An electrical impedance tomography method for stroke monitoring

Abstract

The present invention discloses an electrical impedance tomography method for stroke monitoring, which relates to the field of image processing technology and includes the following steps: continuously collecting brain voltage data of a monitored object in multiple cerebral vascular pulsation cycles, performing time-difference imaging according to the brain voltage data, and obtaining multiple EIT images; obtaining a relative conductivity variation curve over time of each pixel point and a relative conductivity variation curve over time of a global pixel point based on multiple EIT images, obtaining a regression coefficient between the relative conductivity variation curve over time of each pixel point and the relative conductivity variation curve over time of a global pixel point, and generating a functional EIT image through multiple regression coefficients; generating an EIT perfusion image for each cerebral vascular pulsation cycle. The present invention generates a functional EIT image and an EIT perfusion image by analyzing the measurement data of the change process of the starting point and the end point of cerebral vascular diastole, and can be widely used in actual clinical environments.

Description

Technical Field

[0001] The invention relates to the technical field of image processing, and in particular to an electrical impedance tomography method for monitoring cerebral stroke. Background Art

[0002] Stroke, as a serious cerebrovascular disease, is divided into ischemic stroke and hemorrhagic stroke. When a stroke occurs, due to blood vessel blockage or rupture, part of the blood cannot flow to the brain, causing brain cells to gradually die due to lack of oxygen, further affecting the function of the brain area, and may eventually lead to serious disabilities such as hemiplegia, language disorders, and cognitive decline. The condition of stroke is complex and changeable. For patients with hemorrhagic stroke, rebleeding is a serious complication that may cause rapid deterioration of the condition. For patients with ischemic stroke, there is a risk of transformation to hemorrhage during antithrombotic treatments such as thrombolysis and intravascular therapy. This transformation may cause further damage to the originally ischemic brain tissue and even endanger life. Therefore, real-time monitoring of the development of stroke, early warning of secondary injury changes and real-time feedback of intervention effects are of great significance to the treatment of clinical patients. In the prior art, stroke is monitored by obtaining relevant monitoring parameters in the development of stroke.

[0003] Electrical impedance tomography (EIT) is a new medical imaging technology. This technology applies a set of safe excitation currents on the human body surface, measures the voltage changes on the corresponding electrodes simultaneously, and combines with reconstruction algorithms to generate images that can show the impedance distribution inside human tissues. EIT technology has significant advantages such as no radiation damage, high temporal resolution, and high cost-effectiveness of equipment. It has broad application prospects in the fields of chest imaging, brain function imaging, and abdominal organ imaging, and is attracting many research teams to engage in research and exploration in related fields. EIT technology is divided into two forms: static electrical impedance tomography and dynamic electrical impedance tomography. Static EIT aims to reconstruct the absolute impedance value distribution inside the human body. This technology has extremely high accuracy requirements for the boundary shape of the imaging site, the position of the electrode, and the contact impedance between the electrode and the skin. However, these key information are difficult to obtain accurately in practice, resulting in unsatisfactory imaging results. Dynamic EIT focuses on reconstructing the relative conductivity of the image from the two sets of measurement values. This method can compensate for errors and offsets in the data or reconstruction model that do not change with time. Dynamic EIT can work effectively even if the electrode boundary shape is not completely correct, so dynamic EIT is more widely used in current clinical applications and research.

[0004] In the international research field, researchers such as Manwaring and Sadlier have successfully established an intracranial hemorrhage model by injecting blood into the skull of piglets and rats, and used EIT technology to monitor the bleeding process. These research results show that EIT is sensitive to stroke changes and has high temporal resolution, and has the ability to monitor intracranial hemorrhage in real time. In China, researchers Xu Canhua and Dai Meng have also successfully monitored the process of intracerebral hemorrhage and subarachnoid hemorrhage in piglets, further verifying the application potential of EIT in real-time monitoring of intracranial hemorrhage. In addition, the research of Shi Xuetao and others has shown that EIT can quickly and sensitively detect potential cerebral ischemia. In addition, they have also demonstrated the possibility of using EIT to perform real-time imaging of focal cerebral infarction in rabbits.

[0005] Existing studies use dynamic EIT technology to perform differential imaging by comparing the data before and after the monitoring subject's brain hemorrhage or ischemia. Therefore, the existing technology needs to collect the data before and after the hemorrhage of the monitoring subject to obtain the EIT image, and the EIT image can be used to monitor the stroke. However, in actual clinical settings, the stroke of the monitored subject is often sudden, and it is usually impossible to obtain the data before the hemorrhage or ischemia of the monitored subject. Therefore, in actual clinical settings, EIT images for reference are often not generated. Summary of the invention

[0006] The present invention provides an electrical impedance tomography method for stroke monitoring, which solves the problem that the existing dynamic EIT technology is usually unable to obtain data before bleeding or ischemia of the monitored object when facing the challenges of actual clinical environments, so that the final imaging data does not match the actual situation.

