operation of etching is: scanning each pixel in the image with a structural element (generally 3 ⁇ 3 size), and performing an AND operation on each pixel of the structural element with the pixel it covers, if both are 1, Then the pixel is 1, otherwise it is 0.
  • the specific operation of the expansion is: scanning each pixel in the image with a structural element (generally 3 ⁇ 3 size), and performing an AND operation on each pixel of the structural element with the pixel it covers, if both are 0, Then the pixel is 0, otherwise it is 1.
  • the role of corrosion is to eliminate the boundary points of the object, to narrow the target, and to eliminate the noise points smaller than the structural elements; the effect of the expansion is to merge all the background points in contact with the object into the object, so that the target is enlarged, and the cavity in the target can be added.
  • the open operation is a process of first etching and then expanding, which can eliminate fine noise on the image and smooth the boundary of the object.
  • Step 240 Acquire gray information of target area image in the two-dimensional scanned image
  • step 240 acquiring grayscale information of the target region image in the two-dimensional scanned image is: acquiring grayscale information of the body part in the two-dimensional scanned image, and removing CT scan bed information;
  • Step 250 Combine the gray image information of the target area image in the two-dimensional scanned image acquired by each thread, and output a three-dimensional bed-away CT scan image.
  • FIG. 3 is a flowchart of a method for removing a CT image scanning bed according to another embodiment of the present invention.
  • the CT image scan bed removal method of the embodiment of the present invention may be removed for a parallel CT scan bed, or may be removed for a non-parallel CT scan bed. If the CT image scanning bed removal method of the embodiment of the present invention is removed for a non-parallel CT scan bed, the specific method includes the following steps:
  • Step 40 reading a three-dimensional CT image containing scan bed information by using a CT device
  • Step 50 According to the principle of the bed-out algorithm, the three-dimensional CT image of the scanned image is read, the Otsu (OTSU) threshold segmentation is sequentially performed on the three-dimensional CT image, and the foreground image region (including the body part and the scanning bed in the CT image) is extracted, and the morphology is opened.
  • Image segmentation process such as operation;
  • the Otsu (OTSU) threshold segmentation is based on the grayscale characteristics of the image, and the image is divided into two parts: the background and the target.
  • the segmentation that maximizes the variance between classes means that the probability of misclassification is minimal.
  • Morphology is mainly to obtain the topological and structural information of objects, and obtain some more essential forms of objects through some operations of interaction between objects and structural elements.
  • the application in image processing is mainly to use the basic operations of morphology to observe and process images to achieve the purpose of improving image quality.
  • Corrosion and expansion in image morphology can well denoise binary images.
  • the specific operation of etching is: scanning each pixel in the image with a structural element (generally 3 ⁇ 3 size), and performing an AND operation on each pixel of the structural element with the pixel it covers, if both are 1, Then the pixel is 1, otherwise it is 0.
  • the specific operation of the expansion is: scanning each pixel in the image with a structural element (generally 3 ⁇ 3 size), and performing an AND operation on each pixel of the structural element with the pixel it covers, if both are 0, Then the pixel is 0, otherwise it is 1.
  • the role of corrosion is to eliminate the boundary points of the object, to narrow the target, and to eliminate the noise points smaller than the structural elements; the effect of the expansion is to merge all the background points in contact with the object into the object, so that the target is enlarged, and the cavity in the target can be added.
  • the open operation is a process of first etching and then expanding, which can eliminate fine noise on the image and smooth the boundary of the object.
  • Step 60 Acquire a segmentation result map, and output a three-dimensional bed-away CT scan image.
  • FIG. 4 is a schematic structural diagram of a CT image scanning bed removing apparatus according to an embodiment of the present invention.
  • the CT image scanning bed removing apparatus of the embodiment of the present invention includes an image reading module, an image processing module, and an image output module.
  • the image reading module reads the three-dimensional CT image, counts the number of CT device cores, and initializes the sub-algorithm.
  • the image processing module extracts the two-dimensional scanned image from the input three-dimensional CT image, and automatically distributes the two-dimensional scanned image to the kernel through the shared memory, so that the multi-thread parallel processing performs the bed-out operation on the two-dimensional scanned image.
  • the image output module is used to end the parallel operation and output the 3D CT image after going to bed.
  • the image processing module includes an image segmentation module, an image extraction module, an image operation module, an information acquisition module, and an image combining module.
  • the image segmentation module reads the two-dimensional scanned image and performs Otsu (OTSU) threshold segmentation on the read two-dimensional scanned image.
  • the image segmentation module performs Otsu (OTSU) threshold segmentation on the read two-dimensional scanned image according to the principle of the bed-out algorithm.
  • Otsu (OTSU) threshold segmentation divides the image into two parts, the background and the target, according to the grayscale characteristics of the image. The larger the variance between the background and the target, the greater the difference between the two parts that make up the image. When part of the target is divided into the background or part of the background is divided into the target, the difference between the two parts will be smaller. Therefore, the segmentation that maximizes the variance between classes means that the probability of misclassification is minimal.
