Low-field magnetic resonance stroke identification sequence and device with the identification sequence

Abstract

The present invention relates to a low-field magnetic resonance stroke identification sequence and a device having the identification sequence. The inversion recovery sequence is adopted, and the TE adopting system can achieve a minimum or near-minimum TE value, and the influence of the T2 effect on the signal S is minimized; at the same time, a TR value and a corresponding TI value combination are selected to make the cerebral hemorrhage signal a high signal, and the cerebral infarction tissue and the brain substance are equal signals or low signals, so that hemorrhagic stroke can be judged quickly and accurately, and low-field magnetic resonance is adopted, which has high accessibility to patients.

Description

Technical Field

[0001] The invention relates to the technical field of magnetic resonance imaging (MRI), and in particular to a low-field magnetic resonance stroke identification sequence and a device having the low-field magnetic resonance stroke identification sequence. Background Art

[0002] Stroke, also known as cerebral infarction, is an acute cerebrovascular disease, including hemorrhagic stroke and ischemic stroke. Hemorrhagic stroke is also known as cerebral hemorrhage, and ischemic stroke is also known as cerebral ischemia or cerebral infarction. The rescue time for cerebral infarction is very limited. It is generally believed that the best treatment time for stroke is within 4.5 hours. If it exceeds 6 hours, the best rescue time is often missed. Therefore, in the diagnosis and treatment of stroke, judging whether the patient has cerebral hemorrhage or cerebral infarction, or infarction with bleeding, is the key for doctors to decide on the treatment plan. How to quickly and accurately judge the patient's condition is related to the treatment effect and life of the patient.

[0003] Among the existing diagnostic assistance technologies and processes, CT head scan is the first-line diagnostic method used in hospitals for cerebral hemorrhage. Computed tomography (CT) is very sensitive to cerebral hemorrhage and is considered the "gold standard" for diagnosing cerebral hemorrhage. It is a necessary examination method for patients with initial stroke diagnosis and is also an important imaging basis for the existing stroke diagnosis and treatment process. However, CT is not sensitive to acute ischemic stroke, and the lesion image is not obvious. General studies believe that CT can only see cerebral infarction lesions that are more than 24 hours old, and it is difficult to detect earlier cerebral infarctions, which can easily lead to misdiagnosis of early cerebral infarctions.

[0004] Some hospitals also use special superconducting magnetic resonance equipment to identify and examine stroke, improve the detection rate of acute cerebral infarction, and accurately determine the time window of cerebral infarction. Magnetic resonance imaging can improve the detection rate of cerebral hemorrhage, especially the detection rate of post-cerebral infarction hemorrhage and long-term cerebral hemorrhage. However, the current superconducting magnetic resonance equipment is a high-field strength equipment. The commonly used field strength is generally 1.5T and 3.0T. It needs to be installed in a special electromagnetic shielding room. It is extremely unfriendly to stroke patients who are in a hurry to rescue, and it is easy to delay the rescue opportunity. If the patient has metal implants or other medical auxiliary equipment in the body, the metal object will absorb electromagnetic waves during the MRI scan, and the absorbed electromagnetic waves will be converted into heat. The higher the magnetic field strength, the higher the heat generation. Under the action of the strong magnetic field, the heat generation will burn the patient. Therefore, such patients cannot be put on the machine for MRI imaging, which reduces accessibility. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a low-field magnetic resonance stroke identification sequence with high accessibility and the ability to quickly and accurately judge hemorrhagic stroke, as well as a device having the low-field magnetic resonance stroke identification sequence.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a low-field magnetic resonance stroke identification sequence, using an inversion recovery sequence, and its relative signal intensity formula is as follows:

[0007]

[0008] In the formula, S represents the relative signal strength;

[0009] PD stands for tissue proton density;

[0010] TI represents the inversion recovery time of the sequence;

[0011] T1 represents the T1 relaxation time of the tissue;

[0012] TR represents the repetition time of the sequence;

[0013] TE last Represents the last echo time of a multi-echo sequence;

[0014] TE stands for effective echo time;

[0015] T2 represents the T2 relaxation time of the tissue;

[0016] i=1, 2, …, N represents N components, where brain parenchyma and cerebral hemorrhage are single-component models N=1, and cerebral infarction tissue adopts a two-component model, N=2; subscript i represents the value of the i-th tissue;

[0017] The brain parenchyma and cerebral hemorrhage were used as a single-component model, N=1, and the cerebral infarction tissue was used as a two-component model, N=2;

[0018] Among them, the TE system can achieve the minimum or close to the minimum TE value, so that As close to 1 as possible, the T2 effect has the least impact on the signal S;

[0019] The TR value and the corresponding TI value combination are selected so that the cerebral hemorrhage signal is high signal, and the cerebral infarction tissue and brain parenchyma are equal signal or low signal.