[0007] The present invention provides an electrical impedance tomography method for stroke monitoring, comprising the following steps:

[0008] Continuously collecting brain voltage data of the monitored object in multiple cerebral vascular pulsation cycles, performing time-difference imaging based on the brain voltage data, and obtaining multiple EIT images; wherein the EIT image includes multiple pixel points;

[0009] Based on multiple EIT images, a relative conductivity variation curve of each pixel point and a relative conductivity variation curve of the global pixel point are obtained, a regression coefficient between the relative conductivity variation curve of each pixel point and the relative conductivity variation curve of the global pixel point is obtained, and a functional EIT image is generated through multiple regression coefficients;

[0010] For each cerebral vascular pulsation cycle, an EIT perfusion image is generated, and the value of a pixel point in the EIT perfusion image is the relative conductivity between the start point and the end point of the cerebral vascular diastolic period.

[0011] Preferably, the brain voltage data of the monitored object during multiple cerebral vascular pulsation cycles are continuously collected through multiple pairs of electrodes.

[0012] Preferably, the time-difference imaging is performed based on the brain voltage data, as shown in the following formula:

[0013] .

[0014] In the formula, is the relative conductivity value of different pixels between the current time and the reference time, is the voltage change between the current time and the reference time, T is the transpose symbol, is the regularization parameter, W is the system model noise, H is the sensitivity matrix, R is the regularization matrix.

[0015] Preferably, the relative conductivity values ​​of all pixels in the EIT image at different times are added up and averaged to obtain a curve of relative conductivity variation of global pixels over time.

[0016] Preferably, the step of obtaining a regression coefficient between a relative conductivity variation curve of each pixel point and a relative conductivity variation curve of the global pixel point over time is specifically shown in the following formula:

[0017] .

[0018] In the formula, is the pixel point in the EIT image i In time t The relative conductivity value of and It's a pixel i Different regression coefficients of It's a pixel i The fitting error, M is the total number of pixels in the EIT image, m It is m Pixels.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] The present invention first continuously collects the brain voltage data of the monitored object in multiple cerebrovascular pulsation cycles, performs time difference imaging based on the brain voltage data, and obtains multiple EIT images. Then, based on the multiple EIT images, the relative conductivity variation curve of each pixel point over time and the relative conductivity variation curve of the global pixel point over time are obtained, and the regression coefficient between the relative conductivity variation curve of each pixel point over time and the relative conductivity variation curve of the global pixel point over time is obtained, and a functional EIT image is generated through multiple regression coefficients. And for each cerebrovascular pulsation cycle, an EIT perfusion image is generated. The present invention adopts dynamic cerebral blood flow perfusion EIT technology, and generates functional EIT images and EIT perfusion images by analyzing the measurement data of the change process of the starting point and the end point of diastolic period of cerebral blood vessels. It does not need to obtain data before bleeding or ischemia, and can directly image in real time, which can be widely used in actual clinical environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A flow chart of an electrical impedance tomography method for stroke monitoring according to the present invention;

[0023] Figure 2 Schematic diagram of the functional EIT imaging process of the present invention;

[0024] in: Figure 2 (a): Schematic diagram of EIT image sequence, Figure 2 (b): The relative conductivity of a single pixel changes with time. Figure 2 (c): The relative conductivity of the global pixel changes with time. Figure 2 (d): functional EIT diagram;

[0025] Figure 3 It is a schematic diagram of the brain perfusion area of ​​the present invention;

[0026] Figure 4 This is a schematic diagram of the direction and division of the main cerebral arteries;

[0027] in: Figure 4 (a): Schematic diagram of the main arteries of the brain. Figure 4 (b): Schematic diagram of partition;

[0028] Figure 5Schematic diagram of the experimental method and process of Example 2 of the present invention;

[0029] in, Figure 5 (a): 16 copper electrode positions, indwelling needle positions, and intracranial pressure probe positions. Figure 5 (b): Experimental scene diagram, Figure 5 (c): Schematic diagram of experimental time;

[0030] Figure 6 These are the EIT perfusion images of the normal state, the left neck compressed, and the right neck compressed according to the present invention;

[0031] in, Figure 6 (a): Brain EIT perfusion image under normal conditions, Figure 6 (b): EIT perfusion image of the left neck. Figure 6 (c): EIT perfusion image of right neck compression;

[0032] Figure 7 The cerebral perfusion heterogeneity and cerebral perfusion asymmetry index of the control group and the neck compression group of the present invention;

[0033] in, Figure 7 (a): control group, Figure 7 (b): neck compression group;

[0034] Figure 8 The TCD images of the blood flow velocities of the left and right brains in a normal state according to Example 1 of the present invention;

[0035] in, Figure 8 (a): Left cerebral blood flow velocity in normal state, Figure 8 (b): right cerebral blood flow velocity in normal state;

[0036] Fig. 9 The TCD images of the blood flow velocities of the left and right brains under the condition of compressing the left carotid artery according to Example 1 of the present invention;

[0037] Fig. 9 (a): Left cerebral blood flow velocity when the left carotid artery is compressed. Fig. 9 (b): right cerebral blood flow velocity when the left carotid artery is compressed;