  • the image extraction module extracts the target area image information in the two-dimensional scanned image, and extracting the target area image information in the two-dimensional scanned image includes extracting the body part information in the two-dimensional scanned image.
  • the image operation module performs a morphological opening operation on the extracted image information of the target area.
  • Morphology is mainly to obtain the topological and structural information of objects, and obtain some more essential forms of objects through some operations of interaction between objects and structural elements.
  • the application in image processing is mainly to use the basic operations of morphology to observe and process images to achieve the purpose of improving image quality. Corrosion and expansion in image morphology can well denoise binary images.
  • the specific operation of etching is: scanning each pixel in the image with a structural element (generally 3 ⁇ 3 size), and performing an AND operation on each pixel of the structural element with the pixel it covers, if both are 1, Then the pixel is 1, otherwise it is 0.
  • the specific operation of the expansion is: scanning each pixel in the image with a structural element (generally 3 ⁇ 3 size), and performing an AND operation on each pixel of the structural element with the pixel it covers, if both are 0, Then the pixel is 0, otherwise it is 1.
  • the role of corrosion is to eliminate the boundary points of the object, to narrow the target, and to eliminate noise points smaller than the structural elements; the effect of the expansion is to merge all the background points that are in contact with the object into the object. Increase the target to fill in the holes in the target.
  • the open operation is a process of first etching and then expanding, which can eliminate fine noise on the image and smooth the boundary of the object.
  • the information acquisition module acquires grayscale information of the target area image in the two-dimensional scanned image.
  • the information acquisition module acquires grayscale information of the body part in the two-dimensional scanned image, and removes the CT image scanning bed information.
  • the image combining module combines the grayscale information of the target area image in the two-dimensional scanned image acquired by each thread to go to the bed scan bed information.
  • CT image scanning bed removal method of the embodiment of the present invention is verified by clinical experiments. It can be understood that the clinical experiment verification is used to further illustrate the beneficial effects of the present invention, and the embodiments and the scope of protection of the invention are not limited.
  • FIG. 5 is a diagram showing experimental results of a CT image scanning bed removing method according to an embodiment of the present invention.
  • the first line of Fig. 5 gives the original CT image with a scanning bed, where (A) is a three-dimensional image, the thin plate on the left side of the body can be clearly seen, and (B) is a tangential image of the axis, and the scanning bed is approximately Two curved curves, (C) is a sagittal image, the scan bed is approximately a vertical line that is nearly parallel to the body; the second line of Figure 5 gives the result of the algorithm after the bed is removed by the algorithm of the present invention, visually, The scanning bed removal procedure proposed by the invention can remove the scanning bed well and has almost no false erosion.
  • the average consumption time formula for each tomogram is as follows:
  • tc i is the time required for the i-th tomographic image segmentation.
  • the image segmentation accuracy parameter is calculated as follows:
  • the image segmentation error rate parameter is calculated as follows:
  • FIG. 6 is a graph showing accuracy of three-dimensional data segmentation of a CT image scanning bed removal method according to an embodiment of the present invention. Overall, the average accuracy of the segmentation is 99%.
  • the segmentation algorithm of the CT image scanning bed removal method in the embodiment of the present invention is very effective and accurate; the average values of the false positive and the false negative are 0.4% and 1.63%, respectively, indicating the present invention.
  • the bed-out algorithm accurately removes the scanning bed and hardly damages the body part.
  • the CT image scanning bed removal method software of the embodiment of the invention is implemented as Visual Studio 2010 and ITK, and is accelerated by using OpenMP.
  • the experimental machine is 8 cores Cores(TM), clocked at 3.7GHz and 16G memory.
  • the above table compares the manual split time, does not introduce the split run time of this acceleration strategy, and the time to introduce this acceleration strategy.
  • the CT image scanning bed removal method of the embodiment of the present invention can perform a bed-free operation on an image with a single resolution of [512, 512] in 0.29 seconds, and the bed is removed.
  • the speed is not more than 2.72 times of the acceleration, and is more than 400 times of the manual division.
  • the CT image scanning bed removal method of the embodiment of the invention greatly improves the removal speed of the scanning bed and meets the real-time requirement.
  • the image segmentation algorithm used in the CT image scanning bed removal method and apparatus of the embodiment of the present invention is very effective and accurate, and the body bed information is not lost while the scanning bed is removed.
  • the CT image stripping method and apparatus of the embodiment of the present invention are large. The amplitude increases the removal speed of the scanning bed to meet the real-time requirements.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Quality & Reliability (AREA)
  • Apparatus For Radiation Diagnosis (AREA)
  • Image Processing (AREA)

