Angiography image processing method based on dynamic sequence inter-frame difference

Abstract

The invention relates to an angiography image processing method based on dynamic sequence inter-frame difference, which comprises the following steps of: calculating an inter-frame difference image of a current frame It and a previous frame It-1 by inputting dynamic sequence angiography image frames I0, I1... It; and then, calculating a reference difference image of the current frame It and the initial frame I0, and carrying out threshold noise reduction processing on the Dt (x, y) sum to obtain a denoised frame difference image D't and finally synthesize a current frame difference blood vessel image. Therefore, the method has the dose advantage, the single scanning dose is remarkably reduced compared with a traditional scheme, and the vascular signal-to-noise ratio is remarkably increased. Real-time image display can be achieved, acceleration is conducted through a GPU in DSA equipment, the processing speed can reach 30 ms / frame, and the requirement for real-time navigation of interventional operation is met.

Description

Technical Field

[0001] The present invention relates to an image processing method, in particular to an angiography image processing method based on the difference between dynamic sequence frames. Background Art

[0002] As for the existing angiography image processing methods, the traditional vascular subtraction is mainly used to calculate the "maximum filling display" in the blood vessels. t ras t ) registration and subtraction.

[0003] The MIP / MinIP calculation formula is as follows:

[0004]

[0005] or

[0006] in: During implementation, Max or Min calculation needs to be used depending on the type of contrast agent.

[0007] This approach has the following flaws during implementation:

[0008] 1. Differentiating contrast media types: When calculating the maximum vascular filling indication based on the contrast media type, choose MIP / MinIP. This requires experienced personnel to make timely judgments based on the current data, requiring a high level of manual intervention and experience. If the operating personnel are required to participate in the process, it can easily distract them and affect the final treatment outcome.

[0009] 2. Mask dependency: The mask acquisition process requires X-ray imaging, which increases the complexity of the system. Consequently, implementation costs are high and the processing is complex.

[0010] 3. MIP / MinIP requires noise accumulation calculation: During this process, the maximum / minimum density projection calculation is involved, and noise accumulates. To achieve better results, a high signal-to-noise ratio image input is required, resulting in a high radiation dose.

[0011] 4. Low utilization of dynamic information: Single-frame mask cannot utilize time series features and has insufficient contrast of small blood vessels.

[0012] In light of these limitations, our authors have actively researched and innovated, aiming to develop an angiography image processing method based on inter-frame differences in dynamic sequences. This method can directly extract the dynamic filling process of blood vessels, avoiding the potential artifacts introduced by mask acquisition and eliminating the radiation dose required. This method can achieve frame-by-frame noise reduction before synthesizing vascular structures, significantly improving the signal-to-noise ratio of images of maximum blood vessel filling. Summary of the Invention

[0013] In order to solve the above technical problems, the purpose of the present invention is to provide a method for processing angiography images based on the difference between frames of a dynamic sequence.

[0014] The present invention provides a method for processing angiographic images based on dynamic sequence frame differences, comprising the following steps:

[0015] Step 1: Input dynamic sequence angiography image frames I0, I1...I t , where I0 is the starting frame and t is the frame number;

[0016] Step 2: Calculate the current frame I t With the previous frame I t-1 Inter-frame difference image: D t (x,y)=|I t (x,y)-I t-1 (x,y)|, where (x,y) represents the pixel coordinates, I t (x,y) is the grayscale value of the pixel at coordinate (x,y) in the tth frame;

[0017] Step 3: Calculate the current frame I t The baseline difference image with the starting frame I0: Where I0(x,y) is the grayscale value of the pixel at the coordinate (x,y) of the starting frame;

[0018] Step 4: D t (x,y) and Perform threshold noise reduction processing to obtain the noise-reduced frame difference image D′ t and

[0019] Step 5: Synthesize the current frame differential blood vessel image: Among them, α t , β t D′ t and The weight coefficient, DV t is the vascular structure feature image of the current frame.

[0020] Furthermore, in the above-mentioned angiography image processing method based on the difference between frames of dynamic sequence, the threshold noise reduction processing in step 4 includes: statistically calculating D t (x,y) and noise level; setting a noise threshold according to noise statistics; and performing noise reduction on the difference image based on the noise threshold.

