CN108852385B - X-ray radiography method and dynamic radiograph reading method based on X-ray radiography - Google Patents

X-ray radiography method and dynamic radiograph reading method based on X-ray radiography Download PDF

Info

Publication number
CN108852385B
CN108852385B CN201810203889.1A CN201810203889A CN108852385B CN 108852385 B CN108852385 B CN 108852385B CN 201810203889 A CN201810203889 A CN 201810203889A CN 108852385 B CN108852385 B CN 108852385B
Authority
CN
China
Prior art keywords
contrast
signal
characteristic
ray
frequency
Prior art date
Legal status (The legal status 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 status listed.)
Active
Application number
CN201810203889.1A
Other languages
Chinese (zh)
Other versions
CN108852385A (en
Inventor
肖体乔
王飞翔
陈荣昌
王玉丹
杜国浩
邓彪
谢红兰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Institute of Applied Physics of CAS
Original Assignee
Shanghai Institute of Applied Physics of CAS
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 Shanghai Institute of Applied Physics of CAS filed Critical Shanghai Institute of Applied Physics of CAS
Priority to CN201810203889.1A priority Critical patent/CN108852385B/en
Publication of CN108852385A publication Critical patent/CN108852385A/en
Application granted granted Critical
Publication of CN108852385B publication Critical patent/CN108852385B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/48—Diagnostic techniques
    • A61B6/481—Diagnostic techniques involving the use of contrast agents
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/50—Clinical applications
    • A61B6/504—Clinical applications involving diagnosis of blood vessels, e.g. by angiography
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
    • A61B6/5258—Devices using data or image processing specially adapted for radiation diagnosis involving detection or reduction of artifacts or noise
    • A61B6/5264—Devices using data or image processing specially adapted for radiation diagnosis involving detection or reduction of artifacts or noise due to motion
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B27/00—Optical systems or apparatus not provided for by any of the groups G02B1/00 - G02B26/00, G02B30/00
    • G02B27/02—Viewing or reading apparatus
    • G02B27/022—Viewing apparatus
    • G02B27/023—Viewing apparatus for viewing X-ray images using image converters, e.g. radioscopes

Abstract

The present invention provides an X-ray radiography method including: an original contrast image sequence; selecting two characteristic points on one frame of an original contrast image sequence, extracting time sequence signals of the two characteristic points and carrying out Fourier transform on the time sequence signals into frequency domain signals; carrying out characteristic analysis on the frequency domain signals of the two characteristic points, and selecting at least two maximum module value differences of the two characteristic points under the same frequency; traversing all the pixel points, performing Fourier band-pass filtering processing on the time sequence signal of each pixel point by taking 0Hz and one of the characteristic frequencies as a reference, and calculating the modulation depth of each pixel point through the obtained signal; and outputting an image with the modulation depth as the contrast imaging parameter. In addition, a dynamic radiographing method based on the X-ray contrast method is also provided. The X-ray radiography method and the dynamic radiograph reading method based on the X-ray radiography realize high signal-to-noise ratio imaging under low X-ray dosage, omit subtraction and interframe registration and realize the effect of image fusion.