Abstract

一种CT图像扫描床去除方法及装置。该CT图像扫描床去除方法，包括：步骤a：通过图像处理设备主线程读取三维CT图像，统计CT设备内核数量，并初始化子算法（10）；步骤b：图像处理设备主线程从输入的三维CT图像中提取二维扫描图像，通过共享内存，图像处理设备自动将二维扫描图像分配给内核，实现多线程并行处理对二维扫描图像进行去床操作（20）；步骤c：图像处理设备结束并行操作，输出去床后的三维CT图像（30）。所述CT图像扫描床去除方法及装置采用的图像分割算法有效和精确，去除扫描床的同时不丢失身体部位的数据信息。

Description

一种CT图像扫描床去除方法及装置 技术领域
本申请涉及图像分割技术领域，特别涉及一种CT图像扫描床去除方法及装置。
背景技术
随着医学影像技术的发展，CT摄影技术逐渐成熟，被扫描的图像更加清晰，产生的数据量也不断增大。现今较常用的CT图像的空间分辨率是512*512，仅一张扫描图像(slice)就有26万多个像素点，若灰度值以8字节进行存储，其数据量达到2兆字节，而对于一个CT体数据，其扫描的slice数一般大于100张，也就是说一个常用的三维CT扫描图像的数据量要超过200兆字节，待处理图像庞大的数据量和现有医学图像分割算法的局限性等严重影响临床治疗的效率，图像分割的加速是临床诊断实时性的基础。
图像分割中常见的加速方法主要有硬件加速和软件加速。硬件加速是利用高配置设备的大内存、大容量、多CPU等特点提升图像分割的速度，但是其缺陷在于，(1)需根据实际的应用，设计硬件，增加设备成本，而且维修费用较高、难度较大；(2)现有的图像分割算法局限性的影响，加速效果不明显。软件加速是通过深入研究图像分割算法原理，进行算法本身的简化或创新，如减少内循环或降采样预处理图像等，但是其缺陷在于，(1)需要研究算法的本质，从算法的基本原理出发重新编写代码，受算法复杂性、多样性的影响，极难实现且耗时；(2)提速有限，如图像分割过程中不可避免的图像预处理、灰度统计或多层循环等。这种方法适用于研究者进行图像分割算法创新，且局限性较大。
CT扫描床是用来配合扫描装置完成全方位的扫描，扫描床具有上下前后等移动功能，在使用中根据不同的扫描位置来调整扫描床以达到合理的位置。但是在实践中，拍摄的CT图像通常含有扫描床的影像，扫描床的影像会对CT图像产生干扰，影响临床诊断的准确性，因此去除CT扫描床是CT图像处理的基础。目前，CT扫描床去除方法为通过具有内置去床算法的CT设备实现。具有内置去床算法的CT设备是厂家根据设备中扫描床的型号特点，添加去床程序，拍摄的结果图像直接为去床后的CT图像。但是采用具有内置去床算法的CT设备由于内置去床算法不具有通用性，不同厂家之间的设备算法不同，不能通用，另外，去床算法不可见，研究者和医生不能根据实际需要修改算法；另外，具有内置去床算法的CT设备通常采用硬件加速或者软件加速，其加速效果不明显，去除扫描床影像耗时长，且去床效果不好。
发明内容
本申请提供了一种CT图像扫描床去除方法及装置，以解决现有技术中内置去床算法不具有通用性、去除扫描床影像耗时长且去床效果不好的技术问题。
为了解决上述问题，本发明提供了如下技术方案：一种CT图像扫描床去除方法，包括：
步骤a：通过图像处理设备主线程读取三维CT图像，统计CT设备内核数量，并初始化子算法；
步骤b：图像处理设备主线程从输入的三维CT图像中提取二维扫描图像，通过共享内存，图像处理设备自动将二维扫描图像分配给内核，实现多线程并行处理对二维扫描图像进行去床操作；
步骤c：图像处理设备结束并行操作，输出去床后的三维CT图像。
本发明实施例采取的技术方案还包括：所述步骤b具体包括：
步骤b1：从输入的三维CT图像中提取二维扫描图像，读取二维扫描图像，对读取的二维扫描图像进行分割；
步骤b2：提取二维扫描图像中的目标区域图像信息；
步骤b3：对提取的目标区域图像信息进行形态学开运算；
步骤b4：获取二维扫描图像中的目标区域图像灰度信息；
步骤b5：将各线程获取的二维扫描图像中的目标区域图像灰度信息进行结合，去床扫描床信息。
本发明实施例采取的技术方案还包括：在所述步骤b1中，对读取的二维扫描图像进行分割采用大津阈值分割。
本发明实施例采取的技术方案还包括：在所述步骤b2中，提取二维扫描图像中的目标区域图像信息包括提取二维扫描图像中的身体部分信息。
本发明实施例采取的技术方案还包括：在所述步骤b3中，获取二维扫描图像中的目标区域图像灰度信息为：获取二维扫描图像中身体部分的灰度信息，去除CT图像扫描床信息。
本发明实施例采取的另一技术方案为：一种CT图像扫描床去除装置，包括图像读取模块、图像处理模块和图像输出模块，所述图像读取模块读取三维CT图像，统计CT设备内核数量，并初始化子算法，所述图像处理模块从输入的三维CT图像中提取二维扫描图像，通过共享内存，自动将二维扫描图像分配给内核，实现多线程并行处理对二维扫描图像进行去床操作，所述图像输出模块用于结束并行操作，输出去床后的三维CT图像。
本发明实施例采取的技术方案还包括：所述图像处理模块包括图像分割模块、图像提取模块、图像运算模块、信息获取模块和图像结合模块，所述图像分割模块读取二维扫描图像，对读取的二维扫描图像进行阈值分割，所述图像提取模块提取二维扫描图像中的目标区域图像信息，所述图像运算模块对提取 的目标区域图像信息进行形态学开运算，所述信息获取模块获取二维扫描图像中的目标区域图像灰度信息，所述图像结合模块将各线程获取的二维扫描图像中的目标区域图像灰度信息进行结合，去床扫描床信息。
本发明实施例采取的技术方案还包括：所述图像分割模块对读取的二维扫描图像进行阈值分割采用大津阈值分割。
本发明实施例采取的技术方案还包括：所述图像提取模块提取二维扫描图像中的目标区域图像信息具体为：提取二维扫描图像中的目标区域图像信息包括提取二维扫描图像中的身体部分信息。
