WO2020172999A1 - 冠状动脉造影图像序列的质量评分方法和装置 - Google Patents

冠状动脉造影图像序列的质量评分方法和装置 Download PDF

Info

Publication number
WO2020172999A1
WO2020172999A1 PCT/CN2019/086605 CN2019086605W WO2020172999A1 WO 2020172999 A1 WO2020172999 A1 WO 2020172999A1 CN 2019086605 W CN2019086605 W CN 2019086605W WO 2020172999 A1 WO2020172999 A1 WO 2020172999A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
frame
quality
quality score
calculate
Prior art date
Application number
PCT/CN2019/086605
Other languages
English (en)
French (fr)
Inventor
霍云飞
王鹏
王之元
刘广志
曹文斌
徐磊
Original Assignee
苏州润迈德医疗科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 苏州润迈德医疗科技有限公司 filed Critical 苏州润迈德医疗科技有限公司
Publication of WO2020172999A1 publication Critical patent/WO2020172999A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0012Biomedical image inspection
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10116X-ray image
    • G06T2207/10121Fluoroscopy
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30004Biomedical image processing
    • G06T2207/30048Heart; Cardiac
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30004Biomedical image processing
    • G06T2207/30101Blood vessel; Artery; Vein; Vascular
    • G06T2207/30104Vascular flow; Blood flow; Perfusion
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30168Image quality inspection