[0020] As an improvement of the above technical solution, the selection of the TR value and the corresponding TI value combination complies with the principle of the fastest clinical scanning speed.

[0021] As an improvement of the above technical solution, the corresponding relationship between TR and TI conforms to the following fitting results:

[0022] TR = 0.0006925*TI 2 +0.7426*TI+67.78.

[0023] As an improvement of the above technical solution, the TI is fixed and the TR value fluctuates by 10%; or the TR is fixed and the TI fluctuates by 10%.

[0024] As an improvement of the above technical solution, wherein the low field is 0.23T, the TE is 24 milliseconds, and the TI is 685 milliseconds when the TR is 900 milliseconds.

[0025] As an improvement of the above technical solution, wherein the low field is 0.23T, the TE is 24 milliseconds, and the TI is 800 milliseconds when the TR is 1100 milliseconds.

[0026] As an improvement of the above technical solution, the low field is 0.23T, the TE is 24 milliseconds, and the TI is 1000 milliseconds when the TR is 1500 milliseconds.

[0027] As an improvement of the above technical solution, the PD is obtained by the signal intensity of the proton weighted image.

[0028] As an improvement of the above technical solution, a series of inversion recovery pulses with different TI values ​​are applied before the fast spin echo sequence. At this time, the signal size of the region of interest is related to the T1 value, and the T1 measurement value is obtained by fitting using the following formula:

[0029]

[0030] in, is the tissue signal strength;

[0031] α is the weight coefficient;

[0032] τ is the TI value of the scanning sequence;

[0033] TR is the repetition time;

[0034] TE last is the last echo time;

[0035] T1 is the T1 relaxation time of the tissue.

[0036] As an improvement of the above technical solution, a fast spin echo sequence is used, and other parameters except TE time are kept unchanged. The correlation between TE and T2 is used to fit the following formula to obtain the T2 measurement value:

[0037]

[0038] in, is the tissue signal strength;

[0039] αis the weight coefficient;

[0040] β is a series of echo times;

[0041] T2 is the T2 relaxation time of the tissue.

[0042] To solve the above technical problems, the present invention also provides the following technical solutions: a low-field magnetic resonance stroke identification device, comprising a magnetic resonance scanner; the magnetic resonance scanner executes any of the above-mentioned low-field magnetic resonance stroke identification sequences.

[0043] Compared with the prior art, the present invention uses an inversion recovery sequence, so that the TE using system can reach the minimum or near-minimum TE value, and the influence of the T2 effect on the signal S is minimized; at the same time, the TR value and the corresponding TI value combination are selected to make the cerebral hemorrhage signal a high signal, and the cerebral infarction tissue and brain parenchyma an equal signal or a low signal, thereby being able to quickly and accurately judge hemorrhagic stroke, and using low-field magnetic resonance imaging, which is highly accessible to patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to make the purpose, technical solution and beneficial effects of the present invention clearer, the present invention provides the following drawings for illustration:

[0045] Attached Figure 1a : T2 image at the beginning of the pig cerebral hemorrhage experiment (0h);

[0046] Attached Figure 1b : FLAIR image at the beginning of the pig intracerebral hemorrhage experiment (0h);

[0047] Attached Figure 1c : DWI image at the beginning of the pig intracerebral hemorrhage experiment (0h);

[0048] Attached Figure 1d : T1 image at the beginning of the pig cerebral hemorrhage experiment (0h);

[0049] Attached Figure 2a :T2 image at the end of the pig brain hemorrhage experiment (17h);

[0050] Attached Figure 2b : FLAIR images at the end of the pig cerebral hemorrhage experiment (17h);

[0051] Attached Figure 2c : DWI images at the end of the pig cerebral hemorrhage experiment (17h);

[0052] Attached Figure 2d :T1 image at the end of the pig brain hemorrhage experiment (17h);

[0053] Attached Figure 3a : T1 relaxation time curve of pig cerebral hemorrhage test;

[0054] Attached Figure 3b : T2 relaxation time curve of pig cerebral hemorrhage test;

[0055] Attached Figure 3c : PD normalized curve of pig cerebral hemorrhage test;

[0056] Attached Figure 3d : T1-weighted curve of pig cerebral hemorrhage test;

[0057] Attached Figure 3e : T2-weighted curve of pig cerebral hemorrhage test;

[0058] Attached Figure 3f : FLAIR curve of pig cerebral hemorrhage test;

[0059] Attached Figure 4a : T2 image at the beginning of the pig brain ischemia experiment (0h);

[0060] Attached Figure 4b : FLAIR images at the beginning of the pig brain ischemia experiment (0h);