本发明实施例采取的技术方案还包括：所述信息获取模块获取二维扫描图像中的目标区域图像灰度信息具体为：获取二维扫描图像中身体部分的灰度信息，去除CT图像扫描床信息。
本发明实施例的CT图像扫描床去除方法及装置采用的图像分割算法非常有效和精确，去除扫描床的同时不丢失身体部位信息；另外，本发明实施例的CT图描床去除方法及装置大幅度提高扫描床去除速度，满足实时性要求。
附图说明
图1是本发明实施例的CT图像扫描床去除方法的流程图；
图2是本发明实施例的CT图像扫描床去除方法对二维扫描图像进行去床操作的流程图；
图3是本发明另一实施例的CT图像扫描床去除方法的流程图；
图4是本发明实施例的CT图像扫描床去除装置的结构示意图；
图5是本发明实施例的CT图像扫描床去除方法的实验结果图；
图6是本发明实施例的CT图像扫描床去除方法三维数据分割的准确性度 量。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅用以解释本发明，并不用于限定本发明。
请参阅图1，图1是本发明实施例的CT图像扫描床去除方法的流程图。本发明实施例的CT图像扫描床去除方法包括：
步骤10：通过图像处理设备(主线程)读取三维CT图像，统计CT设备内核数量，并初始化子算法；
在步骤10中，读取三维CT图像的图像处理设备可以设置在CT设备中，也可以设置在CT设备外或者独立于CT设备。
步骤20：图像处理设备主线程从输入的三维CT图像中提取二维扫描图像，通过共享内存，图像处理设备自动将二维扫描图像分配给内核，实现多线程并行处理对二维扫描图像进行去床操作；
步骤30：图像处理设备(主线程)结束并行操作，输出去床后的三维CT图像。
请参阅图2，图2是本发明实施例的CT图像扫描床去除方法对二维扫描图像进行去床操作的流程图。本发明实施例的CT图像扫描床去除方法对二维扫描图像进行去床操作具体包括：
步骤210：从输入的三维CT图像中提取二维扫描图像，读取二维扫描图像，对读取的二维扫描图像进行大津(OTSU)阈值分割；
在步骤210中，根据去床算法原理，对读取的二维扫描图像进行大津(OTSU)阈值分割。大津(OTSU)阈值分割是按图像的灰度特性,将图像分成背景 和目标两部分。背景和目标之间的类间方差越大,说明构成图像的两部分的差别越大,当部分目标错分为背景或部分背景错分为目标都会导致两部分差别变小。因此,使类间方差最大的分割意味着错分概率最小。
步骤220：提取二维扫描图像中的目标区域图像信息；
在步骤220中，提取二维扫描图像中的目标区域图像信息包括提取二维扫描图像中的身体部分信息。
步骤230：对提取的目标区域图像信息进行形态学开运算；
在步骤230中，形态学主要是获取物体拓扑和结构信息，通过物体和结构元素相互作用的某些运算，得到物体更本质的形态。在图像处理中的应用主要是：利用形态学的基本运算，对图像进行观察和处理，从而达到改善图像质量的目的。图像形态学中的腐蚀和膨胀能很好的对二值图像进行减噪处理。腐蚀的具体操作是：用一个结构元素(一般是3×3的大小)扫描图像中的每一个像素，用结构元素中的每一个像素与其覆盖的像素做“与”操作，如果都为1，则该像素为1，否则为0。膨胀的具体操作是：用一个结构元素(一般是3×3的大小)扫描图像中的每一个像素，用结构元素中的每一个像素与其覆盖的像素做“与”操作，如果都为0，则该像素为0，否则为1。腐蚀的作用是消除物体边界点，使目标缩小，可以消除小于结构元素的噪声点；膨胀的作用是将与物体接触的所有背景点合并到物体中，使目标增大，可添补目标中的空洞。开运算是先腐蚀后膨胀的过程，可以消除图像上细小的噪声，并平滑物体边界。
步骤240：获取二维扫描图像中的目标区域图像灰度信息；
在步骤240中，获取二维扫描图像中的目标区域图像灰度信息为：获取二维扫描图像中身体部分的灰度信息，去除CT图像扫描床信息；
步骤250：将各线程获取的二维扫描图像中的目标区域图像灰度信息进行结合，输出三维去床CT扫描图像。
请参阅图3，图3是本发明另一实施例的CT图像扫描床去除方法的流程 图。本发明实施例的CT图像扫描床去除方法可以针对并行CT扫描床进行去除，也可以针对非并行CT扫描床进行去除。本发明实施例的CT图像扫描床去除方法如果针对非并行CT扫描床进行去除，则具体方法包括以下步骤：
步骤40：通过CT设备读入含有扫描床信息的三维CT图像；
步骤50：根据去床算法原理，读取扫描图三维CT图像，对三维CT图像依次进行大津(OTSU)阈值分割、提取前景图像区域(包括CT图像中的身体部位和扫描床)、形态学开运算等图像分割过程；
其中，大津(OTSU)阈值分割是按图像的灰度特性,将图像分成背景和目标两部分。背景和目标之间的类间方差越大,说明构成图像的两部分的差别越大,当部分目标错分为背景或部分背景错分为目标都会导致两部分差别变小。因此,使类间方差最大的分割意味着错分概率最小。形态学主要是获取物体拓扑和结构信息，通过物体和结构元素相互作用的某些运算，得到物体更本质的形态。在图像处理中的应用主要是：利用形态学的基本运算，对图像进行观察和处理，从而达到改善图像质量的目的。图像形态学中的腐蚀和膨胀能很好的对二值图像进行减噪处理。腐蚀的具体操作是：用一个结构元素(一般是3×3的大小)扫描图像中的每一个像素，用结构元素中的每一个像素与其覆盖的像素做“与”操作，如果都为1，则该像素为1，否则为0。膨胀的具体操作是：用一个结构元素(一般是3×3的大小)扫描图像中的每一个像素，用结构元素中的每一个像素与其覆盖的像素做“与”操作，如果都为0，则该像素为0，否则为1。腐蚀的作用是消除物体边界点，使目标缩小，可以消除小于结构元素的噪声点；膨胀的作用是将与物体接触的所有背景点合并到物体中，使目标增大，可添补目标中的空洞。开运算是先腐蚀后膨胀的过程，可以消除图像上细小的噪声，并平滑物体边界。