[0038] Fig.10 The TCD images of the blood flow velocities of the left and right brains when the right carotid artery is compressed in Example 1 of the present invention;

[0039] in, Fig.10 (a): Left cerebral blood flow velocity when the right carotid artery is compressed. Fig.10 (b): right cerebral blood flow velocity when the right carotid artery is compressed;

[0040] Fig.11 The EIT perfusion images under different states of the present invention and the monitoring results of the OSTU algorithm;

[0041] in, Fig.11 (a): EIT perfusion image with left neck compressed. Fig.11 (b): OSTU algorithm monitoring results when the left neck is compressed. Fig.11 (c): EIT perfusion image with right neck compressed. Fig.11 (d): OSTU algorithm monitoring results when the right neck is compressed;

[0042] Fig.12 The changes of EIT transmission impedance, intracranial pressure and mean arterial pressure of the present invention;

[0043] in, Fig.12 (a): EIT transmission impedance change, Fig.12 (b): Changes in intracranial pressure, Fig.12 (c): changes in mean arterial pressure;

[0044] Fig.13 The cerebral perfusion heterogeneity index and cerebral perfusion asymmetry index of the normal group and the blood injection group of the present invention;

[0045] in, Fig.13 (a): Cerebral perfusion heterogeneity index of normal group and blood injection group, Fig.13 (b): Cerebral perfusion asymmetry index of normal group and blood injection group;

[0046] Fig.14 This is a cerebral hemorrhage area map according to Example 2 of the present invention;

[0047] in, Fig.14 (a): EIT perfusion image at time t1, Fig.14 (b): OSTU algorithm monitoring results at time t1, Fig.14 (c): EIT perfusion image at time t2, Fig.14 (d): OSTU algorithm monitoring results at time t2, Fig.14 (e): EIT perfusion image at time t3, Fig.14 (f): OSTU algorithm monitoring result at time t3;

[0048] Fig.15 This is a schematic diagram of the results of Example 3 of the present invention;

[0049] in, Fig.15 (a): EIT perfusion image, Fig.15 (b): OSTU monitoring results, Fig.15(c): CT imaging results. DETAILED DESCRIPTION

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

[0051] The present invention provides an electrical impedance tomography method for stroke monitoring, wherein the brain voltage data of the monitored object is collected by an electrical impedance tomography device EC-100 PRO, and then a brain EIT image reconstruction template is established based on the CT image, and then an algorithm is used to reconstruct the EIT image. Then, through the EIT image sequence, the relative conductivity change curve of the brain area corresponding to each pixel point and the global pixel point relative conductivity change curve over time are calculated, and the linear regression coefficient of the two is calculated to obtain a regression functional EIT image. On this basis, the intracranial area where the first 80% of the pixels of the maximum regression functionality in the EIT image are located is obtained, and based on the symmetry of the brain, the area is mirrored and merged to obtain a brain perfusion area. Subsequently, based on the brain perfusion area, the brain perfusion inhomogeneity and the symmetry index of the bilateral brain perfusion parameters are calculated from the EIT perfusion image, and the two parameters are extracted. Specifically, refer to Figure 1 , including the following steps:

[0052] Step 1: Collect the brain voltage data of the monitored object in each beating cycle, and perform time-difference imaging based on the brain voltage data to obtain multiple EIT images of the monitored object.

[0053] Through 16 electrodes attached to the head of the monitored object, select one pair of electrodes in turn to apply excitation current, and measure the voltage signal on each adjacent electrode pair. After the excitation electrode is switched once, EIT imaging data, that is, one frame of data, is obtained. By measuring the data of two frames at different times, the relative conductivity of the two times is reconstructed. Then, a reconstruction template that conforms to the actual electrode distribution position is used to perform forward problem calculation on the EIT imaging data, and an inverse problem calculation on the reconstruction template, and the relative conductivity of the monitored object is obtained through formula (1):

[0054] (1)

[0055] In the formula, is the relative conductivity value between the current time and the reference time, is the voltage change between the current time and the reference time, T is the transpose symbol, is the regularization parameter,W is the system model noise, H is the sensitivity matrix, R is the regularization matrix. , W , H and R All are existing related parameters. Among them, current time and reference time are two concepts in differential EIT. Reference time usually refers to the voltage data collected at a certain moment, which serves as the benchmark for subsequent measurements. The current time refers to the voltage data at the current moment, which is used to compare with the reference time to reconstruct the relative conductivity image. In the EIT imaging process, by selecting the data at a specific moment as the reference time, the data at subsequent moments is used as the current time, and it is subtracted from the reference time to calculate the relative conductivity. This method can effectively reduce the impact of various interference factors on the imaging results and improve the imaging quality.

[0056] Step 2: Generate fEIT image, that is, functional EIT image.