[0061] Attached Figure 4c : DWI images of pig brain ischemia test at the beginning (0h);

[0062] Attached Figure 4d : T1 image at the beginning of the pig brain ischemia experiment (0h);

[0063] Attached Figure 5a : T2 image at the end of the pig brain ischemia experiment (24h);

[0064] Attached Figure 5b : FLAIR images at the end of the pig brain ischemia experiment (24h);

[0065] Attached Figure 5c : DWI images of pig brain ischemia experiment at the end (24h);

[0066] Attached Figure 5d : T1 image at the end of the pig brain ischemia experiment (24h);

[0067] Attached Figure 6a : T1 relaxation time curve of pig brain ischemia test;

[0068] Attached Figure 6b : T2 relaxation time curve of pig brain ischemia test;

[0069] Attached Figure 6c : PD normalized curve of pig brain ischemia test;

[0070] Attached Figure 6d : T1-weighted curve of pig brain ischemia test;

[0071] Attached Figure 6e: T2-weighted curve of pig brain ischemia test;

[0072] Attached Figure 6f : FLAIR curve of pig brain ischemia test;

[0073] Attached Figure 7a :TE 24ms, TR 900ms fitting diagram;

[0074] Attached Figure 7b : TE 24ms, TR 1100ms fitting diagram;

[0075] Attached Figure 7c : TE 24ms, TR 1500ms fitting diagram;

[0076] Attached Figure 8 : TR / TI relationship curve, unit: ms;

[0077] Attached Figure 9a :TR / TI=900 / 685ms hemorrhage image, where the red circle indicates the hemorrhage area and the blue circle indicates the brain parenchyma area;

[0078] Attached Figure 9b : TR / TI=1100 / 800ms hemorrhage image, where the red circle indicates the hemorrhage area and the blue circle indicates the brain parenchyma area;

[0079] Attached Fig.9c :TR / TI=1500 / 1000ms hemorrhage image, where the red circle indicates the hemorrhage area and the blue circle indicates the brain parenchyma area;

[0080] Attached Fig.10a :DWI image of the brain in case 1,

[0081] Attached Fig.10b : ADC image of the brain in case 1;

[0082] Attached Fig.10c : FLAIR images of the brain in case 1;

[0083] Attached Fig.10d : PD images of the brain in case 1;

[0084] Attached Fig.11a :DWI image of the brain in case 2,

[0085] Attached Fig.11b : ADC image of the brain in case 2;

[0086] Attached Fig.11c : FLAIR images of the brain in case 2;

[0087] Attached Fig.11d : PD images of the brain in case 2;

[0088] Attached Fig.11e : Typical image mosaic of brain DWI image, ADC image, FLAIR image and PD image in case 2;

[0089] Attached Fig.12a :DWI image of the brain in case 3,

[0090] Attached Figure 12b : ADC image of the brain in case 3;

[0091] Attached Fig.12c : FLAIR images of the brain in case 3;

[0092] Attached Fig.12d : PD images of the brain in case 3;

[0093] Attached Fig.13a :DWI image of the brain in case 4,

[0094] Attached Fig.13b : ADC image of the brain in case 4;

[0095] Attached Fig.13c : FLAIR images of the brain in case 4;

[0096] Attached Fig.13d : PD images of the brain in case 4;

[0097] Attached Fig.14a :DWI image of the brain in case 5,

[0098] Attached Fig.14b : ADC image of the brain in case 5;

[0099] Attached Fig.14c : FLAIR images of the brain in case 5;

[0100] Attached Fig.14d : PD images of the brain in case 5;

[0101] Attached Fig.15a :DWI image of the brain in case 6,

[0102] Attached Fig.15b : ADC image of the brain of case 6;

[0103] Attached Fig.15c : FLAIR images of the brain in case 6;

[0104] Attached Fig.15d : PD images of the brain in case 6;

[0105] In the above case images, the red circle indicates that the highlighted signal area is the lesion area;

[0106] DWI image: A magnetic resonance imaging sequence in clinical imaging diagnosis. High signals annotated in the image are generally considered to be caused by abnormal lesions.

[0107] ADC image: A magnetic resonance sequence in clinical imaging diagnosis, which is a calculated value image of DWI. The comprehensive diagnosis of the signal of the lesion area corresponding to the DWI image can determine whether the lesion is an acute stroke;

[0108] FLAIR image: A magnetic resonance imaging sequence image in clinical imaging diagnosis. While other tissues in the image maintain T2 images, water has a low signal. The highlighted part is generally considered to be caused by nodular lesions.