Description

X-ray radiography method and dynamic radiograph reading method based on X-ray radiography
Technical Field
The invention relates to an X-ray angiography method based on multi-frame medical images, in particular to an angiography data processing method and a dynamic radiograph reading method based on space-time filtering.
Background
X-ray angiography methods are X-ray imaging methods that remove background structures from the final image, and are largely classified into two types. One is the K-edge subtraction angiography method, which is to obtain X-ray images of the sample at K-edge energies slightly below and slightly above the contrast agent, respectively. Then, the two images are subjected to difference, and a blood vessel image only containing contrast agent information can be obtained by utilizing the strong absorption effect of the contrast agent on the X-ray on the K-edge energy; the second is a single-energy time subtraction technique, which collects a mask image containing a background structure at an energy slightly higher than the K edge of the contrast agent. Then, a contrast agent is injected, followed by the acquisition of images of the blood vessels and perfused tissue containing the contrast agent. The mask image is subtracted from the image containing the injected blood vessels and perfused tissue. Thus, a blood vessel image containing only contrast agent information is obtained; currently, a single-energy time subtraction technique is commonly used to perform examination and diagnosis of vascular disease, measurement of vascular position, surgical planning of interventional therapy, biomedicine and other research, and multi-frame angiographic images are realized through a radiology imaging device and a subsequent data processing method thereof. However, direct subtraction often results in unacceptable artifacts due to unavoidable breathing, heartbeat, and other partial tissue movements of the human (or other living organism), stress of the blood vessels to the contrast agent, and vibration of external equipment.
In the patent document CN 106999129A, CN 101822545B, CN 102103745A, CN 102663709A, CN 1897033A, CN 101799918B, the software implementation of the apparatus and method of the digital subtraction angiography technique, and the subsequent vessel enhancement, vessel segmentation, elimination of motion artifacts, image fusion of different region angiography, etc. are respectively involved.
In CN 106999129A, the method additionally acquires a complete respiratory cycle image sequence and a heartbeat cycle image sequence that do not contain a contrast agent, and in a subsequent subtraction operation, the respiratory artifact and the heartbeat artifact are respectively subtracted from a mask and a filling slice respectively by matching the respiratory phase and the heartbeat phase, so as to achieve improvement of the quality of the subtracted image; the method in CN 101822545B belongs to an image registration technique, which makes the image registration of the digital subtraction operation mask and the filling film more accurate, thereby removing the motion artifact. The methods in CN 106999129A and CN 101822545B are to solve the problem of eliminating motion artifacts, but the operations are complex and have high requirements on the quality of original data.
In CN 102103745A, the method of computer software is used to realize digital subtraction operation, and dynamic image reading effect of the subtracted image can be realized, but the digital subtraction imaging technology is not substantially improved.
The method in CN 102663709A is a blood vessel enhancement algorithm, an X-ray coronary angiography image is decomposed into a series of intrinsic mode functions by using an empirical mode, background noise is removed by using a distribution rule of noise, some specific layer intrinsic mode functions are further selected for weighting to construct a coronary artery image, and then a blood vessel structure in the image is enhanced by using a blood vessel measurement function based on a Hessian matrix characteristic value, so that the visual effect of the angiography image is improved; CN 1897033 a proposes a local threshold blood vessel segmentation method based on a priori knowledge of blood vessel diameter and an overlap blocking technique to extract a target blood vessel for the case that a digital subtraction image cannot completely eliminate noise signals caused by human tissues and the background and the target are mixed together. CN 102663709 a relates to vessel enhancement, and CN 1897033 a relates to vessel segmentation, which are all performed after the digital subtraction imaging operation, in order to further improve the quality of the vessel image.
For the case of discontinuous distribution of contrast agent in blood vessels, CN 101799918B relates to image fusion and is based on digital subtraction imaging operation. The method in CN 101799918B provides an adaptive fusion method of digital subtraction images through dynamic blurring technique and Ridgelet transformation theory, aiming at that in digital subtraction imaging, a single subtraction image can only represent the local blood vessel image condition. With the data processing method proposed in this patent document, the imaging quality can be improved to some extent.
In summary, the X-ray contrast methods disclosed in these patent documents are based on the conventional digital subtraction imaging technology, and it is often difficult to solve the following problems:
1) the traditional X-ray radiography method based on the digital subtraction imaging technology is contrast agent subtraction imaging, an image with an undesirable radiography effect is generally subjected to imaging quality improvement by adopting an algorithm with an enhanced structure to be tested, the algorithm is too complex and has limited promotion effect, the algorithm is strongly dependent on high radiation dose, and the radiography effect in an original radiography image sequence is not ideal due to low radiation dose or low contrast agent concentration, so that the modulation depth is influenced, and the resolution of the image is reduced.
2) The conventional X-ray radiography method based on the digital subtraction imaging technology requires a subtraction operation to obtain a blood vessel image, and the key step is to find a proper mask and a proper filling film. However, the signal-to-noise ratio of the imaging is greatly affected because the inevitable movement of the sample in the contrast process causes the displacement between the mask and the filling film. The general method of using frame-to-frame registration solves the problem, and sometimes an invasive external feature point is introduced to complete frame-to-frame matching. This greatly increases the complexity of the algorithm and has limited improvement in image quality.