[0057] Reference Figure 2 , EIT images represent the changes in brain tissue impedance over time. Since it can capture spatial information with high temporal resolution, a sequence of EIT images over a period of time may contain valuable functional information that affects clinical decision-making. By extracting the information and presenting it in the form of an image, a functional EIT image can be obtained. Among them, the EIT image sequence is composed of multiple EIT images generated over time. Therefore, multiple EIT images of multiple beating cycles are continuously collected, and the beating cycle starts from the starting point of the current cerebral vasodilation until the starting point of the next cerebral vasodilation. The EIT image contains a certain number of pixels, and the relative conductivity value of each pixel changes with time to form a curve, that is, a relative conductivity change curve over time. If the relative conductivity values ​​of all pixels are added up and averaged, a curve of the relative conductivity value of a global pixel changing with time can be obtained. The present invention calculates fEIT by a linear regression formula. In the fEIT image, the value of each pixel is the regression coefficient of the following linear regression formula:

[0058] (2)

[0059] In the formula, is the pixel point in the EIT image i In time t The relative conductivity value of and Pixel i Different regression coefficients of Pixel i The fitting error,M is the total number of pixels of the EIT image, m For the m Pixels. Figure 2 (d) is the fEIT image, and the value of each pixel in it is the regression coefficient between the relative conductivity change curve of the pixel and the global pixel relative conductivity change curve over time.

[0060] In this embodiment, α i is drawn at the pixel point in the fEIT image i The value on. Figure 2 In the , a specific period of time of EIT image sequence was selected for imaging processing. Through analysis, the relative conductivity curve of a single pixel and the global pixel relative conductivity curve were extracted. When the relative conductivity curve of a pixel shows a high correlation with the global pixel relative conductivity curve, its regression coefficient is large, which usually means that the pixel is more likely to belong to the perfusion area.

[0061] For each pulsation cycle, an EIT perfusion image is generated. Each pixel of these EIT perfusion images represents the relative conductivity between the start and end of diastolic phase of the same region of blood vessels.

[0062] In actual clinical settings, fEIT images and EIT perfusion images are used as references to monitor stroke.

[0063] Step 3: Determine the cerebral perfusion area based on the fEIT image.

[0064] Since there is no detection and imaging method for brain perfusion area at present, the present invention innovatively proposes a method that can effectively obtain brain perfusion area.

[0065] Functional regions of interest (fROIs) are defined by analyzing the functional behavior of pixels during perfusion (based on fEIT images). They are calculated by taking a threshold of 20% of the maximum value of the pixels in the fEIT image and defining all pixels with values ​​exceeding this threshold as "perfused areas". By automatically excluding these non-perfused brain areas from the analysis, fROIs may lead to an underestimation of the actual perfusion heterogeneity. To overcome this limitation, the brain perfusion areas defined by the fEIT images are mirrored from left to right and from right to left, assuming relative symmetry between the left and right brain in the EIT electrode plane, and they are merged by a logical OR operation.

[0066] Step 4: Obtain the cerebral perfusion heterogeneity index and cerebral perfusion asymmetry index based on the cerebral perfusion area.

[0067] Stroke is a lack of blood supply to the brain due to blockage or rupture of cerebral blood vessels. When cerebral blood vessels are blocked, blood flow at the blocked site is reduced or interrupted, resulting in hypoxia and insufficient nutrient supply to brain cells in that area. Therefore, the blood flow distribution under stroke is usually uneven, which may lead to increased blood flow in some areas and decreased blood flow in other areas. This uneven distribution of blood flow is called uneven cerebral perfusion. After a stroke, the degree of uneven cerebral perfusion depends on a variety of factors, including the size, location and severity of the infarction, as well as the body's self-regulation ability. Severe stroke may cause ischemia and necrosis of a large area of ​​brain tissue, thereby aggravating uneven cerebral perfusion. In addition, inflammatory response and edema after stroke may further affect blood flow distribution. Uneven cerebral perfusion is of great significance to the patient's rehabilitation and prognosis. It may lead to sequelae such as neurological impairment, cognitive impairment and movement disorders. Therefore, in the treatment of stroke, restoring the uniform distribution of cerebral blood flow is an important goal. By using drugs, surgery or other treatment methods, the distribution of cerebral blood flow can be improved, the degree of cerebral perfusion inhomogeneity can be reduced, and thus the patient's rehabilitation and recovery of function can be facilitated. Therefore, the present invention proposes an index for evaluating cerebral perfusion inhomogeneity.

[0068] In order to accurately calculate the heterogeneity and asymmetry of cerebral perfusion, the cerebral perfusion area must be clearly defined in the EIT perfusion images.