[0109] PD image: A magnetic resonance imaging sequence used in clinical diagnostic imaging. This type of image is mainly used to determine whether a suspected stroke lesion is a cerebral ischemic or hemorrhagic lesion. DETAILED DESCRIPTION

[0110] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0111] 1. About Field Strength B0

[0112] The Magnetic Resonance Imaging (MRI) market is overwhelmingly dominated by high-field systems, especially for medical or clinical MRI applications. The general trend in medical imaging is to produce MRI scanners with increasingly larger field strengths, with the vast majority of clinical MRI scanners operating at 1.5T or 3T, and higher field strengths of 7T and 9T used in research environments. "High field" generally refers to MRI systems currently used in clinical settings, and more specifically, to MRI systems that operate with a main magnetic field (i.e., B0 field) of 1.0T or more. Clinical systems that operate between 0.5T and 1.0T are also generally described as "mid-field". Field strengths between approximately 0.3T and 0.5T are described as "medium-low field". In contrast, "low field" generally refers to MRI systems that operate with a B0 field in the range between approximately 0.18T and 0.3T. Low-field MRI systems that operate with a B0 field of less than 0.18T are referred to as "extremely low field".

[0113] 2. MRI manifestations of cerebral hemorrhage and cerebral ischemia (cerebral infarction)

[0114] Studies have shown that the magnetic resonance characteristics of cerebral hemorrhage and cerebral ischemia are usually complex and changeable, and this performance is not only related to time, but also has a great relationship with field strength.

[0115] For ultra-acute cerebral hemorrhage (within 6 hours), the increase of paramagnetic substances such as deoxyhemoglobin in red blood cells causes the local T2* effect to shorten, so it usually presents isosignal or low signal characteristics in high-field MRI. Since the magnetic sensitivity effect is positively correlated with the square of the field strength, ultra-acute hemorrhage is usually less affected by this in low-field MRI, and manifests as continuous high signals in DWI and FLAIR.

[0116] For hyperacute cerebral ischemia, magnetic resonance imaging usually shows high signals in DWI (diffusion weighted imaging) and FLAIR and other signals. As time goes by, due to damage to the blood-brain barrier, etc., the FLAIR in the infarction area will gradually become a high signal.

[0117] Therefore, it can be seen that both cerebral hemorrhage and cerebral ischemia will have a period in low-field MRI images, where both DWI and FLAIR show high signals at the same time, making it impossible to distinguish them. To this end, the following animal experiment was conducted to find a low-field MRI stroke identification sequence that can quickly and accurately determine hemorrhagic stroke.

[0118] 1. T1 and T2 quantitative analysis methods

[0119] T1 value measurement: Before the fast spin echo sequence, a series of inversion recovery pulses with different TI values ​​are applied. At this time, the signal size of the region of interest is related to the T1 value. The following formula is used for fitting to obtain the T1 measurement value.

[0120]

[0121] in, is the tissue signal strength;

[0122] α is the weight coefficient;

[0123] τ is the TI value of the scanning sequence;

[0124] TR is the repetition time;

[0125] TE last is the last echo time;

[0126] T1 is the T1 value of the tissue.

[0127] The fixed scanning parameters of this sequence are: the first round of scanning TR 4000ms, the last echo time TE 107ms, and the values ​​of a series of TI values ​​τ are [100150 200 250 400 600 800 1000]ms; the second round of scanning TR 10000ms, the last echo time TE 107ms, and the values ​​of a series of TI values ​​τ are [3500 4000 4500]ms. Correspondingly, a series of Value, by comparing data τ with data The above formula is used for numerical fitting to obtain the final T1 measurement value.

[0128] T2 value measurement: Use a fast spin echo sequence, keep other parameters except TE time unchanged, use the correlation between TE and T2, and use the following formula for fitting to obtain the T2 measurement value:

[0129]

[0130] in, is the tissue signal strength;

[0131] α is the weight coefficient;

[0132] β is a series of echo times;

[0133] T2 is the T2 value of the tissue;

[0134] The scanning parameters of this sequence are: TR 4000ms, a series of echo time β values ​​are [42, 56, 70, 84, 98, 112]ms, and a series of corresponding value, by comparing the data β with the data The above formula is used for numerical fitting to obtain the final T2 calculation value.

[0135] PD (proton density) value measurement: According to the principle of magnetic resonance, the present invention directly uses the signal intensity of the PD-weighted image to measure the relative PD value between tissues. This value is the signal value of the PD-weighted image, not the absolute proton density of the tissue. Because the signal value is of a huge magnitude, the present invention normalizes its magnitude to the single-digit level, which does not affect the expression of the trend in the diagram. The scanning parameters of this sequence are: TE is the minimum TE of the system, TR 10000ms.

[0136] Selection of the region of interest: The basal ganglia region is selected for estimation of the T1 and T2 values ​​of the brain parenchyma. The signal in this region is relatively uniform, is less affected by signals such as cerebrospinal fluid, has a relatively unobvious volume effect, and can reflect the signal characteristics of the brain parenchyma.