[0021] Furthermore, the above-mentioned angiography image processing method based on the difference between frames of a dynamic sequence further includes step six, calculating the maximum blood vessel filling image: Among them, k is the grayscale adjustment coefficient, which is used to adjust the grayscale of the blood vessel image to the display range, n is the final frame number, Indicates the image of maximum vascular filling.

[0022] Furthermore, in the above-mentioned angiography image processing method based on the difference between frames of a dynamic sequence, in step six, Can be converted into Can cache the maximum blood vessel maximum filling image of the previous frame Quickly calculate the current At the same time The cache is used for calculating the maximum filling image of the largest blood vessel in the next frame.

[0023] Furthermore, in the above-mentioned angiography image processing method based on the difference between frames of a dynamic sequence, in step six, the grayscale adjustment coefficient k is used to adjust The grayscale value is mapped to the visual range of the medical display device.

[0024] Furthermore, the above-mentioned angiography image processing method based on the difference between dynamic sequence frames further includes step seven, based on DV t Sequential calculation of time density curve;

[0025] Step 8: Generate the maximum blood flow time image based on the time density curve;

[0026] Step 9: Display the maximum blood flow time image as a blood flow color map through pseudo-color rendering.

[0027] Furthermore, in the above-mentioned angiography image processing method based on the difference between frames of a dynamic sequence, in step eight, the DV of all frames previously outputted are converted into t The results are cached, and the time density curve is calculated to obtain the maximum blood flow time image.

[0028] Furthermore, in the above-mentioned angiography image processing method based on the difference between frames of a dynamic sequence, the starting frame I0 is an image frame in the dynamic sequence in which the contrast agent has not reached the target blood vessel area.

[0029] Furthermore, in the above-mentioned angiography image processing method based on dynamic sequence frame difference, the dynamic sequence angiography image is an image acquired by an X-ray angiography device; the weight coefficient α t and β t Dynamic adjustment according to the dynamic filling stage of the blood vessels, increasing α in the initial filling stage of the contrast agent t value, increasing β in the stable filling stage t value.

[0030] By means of the above solution, the present invention has at least the following advantages:

[0031] 1. It has a dose advantage. The single scan dose is significantly reduced compared with the traditional scheme, and the vascular signal-to-noise ratio is significantly improved.

[0032] 2. Real-time image display can be achieved, and through GPU acceleration in the DSA device, the processing speed can reach 30ms / frame, meeting the real-time navigation requirements of interventional surgery.

[0033] 3. The implementation process is simple and easy to operate, avoiding the need to select different MIP / MinIP treatment plans based on the type of contrast agent, and eliminating the need for users to make comparative selections.

[0034] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention and implement it according to the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a schematic diagram of the implementation process of the angiography image processing method based on dynamic sequence frame differences.

[0036] Figure 2 The angiography image is obtained by the angiography image processing method based on the difference between dynamic sequence frames.

[0037] Figure 3 It is a diagram of the color atlas of blood flow. DETAILED DESCRIPTION

[0038] The following embodiments of the present invention are described in further detail with reference to the accompanying drawings and examples. The following examples are used to illustrate the present invention but are not intended to limit the scope of the present invention.

[0039] like Figures 1 to 2A method for processing angiographic images based on the difference between frames of a dynamic sequence is provided, which is distinctive in that it comprises the following steps:

[0040] Step 1: Input dynamic sequence angiography image frames I0, I1...I t , where I0 is the starting frame and t is the frame number. In the initial stage, the starting frame I0 is the image frame in the dynamic sequence where the contrast agent has not reached the target blood vessel area. It is not processed and can be cached internally.

[0041] Step 2: Calculate the current frame I t With the previous frame I t-1 Inter-frame difference image: D t (x,y)=|I t (x,y)-I t-1 (x,y)|. Where (x,y) represents the pixel coordinates, I t (x, y) is the grayscale value of the pixel at the coordinate (x, y) of the t-th frame. During implementation, the pixel grayscale value can be selected from the current frame image I t With the previous frame image I t-1 Calculate the absolute value of the corresponding pixel value difference point by point.