Definitions

  • the invention relates to the technical field of medical image processing, in particular to a method and device for scoring the quality of coronary angiography image sequences, which can be applied to the clinical diagnosis of X-ray coronary angiography images.
  • Coronary atherosclerosis is the main cause of heart damage and myocardial infarction. Accurate diagnosis and correct positioning and quantification are very important. Coronary angiography can provide fast, high-quality blood vessel image sequences. It is currently one of the main basis for the diagnosis and treatment of coronary heart disease widely used in clinical medicine, and is known as the "gold standard" for the diagnosis of coronary heart disease. In recent years, coronary angiography image analysis has attracted the attention of the majority of heart disease researchers, and has become a research hotspot at home and abroad.
  • the purpose of the present invention is to provide a method and device for scoring the quality of coronary angiography image sequences, which can quantify the image quality and calculate and rank the quality scores of each frame of the images in the sequence.
  • the image quality score not only can the quality of coronary angiography images be judged intuitively and quickly, but also the optimal frame in the coronary angiography image sequence can be automatically recommended, which is convenient for the subsequent analysis of coronary angiography images and shortens the manual processing time. Reduce the error caused by artificial subjective factors to the analysis results.
  • a method for scoring the quality of coronary angiography image sequences including the following steps:
  • S04 Calculate the quality score of each frame of contrast image according to the quality score and blood vessel responsivity of the optimal frame image.
  • the step S01 includes the following steps:
  • ⁇ 1 , ⁇ 2 are the two eigenvalues of the Hessian matrix; ⁇ is the distinguishing parameter between linear and blocky, c is the smoothness parameter of the linear object, and ⁇ is the clarity parameter of the linear object.
  • the step S03 specifically includes:
  • SSIM is the structural similarity
  • the quality score cl of each frame of contrast image in the step S04 is:
  • NRSS represents the image quality score value of the optimal frame
  • H opt is the blood vessel responsivity of the optimal frame
  • H i represents the blood vessel responsivity of the i-th frame image.
  • the invention also discloses a quality scoring device for coronary angiography image sequences, including:
  • a blood vessel response degree calculation module calculate the blood vessel response degree of each frame of the coronary angiography image sequence
  • An optimal frame image judgment module compare the vascular responsivity of each frame of contrast image to obtain the optimal frame of the contrast image sequence;
  • An optimal frame image quality score calculation module use image gradient information to calculate the quality score of the optimal frame image
  • a contrast image quality score calculation module calculate the quality score of each frame of contrast image according to the quality score of the optimal frame image and the vascular response.
  • the calculation method of the blood vessel responsivity calculation module includes the following steps:
  • ⁇ 1 , ⁇ 2 are the two eigenvalues of the Hessian matrix; ⁇ is the distinguishing parameter between linear and blocky, c is the smoothness parameter of the linear object, and ⁇ is the clarity parameter of the linear object.
  • the calculation method of the optimal frame image quality score calculation module specifically includes:
  • SSIM is the structural similarity
  • the quality score cl of each frame of the contrast image in the contrast image quality score calculation module is:
  • NRSS represents the image quality score value of the optimal frame
  • H opt is the blood vessel responsivity of the optimal frame
  • H i represents the blood vessel responsivity of the i-th frame image.
  • the method of the invention can quantify the image quality, and can calculate and sort the quality score of each frame of the image in the sequence.
  • image quality scoring not only can the quality of coronary angiography images be judged intuitively and quickly, but also the optimal frame in the coronary angiography image sequence can be automatically recommended, which is convenient for the subsequent analysis of coronary angiography images and shortens the manual processing time. Reduce the error caused by artificial subjective factors to the analysis results, and improve the diagnostic efficiency and accuracy.
  • FIG. 1 is a flowchart of the method for scoring the quality of coronary angiography image sequences of the present invention
  • Figure 2 is the original coronary angiography image sequence
  • Figure 3 is a schematic diagram of the quality score of the coronary angiography image sequence in Figure 2;
  • Figure 4 is a sequence of coronary angiography images rearranged from small to large according to the image quality score
  • Fig. 5 is the optimal frame of the coronary angiography image sequence in Fig. 2.
  • the quality scoring method of coronary angiography image sequence firstly, calculate the vascular response of each frame of contrast image; secondly, compare the vascular response of each frame of contrast image to determine the best contrast image sequence Frame; Then, because the gradient information contains edge information, the image gradient information is used to calculate the quality score of the optimal frame image; Finally, the quality score of each frame of contrast image is calculated according to the quality score of the optimal frame image and the vascular response .
  • Each specific step includes:
  • the first step is to calculate the vascular responsivity of each frame of the coronary angiography image sequence.
  • g(x,y) is the Gaussian convolution template
  • I(x,y) is the coronary angiography image.
  • K (I xx +I yy )/2
  • is used to adjust the difference between linear and blocky
  • c is a parameter that controls the smoothness of linear objects
  • is a parameter that controls the clarity of linear objects. The greater the responsiveness, the greater the possibility that the current position is the blood vessel area.
  • the vascular response is expressed as ⁇ V i
  • i 0,1,2,...,f ⁇ , where f is the coronary angiography image sequence The total number of frames.
  • the second step is to calculate the optimal frame of the tubular angiography image sequence.
  • the third step is to calculate the image quality score of the optimal frame.
  • the Sobel operator Utilizing the human eye's most sensitive feature of horizontal and vertical edge information, the Sobel operator is used to extract the reference image Ir and the gradient images G r and G of the optimal frame I respectively.
  • the image quality score calculation method of the optimal frame can be expressed as:
  • SSIM is the structural similarity, and its expression is as follows:
  • the fourth step is to calculate the quality score cl of each frame of image.
  • NRSS represents the image quality score value of the optimal frame
  • H opt is the blood vessel responsivity of the optimal frame
  • H i represents the blood vessel responsivity of the i-th frame image.
  • the score range is [0,100].
  • the score value is 85-100, it indicates that the quality of coronary angiography image is relatively high and can meet the requirements of lesion analysis; when the score value is 70-85, it indicates that the quality of coronary angiography image is low and can Meet the requirements of part of the lesion analysis, but the results of the lesion analysis may have relatively large errors; when the score value is 0 to 70, it indicates that the quality of coronary angiography images is poor and cannot meet the requirements of lesion analysis.