[0137] 2. Test equipment

[0138] The equipment used in the present invention is the mobile head and neck magnetic resonance system ACUTA Elfin manufactured by Foshan Regato Medical Technology Co., Ltd. The basic parameters are as follows:

[0139] Nominal B0 value: 0.23T±0.01T

[0140] Maximum spatial encoding gradient: 25mT / m

[0141] Maximum gradient switching rate: 60T / m / s.

[0142] 3. Bleeding test results

[0143] In this experiment, pig venous blood was collected and injected into the basal ganglia of the pig brain. The MRI showed that the bleeding focus was clearly visible. Starting from the blood injection, a round of MRI sequence scanning was completed every hour to continuously evaluate the signal changes of the bleeding focus within 17 hours.

[0144] See attached Figures 1a-3f , Using the above measurement method, the hourly measurement values ​​were continuously calculated, and it was found that the changes were not obvious within 17 hours, so the 17-hour data were averaged to obtain the average value for subsequent calculations. By directly measuring the signal intensity of T1-weighted, T2-weighted, and FLAIR images, it can be found that under 0.23T magnetic resonance, the hemorrhage T1WI showed a slightly low signal within 17 hours, and T2WI, FLAIR, and DWI showed high signals, and the signal intensity did not change significantly over time.

[0145] According to the above measurement method, the average values ​​of T1, T2, and proton density PD of brain parenchyma and hemorrhage are as follows:

[0146] Table 1

[0147]

[0148] 4. Ischemia Experiment Results

[0149] The cerebral ischemic area analyzed in this experiment is located in the medulla oblongata. The cerebral ischemic model also starts with the formation of cerebral infarction and continuously evaluates the signal changes of the bleeding focus within 24 hours.

[0150] See attached Figures 4a-6f Through 24-hour monitoring, it was found that under 0.23T magnetic resonance imaging, the T2WI, FLAIR, and DWI signals of cerebral ischemia continued to increase, while the T1-weighted signal continued to decrease.

[0151] According to the above measurement method, the normalized values ​​of T1, T2, and PD measured hourly are shown in Figure 6a-6cThe red line, because its data is not stable over time, the present invention believes that it is impossible to use a single component model to explain the phenomenon of cerebral infarction at different onset times. In the literature, Sean CL Deoni et al. also proposed that in magnetic resonance scanning, there are multi-component analysis models for the characteristics of tissues such as T1. According to the results of this experiment and the physiological process of cerebral infarction, the present invention believes that after cerebral infarction occurs, there are water molecules that can be developed by magnetic resonance in the lesion, which gradually change from brain parenchyma characteristics to edema-like characteristics (long T1, long T2), and the proportion of this transformation increases over time. Part of this water molecule is transformed from water molecules in the brain parenchyma, and part may be water entering from the outside of the lesion, and it continues to increase over time. Therefore, the present invention believes that the physiological process of cerebral infarction tissue can be simulated by using two components, brain parenchyma and edema-like, and the proportion of these two components changes over time. In addition, since the acquisition layer thickness of this experiment is 7mm, and the pig brain is relatively small, the thickness of the brain tissue slice at the medulla oblongata is less than 7mm, and the scanning layer slice at this position is mixed with cerebrospinal fluid signals, so the simulation calculation model of this experiment is three different components, namely brain parenchyma, cerebrospinal fluid and edema-like. Due to the entry of external water, the content of brain parenchyma decreases over time, and its PD value decreases. The content of edema-like substances increases over time, which is manifested as an increase in its PD value. Since cerebrospinal fluid is a real tissue in the slice, its content remains unchanged.

[0152] Therefore, the simulation formula described in measurement method 1 is changed to a multi-component formula in the cerebral infarction test, as follows:

[0153] Multi-component T1 acquisition sequence signal fitting formula:

[0154]

[0155] The fitting formula for multi-component T2 acquisition sequence signal is:

[0156]

[0157] The fitting formula for multi-component T1-weighted imaging and FLAIR imaging signals is:

[0158]

[0159] The signal fitting formula for multi-component PD-weighted imaging and T2-weighted imaging is:

[0160]

[0161] Where i = 1, 2, …, N represents N components. According to the above description, N for the cerebral infarction test is 3. The meanings of T1, T2, and PD are the same as above, and the subscript i represents the value of the i-th tissue.

[0162] Among them, the expressions of the PD values ​​of the three tissues changing with time are as follows:

[0163] The brain parenchyma content decreases with time t (unit: hours), and the expression is:

[0164]

[0165] The cerebrospinal fluid content remains unchanged and is expressed as:

[0166]

[0167] The edema content increases with time t (unit: hour), and the expression is:

[0168]

[0169] Substitute the PD formulas of the three tissues into the above formula for calculation.