3) In the traditional X-ray contrast method based on the digital subtraction imaging technology, the selected contrast agent in the filling slice can not be completely perfused into all blood vessels but is partially perfused. Therefore, a single contrast image can only represent the local vessel imaging condition. Doctors are time consuming, labor intensive, and prone to missed diagnoses when viewing multiple contrast image sequences. Therefore, each digital subtraction angiography image is fused into a complete image, so that the complete image contains the whole structure of the blood vessel, and the method has important significance for medical auxiliary diagnosis. However, the image fusion technique is not only complicated in operation, but also introduces new noise.
4) In the traditional X-ray radiography method based on the digital subtraction imaging technology, a proper mask cannot be found for any filling film, and blood vessels in the filling film have large displacement. So that a good dynamic interpretation and playback of the digital subtraction angiography process cannot be performed.
Disclosure of Invention
The invention provides an X-ray radiography method based on space-time filtering and a dynamic radiography method based on X-ray radiography, which are used for realizing high signal-to-noise ratio imaging under low X-ray dosage, omitting subtraction and interframe registration and realizing the effect of image fusion.
In order to achieve the above object, the present invention provides an X-ray imaging method including: step S1: acquiring an original contrast image sequence of N frames by using sampling frequency Fs, and taking the original contrast image sequence as a current contrast image sequence; step S2: in the original contrast image sequenceSelecting blood vessel characteristic points and background characteristic points on one frame of the line, extracting time sequence signals of the two characteristic points and Fourier transforming the time sequence signals into frequency domain signals; step S3: carrying out feature analysis on the frequency domain signals of the blood vessel feature points and the background feature points, sequentially selecting at least two maximum module value differences of the two feature points under the same frequency, and taking the frequency corresponding to the ith maximum module value difference as the feature frequency xi x Fs/N Hz; step S4: traversing all pixel points of the contrast region in the current contrast image sequence, carrying out Fourier band-pass filtering processing on the time sequence signal of each pixel point by taking 0Hz and one of the characteristic frequencies xi x Fs/N Hz as the reference, and calculating the modulation depth MD of each pixel point through the obtained signali(ii) a Step S5: outputting to modulate depth MDiAn angiographic image being an angiographic imaging parameter; wherein xi is a positive integer.
The step S1 includes: under the irradiation of an X-ray light source, an original contrast image sequence is acquired by taking the period from before the contrast agent is not perfused as a starting frame to after the contrast agent is perfused as an ending frame.
The energy of the X-ray is higher than the absorption edge of the contrast agent, and the photon flux is 109—1010photonssecond‐ 1mm‐2。
The contrast agent is a non-ionic type iodized salt contrast agent, and the non-ionic type iodized salt contrast agent is one selected from the group consisting of iofelone, iohexol, iopromide, ioversol, iodixanol and diatrizoate.
The dosage of the contrast agent is 100-500 microliter, and the concentration is 150-350280 mg/ml‐1。
The modulus differences selected in step S3 are 2-4, and the number of the characteristic frequencies is equal to the number of the modulus differences.
The modulation depth MD of each pixel pointiCalculated by the following formula:
Figure BDA0001595324360000041
wherein S is0(x,y,t)、Sxi(x, y, t) are respectively a direct current signal obtained after Fourier band-pass filtering processing is carried out on the time sequence signal of each pixel point and an alternating current signal corresponding to xi x Fs/N Hz frequency;
Mean(abs(Sxi(x, y, t))) is the average of the absolute values of the alternating signals corresponding to the frequencies xi x Fs/N Hz;
Mean(abs(S0(x, y, t))) is the average of the absolute values of the dc signals.
The step S4 further includes: before traversing all pixel points of the contrast region, denoising each frame image, wherein the denoising comprises median denoising, Gaussian denoising or wiener denoising.
Further, the invention also provides a dynamic radiographing method based on X-ray contrast, which comprises the following steps:
step S1': using the X-ray imaging method according to claim 1, the characteristic frequency xi X Fs/N Hz is obtained and the modulation depth MD is obtainediTaking j as 1 as a contrast image corresponding to the first N frames of the original contrast image sequence, wherein j is a contrast image of a contrast imaging parameter; step S2': selecting the first N-j frames of the original contrast image sequence as a current contrast image sequence, and calculating the characteristic frequency xi x Fs/(N-j) Hz corresponding to the current contrast image sequence according to the characteristic frequency obtained in the step S1'; step S3': traversing all pixel points of the contrast region in the current contrast image sequence, performing Fourier band-pass filtering processing on the time sequence signal of each pixel point by taking 0Hz and the characteristic frequency xi x Fs/(N-j) Hz obtained in the step S2' as the reference, and calculating the modulation depth MD of each pixel point according to the obtained signali(ii) a Is obtained to modulate the depth MDiThe contrast image which is a contrast imaging parameter is used as a contrast image corresponding to the first N-j frame of the original contrast image sequence; step S4': when N-j is N, sequentially playing the contrast images corresponding to the previous N, N +1 …, N-1, N-2 and N frames; otherwise, j +1 is updated to j and returns to step S2'.
When i is 1, the modulation depth MD1The contrast image, which is a contrast imaging parameter, is an angiographic image.
Compared with the existing X-ray radiography method based on the digital subtraction technology, the X-ray radiography method based on the space-time filtering has the following characteristics and beneficial effects.
1) The imaging parameter adopted by the invention is the modulation depth, namely the ratio of the characteristic frequency signal of contrast agent perfusion to the 0-frequency direct current signal, the algorithm is simpler, and even if the contrast effect in the original contrast image sequence is not ideal due to low radiation dose or low contrast agent concentration, the modulation depth is hardly influenced, so that low radiation dose imaging can be realized, and the resolution is higher compared with the traditional digital subtraction radiography method. According to the characteristics, the angiography method is not only suitable for synchrotron radiation X-ray light sources and hospital radiological imaging equipment, but also suitable for laboratory X-ray light tube light sources.