[0042] Step 3: Calculate the current frame I t The baseline difference image with the starting frame I0: Among them, I0(x,y) is the grayscale value of the pixel at the coordinate (x,y) of the starting frame. It can be understood that the current frame image I t Calculate the absolute value of the corresponding pixel value difference with the starting frame image I0 point by point.

[0043] That is to say, starting from I1 to the last frame, each frame I t , respectively calculate: (1) and the previous frame I t_1 The difference between (absolute value): D t (x,y)=|I t (x,y)-I t-1 The difference (absolute value) between (x,y)|.(2) and the first frame I0: In short, during the implementation of the present invention, the current frame I is used t With the previous frame I t-1 Inter-frame difference image D t and the current frame I t The baseline differential image with the starting frame I0 Perform frame-by-frame difference analysis of dynamic sequences. More differential data can be introduced here: such as the current frame I t With the previous i(i>0 and i< t -1) Frame I t-iThis will increase the computational complexity and the difficulty of setting the superposition weight coefficient.

[0044] Step 4: D t (x,y) and Perform threshold noise reduction processing to obtain the noise-reduced frame difference image D′ t and During this period, the threshold noise reduction process includes: statistical D t (x,y) and The noise level is set according to the noise statistics, and the noise threshold is set based on the noise threshold to perform noise reduction on the difference image. During the actual implementation, the noise reduction process may include a combination of threshold noise reduction and directional filtering, or a dynamic threshold adjustment (based on a local window), etc., which will not be described in detail here. In short, based on the D t (x,y) and Threshold noise reduction is performed to suppress the current frame difference image noise. While threshold noise reduction is employed to ensure the real-time performance of the entire algorithm, more approaches, or combinations of approaches, can be employed to achieve even better noise reduction results. For example, combining threshold noise reduction with directional filtering, or employing dynamic threshold adjustment (based on a local window), all of these different noise reduction schemes can be applied in the present invention and will not be detailed here.

[0045] Step 5: Synthesize the current frame differential blood vessel image: Among them, α t , β t D′ t and The weight coefficient, DV t is the characteristic image of the vascular structure of the current frame. In conjunction with a preferred embodiment of the present invention, the weight coefficient is related to the relative acquisition frequency of blood flow (the difference in contrast agent between two adjacent frames). t =0.5,β t =0.5 can achieve the ideal effect. In the case of slow blood flow or high acquisition frame rate, α can be reduced t Increase β t Get better results, and vice versa. Figure 2 In the actual processing of angiographic images, α t =0.5,β t = 0.5 parameters for calculation. During the implementation, the present invention adopts a linear calculation method. Of course, other methods can also be used to synthesize the current frame differential blood vessel image, such as similar I will not go into details here.

[0046] Step 6: Calculate the maximum blood vessel filling image: Among them, k is the grayscale adjustment coefficient, which is used to adjust the grayscale of the blood vessel image to the display range, n is the final frame number, Indicates the maximum filling image of blood vessels. During implementation, in order to facilitate data processing, Can be converted into This means that the maximum blood vessel filling image of the previous frame can be cached Then, quickly calculate the current At the same time, The cache is used for calculating the maximum filling image of the largest blood vessel in the next frame.

[0047] Step 7, based on DV t Sequential calculation of time density curves.

[0048] Step 8: Generate the maximum blood flow time image based on the time density curve. During this period, the DV of all frames previously output t The results are cached, and the time density curve is calculated to obtain the maximum blood flow time image.

[0049] Step 9: Display the maximum blood flow time image as a blood flow color map through pseudo color rendering. The final image is as follows Figure 3 Of course, when the final result is displayed, grayscale flipping and brightness and contrast adjustment are performed as needed. These operations belong to common image display processing and will not be described in detail here.

[0050] In short, during the implementation of the present invention, the present invention and the built-in program of the DSA device are combined to establish the following processing steps to achieve convenient operation.

[0051] Input a frame of angiography data: image data collected by the DSA device during the angiography phase.

[0052] Calculate the difference from the starting frame:

[0053] Calculate the difference with the previous frame: D t (x,y)=|I t (x,y)-I t_1 (x,y)|.

[0054] Noise suppression: The noise level of the difference image is statistically analyzed, and the noise threshold is set to achieve noise reduction (threshold noise reduction processing), and the noise-reduced frame difference image D′ is obtained. t and

[0055] Differential superposition: The differential blood vessel image corresponding to the current frame.