[0170] The scanning parameters of the T1 measurement acquisition sequence are as follows: the first round of scanning TR is 4000ms, the last echo time TE is 107ms, and the values ​​of a series of TI values ​​τ are [100150 200 250 400 600 800 1000]ms respectively; the second round of scanning TR is 10000ms, the last echo time TE is 107ms, and the values ​​of a series of TI values ​​τ are [3500 4000 4500]ms respectively. Value, by comparing data τ with data The above formula is used for numerical fitting to obtain the final T1 measurement value.

[0171] T2 measurement acquisition sequence scanning parameters: TR 4000ms, a series of echo time τ values ​​of [42, 56, 70, 84, 98, 112]ms, correspondingly a series of value, by comparing the data β with the data The above formula is used for numerical fitting to obtain the final T2 measurement value.

[0172] T1-weighted image scanning parameters: TR 1420ms, TE 24.3ms, TI 425ms.

[0173] FLAIR image scanning parameters: TR 4380ms, TE 104ms, TI 1600ms.

[0174] PD-weighted image scanning parameters: TR 10000ms, TE 21ms.

[0175] T2-weighted image scanning parameters: TR 2800ms, TE 104ms.

[0176] In the three components of this experiment, T1 and T2 of brain parenchyma and cerebrospinal fluid have been measured by formula 1 and remain unchanged. The unknown variable in the above formula is the time change rate of brain parenchyma PD b. 脑实质 , PD value of cerebrospinal fluid, T1 and T2 values ​​of edema-like tissue, and time rate of change of PDb 类水肿 By integrating the above six sets of data into their respective formulas and performing numerical fitting, we obtain the following values:

[0177] Table 2

[0178] Brain parenchyma Cerebrospinal fluid Edema T1 (ms) 420 4200 2200 T2 (ms) 105 2200 500 PD (normalized) 1.15-0.00625t 0.15 0.00167t

[0179] Note: 1. The unit of time t is hour (h)

[0180] 2. Due to insufficient anesthesia of the pigs in the 6th hour, movement dislocation occurred. The experimental data of the first 6 hours were not included in the fitting calculation, and the subsequent values ​​were highly correlated with the fitting values.

[0181] 5. Sequence parameters for intracerebral hemorrhage identification

[0182] According to the data obtained in Sections 3 and 4, the inversion recovery sequence is used to identify cerebral hemorrhage, and the relative signal intensity formula is as follows:

[0183]

[0184] This formula is used for simulation calculation. Considering the actual scanning, the influence of T2 effect on the signal is reduced as much as possible. The TE of the scanning sequence adopts the minimum TE value of the system, so that As close to 1 as possible, the T2 effect has the least effect on the signal, and the contrast between cerebral hemorrhage and cerebral infarction tissue and brain parenchyma is optimal. The longer the TE, the closer the contrast between cerebral hemorrhage and cerebral infarction and brain parenchyma is, which is not conducive to discrimination. However, a TE slightly higher than the minimum TE value can be appropriately selected to distinguish the contrast between cerebral hemorrhage and cerebral infarction and brain parenchyma. Given a certain TR value, numerical fitting is performed according to the formula with TI and relative signal intensity S. Among them, brain parenchyma and cerebral hemorrhage are single-component models N=1, and T1, T2, and PD values ​​use the parameters in Table 1; since there is no cerebrospinal fluid doping in cerebral infarction tissue in the cerebral ischemia model, cerebrospinal fluid is removed, and a two-component model is used, N=2, and the simulated two components are brain parenchyma and edema-like tissue. T1, T2, and PD values ​​use the data in Table 2. The cerebral ischemia model simulates the signal changes every hour within 24 hours, and the PD value within 24 hours uses the initial value and linear change coefficient in Table 2.

[0185] Through fitting calculations, it was found that for any given TR value, there is a critical point for signal intensity. This critical point makes the signals of the brain parenchyma and the 1-24 hour cerebral infarction tissue equal, and the hemorrhage signal is greater than the signals of the brain parenchyma and the 1-24 hour cerebral infarction tissue. This signal intensity is manifested in the image as bright signals for cerebral hemorrhage and equal signals for the brain parenchyma and the 1-24 hour cerebral infarction tissue. Figure 7a -c Taking TR of 900ms, 1100ms, and 1500ms as an example, the solid lines of three colors in the legend correspond to the cerebral hemorrhage, brain parenchyma, and cerebral infarction signals at the 24th hour, respectively, and the dotted lines correspond to the cerebral infarction signals from the 1st hour to the 23rd hour after the onset of cerebral infarction. It can be concluded that the critical point is TI 685ms when TR 900ms, TI 800ms when TR 1100ms, and TI 1000ms when TR 1500ms.