步骤60：获取分割结果图，输出三维去床CT扫描图像。
请参阅图4，图4是本发明实施例的CT图像扫描床去除装置的结构示意 图。本发明实施例的CT图像扫描床去除装置包括图像读取模块、图像处理模块和图像输出模块。图像读取模块读取三维CT图像，统计CT设备内核数量，并初始化子算法。图像处理模块从输入的三维CT图像中提取二维扫描图像，通过共享内存，自动将二维扫描图像分配给内核，实现多线程并行处理对二维扫描图像进行去床操作。图像输出模块用于结束并行操作，输出去床后的三维CT图像。图像处理模块包括图像分割模块、图像提取模块、图像运算模块、信息获取模块和图像结合模块。图像分割模块读取二维扫描图像，对读取的二维扫描图像进行大津(OTSU)阈值分割。图像分割模块根据去床算法原理，对读取的二维扫描图像进行大津(OTSU)阈值分割。大津(OTSU)阈值分割是按图像的灰度特性,将图像分成背景和目标两部分。背景和目标之间的类间方差越大,说明构成图像的两部分的差别越大，当部分目标错分为背景或部分背景错分为目标都会导致两部分差别变小。因此,使类间方差最大的分割意味着错分概率最小。
图像提取模块提取二维扫描图像中的目标区域图像信息，提取二维扫描图像中的目标区域图像信息包括提取二维扫描图像中的身体部分信息。
图像运算模块对提取的目标区域图像信息进行形态学开运算。形态学主要是获取物体拓扑和结构信息，通过物体和结构元素相互作用的某些运算，得到物体更本质的形态。在图像处理中的应用主要是：利用形态学的基本运算，对图像进行观察和处理，从而达到改善图像质量的目的。图像形态学中的腐蚀和膨胀能很好的对二值图像进行减噪处理。腐蚀的具体操作是：用一个结构元素(一般是3×3的大小)扫描图像中的每一个像素，用结构元素中的每一个像素与其覆盖的像素做“与”操作，如果都为1，则该像素为1，否则为0。膨胀的具体操作是：用一个结构元素(一般是3×3的大小)扫描图像中的每一个像素，用结构元素中的每一个像素与其覆盖的像素做“与”操作，如果都为0，则该像素为0，否则为1。腐蚀的作用是消除物体边界点，使目标缩小，可以消除小于结构元素的噪声点；膨胀的作用是将与物体接触的所有背景点合并到物体中， 使目标增大，可添补目标中的空洞。开运算是先腐蚀后膨胀的过程，可以消除图像上细小的噪声，并平滑物体边界。
信息获取模块获取二维扫描图像中的目标区域图像灰度信息。信息获取模块获取二维扫描图像中身体部分的灰度信息，去除CT图像扫描床信息。
图像结合模块将各线程获取的二维扫描图像中的目标区域图像灰度信息进行结合，去床扫描床信息。
对本发明实施例的CT图像扫描床去除方法进行临床实验验证，可以理解，临床实验验证用于进一步说明本发明的有益效果，对发明的实施方式和保护范围并不构成限制。
请参阅图5，图5是本发明实施例的CT图像扫描床去除方法的实验结果图。图5第一行给出原始的带有扫描床的CT图像，其中(A)为三维图像，可以清晰的看到身体左侧的薄板，(B)为轴面切向图像，扫描床近似为两条弯曲的曲线，(C)为矢状面图像，扫描床近似为一条垂线与身体近乎平行；图5第二行给出经本发明算法去床后的结果图，从视觉上，本发明提出的扫描床去除程序能够很好地去除扫描床，且几乎没有误侵蚀现象。
每个断层图像的平均消耗时间公式如下：
Figure PCTCN2016087435-appb-000001
其中tci为第i个断层图像分割所需要的时间。
图像分割准确率参数计算公式如下：
Figure PCTCN2016087435-appb-000002
图像分割误差率参数计算公式如下：
假阳性(false positive,FP)：表示本发明提出的算法未能成功去除扫描床的误差率。
Figure PCTCN2016087435-appb-000003
假阴性(false negative,FN)：表示本发明提出的算法对身体部位(mask)的误侵蚀率。
Figure PCTCN2016087435-appb-000004
其中|·|用来统计三维数据内的点个数，G为手动分割的金标准，而S为分割结果。
请参阅图6，图6是本发明实施例的CT图像扫描床去除方法三维数据分割的准确性度量。整体上来看，分割的平均精度达到99％，本发明实施例的CT图像扫描床去除方法分割算法非常有效和精确；假阳性和假阴性的平均值分别为0.4％和1.63％，说明本发明的去床算法能精确的去除扫描床，而且几乎不损伤身体部位。
本发明实施例的CT图像扫描床去除方法软件实现为Visual Studio 2010、ITK，采用OpenMP进行加速实现。实验机器为8核
Figure PCTCN2016087435-appb-000005
 Cores(TM)，主频3.7GHz，内存16G。
Figure PCTCN2016087435-appb-000006
上述表格比较了手动分割时间，没有引入本加速策略的分割运行时间，以及引入本加速策略的时间。通过分析发现，本发明实施例的CT图像扫描床去除方法能在0.29秒内对单张分辨率为[512,512]的图像进行去床操作，其去床 的速度是未加速的2.72倍，是手动分割的400倍以上，本发明实施例的CT图像扫描床去除方法大幅度提高扫描床去除速度，满足实时性要求。
本发明实施例的CT图像扫描床去除方法及装置采用的图像分割算法非常有效和精确，去除扫描床的同时不丢失身体部位信息；另外，本发明实施例的CT图描床去除方法及装置大幅度提高扫描床去除速度，满足实时性要求。
虽然本发明参照当前的较佳实施方式进行了描述，但本领域的技术人员应能理解，上述较佳实施方式仅用来说明本发明，并非用来限定本发明的保护范围，任何在本发明的精神和原则范围之内，所做的任何修饰、等效替换、改进等，均应包含在本发明的权利保护范围之内。