[0069] Based on the determination of the cerebral perfusion area in the previous step, the median of each EIT perfusion image was calculated, and the sum of the absolute differences between the median and each pixel was considered to be the change in blood flow distribution in the entire perfusion area. In order to make the index universal and comparable between individuals, the cerebral perfusion heterogeneity index was normalized by dividing by the sum of the relative conductivity values ​​within the cerebral perfusion area, as shown in formula (3):

[0070] (3)

[0071] In the formula, is the cerebral perfusion heterogeneity index, ROI It is the area of ​​cerebral perfusion. is the relative conductivity value between the current time and the reference time, x and y are the horizontal and vertical coordinates of the pixel point, In the brain perfusion area The median value 。

[0072] Asymmetry of cerebral perfusion is a manifestation of unevenness. The main reason for left-right asymmetry of cerebral perfusion caused by stroke is the rupture or blockage of cerebral blood vessels, which will lead to a decrease in blood flow on one side of the brain, resulting in insufficient blood supply to the brain tissue on that side, causing ischemia and hypoxia of the brain tissue. This state of ischemia and hypoxia can cause complications such as cerebral edema, cerebral aneurysm or thrombosis, further aggravating the asymmetry of cerebral perfusion and ultimately affecting brain function. It has been proposed that the resistivity asymmetry index is intended to assess the differences between the left and right brains. However, the EIT perfusion images obtained from different patients cannot be directly compared because they only show relative conductivity values. The definition of this index is based on the basic physiological properties of the brain. Therefore, the present invention has developed an asymmetry index based on EIT perfusion images, which can be used for comparison between different individuals.

[0073] The calculation of cerebral perfusion asymmetry index is based on monitoring of cerebral perfusion areas, such as Figure 3 As shown, at this time, the left and right brain perfusion areas are symmetrical about the midline. Then the EIT perfusion image is normalized. The calculation method is shown in formula (4):

[0074] (4)

[0075] In the formula, AI is the asymmetry index, It is normalized , The left area. The right area. is the horizontal coordinate of the midline dividing the left and right hemispheres of the brain. and The left and right areas are i The number of pixels in a row, Indicates i The vertical coordinate of the row pixel, x is the horizontal coordinate of the pixel.

[0076] Reference Figure 4 Considering the direction and blood supply area distribution of the anterior cerebral artery, middle artery, posterior cerebral artery and arterial ring, the present invention divides the human brain into two hemispheres, and divides each hemisphere into the anterior target area, the middle target area and the posterior target area. Since it is difficult to clearly define the edge position of each area, the present invention uses a simple straight line segmentation to define the area edge. ROI 1. ROI 2. ROI 3. ROI 4. ROI 5 and ROI6 respectively correspond to the left anterior cerebral artery, the left middle cerebral artery, the left posterior cerebral artery, the right anterior cerebral artery, the right middle cerebral artery and the right posterior cerebral artery (in a mirror image relationship with the actual one), that is, multiple sub-cerebral perfusion areas.

[0077] The calculation formula is as follows:

[0078] (5)

[0079] (6)

[0080] In the formula, For the n The average reconstruction value of each sub-cerebral perfusion area, is the relative conductivity value between the current time and the reference time, For the n The number of pixels in the sub-cerebral perfusion area, For the j The percentage of brain perfusion area For the j The average reconstruction value of each sub-cerebral perfusion area.

[0081] The OSTU algorithm (maximum inter-class variance method) is widely regarded as an efficient image segmentation technique, especially when binarization is performed. The core of the OSTU algorithm is to calculate the grayscale histogram of the image and find a threshold based on the histogram so that the image is divided into two categories (usually foreground and background) with the largest inter-class variance between the two categories. The basis of this method is that if the internal pixel values ​​of the foreground and background are relatively uniform and the difference between the two is large, then the boundary between them should be very obvious, so that the target of interest can be effectively segmented. This method is not only computationally efficient, but also generally applicable to many types of images. The advantage of the OSTU algorithm is that it is highly automated and users do not need to manually set the threshold, which greatly simplifies the operation process and reduces errors caused by human factors. At the same time, because the algorithm is based on mathematical statistical principles, its results are highly reliable and scientific.

[0082] In EIT perfusion images, the difference between the lesion area and normal tissue is often manifested as a significant difference in pixel value. Since stroke can cause a decrease in blood flow in local brain tissue, this change can be reflected by the low perfusion area in the brain EIT perfusion image. The OSTU algorithm uses this feature to calculate the grayscale histogram of the image and find an optimal threshold to divide the image into two categories: low perfusion area and normal perfusion area. In clinical practice, the low perfusion area and normal perfusion area can be used as a reference to distinguish the lesion area.

[0083] Specifically, the OSTU algorithm first counts the number of pixels at each gray level in the brain EIT perfusion image to form a grayscale histogram. Then, the algorithm tries different thresholds to divide the image into two categories, one is the pixels with grayscale values ​​lower than the threshold (assumed to be low perfusion areas), and the other is the pixels with grayscale values ​​higher than the threshold (assumed to be normal perfusion areas). For each threshold, the algorithm calculates the average value and probability of the two categories of pixels, and then obtains the inter-class variance. The goal of the OSTU algorithm is to find the threshold that maximizes the inter-class variance, because this threshold can maximize the distinction between low perfusion areas and normal perfusion areas.

[0084] Example 1

[0085] Experimental study on imaging monitoring of cerebral blood perfusion status in healthy volunteers.