[0186] A series of TR [900 1000 1100 1200 1300 1400 1500 1750 20003000ms] were selected for simulation, and a series of critical points TI were obtained. The quadratic polynomial was used for fitting. Figure 8 , it can be concluded that the optimal correspondence between TR and TI conforms to the following fitting results:

[0187] TR = 0.0006925*TI 2 +0.7426*TI+67.78

[0188] Generally speaking, if TI is fixed and TR value fluctuates by 10%, or if TR is fixed and TI fluctuates by 10%, it will have little impact on image quality.

[0189] Generally, the selection principle of the combination of TR value and corresponding TI value is to make the signal of cerebral hemorrhage high signal, and the signal of cerebral infarction tissue and brain parenchyma equal or low signal. The selection of TR and TI conforms to the principle of fastest clinical scanning speed. The scanning time requirement can also be appropriately relaxed if it is acceptable to the patient, so that the contrast between cerebral hemorrhage and cerebral infarction tissue and brain parenchyma is stronger and easier to identify.

[0190] Clinical results TR900, 1100 and 1500ms were selected for testing. The results are shown in Figure 9a -c, measure the grayscale values ​​of the hemorrhage (marked in red) and the contralateral normal brain tissue (marked in blue), and compare them with the simulation results. The results are shown in the following table

[0191]

[0192] It can be seen that the test results are in good agreement with the fitted values.

[0193] 5. Stroke Cases

[0194] The following clinical trial was conducted using a device with a low-field MRI stroke identification sequence, which verified that the low-field MRI stroke identification sequence can quickly and accurately determine hemorrhagic stroke. The device used was the mobile head and neck MRI system ACUTA Elfin manufactured by Foshan Regato Medical Technology Co., Ltd. The TR / TI combination used in the following case was 1100 / 800ms, and TE was the system's minimum value of 24ms.

[0195] Case 1: A 47-year-old male patient presented with left limb weakness for 24 hours. After admission, a CT scan confirmed cerebral infarction. He then entered the mobile head and neck MRI system ACUTA Elfin for scanning. Figures 10a-10d The images shown in the figure show that the DWI image shows multiple bright signals in the right frontal lobe and basal ganglia, the ADC corresponding lesion area is low signal, the frontal lobe is isosignal, the FLAIR corresponding lesion area is high signal, and the PD corresponding lesion area is isosignal. Based on the images and combined with the clinical comprehensive judgment, this case is diagnosed as multiple cerebral infarctions in the acute / subacute stage.

[0196] Case 2: A 64-year-old female patient presented with slurred speech and general weakness for 24.6 hours. She was scanned by the mobile head and neck MRI system ACUTA Elfin. Figures 11a-11e The images shown in the figure show that the DWI image shows a bright signal on the upper edge of the left basal ganglia, the ADC corresponding to the lesion area is a low signal, the FLAIR corresponding to the lesion area is a high signal, and the PD corresponding to the lesion area is an equal signal. Based on the image and combined with the clinical comprehensive judgment, this case is acute cerebral infarction. However, CT suggests the possibility of cerebral infarction, and further examination is recommended, but an accurate judgment cannot be made.

[0197] Case 3: A 50-year-old male patient presented with dizziness and sudden weakness of the left limbs for 29 hours. After admission, he was scanned by the mobile head and neck MRI system ACUTA Elfin. Figures 12a-12d The images shown in the figure show that the DWI image shows a bright signal in the right brainstem, the ADC corresponding to the lesion area is a low signal, and the FLAIR corresponding to the lesion area is a high signal. The initial diagnosis is a stroke, and the PD corresponding to the lesion area is a bright signal. Based on the image and combined with the clinical comprehensive judgment, this case is an acute cerebral hemorrhage. At the same time, the CT report is acute cerebral hemorrhage, which is consistent with the imaging results of the mobile head and neck magnetic resonance system ACUTA Elfin.

[0198] Case 4: A 50-year-old male patient presented with numbness on the right side of his body. He was admitted to the hospital and scanned with the mobile head and neck MRI system ACUTA Elfin. Figures 13a-13dThe images shown in the figure show that the DWI image shows bright signals in the left basal ganglia and corona radiata, the ADC corresponding to the lesion area is low signal, and the FLAIR corresponding to the lesion area is high signal. It is initially judged to be a stroke, and the PD corresponding to the lesion area is a bright signal. Based on the image and combined with the clinical comprehensive judgment, this case is an acute cerebral hemorrhage. At the same time, the CT report is acute cerebral hemorrhage, which is consistent with the imaging results of the mobile head and neck magnetic resonance system ACUTA Elfin.