2) In the angiography method, subtraction operation is not needed, but the characteristic signal of contrast agent perfusion and the characteristic signal of biological motion noise are distinguished from each other in a frequency domain, so that the obtained angiography image only contains blood vessel information, and the motion information of background tissues can be imaged independently. Therefore, inter-frame registration is not needed, the complexity of the algorithm is greatly reduced, the imaging signal-to-noise ratio of the angiography method is higher, the elimination of the motion artifact is adaptively realized, and the method is more suitable for being applied to X-ray angiography imaging with obvious motion artifacts.
3) The imaging parameter of the angiography method is the modulation depth MDiThe method makes full use of the information of the whole contrast agent perfusion sequence, so that the high-quality blood vessel integral structure can be obtained without using an image fusion technology. The image fusion effect can be automatically realized under the condition that the distribution and dispersion of the contrast agent in the blood vessel are discontinuous.
4) The invention can provide an angiographic image sequence of a contrast agent perfusion process, and is particularly suitable for application in X-ray angiographic imaging with significant motion artifacts. The dynamic radiograph reading and playing of the angiography image sequence has continuous and clear dynamic pictures and stable and non-drifting blood vessel tracks. Because the signal at the equilibrium location of the blood vessel is more dominated by the step signal, the signal at the non-equilibrium location is more dominated by the period-like signal. This is easily distinguished in the frequency domain, which does not interfere with the extraction of contrast agent perfusion information. Thus, even if the blood vessel in the sample is periodically displaced, the blood vessel in the image is stabilized at the equilibrium position of the blood vessel without searching for a corresponding mask.
Drawings
FIG. 1 is an image of one frame of an original contrast image sequence according to one embodiment of the present invention.
Fig. 2A-2C are graphs of time series signals corresponding to 0Hz, 1 × Fs/N Hz, 3 × Fs/NHz of the blood vessel feature points v shown in fig. 1.
FIG. 2D-FIG. 2F are graphs of time series signals corresponding to point b shown in FIG. 1, 0Hz, 1 x Fs/N Hz, and 3 x Fs/N Hz.
Fig. 3A-3B are contrast images of the original contrast image sequence shown in fig. 1 with MD1 and MD2 as imaging parameters.
Fig. 4 is a conventional time subtraction angiography image.
Detailed Description
The present invention will be further described with reference to the following specific examples. It should be understood that the following examples are illustrative only and are not intended to limit the scope of the present invention.
According to one embodiment of the invention, the X-ray radiography method of the invention is based on spatial-temporal filtering, and specifically comprises:
step S1: acquisition of raw contrast data. Similar to the conventional X-ray digital subtraction angiography method, under the irradiation of an X-ray light source, an original contrast image sequence of tens to hundreds of frames is acquired by perfusing a contrast medium to a contrast object from a start frame before the non-perfusion of the contrast medium to an end frame after the perfusion of the contrast medium.
In this embodiment, the subject is an experimental mouse, and an angiographic tube is inserted into a bifurcation between an External Carotid Artery (ECA) and a Common Carotid Artery (CCA) during preparation of an animal, wherein the angiographic region is a mouse brain; contrast agent 180. mu.l, concentration 280mgIml‐1Iobipiro (Iopamiro) contrast agents. Wherein the contrast agent is usedThe amount can be regulated within 100-500 microliter range according to biological sample, imaging region, X-ray photon flux, etc., and the concentration of contrast agent can be regulated within 150-350 mgI ml‐1Specific adjustment within the range. The iobipiro contrast agent is a non-ionic type iodine salt contrast agent, and can be replaced by other common non-ionic type iodine salt contrast agents, such as one of iohexol, iopromide, ioversol, iodixanol, diatrizoate and the like. The angiography tube is injected into an External Carotid Artery (ECA), although the contrast ratio is improved by the contrast agent, the contrast agent and the biological sample tissues are interlaced and superposed together under the condition of no subtraction operation due to the influence of bones or other thick soft tissues of the biological sample, and the contrast ratio of the acquired angiography image sequence is relatively low due to the influence of factors such as low concentration of the contrast agent, low photon flux of X-rays and low detection efficiency of a detector; the perfusion rate was controlled by a mini-syringe pump model LSP 01-1A manufactured by Baoding Lange constant flow Pump, Inc. (Longerpump), with an injection rate of 133.3 μ ls‐1. The X-ray source can be synchrotron radiation X-ray, laboratory X-ray machine, hospital radiological medical imaging equipment, in this embodiment, the X-ray source is a laboratory light source of Shanghai synchrotron radiation light source BL13W laboratory station; the energy of the X-ray is determined by the type of the contrast agent used, and generally is slightly higher than the absorption edge of the contrast agent, and in this embodiment, the energy of the X-ray is 33.3 keV; the photon flux is determined according to the operating state of the X-ray device, the size of the acceptable radiation dose for imaging and other factors, and the imaging quality can be correspondingly improved by improving the photon flux within the acceptable radiation dose range. Photon flux at 109—1010photonssecond‐1mm‐2All within the range, the photon flux is 2.38 × 1010 photossecond in the present embodiment‐1mm‐2. The detector is a PCOX ray CCD (pixel size 6.5X 6.5 μm, field size 13X 13mm, PCO-TECH Inc, Germany); the detector exposure time is 10 milliseconds, and the acquisition frame rate, i.e., the sampling frequency Fs, is 100 fps.
Furthermore, the raw data may also be obtained by truncating a portion of the sequence corresponding to the contrast agent perfusion from a conventional time-subtracted angiographic sequence.
With respect to the contrast image sequence obtained in step S1, the blood vessel portion and the background tissue portion have different time-series variation characteristics due to the absorption of X-rays by the contrast agent. The background tissue is ideally a series of static signals, but in practice it is a group of periodic-like oscillation signals or other types of clutter in time sequence due to unavoidable respiration, heartbeat, and other tissue motion, vascular stress to contrast agents, and vibration of external equipment. The corresponding vessel region, due to the perfusion of the contrast agent, should ideally be a set of signals that varies with the filling and concentration of the contrast agent. In practice, it is a set of periodic-like signals superimposed on a step signal. The signal corresponding to the vessel region is thus distinguished from the background tissue signal only in that the vessel signal has an additional step signal, which corresponds to the vessel information with the contrast agent perfusion.