Claims (10)

  1. 一种CT图像扫描床去除方法，包括：
    步骤a：通过图像处理设备主线程读取三维CT图像，统计CT设备内核数量，并初始化子算法；
    步骤b：图像处理设备主线程从输入的三维CT图像中提取二维扫描图像，通过共享内存，图像处理设备自动将二维扫描图像分配给内核，实现多线程并行处理对二维扫描图像进行去床操作；
    步骤c：图像处理设备结束并行操作，输出去床后的三维CT图像。
  2. 根据权利要求1所述的CT图像扫描床去除方法，其特征在于，所述步骤b具体包括：
    步骤b1：从输入的三维CT图像中提取二维扫描图像，读取二维扫描图像，对读取的二维扫描图像进行分割；
    步骤b2：提取二维扫描图像中的目标区域图像信息；
    步骤b3：对提取的目标区域图像信息进行形态学开运算；
    步骤b4：获取二维扫描图像中的目标区域图像灰度信息；
    步骤b5：将各线程获取的二维扫描图像中的目标区域图像灰度信息进行结合，去床扫描床信息。
  3. 根据权利要求2所述的CT图像扫描床去除方法，其特征在于，在所述步骤b1中，对读取的二维扫描图像进行分割采用大津阈值分割。
  4. 根据权利要求2所述的CT图像扫描床去除方法，其特征在于，在所述步骤b2中，提取二维扫描图像中的目标区域图像信息包括提取二维扫描图像中的身体部分信息。
  5. 根据权利要求1所述的CT图像扫描床去除方法，其特征在于，在所述步骤b3中，获取二维扫描图像中的目标区域图像灰度信息为：获取二维扫描图像中身体部分的灰度信息，去除CT图像扫描床信息。
  6. 一种CT图像扫描床去除装置，其特征在于：包括图像读取模块、图像处理模块和图像输出模块，所述图像读取模块读取三维CT图像，统计CT设备内核数量，并初始化子算法，所述图像处理模块从输入的三维CT图像中提取二维扫描图像，通过共享内存，自动将二维扫描图像分配给内核，实现多线程并 行处理对二维扫描图像进行去床操作，所述图像输出模块用于结束并行操作，输出去床后的三维CT图像。
  7. 根据权利要求6所述的CT图像扫描床去除装置，其特征在于，所述图像处理模块包括图像分割模块、图像提取模块、图像运算模块、信息获取模块和图像结合模块，所述图像分割模块读取二维扫描图像，对读取的二维扫描图像进行阈值分割，所述图像提取模块提取二维扫描图像中的目标区域图像信息，所述图像运算模块对提取的目标区域图像信息进行形态学开运算，所述信息获取模块获取二维扫描图像中的目标区域图像灰度信息，所述图像结合模块将各线程获取的二维扫描图像中的目标区域图像灰度信息进行结合，去床扫描床信息。
  8. 根据权利要求7所述的CT图像扫描床去除装置，其特征在于，所述图像分割模块对读取的二维扫描图像进行阈值分割采用大津阈值分割。
  9. 根据权利要求7或8所述的CT图像扫描床去除装置，其特征在于，所述图像提取模块提取二维扫描图像中的目标区域图像信息具体为：提取二维扫描图像中的目标区域图像信息包括提取二维扫描图像中的身体部分信息。
  10. 根据权利要求7或8所述的CT图像扫描床去除装置，其特征在于，所述信息获取模块获取二维扫描图像中的目标区域图像灰度信息具体为：获取二维扫描图像中身体部分的灰度信息，去除CT图像扫描床信息。
PCT/CN2016/087435 2016-05-12 2016-06-28 一种ct图像扫描床去除方法及装置 WO2017193461A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/183,758 US20190073752A1 (en) 2016-05-12 2018-11-08 Method and device for removing scanning bed from ct image