[0199] Case 5: A 59-year-old female patient was admitted to the hospital with abnormal vision and headache for 24 hours. She then entered the mobile head and neck MRI system ACUTA Elfin for scanning. Figures 14a-14d The images shown in the figure show that the DWI image shows a bright signal under the left posterior ventricle, the ADC corresponding lesion area is a low signal, and the FLAIR corresponding lesion area is a high signal. It is initially judged to be a stroke. The PD corresponding lesion area is a bright signal. Based on the image and combined with the clinical comprehensive judgment, this case is subacute cerebral hemorrhage. At the same time, the CT report was indeterminate, but after clinical verification, this case was subacute cerebral hemorrhage.

[0200] Case 6: A 43-year-old male patient was admitted to the hospital for the first time with sudden left limb weakness 5.5 hours ago without obvious cause. He then entered the mobile head and neck magnetic resonance imaging system ACUTA Elfin device for scanning, and the following results were obtained: Figures 15a-15d The images shown in the figure show that the DWI image shows a large area of ​​bright signals in the right basal ganglia, insula, and corona radiata, the ADC corresponding to the lesion area is low signal, and the FLAIR corresponding to the lesion area is high signal. It is initially judged to be a stroke, and the PD corresponding to the lesion area is a bright signal. Based on the image and combined with the clinical comprehensive judgment, this case is acute cerebral hemorrhage. At the same time, the CT report is acute cerebral hemorrhage, which is consistent with the imaging results of the mobile head and neck magnetic resonance system ACUTA Elfin.

Claims

1. A method for generating a low-field magnetic resonance stroke identification sequence, characterized in that: The method uses an inversion recovery sequence, and the relative signal intensity formula is as follows: In the formula, S represents the relative signal strength; PD stands for tissue proton density; TI represents the inversion recovery time of the sequence; T1 represents the T1 relaxation time of the tissue; TR represents the repetition time of the sequence; TE last Represents the last echo time of a multi-echo sequence; TE stands for effective echo time; T2 represents the T2 relaxation time of the tissue; i=1, 2, …, N represents N components, where brain parenchyma and cerebral hemorrhage are single-component models N=1, and cerebral infarction tissue adopts a two-component model, N=2; subscript i represents the value of the i-th tissue; The brain parenchyma and cerebral hemorrhage were used as a single-component model, N=1, and the cerebral infarction tissue was used as a two-component model, N=2; Among them, the TE system can achieve the minimum or close to the minimum TE value, so that As close to 1 as possible, the T2 effect has the least impact on the signal S; The TR value and the corresponding TI value combination are selected so that the cerebral hemorrhage signal is high signal, and the cerebral infarction tissue and brain parenchyma are equal signal or low signal.

2. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: The selection of the TR value and the corresponding TI value combination complies with the clinical principle of the fastest scanning speed.

3. The low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: The corresponding relationship between TR and TI is consistent with the following fitting results: TR = 0.0006925*TI 2 +0.7426*TI+67.

78.

4. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 3, characterized in that: The TI is fixed and the TR value fluctuates by 10%; or the TR is fixed and the TI fluctuates by 10%.

5. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: The low field is 0.23 T, the TE is 24 milliseconds, and the TI is 685 milliseconds when the TR is 900 milliseconds.

6. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: The low field is 0.23 T, the TE is 24 milliseconds, and the TI is 800 milliseconds when the TR is 1100 milliseconds.

7. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: The low field is 0.23 T, the TE is 24 milliseconds, the TR is 1500 milliseconds and the TI is 1000 milliseconds.

8. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: The PD is obtained by the signal intensity of the proton-weighted image.

9. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: Before the fast spin echo sequence, a series of inversion recovery pulses with different TI values ​​are applied. At this time, the signal size of the region of interest is related to the T1 value. The following formula is used for fitting to obtain the T1 measurement value: in, is the tissue signal strength; α is the weight coefficient; τ is the TI value of the scanning sequence; TR is the repetition time; TE last is the last echo time; T1 is the T1 relaxation time of the tissue.

10. The method for generating a low-field magnetic resonance stroke identification sequence according to claim 1, characterized in that: Using a fast spin echo sequence, keeping all parameters except TE time unchanged, using the correlation between TE and T2, the following formula is used for fitting to obtain the T2 measurement value: in, is the tissue signal strength; α is the weight coefficient; β is a series of echo times; T2 is the T2 relaxation time of the tissue.

11. A low-field magnetic resonance stroke identification device, comprising a magnetic resonance scanner; characterized in that: The magnetic resonance scanner executes a low-field magnetic resonance stroke identification sequence generated by the method according to any one of claims 1-8.

Images