[0086] After the formation of a stroke lesion, the brain is not supplied with enough blood due to rupture or blockage of blood vessels, which in turn causes a decrease in cerebral blood flow. The degree of cerebral blood flow reduction is related to the size, location and severity of the lesion. In order to simulate the distribution of cerebral blood flow when a stroke occurs on one side of the brain, it can be observed by compressing the unilateral carotid artery of healthy volunteers. The carotid artery compression experiment is a commonly used clinical examination method used to assess the degree of carotid artery stenosis or blockage. This experiment observes changes in the patient's cerebral blood flow by compressing a unilateral carotid artery. When compressing one side of the carotid artery, blood flow is obstructed due to the compression of the carotid artery on that side, resulting in a decrease in cerebral blood flow on that side. It can provide a relatively safe way to observe the effect of carotid artery compression on cerebral blood flow in a controlled environment using TCD as a control method. Compressing the carotid artery by external force can reduce the amount of blood supplied to the brain, which is similar to the decrease in cerebral blood flow caused by rupture or blockage of blood vessels during a stroke. The experimental steps are as follows:

[0087] S1: Screening and preparation of monitoring subjects: First, appropriate monitoring subjects are screened through a detailed medical history and physical examination. This includes assessing the monitoring subject's cardiovascular health, the condition of the carotid artery, and any possible risk factors such as high blood pressure, diabetes, or high cholesterol.

[0088] S2: Scalp Cleansing: Wash the subject's hair and keep the scalp dry before the experiment. This is to ensure that any monitoring equipment used in the experiment, such as ultrasound probes or blood flow detectors, can be placed correctly and work properly. A clean scalp helps improve the sensitivity and accuracy of the monitoring equipment.

[0089] S3: Environmental preparation: Ensure that the experiment is conducted in a quiet and temperature-controlled environment, which helps the monitored subjects relax and reduces the impact of external variables on the experimental results.

[0090] S4: Rest time: Before the experiment begins, give the subject enough time (usually 10 minutes) to adapt to the environment and relax. This helps stabilize heart rate and blood pressure, thereby obtaining more accurate baseline measurement data.

[0091] S5: Carotid artery compression: Press the left carotid artery of each subject with moderate pressure for 60 seconds to simulate blood flow blockage during stroke. Then give a 5-minute rest period, during which the subject's response and changes in cerebral blood flow are monitored. Afterwards, repeat the same operation on the contralateral side.

[0092] like Figure 6 As shown, EIT perfusion images of the brain were obtained in normal state, when the left carotid artery was compressed, and when the right carotid artery was compressed by electrical impedance tomography. These images reveal significant differences in brain blood perfusion under different experimental conditions. Under normal conditions, EIT perfusion images show that the blood flow distribution of brain tissue is relatively uniform and symmetrical. However, when the left carotid artery is compressed, the situation changes significantly. Due to the temporary occlusion of the left carotid artery, the blood flow in the brain on this side is reduced, which is manifested as a weakened perfusion signal in the left brain area on the perfusion EIT image. Relatively speaking, the right carotid artery is not affected and continues to remain unobstructed, so the blood perfusion of the right brain tissue appears relatively normal on the image, forming a sharp contrast with the left side. Similarly, when the right carotid artery is compressed, a similar asymmetric perfusion pattern is observed. At this time, the blood flow of the right brain tissue is restricted, while the left side remains normal, thus showing a difference in left-right contrast on the EIT image. These observations not only verify the sensitivity of EIT technology in detecting changes in cerebral blood flow, but also provide a non-invasive and effective tool for real-time monitoring of cerebral perfusion status in clinic.

[0093] Figure 7The data analysis presented in the present study reveals the heterogeneity and asymmetry of cerebral perfusion in different states. These findings are of great significance for a deeper understanding of the complex changes in cerebral hemodynamics. The significant difference in cerebral perfusion parameters between the normal state and the cervical compression state (P<0.001) clearly shows that the P value is a statistical measure of the probability of the observed result occurring when the hypothesis is true. The smaller the P value, the less likely the result is caused by random errors alone. Under brief carotid compression, cerebral blood flow will be significantly affected, resulting in a decrease in blood flow on one side, which can be observed by EIT. This effect is likely to be caused by the decrease in blood flow caused by carotid compression, which in turn affects the corresponding brain regions. The cerebral perfusion zoning percentage data presented in Table 1 provide intuitive quantitative evidence for the changes in cerebral blood flow distribution under different states. Under normal conditions, the observation of the perfusion zoning percentage of each brain region shows a relatively balanced state, indicating that the blood supply to various brain regions maintains a stable balance without external interference. However, when the left carotid artery is pressed, the situation changes significantly. The data show that the compressed side (left side) has a relatively balanced blood flow distribution. ROI 1. ROI 2 and ROI The proportion of 3 in the left cerebral region was significantly lower than that in the right cerebral region (P<0.001). Specifically, due to the compression of the left carotid artery, the amount of blood it delivers to the brain is reduced, resulting in a decrease in the proportion of perfusion zonation in the left cerebral region. Similarly, when pressing the right carotid artery, it can be reasonably inferred that the right ROI 4. ROI 5 and ROI The proportion of 6 will also decrease accordingly, showing asymmetry similar to that on the left.