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610319007.9A CN105931251A (zh) 2016-05-12 2016-05-12 一种ct图像扫描床去除方法及装置
CN201610319007.9 2016-05-12

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/183,758 Continuation US20190073752A1 (en) 2016-05-12 2018-11-08 Method and device for removing scanning bed from ct image

Publications (1)

Publication Number Publication Date
WO2017193461A1 true WO2017193461A1 (zh) 2017-11-16

Family

ID=56835892

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/087435 WO2017193461A1 (zh) 2016-05-12 2016-06-28 一种ct图像扫描床去除方法及装置

Country Status (3)

Country Link
US (1) US20190073752A1 (zh)
CN (1) CN105931251A (zh)
WO (1) WO2017193461A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108492299A (zh) * 2018-03-06 2018-09-04 天津天堰科技股份有限公司 一种三维图像的切割方法

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106952264B (zh) * 2017-03-07 2020-07-10 青岛海信医疗设备股份有限公司 三维医学目标的切割方法及装置
CN110335284A (zh) * 2019-07-11 2019-10-15 上海昌岛医疗科技有限公司 一种病理图像的去除背景的方法
CN110992331A (zh) * 2019-11-27 2020-04-10 中国地质大学（武汉） 一种二维多孔介质孔隙结构特征的定量评价装置及方法
CN111127475A (zh) * 2019-12-04 2020-05-08 上海联影智能医疗科技有限公司 Ct扫描图像处理方法、系统、可读存储介质和设备
CN113077474B (zh) * 2021-03-02 2024-05-17 心医国际数字医疗系统(大连)有限公司 基于ct影像的床板去除方法、系统、电子设备及存储介质