Step S2: two characteristic points are selected on one frame of an original contrast image sequence, time sequence signals of the two characteristic points are extracted, and the time sequence signals are Fourier transformed into frequency domain signals.
One frame of the original contrast image sequence is shown in fig. 1. And selecting the following two feature points on the frame of the original contrast image sequence, wherein one of the two feature points selects a blood vessel region pixel point obviously positioned on a blood vessel as a blood vessel feature point v. And secondly, selecting background tissue pixel points with motion representativeness as background feature points b. The time sequence signals of the two characteristic points are extracted, and the signal at the background characteristic point b can be found to be the superposition of a periodic signal and a static direct current signal. The signal at the vessel feature v is superimposed on a step signal like a static dc signal and a partially periodic signal. And then performing Fourier frequency domain transformation on the two extracted time sequence signals by adopting a Fast Fourier Transform (FFT) method.
Setting the sampling frequency of the frequency domain signal obtained by the method as Fs; the number of samples is the same as the number of frames,setting to be N; then, the frequency of 0Hz in the frequency domain corresponds to Direct Current (DC) signals, and the frequencies of 1 × Fs/N Hz, 2 × Fs/N Hz, 3 × Fs/N Hz, …, i × Fs/N Hz, … (N-1) × Fs/N Hz correspond to high frequency Alternating Current (AC) signals. The corresponding module value of the blood vessel characteristic point v at the frequency i x Fs/N Hz is AviI.e. the model value corresponding to the blood vessel characteristic point v comprises Av1、Av2、Av3、…Avi…AvN‐1(ii) a The corresponding mode value of the background characteristic point b at the frequency i x Fs/N Hz is AbiThat is, the module value corresponding to the background feature point b includes Ab1、Ab2、Ab3、…Abi…AbN‐1。
Step S3: and (3) carrying out characteristic analysis on the frequency domain signals of the blood vessel characteristic point v and the background characteristic point b, wherein the module value difference of the two characteristic points b and v under the same frequency changes along with the change of the frequency, sequentially selecting 2 module value differences from large to small, and taking the corresponding frequency as the characteristic frequency.
In this embodiment, only 2 frequencies with the largest module value difference are selected as the characteristic frequencies, where the two characteristic frequencies x1 × Fs/N Hz are 1 × Fs/N Hz and x2 × Fs/N Hz is 3 × Fs/N Hz, where the characteristic frequency corresponding to the largest module value difference is represented by x1 × Fs/N Hz, and has a size of 1 × Fs/N Hz, and corresponds to a blood vessel; the characteristic frequency corresponding to the next largest module difference is represented by x2 Fs/N Hz, and is 3 Fs/N Hz, corresponding to the moving background tissue.
The difference in the modulus is related to, among other things, the intensity of the X-rays, the exposure time of the detector, the contrast agent dose, and the thickness of the sample type. In practice, the original contrast image sequence with better image quality is acquired as much as possible under the acceptable radiation dose of the sample, and the module values generally have obvious distinguishable differences, namely, the module values are 5-20 times. Therefore, in general, the frequencies corresponding to 2-4 modulo differences whose modulo difference is greater than 5 times the modulo value may be sequentially used as the characteristic frequencies (the frequency corresponding to the ith largest modulo difference is used as the characteristic frequency xi x Fs/N Hz), rather than being limited to 2, so as to easily distinguish the blood vessel signal from the background tissue signal. The characteristic frequency x1 x Fs/N Hz corresponding to the maximum module value difference is the optimal frequency of the contrast agent perfusion signal, and the module value corresponding to the blood vessel characteristic point v at the frequency is certainly larger than the module value corresponding to the background characteristic point b; the other characteristic frequencies often correspond to motion noise or non-optimal vessel signals or superposition of the motion noise and the non-optimal vessel signals, and the difference between the module value corresponding to the vessel characteristic point v and the module value corresponding to the background characteristic point b is relatively small and has an uncertain magnitude relation.
In most cases, the contrast agent is a continuous perfusion with a constant concentration, so that in the original sequence of contrast images, the signal of the characteristic point v of the blood vessel corresponds to a step signal with a characteristic frequency of 1 x Fs/N Hz. If the contrast agent is perfused in a large or intermittent concentration change or has special perfusion characteristics in some local special blood vessels, the contrast agent perfusion signal is not a step signal, and the characteristic frequency of the blood vessels may be other frequencies.
Step S4: under the condition that i is 1 or 2, traversing all pixel points of an imaging area in an original imaging image sequence, performing Fourier band-pass filtering processing on a time sequence signal of each pixel point by taking 0Hz and one of characteristic frequencies as references, and calculating an imaging parameter of each pixel point, namely Modulation Depth (MD)i)。
If there are many spurious noises in the original image of the contrast image sequence, denoising processing, such as median denoising, gaussian denoising, wiener denoising, may be performed on each frame image before traversing all the pixel points of the contrast region, i.e., before calculating the contrast imaging parameters of each point.