[0094] Table 1 The proportion of cerebral perfusion regions under three conditions

[0095]

[0096] Reference Figure 8-Figure 10, among which, under normal conditions, the peak systolic blood flow velocity of the left brain is 99, the diastolic blood flow velocity of the left brain is 40, and the ratio of the peak systolic blood flow velocity to the peak diastolic blood flow velocity of the left brain is 2.45. The peak systolic blood flow velocity of the right brain is 99, the diastolic blood flow velocity of the right brain is 40, and the ratio of the peak systolic blood flow velocity to the peak diastolic blood flow velocity of the right brain is 2.45. Under the condition of compressing the left carotid artery, the peak systolic blood flow velocity of the left brain is 45, the diastolic blood flow velocity of the left brain is 28, and the ratio of the peak systolic blood flow velocity to the peak diastolic blood flow velocity of the left brain is 1.61. The peak systolic blood flow velocity of the right brain is 128, the diastolic blood flow velocity of the right brain is 54, and the ratio of the peak systolic blood flow velocity to the peak diastolic blood flow velocity of the right brain is 2.39. When the right carotid artery is compressed, the peak systolic blood flow velocity of the left brain is 111, the diastolic blood flow velocity of the left brain is 44, and the ratio of the peak systolic blood flow velocity to the peak diastolic blood flow velocity of the left brain is 2.53. The peak systolic blood flow velocity of the right brain is 27, the diastolic blood flow velocity of the right brain is 20, and the ratio of the peak systolic blood flow velocity to the peak diastolic blood flow velocity of the right brain is 1.38. Therefore, under normal conditions, the difference in blood flow velocity between the left and right brains is small. After compressing the carotid artery, the blood flow velocity of the compressed side of the brain is significantly lower than that of the non-compressed side.

[0097] Fig.11 The cerebral EIT perfusion images and the low perfusion areas accurately monitored by the OSTU algorithm are presented. From these images, the obvious correspondence between the low perfusion areas and the neck compression areas can be clearly observed. This intuitive demonstration of the efficiency and stability of the OSTU algorithm in monitoring areas of insufficient cerebral blood flow.

[0098] Example 2

[0099] Imaging experimental study based on animal model of cerebral hemorrhage.

[0100] This example constructs a pig cerebral hemorrhage model to study the changes in cerebral blood perfusion during cerebral hemorrhage. Due to the similarities between pigs and humans, the data obtained from the pig model are highly correlated with human pathological conditions, which helps to transform research results into clinical applications.

[0101] The experimental method of the cerebral hemorrhage model is to inject autologous non-anticoagulated blood into the brain parenchyma of domestic pigs in batches to prepare the cerebral hemorrhage model, and use invasive intracranial pressure (ICP) monitoring as a control, and monitor it through an intracranial pressure monitor. Figure 5After connecting the electrodes to the EIT system, the EIT baseline data was collected through the EIT data acquisition system, and ICP data collection was started at the same time. About 10 minutes later, 2 ml of blood was drawn through the indwelling needle pre-placed in the femoral artery, and then slowly and continuously injected into the brain parenchyma area through the indwelling needle pre-placed in the skull within 1 minute. This process was repeated about 10 minutes later, for a total of 5 times, and its time series is shown in the figure. Figure 5 Norepinephrine was used to maintain blood pressure throughout the operation.

[0102] exist Fig.12 The changes in EIT transmission impedance, intracranial pressure, and mean arterial pressure during the entire experiment are shown in Figure 1. Five blood injections were performed during the experiment, and each injection caused the EIT transmission impedance to rise rapidly. At the same time, the intracranial pressure gradually decreased after a rapid rise. However, it is worth noting that the mean arterial pressure remained stable throughout the experiment without significant fluctuations.

[0103] Fig.13 The cerebral perfusion heterogeneity index and cerebral perfusion asymmetry index of the normal group and the group after blood injection are given. It can be clearly seen from the figure that these parameters of the blood injection group are significantly higher than those of the normal group, which shows that these parameters can effectively distinguish the normal group from the group after blood injection. This distinction is of great significance for clinical diagnosis and treatment.

[0104] Fig.14 The brain EIT perfusion images and the cerebral hemorrhage area obtained by the OSTU algorithm were presented. This not only proves the sensitivity and accuracy of the OSTU algorithm in detecting cerebral hemorrhage, but also further strengthens the clinical application potential of EIT technology as a non-invasive monitoring tool, enabling doctors to monitor cerebral hemorrhage events in a timely and accurate manner and optimize treatment plans.

[0105] Example 3

[0106] An experimental study based on imaging of clinical patients with stroke.

[0107] The following are the experimental steps for using EIT technology to evaluate cerebral blood flow distribution in stroke patients:

[0108] S6: Patient screening and preparation: First, appropriate patients are screened through a detailed medical history and physical examination. This includes assessing the patient's cardiovascular health, previous stroke history, and any possible risk factors such as high blood pressure, diabetes, or high cholesterol.

[0109] S7: Scalp Cleansing: Wash the patient’s hair before the experiment and keep the scalp dry. This is to ensure that any monitoring equipment used in the experiment, such as EIT electrodes, is properly placed and works properly. A clean scalp helps improve the sensitivity and accuracy of the monitoring equipment.