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1555028A (zh) * 2003-12-23 2004-12-15 沈阳东软数字医疗系统股份有限公司 医学图像分割中关于皮肤的自动提取方法
CN1657011A (zh) * 2005-03-22 2005-08-24 东软飞利浦医疗设备系统有限责任公司 一种自动去除黑心伪影的x－射线计算机层析成像机
CN101710420A (zh) * 2009-12-18 2010-05-19 华南师范大学 一种医学图像反分割方法
CN101721222A (zh) * 2009-09-16 2010-06-09 戴建荣 一种修正床板和摆位辅助装置对图像质量影响的方法
US20130336587A1 (en) * 2012-06-15 2013-12-19 Seoul National University R&Db Foundation Region growing apparatus and method using multi-core
CN104240198A (zh) * 2014-08-29 2014-12-24 西安华海盈泰医疗信息技术有限公司 一种ct图像中床板的去除方法及系统
CN104463840A (zh) * 2014-09-29 2015-03-25 北京理工大学 基于pet/ct影像的发热待查计算机辅助诊断方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102324090B (zh) * 2011-09-05 2014-06-18 东软集团股份有限公司 一种从cta图像中去除扫描床的方法及装置
CN103784158B (zh) * 2012-10-29 2016-08-03 株式会社日立制作所 Ct装置及ct图像生成方法
CN103886621B (zh) * 2012-11-14 2017-06-30 上海联影医疗科技有限公司 一种自动提取床板的方法
CO7020178A1 (es) * 2014-05-14 2014-08-11 Leon Ricardo Antonio Mendoza Método para la segmentación y cuantificación automática de tejidos corporales
TW201736865A (zh) * 2016-04-13 2017-10-16 Nihon Medi-Physics Co Ltd 來自核子醫學影像的生理累積之自動去除及ｃｔ影像之自動分段
TW201737206A (zh) * 2016-04-13 2017-10-16 Nihon Medi-Physics Co Ltd Ｃｔ影像中的骨區域之自動推定

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1555028A (zh) * 2003-12-23 2004-12-15 沈阳东软数字医疗系统股份有限公司 医学图像分割中关于皮肤的自动提取方法
CN1657011A (zh) * 2005-03-22 2005-08-24 东软飞利浦医疗设备系统有限责任公司 一种自动去除黑心伪影的x－射线计算机层析成像机
CN101721222A (zh) * 2009-09-16 2010-06-09 戴建荣 一种修正床板和摆位辅助装置对图像质量影响的方法
CN101710420A (zh) * 2009-12-18 2010-05-19 华南师范大学 一种医学图像反分割方法
US20130336587A1 (en) * 2012-06-15 2013-12-19 Seoul National University R&Db Foundation Region growing apparatus and method using multi-core
CN104240198A (zh) * 2014-08-29 2014-12-24 西安华海盈泰医疗信息技术有限公司 一种ct图像中床板的去除方法及系统
CN104463840A (zh) * 2014-09-29 2015-03-25 北京理工大学 基于pet/ct影像的发热待查计算机辅助诊断方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108492299A (zh) * 2018-03-06 2018-09-04 天津天堰科技股份有限公司 一种三维图像的切割方法

Also Published As

Publication number Publication date
CN105931251A (zh) 2016-09-07
US20190073752A1 (en) 2019-03-07

Similar Documents

Publication Publication Date Title
WO2017193461A1 (zh) 一种ct图像扫描床去除方法及装置
US10580137B2 (en) Systems and methods for detecting an indication of malignancy in a sequence of anatomical images
Tripathy et al. Unified preprocessing and enhancement technique for mammogram images
Kumar et al. Descriptive analysis of dental X-ray images using various practical methods: A review
Kainz et al. Fast fully automatic brain detection in fetal MRI using dense rotation invariant image descriptors
JP2023517058A (ja) 画像処理に基づく腫瘍の自動検出
Ribeiro et al. Handling inter-annotator agreement for automated skin lesion segmentation
Zhang et al. Medical image segmentation based on watershed and graph theory
Sagar et al. Color channel based segmentation of skin lesion from clinical images for the detection of melanoma
Zebari et al. Suspicious region segmentation using deep features in breast cancer mammogram images
TWI587844B (zh) 醫療影像處理裝置及其乳房影像處理方法
Mahmood et al. Ultrasound liver image enhancement using watershed segmentation method
CN116823701A (zh) 基于ai的医学影像分析处理方法
Nair et al. Modified level cut liver segmentation from ct images
Nurhayati et al. Stroke identification system on the mobile based CT scan image
Jamil et al. Adaptive thresholding technique for segmentation and juxtapleural nodules inclusion in lung segments
Rad et al. Level set and morphological operation techniques in application of dental image segmentation
CN111161256A (zh) 图像分割方法、图像分割装置、存储介质及电子设备
Elmorsy et al. K3. A region growing liver segmentation method with advanced morphological enhancement
EGA et al. Study on Image Processing Method and Data Augmentation for Chest X-Ray Nodule Detection with YOLOv5 Algorithm
Noviana et al. Axial segmentation of lungs CT scan images using canny method and morphological operation
Yan et al. Automatic detection and localization of pulmonary nodules in ct images based on yolov5
Yan et al. Feature extraction and analysis on X-ray image of Xinjiang Kazak Esophageal cancer by using gray-level histograms
Macho et al. Segmenting Teeth from Volumetric CT Data with a Hierarchical CNN-based Approach.
Rehman et al. Edge of discovery: Enhancing breast tumor MRI analysis with boundary-driven deep learning

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

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

Ref document number: 16901416

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 16901416

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 29.03.2019)

122 Ep: pct application non-entry in european phase

Ref document number: 16901416

Country of ref document: EP

Kind code of ref document: A1