Taking any pixel point s (x, y) as an example, the contrast imaging parameters of the point are calculated. Firstly, a time sequence signal of the pixel point is taken out, and Fourier band-pass filtering processing is carried out on the time sequence signal by taking 0Hz, x1 x Fs/N Hz and x2 x Fs/N Hz as references. The signal after bandpass filtering is S0(x,y,t)、Sx1(x,y,t)、Sx2(x, y, t). Wherein S is0(x, y, t) is a DC signal, Sx1(x,y,t)、Sx2And (x, y, t) is an alternating current signal with corresponding frequency. For example, in this embodiment, the time series signals of the blood vessel feature point v and the background feature point b are subjected to fourier band-pass filtering processing with reference to 0Hz, x1 × Fs/N Hz, and x2 × Fs/N Hz. The v point signal after the band-pass filtering is Sv0(x,y,t)、Svx1(x,y,t)、Svx2(x, y, t) wherein S is as shown in FIG. 2av0(x, y, t) is a direct current signal; as shown in FIGS. 2b and 2c, the AC signal with the corresponding frequency is Svx1(x,y,t)、Svx2(x, y, t). The b point signal after the band-pass filtering is Sb0(x,y,t)、Sbx1(x,y,t)、Sbx2(x, y, t) wherein, as shown in FIG. 2d, Sb0(x, y, t) is a direct current signal. As shown in FIGS. 2e and 2f, the AC signal with the corresponding frequency is Sbx1(x,y,t)、Sbx2(x, y, t). The alternating signal corresponds to a perfusion signal of the contrast agent or a motion noise signal of the background tissue.
Then, calculating the output contrast imaging parameter of the pixel point as Modulation Depth (MD)i) The formula is as follows:
Figure BDA0001595324360000091
Mean(abs(Sxi(x, y, t))) is the average of the absolute values of the alternating signal corresponding to the frequency xi x Fs/N Hz. Mean (abs (S)0(x, y, t))) is the average of the absolute values of the dc signals. Denominator Mean (abs (S)0(x, y, t))) to normalize different regions of the image. This is due to inhomogeneity of the incident X-ray and the differences in density, thickness, and elemental distribution of different regions of the contrast image. The pixel points in different areas have pixel value distribution with larger difference. Resulting in different areas of the blood vessels with different contrast effects. Not conducive to overall angiographic imaging. Therefore, normalization with a dc signal can remove the effect of this factor.
When the total number of eigenfrequencies obtained in step S3 is not 2, i in step S4 may be any number from 1 to the total number of eigenfrequencies, and is not limited to 1 or 2.
Step S5: output in MD1、MD2Is a contrast image of the imaging parameter.
Fig. 3a and 3b show contrast images with MD1 and MD2 as imaging parameters in the present embodiment. Fig. 3a mainly contains blood vessel information, which corresponds to the characteristic frequency corresponding to the maximum module value difference, i.e. x1 × Fs/NHz. This portion of the signal corresponds to the perfusion signal of the vessel, and the signal intensity of this component in the vessel region pixels is 10 times that of its background tissue. Fig. 3b shows mainly moving background tissue, corresponding to the characteristic frequency x2 x Fs/N Hz corresponding to the next largest module difference, the partial signal is mainly respiratory motion. This example successfully distinguishes between vascular signals and motion noise.
Generally, the time-series signal of the pixel of the blood vessel region in the contrast image is ideally a step signal due to the perfusion of the contrast agent, and the main characteristic signal after frequency-domain filtering is distributed at the characteristic frequency x1 x Fs/N Hz corresponding to the maximum module value difference, and correspondingly, the main characteristic of the time-series signal of the background region in the contrast image is distributed at other frequencies, which have a certain difference from the blood vessel perfusion frequency. The method can easily distinguish the contrast agent signal from the motion noise of the background tissue, thereby achieving the effect of inhibiting the noise and improving the imaging resolution. Therefore, in these contrast images, the MD is general1For angiographic images, MD2High frequency moving tissue images of corresponding frequencies, such as moving tissue images affected by breathing, moving tissue images affected by muscle twitches, and the like. This demonstrates that the method of the present invention can easily distinguish between contrast agent signals and background tissue motion noise that are indistinguishable by conventional digital subtraction techniques. In contrast, this example shows a conventional time-subtracted angiographic image as shown in fig. 4, where the contrast agent signal and the motion noise of the background tissue are far less sharp than in the present invention and are also indistinguishable.
In addition, the invention also provides a dynamic radiographing method based on X-ray radiography, which comprises the following steps:
step S1': taking i-1, the characteristic frequency X1 Fs/NHz was obtained using the X-ray imaging method described above, and obtained in MD1Taking an angiography image which is an angiography imaging parameter as an angiography image corresponding to the previous N frames of the original angiography image sequence, and taking j as 1;
step S2': selecting the first N-j frames of the original contrast image sequence as the current contrast image sequence, and calculating the characteristic frequency xi x Fs/(N-j) Hz corresponding to the current contrast image sequence according to the characteristic frequency obtained in the step S1'; the characteristic frequency is replaced by x1 Fs/(N-j) Hz from x1 Fs/N Hz, respectively, because the characteristic frequency changes correspondingly in time series signals of different lengths corresponding to the step signal of the contrast agent signal.
Step S3': repeating steps S4 and S5 of the X-ray contrast method described above, specifically including: traversing all pixel points of the contrast region in the current contrast image sequence, performing Fourier band-pass filtering processing on the time sequence signal of each pixel point by taking 0Hz and the characteristic frequency x1 x Fs/(N-j) Hz obtained in the current step S2' as the reference, and calculating the modulation depth MD of each pixel point according to the obtained signal1(ii) a Is obtained to modulate the depth MD1The angiographic image which is the imaging parameter of the angiography is taken as an angiographic image corresponding to the first N-j frame of the original angiographic image sequence;
step S4': when N-j is N, sequentially playing the contrast images corresponding to the previous N, N +1 …, N-1, N-2 and N frames; otherwise, j +1 is updated to j and returns to step S2'. Where n is the number of image frames that just perfuse the contrast agent into the contrast region.
Therefore, the dynamic radiograph reading playing in the radiography process can be realized, the dynamic picture is continuous and clear, and the blood vessel track is stable and does not shake.
The above embodiments are merely preferred embodiments of the present invention, which are not intended to limit the scope of the present invention, and various changes may be made in the above embodiments of the present invention. All simple and equivalent changes and modifications made according to the claims and the content of the specification of the present application fall within the scope of the claims of the present patent application. The invention has not been described in detail in order to avoid obscuring the invention.