[0110] S8: Environmental preparation: Ensure that the experiment is conducted in a quiet, temperature-controlled environment, which helps patients relax and reduces the influence of external variables on the experimental results.

[0111] S9: Data collection: Use EIT equipment to collect cerebral blood flow data. The patient needs to lie flat on the examination bed with the head placed on the EIT electrode array. The current is transmitted to the brain tissue through the electrodes, and then the distribution of the current is measured through other electrodes. These data will be recorded and analyzed.

[0112] S10: CT control: A brain CT scan is performed before and after EIT data acquisition. The patient needs to lie flat on the examination table of the CT scanner with the head fixed on a bracket to ensure image clarity and consistency. The CT scan will provide detailed images of brain structures for comparison and analysis with the EIT data.

[0113] Table 2 below shows in detail the data of five stroke patients on three key physiological parameters: cerebral perfusion heterogeneity, cerebral perfusion asymmetry, and lesion zoning ratio. These parameters provide important information for a deep understanding of the blood flow pattern and structural damage in the patient's brain, which is crucial for understanding the physiological changes of stroke.

[0114] The data showed that the average value of cerebral perfusion heterogeneity was 0.63, which was significantly higher than 0.51 in healthy people. This indicates that in stroke patients, the blood flow distribution in the brain is more unbalanced, which may be due to vascular occlusion or hemorrhage, resulting in reduced blood flow in some areas, while other areas have increased blood flow due to compensatory mechanisms. In addition, the average value of cerebral perfusion asymmetry was 1.00, which was also significantly higher than 0.74 in healthy people, reflecting that the abnormal blood flow in one cerebral hemisphere caused by stroke may lead to impaired brain function on that side. However, the average value of the proportion of lesion zoning was 0.11, which was lower than 0.16 in normal people, indicating that although patients had local brain damage, the proportion of affected areas was small. Changes in these parameters reveal significant changes in cerebral blood flow and structure in stroke patients. These changes are directly related to the severity of the disease and play an important role in the formulation of treatment plans and monitoring of efficacy.

[0115] Table 2 Comparison of patient parameters

[0116]

[0117] Fig.15 The EIT perfusion images of one patient, the results of the maximum intraclass variance method, and the CT images of the patient are presented.

[0118] In the present invention, for the first time, based on the impact of stroke on cerebral blood perfusion distribution and based on the significant difference in the electrical conductivity characteristics of blood and brain tissue, an attempt is made to monitor cerebral blood perfusion status through EIT as a reference to indirectly monitor stroke, thus providing a new perspective and method for stroke monitoring.

[0119] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0120] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An electrical impedance tomography method for stroke monitoring, characterized in that: The following steps are involved: Continuously collecting brain voltage data of the monitored object in multiple cerebral vascular pulsation cycles, performing time-difference imaging based on the brain voltage data, and obtaining multiple EIT images; wherein the EIT image includes multiple pixel points; Based on multiple EIT images, a relative conductivity variation curve of each pixel point and a relative conductivity variation curve of the global pixel point are obtained, a regression coefficient between the relative conductivity variation curve of each pixel point and the relative conductivity variation curve of the global pixel point is obtained, and a functional EIT image is generated through multiple regression coefficients; For each cerebral vascular pulsation cycle, an EIT perfusion image is generated, wherein the value of a pixel point in the EIT perfusion image is the relative conductivity between the starting point and the end point of the cerebral vascular diastolic period; The regression coefficient between the relative conductivity variation curve of each pixel point and the relative conductivity variation curve of the global pixel point is obtained, which is specifically shown in the following formula: ； In the formula, is the pixel point in the EIT image i In time t The relative conductivity value of and It's a pixel i Different regression coefficients of It's a pixel i The fitting error, M is the total number of pixels in the EIT image, m It is m pixels; α i is the pixel point drawn in the functional EIT image i The value on Determining the cerebral perfusion area based on functional EIT images includes the following steps: The threshold was 20% of the maximum pixel value of the functional EIT image; All pixels in the functional EIT image whose pixel values ​​exceed the threshold are defined as the perfusion area; The perfusion regions were merged by left and right mirroring through a logical OR operation to obtain the final brain perfusion region.

2. The electrical impedance tomography method for stroke monitoring according to claim 1, characterized in that: The brain voltage data of the monitored subject during multiple cerebrovascular pulsation cycles are continuously collected through multiple pairs of electrodes.

3. The electrical impedance tomography method for stroke monitoring according to claim 2, characterized in that: The time difference imaging is performed according to the brain voltage data, as shown in the following formula: ； In the formula, is the relative conductivity value of different pixels between the current time and the reference time, is the voltage change between the current time and the reference time, T is the transpose symbol, is the regularization parameter, W is the system model noise, H is the sensitivity matrix, R is the regularization matrix.

4. The electrical impedance tomography method for stroke monitoring according to claim 1, characterized in that: The relative conductivity values ​​of all pixels in the EIT image at different times are added up and averaged to obtain the time-varying curve of the relative conductivity of the global pixel.

Images