[0056] Differential vascular image: Differential vascular image cache of all frames, used to calculate time density curve, requires input of frame rate information (for drawing time axis).

[0057] Differential accumulation output: Right now The current differential vascular image is superimposed with the differential accumulation of the previous frame.

[0058] "Maximum filling display" image: that is, differential accumulation output. It is necessary to temporarily save the maximum filling image output of the current frame for differential accumulation input of the next frame.

[0059] Calculate the density time curve: Based on the differential vascular image sequence, interpolate on the time axis to find the time corresponding to the maximum density and obtain the maximum blood flow time image.

[0060] “Blood flow color atlas”: Displays the “maximum blood flow time image” as a “blood flow color atlas” using pseudo-color.

[0061] In addition, the indicated orientations or positional relationships described in the present invention are all based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or structure referred to must have a specific orientation or operate with a specific orientation structure. Therefore, it cannot be understood as a limitation on the present invention.

[0062] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for processing angiographic images based on the difference between frames of a dynamic sequence, characterized in that The following steps are involved: Step 1: Input dynamic sequence angiography image frames I0, I1...I t , where I0 is the starting frame and t is the frame number; Step 2: Calculate the current frame I t With the previous frame I t-1 Inter-frame difference image: D t (x,y)=|I t (x,y)-I t-1 (x,y)|, where (x,y) represents the pixel coordinates, I t (x,y) is the grayscale value of the pixel at coordinate (x,y) in the tth frame; Step 3: Calculate the current frame I t The baseline difference image with the starting frame I0: Where I0(x,y) is the grayscale value of the pixel at the coordinate (x,y) of the starting frame; Step 4: D t (x,y) and Perform threshold noise reduction processing to obtain the noise-reduced frame difference image D' t and Step 5: Synthesize the current frame differential blood vessel image: Among them, α t , β t D' t and The weight coefficient, DV t is the vascular structure feature image of the current frame.

2. The angiography image processing method based on dynamic sequence frame difference according to claim 1, characterized in that: The threshold noise reduction process in step 4 includes: t (x,y) and noise level; setting a noise threshold according to noise statistics; and performing noise reduction on the difference image based on the noise threshold.

3. The angiography image processing method based on dynamic sequence frame difference according to claim 1, characterized in that: The sixth step is to calculate the maximum blood vessel filling image: Among them, k is the grayscale adjustment coefficient, which is used to adjust the grayscale of the blood vessel image to the display range, n is the final frame number, Indicates the image of maximum vascular filling.

4. The angiography image processing method based on dynamic sequence frame difference according to claim 3, characterized in that: In the step six, Can be converted into Able to cache the maximum blood vessel filling image of the previous frame Calculate the current At the same time The cache is used for calculating the maximum filling image of the largest blood vessel in the next frame.

5. The angiography image processing method based on dynamic sequence frame difference according to claim 3, characterized in that: In step 6, the grayscale adjustment coefficient k is used to adjust The grayscale value is mapped to the visual range of the medical display device.

6. The angiography image processing method based on dynamic sequence frame difference according to claim 1, characterized in that: Also includes step seven, based on DV t Sequential calculation of time density curve; Step 8: Generate the maximum blood flow time image based on the time density curve; Step 9: Display the maximum blood flow time image as a blood flow color map through pseudo-color rendering.

7. The angiography image processing method based on dynamic sequence frame difference according to claim 6, characterized in that: In the step eight, all the previously output DV t The results are cached, and the time density curve is calculated to obtain the maximum blood flow time image.

8. The angiography image processing method based on dynamic sequence frame difference according to claim 1, characterized in that: The starting frame I0 is an image frame in the dynamic sequence where the contrast agent has not yet reached the target blood vessel area.

9. The angiography image processing method based on dynamic sequence frame difference according to claim 1, characterized in that: The dynamic sequence angiography image is an image collected by an X-ray angiography device; the weight coefficient α t and β t Dynamic adjustment according to the dynamic filling stage of the blood vessels, increasing α in the initial filling stage of the contrast agent t value, increasing β in the stable filling stage t value.

Images