Claims (5)

1. An X-ray imaging method comprising:
step S1: acquiring an original contrast image sequence of N frames by using sampling frequency Fs, and taking the original contrast image sequence as a current contrast image sequence;
step S2: selecting a blood vessel characteristic point and a background characteristic point on one frame of an original contrast image sequence, extracting time sequence signals of the two characteristic points and Fourier transforming the time sequence signals into frequency domain signals;
step S3: carrying out feature analysis on the frequency domain signals of the blood vessel feature points and the background feature points, sequentially selecting at least two maximum module value differences of the two feature points under the same frequency, and taking the frequency corresponding to the ith maximum module value difference as the feature frequency xi x Fs/N Hz;
step S4: traversing all pixel points of the contrast region in the current contrast image sequence, carrying out Fourier band-pass filtering processing on the time sequence signal of each pixel point by taking 0Hz and one of the characteristic frequencies xi x Fs/N Hz as the reference, and calculating the modulation depth MD of each pixel point through the obtained signali；
Step S5: outputting to modulate depth MDiAn angiographic image being an angiographic imaging parameter;
wherein xi is a positive integer;
the modulation depth MD of each pixel pointiCalculated by the following formula:
Figure FDA0003271010630000011
wherein S is0(x,y,t)、Sxi(x, y, t) are respectively a direct current signal obtained after Fourier band-pass filtering processing is carried out on the time sequence signal of each pixel point and an alternating current signal corresponding to xi x Fs/N Hz frequency; mean (abs (S)xi(x, y, t))) is the average of the absolute values of the alternating signals corresponding to the frequencies xi x Fs/N Hz; mean (abs (S)0(x, y, t))) is the average of the absolute values of the dc signals.
2. The X-ray imaging method according to claim 1, wherein the modulo differences selected in step S3 are 2-4 and the number of eigenfrequencies is equal to the number of modulo differences.
3. The X-ray contrast method according to claim 1, wherein the step S4 further includes: before traversing all pixel points of the contrast region, denoising each frame image, wherein the denoising comprises median denoising, Gaussian denoising or wiener denoising.
4. A dynamic radiographing method based on X-ray contrast comprises the following steps:
step S1': using the X-ray imaging method according to claim 1, the characteristic frequencies xi X Fs/NHz are obtained and the modulation depth MD is obtainediTaking j as 1 as a contrast image corresponding to the first N frames of the original contrast image sequence, wherein j is a contrast image of a contrast imaging parameter;
step S2': selecting the first N-j frames of the original contrast image sequence as a current contrast image sequence, and calculating the characteristic frequency xi x Fs/(N-j) Hz corresponding to the current contrast image sequence according to the characteristic frequency obtained in the step S1';
step S3': traversing all pixel points of the contrast region in the current contrast image sequence, performing Fourier band-pass filtering processing on the time sequence signal of each pixel point by taking 0Hz and the characteristic frequency xi x Fs/(N-j) Hz obtained in the step S2' as the reference, and calculating the modulation depth MD of each pixel point according to the obtained signali(ii) a Is obtained to modulate the depth MDiThe contrast image which is a contrast imaging parameter is used as a contrast image corresponding to the first N-j frame of the original contrast image sequence;
step S4': when N-j is N, sequentially playing the contrast images corresponding to the previous N, N +1 …, N-1, N-2 and N frames; otherwise, j +1 is updated to j and returns to step S2'.
5. The dynamic radiograph interpretation method based on X-ray contrast, according to claim 4, wherein when i is 1, the modulation depth MD is used1The contrast image, which is a contrast imaging parameter, is an angiographic image.
CN201810203889.1A 2018-03-13 2018-03-13 X-ray radiography method and dynamic radiograph reading method based on X-ray radiography Active CN108852385B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810203889.1A CN108852385B (en) 2018-03-13 2018-03-13 X-ray radiography method and dynamic radiograph reading method based on X-ray radiography