Landscapes

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

Abstract

本发明公开了一种冠状动脉造影图像序列的质量评分方法,包括:计算冠状动脉造影图像序列中每一帧图像的血管响应度;比较每一帧造影图像的血管响应度,得到造影图像序列的最优帧;利用图像梯度信息计算最优帧图像的质量评分;根据最优帧图像的质量评分和血管响应度,计算每一帧造影图像的质量评分。将图像质量量化,可以计算出序列中图像每一帧的质量评分并排序,通过图像质量评分不仅可以直观迅速的判断出冠状动脉造影图像质量的优劣,而且能够自动推荐出冠状动脉造影图像序列中的最优帧,便于后期冠状动脉造影图像分析,缩短人工处理时间,同时减少人工主观因素对分析结果带来的误差。

Description

冠状动脉造影图像序列的质量评分方法和装置 技术领域
本发明涉及医学图像处理技术领域,具体地涉及一种冠状动脉造影图像序列的质量评分方法和装置,可应用于X射冠状动脉造影图像的临床诊断。
背景技术
冠状动脉分布在心脏的表面,分出许多小分支进入心肌,并为心肌供应血液。冠状动脉粥样硬化是造成心脏受损和心肌梗塞的主要原因,准确的诊断并对其进行正确定位和量化非常关键。冠状动脉造影术可以提供快速、高质量的血管图像序列,是目前医学临床广泛采用的诊断和治疗冠心病的主要依据之一,被称为诊断冠心病的“金标准”。近些年来,冠状动脉造影图像分析已经引起广大心脏疾病研究者的注意,并成为国内外研究的热点。
在冠状动脉造影图像分析前,为了能更好的分析血管病变,首先需要挑选高质量的冠状动脉造影图像。如果对质量差的冠状动脉造影图像进行病变分析,会严重影响分析结果的准确度,所以判断冠状动脉造影图像质量优劣是冠状动脉造影图像分析的前提。评价冠状动脉造影图像质量好坏,需要结合造影图像的清晰度和血管的完整度这两个因素。目前,往往需要人工判断冠状动脉造影图像质量是否符合病变分析的要求,这过程不仅需要多年临床经验,耗费大量时间,而且由于人的主观判断标准存在差异,不可避免导致后续图像分析存在误差。本发明因此而来。
发明内容
为了解决上述存在的技术问题,本发明的目的是:提供了一种冠状动脉造影图像序列的质量评分方法和装置,将图像质量量化,可以计算出序列中图像每一帧的质量评分并排序,通过图像质量评分不仅可以直观迅速的判断出冠状动脉造影图像质量的优劣,而且能够自动推荐出冠状动脉造影图像序列中的最优帧,便于后期冠状动脉造影图像分析,缩短人工处理时间,同时减少人工主观因素对分析结果带来的误差。
本发明的技术方案是:
一种冠状动脉造影图像序列的质量评分方法,包括以下步骤:
S01:计算冠状动脉造影图像序列中每一帧图像的血管响应度;
S02:比较每一帧造影图像的血管响应度,得到造影图像序列的最优帧;
S03:利用图像梯度信息计算最优帧图像的质量评分;
S04:根据最优帧图像的质量评分和血管响应度,计算每一帧造影图像的质量评分。
优选的技术方案中,所述步骤S01包括以下步骤:
S11:利用Hessian矩阵计算血管响应度,血管响应度V为:
Figure PCTCN2019086605-appb-000001
其中,
Figure PCTCN2019086605-appb-000002
λ 1,λ 2为Hessian矩阵的两个特征值;β为线状和块状的区别参数,c为线状物体平滑程度参数,γ为线状物体清晰程度参数。
S12:重复上述步骤,计算造影图像序列每一帧的血管响应度,得到血管的响应度为{V i|i=0,1,2,...,f},其中f为冠状动脉造影图像序列的总帧数。
优选的技术方案中,所述步骤S03具体包括:
S31:对最优帧图像I进行高斯平滑滤波得到参考图像I r
S32:分别提取参考图像I r和最优帧图像I的梯度图像G r和G;
S33:将梯度图像G和G r分别划分为一定尺寸的小块,计算每块的方差,分别找出其中N个最大方差,记为{x i|i=1,2,...,N}和{y i|i=1,2,...,N};
S34:计算最优帧图像的质量评分,最优帧图像的质量评分NRSS为:
Figure PCTCN2019086605-appb-000003
其中,SSIM为结构相似度,
Figure PCTCN2019086605-appb-000004
优选的技术方案中,所述步骤S04中每一帧造影图像的质量评分cl为:
cl=NRSS*SSIM(H opt,H i);
其中NRSS表示最优帧图像质量评分值,H opt为最优帧的血管响应度,H i表示第i帧图像的血管响应度。
本发明还公开了一种冠状动脉造影图像序列的质量评分装置,包括:
一血管响应度计算模块:计算冠状动脉造影图像序列中每一帧图像的血管响应度;
一最优帧图像判断模块:比较每一帧造影图像的血管响应度,得到造影图像序列的最优帧;
一最优帧图像质量评分计算模块:利用图像梯度信息计算最优帧图像的质量评分;
一造影图像质量评分计算模块:根据最优帧图像的质量评分和血管响应度,计算每一帧造影图像的质量评分。
优选的技术方案中,所述血管响应度计算模块的计算方法包括以下步骤:
S11:利用Hessian矩阵计算血管响应度,血管响应度V为:
Figure PCTCN2019086605-appb-000005
其中,
Figure PCTCN2019086605-appb-000006
λ 1,λ 2为Hessian矩阵的两个特征值;β为线状和块状的区别参数,c为线状物体平滑程度参数,γ为线状物体清晰程度参数。
S12:重复上述步骤,计算造影图像序列每一帧的血管响应度,得到血管的响应度为{V i|i=0,1,2,...,f},其中f为冠状动脉造影图像序列的总帧数。
优选的技术方案中,所述最优帧图像质量评分计算模块的计算方法具体包括:
S31:对最优帧图像I进行高斯平滑滤波得到参考图像I r
S32:分别提取参考图像I r和最优帧图像I的梯度图像G r和G;
S33:将梯度图像G和G r分别划分为一定尺寸的小块,计算每块的方差, 分别找出其中N个最大方差,记为{x i|i=1,2,...,N}和{y i|i=1,2,...,N};
S34:计算最优帧图像的质量评分,最优帧图像的质量评分NRSS为:
Figure PCTCN2019086605-appb-000007
其中,SSIM为结构相似度,
Figure PCTCN2019086605-appb-000008
优选的技术方案中,所述造影图像质量评分计算模块中每一帧造影图像的质量评分cl为:
cl=NRSS*SSIM(H opt,H i);
其中NRSS表示最优帧图像质量评分值,H opt为最优帧的血管响应度,H i表示第i帧图像的血管响应度。
与现有技术相比,本发明的优点是:
本发明方法可以将图像质量量化,可以计算出序列中图像每一帧的质量评分并排序。通过图像质量评分不仅可以直观迅速的判断出冠状动脉造影图像质量的优劣,而且能够自动推荐出冠状动脉造影图像序列中的最优帧,便于后期冠状动脉造影图像分析,缩短人工处理时间,同时减少人工主观因素对分析结果带来的误差,提高诊断效率及准确率。
附图说明
下面结合附图及实施例对本发明作进一步描述:
图1为本发明冠状动脉造影图像序列的质量评分方法的流程图;
图2为原始冠状动脉造影图像序列;
图3为图2中冠状动脉造影图像序列的质量评分示意图;
图4为根据图像质量评分从小到大重新排列的冠状动脉造影图像序列;
图5为图2中冠状动脉造影图像序列的最优帧。
具体实施方式
为使本发明的目的、技术方案和优点更加清楚明了,下面结合具体实施方式并参照附图,对本发明进一步详细说明。应该理解,这些描述只是示例 性的,而并非要限制本发明的范围。此外,在以下说明中,省略了对公知结构和技术的描述,以避免不必要地混淆本发明的概念。
如图1所示,冠状动脉造影图像序列的质量评分方法,首先,计算每一帧造影图像的血管响应度;其次,比较每一帧造影图像的血管响应度,判断出造影图像序列的最优帧;然后,由于梯度信息中包含边缘信息,所以利用图像梯度信息计算最优帧图像的质量评分;最后,根据最优帧图像的质量评分和血管响应度,计算每一帧造影图像的质量评分。
具体的每一步骤包括:
第一步,计算冠状动脉造影图像序列中每一帧图像的血管响应度。
1.1)从冠状动脉造影图像序列中提取一帧图像。
1.2)图像预处理。
使用3×3均值滤波器降低图像噪声,提高算法准确度。
1.3)利用Hessian矩阵计算血管响应度。
(a)二维Hessian矩阵为:
Figure PCTCN2019086605-appb-000009
其中,
Figure PCTCN2019086605-appb-000010
g(x,y)为高斯卷积模板,I(x,y)为冠状动脉造影图像。
(b)Hessian矩阵的两个特征值λ 1,λ 2可以由下面公式计算:
Figure PCTCN2019086605-appb-000011
其中,K=(I xx+I yy)/2,
Figure PCTCN2019086605-appb-000012
(c)Hessian矩阵对血管的响应度可表示为:
Figure PCTCN2019086605-appb-000013
其中,
Figure PCTCN2019086605-appb-000014
β用于调整线状和块状的区别,c为控制线状物体平滑程度的参数,γ为控制线状物体清晰程度的参数。响应度越大, 表示当前位置是血管区域的可能越大。
重复上述步骤,计算造影图像序列每一帧的血管响应度,血管的响应度表示为{V i|i=0,1,2,...,f},其中f为冠状动脉造影图像序列的总帧数。
第二步,计算管状动脉造影图像序列的最优帧。
比较造影图像序列每一帧的血管响应度,响应度最大的那一帧即为最优帧。
第三步,计算最优帧的图像质量评分。
3.1)构造参考图像。
定义最优帧的图像为I,对图像I进行高斯平滑滤波得到参考图像I r
3.2)提取梯度信息。
利用人眼对水平和垂直方向的边缘信息最为敏感的特性,使用Sobel算子分别提取参考图像I r和最优帧I的梯度图像G r和G。
3.3)梯度图像的方差分析。
将梯度图像G划分为一定尺寸的小块,例如8×8的小块,块间的步长为4,计算每块的方差,方差越大说明梯度信息越丰富。找出其中N个最大方差,记为{x i|i=1,2,...,N},对应的G r中的对应的N个最大方差为{y i|i=1,2,...,N},其中N=32。
3.4)计算最优帧的图像质量评分。
最优帧的图像质量评分计算方式可以表示为:
Figure PCTCN2019086605-appb-000015
其中SSIM为结构相似度,其表达式如下:
Figure PCTCN2019086605-appb-000016
第四步,计算每一帧图像的质量评分cl。
cl=NRSS*SSIM(H opt,H i),
其中NRSS表示最优帧图像质量评分值,H opt为最优帧的血管响应度,H i 表示第i帧图像的血管响应度。
评分取值范围为[0,100],当评分值为85~100,表明冠状动脉造影图像质量比较高,能满足病变分析的要求;当评分值为70~85表明冠状动脉造影图像质量较低,能满足部分病变分析的要求,但是病变分析结果可能误差比较大;当评分值为0~70表明冠状动脉造影图像质量很差,不能满足病变分析的要求。
应当理解的是,本发明的上述具体实施方式仅仅用于示例性说明或解释本发明的原理,而不构成对本发明的限制。因此,在不偏离本发明的精神和范围的情况下所做的任何修改、等同替换、改进等,均应包含在本发明的保护范围之内。此外,本发明所附权利要求旨在涵盖落入所附权利要求范围和边界、或者这种范围和边界的等同形式内的全部变化和修改例。

Claims (8)

  1. 一种冠状动脉造影图像序列的质量评分方法,其特征在于,包括以下步骤:
    S01:计算冠状动脉造影图像序列中每一帧图像的血管响应度;
    S02:比较每一帧造影图像的血管响应度,得到造影图像序列的最优帧;
    S03:利用图像梯度信息计算最优帧图像的质量评分;
    S04:根据最优帧图像的质量评分和血管响应度,计算每一帧造影图像的质量评分。
  2. 根据权利要求1所述的冠状动脉造影图像序列的质量评分方法,其特征在于,所述步骤S01包括以下步骤:
    S11:利用Hessian矩阵计算血管响应度,血管响应度V为:
    Figure PCTCN2019086605-appb-100001
    其中,
    Figure PCTCN2019086605-appb-100002
    λ 1,λ 2为Hessian矩阵的两个特征值;β为线状和块状的区别参数,c为线状物体平滑程度参数,γ为线状物体清晰程度参数。
    S12:重复上述步骤,计算造影图像序列每一帧的血管响应度,得到血管的响应度为{V i|i=0,1,2,...,f},其中f为冠状动脉造影图像序列的总帧数。
  3. 根据权利要求1所述的冠状动脉造影图像序列的质量评分方法,其特征在于,所述步骤S03具体包括:
    S31:对最优帧图像I进行高斯平滑滤波得到参考图像I r
    S32:分别提取参考图像I r和最优帧图像I的梯度图像G r和G;
    S33:将梯度图像G和G r分别划分为一定尺寸的小块,计算每块的方差,分别找出其中N个最大方差,记为{x i|i=1,2,...,N}和{y i|i=1,2,...,N};
    S34:计算最优帧图像的质量评分,最优帧图像的质量评分NRSS为:
    Figure PCTCN2019086605-appb-100003
    其中,SSIM为结构相似度,
    Figure PCTCN2019086605-appb-100004
  4. 根据权利要求1所述的冠状动脉造影图像序列的质量评分方法,其特征在于,所述步骤S04中每一帧造影图像的质量评分cl为:
    cl=NRSS*SSIM(H opt,H i);
    其中NRSS表示最优帧图像质量评分值,H opt为最优帧的血管响应度,H i表示第i帧图像的血管响应度。
  5. 一种冠状动脉造影图像序列的质量评分装置,其特征在于,包括:
    一血管响应度计算模块:计算冠状动脉造影图像序列中每一帧图像的血管响应度;
    一最优帧图像判断模块:比较每一帧造影图像的血管响应度,得到造影图像序列的最优帧;
    一最优帧图像质量评分计算模块:利用图像梯度信息计算最优帧图像的质量评分;
    一造影图像质量评分计算模块:根据最优帧图像的质量评分和血管响应度,计算每一帧造影图像的质量评分。
  6. 根据权利要求5所述的冠状动脉造影图像序列的质量评分装置,其特征在于,所述血管响应度计算模块的计算方法包括以下步骤:
    S11:利用Hessian矩阵计算血管响应度,血管响应度V为:
    Figure PCTCN2019086605-appb-100005
    其中,
    Figure PCTCN2019086605-appb-100006
    λ 1,λ 2为Hessian矩阵的两个特征值;β为线状和块状的区别参数,c为线状物体平滑程度参数,γ为线状物体清晰程度参数。
    S12:重复上述步骤,计算造影图像序列每一帧的血管响应度,得到血管的响应度为{V i|i=0,1,2,...,f},其中f为冠状动脉造影图像序列的总帧数。
  7. 根据权利要求5所述的冠状动脉造影图像序列的质量评分装置,其 特征在于,所述最优帧图像质量评分计算模块的计算方法具体包括:
    S31:对最优帧图像I进行高斯平滑滤波得到参考图像I r
    S32:分别提取参考图像I r和最优帧图像I的梯度图像G r和G;
    S33:将梯度图像G和G r分别划分为一定尺寸的小块,计算每块的方差,分别找出其中N个最大方差,记为{x i|i=1,2,...,N}和{y i|i=1,2,...,N};
    S34:计算最优帧图像的质量评分,最优帧图像的质量评分NRSS为:
    Figure PCTCN2019086605-appb-100007
    其中,SSIM为结构相似度,
    Figure PCTCN2019086605-appb-100008
  8. 根据权利要求5所述的冠状动脉造影图像序列的质量评分装置,其特征在于,所述造影图像质量评分计算模块中每一帧造影图像的质量评分cl为:
    cl=NRSS*SSIM(H opt,H i);
    其中NRSS表示最优帧图像质量评分值,H opt为最优帧的血管响应度,H i表示第i帧图像的血管响应度。
PCT/CN2019/086605 2019-02-28 2019-05-13 冠状动脉造影图像序列的质量评分方法和装置 WO2020172999A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910148601.X 2019-02-28
CN201910148601.XA CN111626974B (zh) 2019-02-28 2019-02-28 冠状动脉造影图像序列的质量评分方法和装置

Publications (1)

Publication Number Publication Date
WO2020172999A1 true WO2020172999A1 (zh) 2020-09-03

Family

ID=72238807

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/086605 WO2020172999A1 (zh) 2019-02-28 2019-05-13 冠状动脉造影图像序列的质量评分方法和装置

Country Status (2)

Country Link
CN (1) CN111626974B (zh)
WO (1) WO2020172999A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112785566B (zh) * 2021-01-15 2024-01-19 湖南自兴智慧医疗科技有限公司 染色体中期图像评分方法、装置、电子设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101996406A (zh) * 2010-11-03 2011-03-30 中国科学院光电技术研究所 无参考结构清晰度图像质量评价方法
CN104732499A (zh) * 2015-04-01 2015-06-24 武汉工程大学 基于多尺度多方向的视网膜图像增强算法
CN106934806A (zh) * 2017-03-09 2017-07-07 东南大学 一种基于结构清晰度的无参考图失焦模糊区域分割方法
CN107145855A (zh) * 2017-04-28 2017-09-08 努比亚技术有限公司 一种无参考质量模糊图像预测方法、终端及存储介质
US20170372155A1 (en) * 2016-06-23 2017-12-28 Siemens Healthcare Gmbh Image Quality Score Using A Deep Generative Machine-Learning Model

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110052035A1 (en) * 2009-09-01 2011-03-03 Siemens Corporation Vessel Extraction Method For Rotational Angiographic X-ray Sequences
US9053551B2 (en) * 2012-05-23 2015-06-09 International Business Machines Corporation Vessel identification using shape and motion mapping for coronary angiogram sequences
EA024855B1 (ru) * 2012-07-10 2016-10-31 Закрытое Акционерное Общество "Импульс" Способ получения субтракционного ангиографического изображения
US9974506B2 (en) * 2013-11-05 2018-05-22 International Business Machines Corporation Associating coronary angiography image annotations with syntax scores for assessment of coronary artery disease
US9508157B2 (en) * 2014-12-12 2016-11-29 Siemens Medical Solutions Usa, Inc. Reconstruction of aneurysm wall motion
CN108475532B (zh) * 2015-12-30 2022-12-27 皇家飞利浦有限公司 医学报告装置
CN106412571B (zh) * 2016-10-12 2018-06-19 天津大学 一种基于梯度相似性标准差的视频质量评价方法
CN108492300B (zh) * 2018-03-16 2021-07-13 上海理工大学 管状结构增强与能量函数结合的肺部血管树分割方法
CN109377481B (zh) * 2018-09-27 2022-05-24 上海联影医疗科技股份有限公司 图像质量评价方法、装置、计算机设备和存储介质
CN108805871B (zh) * 2018-06-14 2021-06-25 艾瑞迈迪医疗科技(北京)有限公司 血管图像处理方法、装置、计算机设备和存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101996406A (zh) * 2010-11-03 2011-03-30 中国科学院光电技术研究所 无参考结构清晰度图像质量评价方法
CN104732499A (zh) * 2015-04-01 2015-06-24 武汉工程大学 基于多尺度多方向的视网膜图像增强算法
US20170372155A1 (en) * 2016-06-23 2017-12-28 Siemens Healthcare Gmbh Image Quality Score Using A Deep Generative Machine-Learning Model
CN106934806A (zh) * 2017-03-09 2017-07-07 东南大学 一种基于结构清晰度的无参考图失焦模糊区域分割方法
CN107145855A (zh) * 2017-04-28 2017-09-08 努比亚技术有限公司 一种无参考质量模糊图像预测方法、终端及存储介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
JIANG, XIANGANG ET AL.: "Enhancement Filter Algorithm of Retina Blood Vessels Based on Hessian Matrix Features", JOURNAL OF EAST CHINA JIAOTONG UNIVERSITY, vol. 30, no. 3, 30 June 2013 (2013-06-30), pages 37 - 43, XP009522913, ISSN: 1005-0523 *

Also Published As

Publication number Publication date
CN111626974A (zh) 2020-09-04
CN111626974B (zh) 2024-03-22

Similar Documents

Publication Publication Date Title
CN103717122B (zh) 眼科诊断支持设备和眼科诊断支持方法
Niemeijer et al. Fast detection of the optic disc and fovea in color fundus photographs
WO2020042406A1 (zh) 一种眼底图像自动分析比对方法及一种存储设备
Liu et al. Automatic whole heart segmentation using a two-stage u-net framework and an adaptive threshold window
WO2022142030A1 (zh) 高血压性视网膜病变的病变特征的测量方法及测量系统
CN107292835B (zh) 一种眼底图像视网膜血管自动矢量化的方法及装置
WO2021208739A1 (zh) 眼底彩照图像血管评估方法、装置、计算机设备和介质
JP7197708B2 (ja) 眼底画像定量分析の前置処理方法および記憶装置
CN112233789A (zh) 一种区域特征融合的高血压视网膜病变分类方法
Oloumi et al. Computer-aided diagnosis of plus disease in retinal fundus images of preterm infants via measurement of vessel tortuosity
CN112288794B (zh) 眼底图像的血管管径的测量方法及测量装置
Hatanaka et al. Improvement of automatic hemorrhage detection methods using brightness correction on fundus images
CN112396565A (zh) 静脉穿刺机器人的图像及视频血管增强与分割方法和系统
Prentasic et al. Weighted ensemble based automatic detection of exudates in fundus photographs
Maqsood et al. Detection of macula and recognition of aged-related macular degeneration in retinal fundus images
WO2020172999A1 (zh) 冠状动脉造影图像序列的质量评分方法和装置
Ding et al. Multi-scale morphological analysis for retinal vessel detection in wide-field fluorescein angiography
Dikkala et al. A comprehensive analysis of morphological process dependent retinal blood vessel segmentation
Niemeijer et al. Automated localization of the optic disc and the fovea
Liu et al. Retinal vessel segmentation using densely connected convolution neural network with colorful fundus images
CN104851103B (zh) 基于sd‑oct视网膜图像的脉络膜血管抽取方法
Sumathy et al. Feature extraction in retinal fundus images
Chi et al. A composite of features for learning-based coronary artery segmentation on cardiac CT angiography
CN112017132A (zh) 基于最大曲率法和多尺度Hessian矩阵的静脉图像增强方法
Prabakar et al. Implementation of stochastic approach for vessel and ridge studies in retinopathy of prematurity screening

Legal Events

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

Ref document number: 19916818

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19916818

Country of ref document: EP

Kind code of ref document: A1