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810203889.1A CN108852385B (en) 2018-03-13 2018-03-13 X-ray radiography method and dynamic radiograph reading method based on X-ray radiography

Publications (2)

Publication Number Publication Date
CN108852385A CN108852385A (en) 2018-11-23
CN108852385B true CN108852385B (en) 2022-03-04

Family

ID=64326063

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810203889.1A Active CN108852385B (en) 2018-03-13 2018-03-13 X-ray radiography method and dynamic radiograph reading method based on X-ray radiography

Country Status (1)

Country Link
CN (1) CN108852385B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1141156A (en) * 1995-05-22 1997-01-29 株式会社岛津制作所 Digital angiographic apparatus
US6005917A (en) * 1995-04-05 1999-12-21 Andersson; Mats Velocity adaptive filtered angiography
CN101052992A (en) * 2004-07-27 2007-10-10 杜尔牙科器械两合公司 Method and device for improving perceptibility different structures on radiographs
CN101236647A (en) * 2007-12-07 2008-08-06 华中科技大学 Digital blood vessel contrast image enhancement method integrating context information
CN101999885A (en) * 2010-12-21 2011-04-06 中国人民解放军国防科学技术大学 Endogenous optical imaging method for automatically separating arteries and veins
CN104517301A (en) * 2014-12-30 2015-04-15 华中科技大学 Method for iteratively extracting movement parameters of angiography image guided by multi-parameter model

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6983182B2 (en) * 1998-04-10 2006-01-03 Wisconsin Alumni Research Foundation Time resolved computed tomography angiography

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6005917A (en) * 1995-04-05 1999-12-21 Andersson; Mats Velocity adaptive filtered angiography
CN1141156A (en) * 1995-05-22 1997-01-29 株式会社岛津制作所 Digital angiographic apparatus
CN101052992A (en) * 2004-07-27 2007-10-10 杜尔牙科器械两合公司 Method and device for improving perceptibility different structures on radiographs
CN101236647A (en) * 2007-12-07 2008-08-06 华中科技大学 Digital blood vessel contrast image enhancement method integrating context information
CN101999885A (en) * 2010-12-21 2011-04-06 中国人民解放军国防科学技术大学 Endogenous optical imaging method for automatically separating arteries and veins
CN104517301A (en) * 2014-12-30 2015-04-15 华中科技大学 Method for iteratively extracting movement parameters of angiography image guided by multi-parameter model

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Template-based CTA to x-ray angio rigid registration of coronary arteries in frequency domain with automatic x-ray segmentation;Timur Aksoy and Gozde Unal;《Medical physics》;20130911;第40卷(第10期);101903-1至101903-15 *

Also Published As

Publication number Publication date
CN108852385A (en) 2018-11-23

Similar Documents

Publication Publication Date Title
US7840255B2 (en) X-ray CT apparatus and myocardial perfusion image generating system
US8498463B2 (en) Mask construction for cardiac subtraction
US8299413B2 (en) Method for pixel shift calculation in digital subtraction angiography and X-ray diagnostic imaging system for generating images in digital subtraction angiography
US9070181B2 (en) System and method for extracting features of interest from an image
Sone et al. Development of a high-resolution digital tomosynthesis system and its clinical application.
US10258301B2 (en) Digital subtraction angiography
US10736594B2 (en) Data-based scan gating
JP2006289094A (en) Method and system for processing noise of prepared image
US20140105477A1 (en) Low dose cardiac ct imaging with time-adaptive filtration
US8897525B2 (en) Multisegment picture reconstruction for cardio CT pictures
Zhao et al. Dual-energy CT imaging using a single-energy CT data is feasible via deep learning
Yan et al. Image quality of automatic coronary CT angiography reconstruction for patients with HR≥ 75 bpm using an AI-assisted 16-cm z-coverage CT scanner
CN111915538B (en) Image enhancement method and system for digital blood vessel subtraction
Kruger et al. Dynamic tomographic digital subtraction angiography using temporal filtration.
CN108852385B (en) X-ray radiography method and dynamic radiograph reading method based on X-ray radiography
JP4864909B2 (en) Image processing device
JP5229496B2 (en) Image processing device
Al-Ameen et al. Prevalent degradations and processing challenges of computed tomography medical images: A compendious analysis
JP2009054013A (en) Image processor
Kim et al. Evaluation of Ultra-low-dose (ULD) Lung Computed Tomography (CT) Using Deep-learning: A Phantom Study
CN117547293A (en) Four-dimensional angiography imaging method based on three-dimensional dynamic data and frequency domain filtering
Taguchi et al. Perfusion from angiogram and a priori (PAP) with temporal regularization
Weir et al. Contrast-to-Noise Ratio Improvements with AIDR 3D
Umarani et al. Bio Image Processing for Cardiac Image Using Virtual Instrumentation
Umarani et al. Bio Image Processing For Cardiac Image Using Virtual

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant