An automatic analysis system and method for MRA brain images

Abstract

The present invention discloses an automatic analysis system for MRA brain images, comprising an image acquisition module, a first database, a signal display module, and a first processor, which is used to intercept the second image; the first processor determines whether the number of closed figures in the second image whose grayscale is greater than a first threshold value exceeds one, and if so, outputs an alarm signal as the first signal, and otherwise outputs the first signal through a second method; the present invention determines whether the blood vessels in the brain image are interconnected or exist by identifying the number of closed figures through the first processor, and if there is only one closed figure, it indicates that the cerebral blood vessels are well connected, and if there are more than one closed figures, it indicates that there are narrow or closed blood vessels in the cerebral blood vessels, thereby improving the speed of patients seeking medical treatment.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging, and in particular to an automatic analysis system for MRA brain images which can automatically screen MRA images, effectively reduce the labor intensity of medical staff, and improve medical treatment efficiency. Background Art

[0002] Magnetic resonance angiography (MRA) is a method of examination that uses electromagnetic waves to produce images of the two-dimensional structure of the body. It is a type of tomographic imaging that uses the magnetic resonance phenomenon to obtain electromagnetic signals from the human body and reconstruct human body information. The grayscale characteristics of magnetic resonance imaging are that the stronger the magnetic resonance signal, the greater the brightness, and the weaker the magnetic resonance signal, the lower the brightness. Conventional MRA has the advantages of being non-invasive, not requiring contrast agents, and having low radiation. It has become one of the indispensable detection technologies in clinical practice, but it only has three colors: white, gray, and black. In addition, due to the convenience of detection, more patients use it, which results in more images. For medical staff and imaging personnel, the workload is large. Once they are not careful, it is easy to cause irreversible consultation time. Patients often need to wait for a long time to get the results.

[0003] Therefore, there is an urgent need for an automatic analysis system for MRA brain images that can automatically screen MRA images, effectively reduce the labor intensity of medical staff, and improve medical efficiency. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide an automatic analysis system for MRA brain images which can automatically screen MRA images, effectively reduce the labor intensity of medical staff, and improve medical efficiency.

[0005] The present invention provides an automatic analysis system for MRA brain images, comprising:

[0006] An image acquisition module, used for acquiring a first image;

[0007] A first database, which is used to store initial threshold values;

[0008] A signal display module, which is used to display the first signal and the received second image;

[0009] The first signal includes an alarm signal and an observation signal;

[0010] a first processor, configured to identify the first image to obtain a first center point, and cut out the second image from the first image with the first center point as the center and a first length as the radius;

[0011] The first processor converts the initial threshold into a first threshold after receiving the first image;

[0012] The first processor determines whether the number of closed figures in the second image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs an alarm signal as the first signal, and if not, outputs the first signal through a second method;

[0013] The first processor outputs the second image corresponding to the alarm signal to the information display module while outputting the alarm signal.

[0014] The present invention provides an automatic analysis system for MRA brain images, wherein the first method is to select the centroid of the first image as the first center point.

[0015] The present invention provides an automatic analysis system for MRA brain images, wherein the second method is to directly output the observation signal.

[0016] The present invention provides an automatic analysis system for MRA brain images, wherein the first processor screens and removes closed figures whose areas do not exceed a ninth threshold, and the ninth threshold is 0.5% of the area of ​​the first image.

[0017] The present invention provides an automatic analysis system for MRA brain images, wherein the step of "identifying the first image to obtain the first center point" comprises the following steps:

[0018] The first processor determines whether a circle with a radius greater than a third length can be placed in a closed figure with a gray value greater than a first threshold in the first image, and if not, selects the centroid of the first image as the first center point, and if yes, draws a first number of first circles in the closed figure with a gray value greater than the first threshold according to the radius from approximately small to large, and each of the first circles does not overlap, and uses the center of the first circle as a preliminary positioning point;

[0019] A first number of the preliminary positioning points of the first circle are used as first positioning points, and a figure formed by connecting the first positioning points with straight lines that do not intersect each other is formed as a first closed figure;

[0020] The first processor determines whether the straight-line distance between the centroid of the first closed figure and the centroid of the first image exceeds a fourth length. If so, the centroid of the first image is selected as the first center point. Otherwise, the midpoint of the line formed by the centroid of the first image and the centroid of the first closed figure is selected as the first center point.

[0021] The present invention provides an automatic analysis system for MRA brain images, which further includes a second processor.

[0022] The information display module is also used to display the received third image;

[0023] The second processor includes an instruction acquisition module, and the instruction acquisition module is used to acquire a start instruction or a stop instruction;

[0024] The second processor starts running or continues running when receiving the start instruction, and stops running when receiving the stop instruction;

[0025] The second method comprises the following steps: s2.1, the first processor sends the start instruction to the second processor;

[0026] S2.2, the second processor increases the first length by the first variable to obtain a second length;

[0027] s2.3 the second processor cuts out the third image from the first image with the first center point as the center and the second length as the radius;

[0028] S2.4, the second processor determines whether the number of closed figures in the third image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs the alarm signal as the first signal; if not, determines whether the second length is greater than the first multiple of the long side of the first image, and if so, outputs the stop instruction and the observation signal as the first signal; otherwise, runs through the third method;

[0029] The second processor outputs the third image corresponding to the alarm signal to the information display module after outputting the alarm signal;

[0030] After outputting the observation signal, the second processor outputs the third image corresponding to the observation signal to the information display module.

[0031] The present invention provides an automatic analysis system for MRA brain images, wherein the third method is: the second processor deletes the third image, deletes the first length, converts the second length to the first length, and jumps to s2.2.

[0032] The present invention provides an automatic analysis system for MRA brain images, wherein the second processor comprises an instruction acquisition module, and the instruction acquisition module is a keyboard and buttons.

[0033] The present invention provides an automatic analysis system for MRA brain images, wherein the first database is electrically connected to the first processor.

[0034] The present invention provides an automatic analysis system for MRA brain images, wherein the image display module is a liquid crystal display screen.

[0035] The present invention provides an automatic analysis system for MRA brain images, wherein the second processor is electrically connected to the image display module.

[0036] The difference between the automatic analysis system of MRA brain images of the present invention and the prior art is that the present invention uses the first processor to intercept the second image and then preferentially detect the central part of the brain with a high priority, and identifies the number of closed figures to determine whether the blood vessels in the brain image are interconnected or exist. If there is only one closed figure, it means that the cerebral blood vessels are well connected. If there are more than one closed figures, it means that there are narrow or closed blood vessels in the cerebral blood vessels. Then, an alarm signal is output as a first signal to remind the doctor to pay attention. If there are, an alarm signal is output to achieve the purpose of reminding the doctor. This not only reduces the workload of radiologists and avoids doctors from looking for narrow or closed parts from the messy cerebral blood vessels, but also promotes faster completion of test results and improves the speed of patients' medical treatment.

[0037] The automatic analysis system of MRA brain images of the present invention is further described below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a first schematic diagram of an automatic analysis system for MRA brain images;

[0039] Figure 2 is a second schematic diagram of an automatic analysis system for MRA brain images;

[0040] Figure 3 is a third schematic diagram of an automatic analysis system for MRA brain images;

[0041] Figure 4 is a fourth schematic diagram of an automatic analysis system for MRA brain images;

[0042] Figure 5 yes Figure 3 A partial enlarged view of

[0043] Figure 6 It is a schematic diagram of the interface of the information display module;

[0044] Figure 7 This is a schematic diagram of a misjudgment situation. DETAILED DESCRIPTION

[0045] like Figures 1 to 7 As shown, see Figure 1 , 2 , 5, 6. An automatic analysis system for MRA brain images, characterized in that: it includes

[0046] An image acquisition module, used for acquiring a first image;

[0047] A first database, which is used to store initial threshold values;

[0048] A signal display module, which is used to display the first signal and the received second image;

[0049] The first signal includes an alarm signal and an observation signal;

[0050] a first processor, configured to identify the first image to obtain a first center point, and cut out the second image from the first image with the first center point as the center and a first length as the radius;

[0051] The first processor converts the initial threshold into a first threshold after receiving the first image;

[0052] The first processor determines whether the number of closed figures in the second image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs an alarm signal as the first signal, and if not, outputs the first signal through a second method;

[0053] The first processor outputs the second image corresponding to the alarm signal to the information display module while outputting the alarm signal.

[0054] The present invention uses the first processor to intercept the second image and then preferentially detect the central part of the brain with a high priority, and determines whether the blood vessels in the brain image are interconnected or exist by identifying the number of closed figures. If there is only one closed figure, it means that the cerebral blood vessels are well connected. If there are more than one closed figures, it means that there are narrow or closed blood vessels in the cerebral blood vessels, and then outputs an alarm signal as a first signal to remind the doctor to pay attention. If there are, an alarm signal is output to achieve the purpose of reminding the doctor. This not only reduces the workload of radiologists and avoids doctors from looking for narrow or closed parts from the messy brain blood vessels, but also promotes faster completion of detection results and improves the speed of patients seeking medical treatment.

[0055] For example, after testing a patient, a first image is obtained, and after intercepting the first image, a second image is obtained. After testing the second image, two white closed figures are found in the image, and the information display module displays an alarm signal, which means that there may be a blockage in the middle of the patient's brain blood vessels. After observing the second image intercepted from the first image, the attending physician quickly finds the blocked area in the second image which is smaller than the first image, and urgently uses thrombolytic agents to stabilize the patient's condition, and consults with relevant doctors to prepare for surgery.

[0056] The first length may be 8% to 12%, preferably 10%, of the long side of the first image.

[0057] The present invention sets the first length to 8% to 12%, preferably 10%, of the long side of the first image, thereby reducing the search range required by the doctor and alleviating the doctor's workload.

[0058] Wherein, the blood vessels are identified as closed figures in the second image. How to identify the blood vessels as closed figures is a prior art and will not be described in detail here.

[0059] The initial threshold may be 208-256, preferably 240.

[0060] The present invention sets the initial threshold to 208-256, preferably 240, thereby screening peripheral blood vessels and capillaries with slower flow rates.

[0061] Among them, the technology for determining a closed figure is prior art and will not be described in detail here.

[0062] Among them, Figure 1 , 2 As shown, the edge line of the second image intercepted from the first image can be used as a component of the closed figure.

[0063] For further explanation of the present invention, see Figure 1 , 2 , 5. The first method is to select the centroid of the first image as the first center point.

[0064] The present invention improves the capture efficiency of the first center point by setting the first method to select the center of mass of the first image as the first center point.

[0065] For further explanation of the present invention, see Figure 1 , 2 , the second method is to directly output the observation signal.

[0066] The present invention sets the second method to directly output the observation signal, thereby achieving the goal of outputting the observation signal when it is determined that there is no blood vessel blockage, allowing the doctor to continue to treat the patient to maintain the current condition.

[0067] As a further explanation of the present invention, the first processor screens and removes closed figures whose areas do not exceed a ninth threshold, and the ninth threshold may be 0.5% of the area of ​​the first image.

[0068] The present invention uses the first processor to screen and eliminate closed figures whose areas do not exceed the ninth threshold value, thereby removing interfering bright spots and other factors in the image.

[0069] For further explanation of the present invention, see Figure 1 , 2, 4, 5, 6, 7, the “identifying the first image to obtain the first center point” comprises the following steps:

[0070] The first processor determines whether a circle with a radius greater than a third length can be placed in a closed figure with a gray value greater than a first threshold in the first image, and if not, selects the centroid of the first image as the first center point, and if yes, draws a first number of first circles in the closed figure with a gray value greater than the first threshold according to the radius from approximately small to large, and each of the first circles does not overlap, and uses the center of the first circle as a preliminary positioning point;

[0071] A first number of the preliminary positioning points of the first circle are used as first positioning points, and a figure formed by connecting the first positioning points with straight lines that do not intersect each other is formed as a first closed figure;

[0072] The first processor determines whether the straight-line distance between the centroid of the first closed figure and the centroid of the first image exceeds a fourth length. If so, the centroid of the first image is selected as the first center point. Otherwise, the midpoint of the line formed by the centroid of the first image and the centroid of the first closed figure is selected as the first center point.

[0073] The present invention selects the central area of ​​a blood vessel with a larger width as a preliminary positioning point to ensure that it is closer to the individual brain center of different patients, and then uses the distance between the center of mass of the first image and each first positioning point as a screening condition to select a figure composed of a first number of first positioning points as a first closed figure. The midpoint of the line formed by the center of mass of the first closed figure and the center of mass of the first image is used as the first center point to reduce the error of different central areas of the brain caused by individual body growth differences, so that the center of the second image is closer to the patient's brain center, thereby enhancing accuracy.

[0074] For example, the first processor selects 16 points with the largest blood vessel volume width in the first image as the first positioning points, and uses the center points of the 10 first circles with the largest blood vessel volume width in the first image as the points constituting the first closed figure, and connects the 10 points in sequence to obtain a decagon as the first closed figure, and uses the midpoint of the line connecting the center of mass of the decagon and the center of mass of the first image as the first center point. The first center point thus obtained is closer to the central area of ​​the patient's brain.

[0075] The third length is 2-4% of the first image.

[0076] The present invention sets the third length to 2-4% of the first image, thereby screening the aorta with a larger width in the brain.

[0077] Among them, the first first circle is the first circle with the largest radius that can be placed in the closed figure whose gray value is greater than the first threshold.

[0078] The first number is an even number, and may be 4, 6, 8, 10 to 2n, where n is a positive integer, and the first number is preferably 4.

[0079] The present invention ensures that a closed figure can be formed by setting the first number to be at least 3, and as the first number increases, the brain center of the patient can be estimated more accurately.

[0080] The present invention sets the first number to an even number so as to keep the center of mass in a relatively central area to the greatest extent possible, thereby avoiding the occurrence of a serious center of mass shift.

[0081] The third length may be 1-3%, preferably 1.5%, of the long side of the first image.

[0082] The present invention sets the third length to 1-3%, preferably 1.5%, of the long side of the first image, thereby achieving the position of the artery.

[0083] The fourth length may be 5-8%, preferably 6%, of the long side of the first image.

[0084] The present invention sets the fourth length to 5-8%, preferably 6%, of the long side of the first image, thereby avoiding the deviation of the first center point caused by individual differences.

[0085] The connection method of “a figure formed by connecting the first number of points with straight lines that do not cross each other is a first closed figure” is to connect them using straight lines, and each of the first positioning points is only the endpoint of two line segments.

[0086] For further explanation of the present invention, see Figure 1 , 2 , 3, 4, 5, 7, and further comprising a second processor,

[0087] The information display module is also used to display the received third image;

[0088] The second processor includes an instruction acquisition module, and the instruction acquisition module is used to acquire a start instruction or a stop instruction;

[0089] The second processor starts running or continues running when receiving the start instruction, and stops running when receiving the stop instruction;

[0090] The second method comprises the following steps: s2.1, the first processor sends the start instruction to the second processor;

[0091] S2.2, the second processor increases the first length by the first variable to obtain a second length;

[0092] s2.3 the second processor cuts out the third image from the first image with the first center point as the center and the second length as the radius;

[0093] S2.4, the second processor determines whether the number of closed figures in the third image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs the alarm signal as the first signal; if not, determines whether the second length is greater than the first multiple of the long side of the first image, and if so, outputs the stop instruction and the observation signal as the first signal; otherwise, runs through the third method;

[0094] The second processor outputs the third image corresponding to the alarm signal to the information display module after outputting the alarm signal;

[0095] After outputting the observation signal, the second processor outputs the third image corresponding to the observation signal to the information display module.

[0096] The present invention uses the second processor to continuously expand the area of ​​the captured image, and then first checks the core area of ​​the brain, thereby eliminating the most serious situation as quickly as possible, and then checks the edge area of ​​the brain, effectively detecting the conditions of all cerebral blood vessels, and then outputting to help doctors quickly find blocked or narrowed blood vessels from a third image that involves a larger area but is smaller than the first image, thereby reducing the difficulty and workload of doctors in reading the first image, helping doctors to quickly make diagnoses, and helping patients to quickly receive corresponding treatments.

[0097] For example, after testing a patient, a first image is obtained, and after intercepting the first image, a second image is obtained. After detecting the second image, a white closed figure is found in the image, indicating that there is no vascular stenosis or blockage so far. After increasing the interception area, a third image is generated. After detecting the third image, two white closed figures are found in the image. The information display module displays an alarm signal, indicating that the patient's brain blood vessels in the middle area may be blocked. After the radiologist observes the third image with a smaller display range than the first image, the suspected stenotic blood vessel area is quickly found and the patient's family and attending physician are notified. The attending physician uses nitroglycerin to temporarily increase the patient's blood vessel width, and prepares for minimally invasive stent surgery after discussing with the patient's family.

[0098] The third method is to output the observation signal as a first signal.

[0099] The first variable may be 3% to 10% of the long side of the first image, preferably 8%.

[0100] The present invention sets the first variable to be 3% to 10%, preferably 8%, of the long side of the first image, thereby achieving the purpose of expanding the search range.

[0101] The first multiple is a value obtained by taking the square root of the sum of the square of the short side and the square of the long side of the first image, that is, the hypotenuse formed by the short side and the long side in the first image.

[0102] The present invention sets the first multiple, thereby reducing the inefficient identification process, avoiding unlimited magnification range identification, and thus reducing the time patients wait for results.

[0103] The relationship between the first variable and the first multiple is set so that the first variable cannot directly jump to the output observation signal.

[0104] As a further explanation of the present invention, the second processor includes an instruction acquisition module, and the instruction acquisition module is a keyboard and a button.

[0105] The present invention arranges the second processor to include an instruction acquisition module, wherein the instruction acquisition module is a keyboard and buttons, so that an operator can input a start instruction or a stop instruction at any time through the instruction acquisition module.

[0106] As a further explanation of the present invention, the first database is electrically connected to the first processor.

[0107] The present invention arranges the first database to be electrically connected to the first processor, thereby helping the first processor to retrieve data more efficiently.

[0108] As a further explanation of the present invention, the image display module is a liquid crystal display screen.

[0109] The present invention sets the image display module as a liquid crystal display screen, thereby facilitating medical staff to observe more intuitively.

[0110] As a further explanation of the present invention, the second processor is electrically connected to the image display module.

[0111] The present invention arranges the second processor to be electrically connected to the image display module, thereby helping the signal to be transmitted to the image display module more quickly for display.

[0112] For further explanation of the present invention, see Figure 1 , also includes a correction module,

[0113] The correction module determines whether the third method has been executed, and if so, no action is taken; otherwise, the first processor and the second processor jump to execute the third method after outputting the alarm signal;

[0114] The third method further comprises the following steps:

[0115] The correction module determines whether the first threshold is less than the third threshold; if so, outputs the stop instruction and the first signal of the observation signal; if not, the correction module reduces the first threshold by a second variable to generate the second threshold, deletes the first threshold, and converts the second threshold into the first threshold;

[0116] The first processor determines whether the number of closed figures in the third image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs the alarm signal and jumps to s2.2, and otherwise jumps to s2.2.

[0117] The present invention can increase the area of ​​closed figures in the image by changing the first threshold value, thereby screening a larger area of ​​brain blood vessel distribution. At the same time, weakening the screening intensity of the image is conducive to enhancing the sensitivity of recognition, so that the image area remains unchanged while the area of ​​the figure in the image increases, thereby making it easier to diagnose areas of insufficient blood supply or vascular stenosis. The grayscale threshold is changed to better eliminate the blood vessels in the front of the central area, thereby avoiding affecting the normal results. Doctors can better observe whether the blood vessels are stenotic or blocked based on multiple images that output alarm signals, and promptly judge whether the situation is serious and give corresponding diagnosis and treatment plans.

[0118] The present invention increases the interception area and reduces the grayscale threshold through the third method, thereby judging the accuracy of the output signal from two aspects, avoiding the misjudgment of the alarm signal caused by incomplete graphics or grayscale critical value.

[0119] The third threshold may be 180-200, preferably 192.

[0120] The present invention sets the third threshold value to 180-200, preferably 192, so as to avoid wasting computing power and time due to excessive reduction of the gray value.

[0121] The second variable may be 5 to 10, preferably 8.

[0122] The present invention sets the second variable to 5 to 10, preferably 8, so as to gradually increase the displayed area, which is more conducive to doctors observing suspicious areas.

[0123] For example, after testing a patient, a first image is obtained, and after intercepting the first image, a second image is obtained. After detecting the second image, a white closed figure is found in the image, indicating that there is no vascular stenosis or blockage so far. After increasing the intercepted area, a third image is generated. After detecting the third image, two white closed figures are found in the image. The information display module displays an alarm signal, indicating that there is no vascular stenosis or blockage so far. The system continues to operate using the third method, outputs the third image to the image display module once after changing the first threshold, and then uses the fifth method to expand the area of ​​the intercepted first image to generate the third image, and determines the number of closed figures in the third image. Finally, after comparing multiple third images, the radiologist finds that the third image output in front shows complete blockage, and the third image output in the back shows a complete closed figure. Therefore, it is determined that this is a condition caused by vascular stenosis. After the radiologist informs the attending physician of the situation, the attending physician first uses nitroglycerin to temporarily increase the patient's blood vessel width, and prepares for minimally invasive stent surgery after discussing with the patient's family.

[0124] For further explanation of the present invention, see Figure 1 , 6 , further comprising a risk warning module, which is used to receive the alarm signal and the observation signal, and output the second signal to the display module through the fourth method;

[0125] The second signal includes a danger signal, an observation signal and a modification signal;

[0126] The fourth method comprises the following steps:

[0127] The risk warning module determines whether the observation signal is received, and if not, does not take any action; if so, determines whether the alarm signal has been received, and if otherwise, outputs the observation signal; if so, determines whether the number of alarm signals received exceeds a fourth threshold, and if otherwise, outputs the modification signal; if so, outputs the danger signal.

[0128] The present invention determines whether the observation signal is received and then determines whether to terminate the output of all first signals through the risk warning module, and then determines whether the alarm signal is received and then determines whether there is a suspected or possible vascular blockage and stenosis. If not, the observation signal is output to allow medical staff to continue medical observation of the patient. If yes, it is determined whether the number of alarm signals is too many. If too many, it means that there is a high possibility of vascular blockage and stenosis, and it is recommended to take relevant emergency treatment methods immediately. If not, it means that the advantage of conventional MRA is that it is non-invasive and does not require contrast agents. This also leads to certain errors due to the limitations of conventional MRA. Therefore, the modification signal is output to suggest modifying the detection method, such as using enhanced MRA or brain CT to re-detect, to ensure the rigor and accuracy of the detection results.

[0129] For example, after the patient is tested, the system detects the first image and outputs 8 alarm signals in total. The number of alarm signals is greater than the fourth threshold value 3, so the danger signal is output to the information display module to remind the doctor that the patient's condition is more serious.

[0130] For example, after the patient is tested, the system detects the first image and outputs 2 alarm signals in total. The number of alarm signals is less than the fourth threshold value 4, so the modification signal is output to the information display module to remind the doctor that the patient's condition cannot be judged using conventional MRA. The doctor recommends that the patient use enhanced MRA with injection of contrast agent for re-judgment to ensure the accuracy of the diagnosis.

[0131] Wherein, the risk processing module clears the record of the first signal received after outputting any one of the danger signal, observation signal and modification signal.

[0132] The fourth threshold may be 1 to 6 times, preferably 2 times.

[0133] The fourth threshold value of the present invention can be set to 1 to 6 times, preferably 2 times. Although conventional MRA is convenient and fast and can screen most cerebrovascular conditions, it is still necessary to provide a certain error tolerance for conventional MRA detection to avoid misdiagnosis caused by detection methods.

[0134] As a further explanation of the present invention, the database stores patient grades, which include primary patients, secondary patients, tertiary patients, and quaternary patients;

[0135] The database stores records of the modification signal and the danger signal, and the first processor determines whether the number of danger signals is greater than a fifth threshold value, and if so, the patient's disease level is a first-level patient; if not, if the number of danger signals is less than the fifth threshold value and greater than a sixth threshold value, and if so, the patient's disease level is a second-level patient; if otherwise, it is determined whether there is a modification signal, and if so, the patient's disease level is a third-level patient, and if not, the patient's disease level is a fourth-level patient;

[0136] The database stores the first variable and the second variable corresponding to disease levels.

[0137] The present invention can output signals more accurately by grading patients. If the number of times the danger signal is output is too much greater than the fifth threshold, it means that the patient is in a very fragile physical condition. If the number of times the danger signal is output is greater than the sixth threshold and less than the fifth threshold, it means that the patient is in a relatively weak physical condition. For patients whose number is less than the sixth threshold, it is judged whether they have modified the signal and then whether there is a potential risk. If so, it means that the patient's physical condition is in a sub-healthy state. Otherwise, it means that the patient's physical condition is better. The first variable and the second variable are adjusted according to the patient's physical condition, so that the image changes of low-level patients (such as first-level patients) are smaller but more sensitive each time compared with high-level patients (such as fourth-level patients), and they are more inclined to output more alarm signals, so that the system can more carefully examine patients with weaker bodies, which makes it more convenient for doctors to conduct detailed investigations on their impact.

[0138] For example, a patient is hospitalized for the first time and has no medical record for his physical condition before this hospitalization. Therefore, there is no record of danger signals and modification signals before this hospitalization. The corresponding disease level of this patient is a level 4 patient, and the corresponding first variable and second variable are used.

[0139] The fifth threshold may be 3 to 6, preferably 5.

[0140] The present invention sets the fifth threshold value to 3-6, preferably 5, so as to achieve a more scientific classification of weaker patients.

[0141] The sixth threshold may be 1 to 3, preferably 2.

[0142] The present invention sets the sixth threshold value to 1-3, preferably 2, so as to achieve a more scientific classification of patients with poor health.

[0143] The relationship between the disease level and the first variable and the second variable is shown in Table 1 below.

[0144] Disease level First variable Second variable Level 1 patients 3％ 5 Secondary patients 5％ 7 Level 3 patients 7％ 9 Level 4 patients 9％ 10

[0145] Table 1

[0146] As a further explanation of the present invention, the first processor obtains the first energy value Z by the following formula according to the number of bad habits X, the patient's age Y, the blood pressure value A, the total cholesterol value B, and the disease level C;

[0147]

[0148] The corresponding initial threshold and first quantity are obtained through the first energy value Z.

[0149] Specifically, the number of bad habits X in the formula is used as the numerator, so that the number of bad habits X is positively correlated with the first energy value Z, and the weight of the number of bad habits X is balanced by but not multiplied by a multiple; the patient's age Y is positively correlated with the first energy value Z as a part of the logarithm in the numerator, thereby increasing the purpose of more cautious testing required for older people; the blood pressure value A and the total cholesterol value B are separately added values, and the relationship between the two is positively correlated with the first energy value Z, and the total cholesterol value B is multiplied by 10 times, thereby enhancing its weight for the first energy value Z, effectively balancing the blood pressure value A. The higher the blood lipids, the slower the flow rate and the easier it is for blood vessels to be blocked. The higher the blood pressure, the faster the flow rate but the more likely the blood vessels are to be damaged. The disease level C is used as the denominator and is negatively correlated with the first energy value Z. The healthier the body, the smaller the first energy value Z, and the more fragile the body, the larger the first energy value Z.

[0150] The present invention calculates the first energy value Z through a formula, and obtains the corresponding initial threshold and the first quantity value through the first energy value Z, wherein the healthier the patient is, the higher the initial threshold, and the weaker the patient is, the smaller the initial threshold. Therefore, patients with weaker bodies will receive more detailed and focused examinations, and when selecting the center point, more positioning points will be selected to form a closed figure, further reducing individual errors.

[0151] For example, a patient aged 80 is a first-level patient, and his bad habits include smoking, alcoholism, and staying up late. His blood pressure A is 140 mm / Hg, his total cholesterol number B is 6 mmol / L, and the first energy value Z obtained by the first formula is 47.52. Therefore, the initial threshold value obtained by referring to Table 2 is 240, and the first number selected is 12, so that the selected central area is more accurate, and the number of judgments and detections is more.

[0152] For example, a 20-year-old patient is a level 4 patient with bad habits including staying up late, blood pressure A is 120 mm / Hg, total cholesterol number B is 4.9 mmol / L, and the first energy value Z obtained by the first formula is 18.76. Therefore, the initial threshold value obtained by referring to Table 2 is 230, and the first number selected is 4, so that the results can be obtained faster.

[0153] Wherein, the blood pressure value A is the systolic pressure.

[0154] Among them, the comparison relationship between the first energy value Z and the initial threshold, the first quantity, and the third length is shown in Table 2 below.

[0155] The first energy value Z Initial Threshold First quantity The third length (0，20] 230 4 4％ (20，35] 235 8 3％ (35，+∞) 240 12 2％

[0156] Table 2

[0157] As a further explanation of the present invention, the information display module is also used to display second information;

[0158] The second information includes: a central nerve threat signal, a middle nerve threat signal, and a peripheral nerve threat signal;

[0159] The first processor determines whether the number of alarm signals of the patient is greater than the seventh threshold. If so, it outputs a central nervous threat signal as the second information to the information display module. If not, it determines whether the number of alarm signals is less than the eighth threshold. If so, it outputs a peripheral nervous threat signal as the second information. If otherwise, it outputs a central nervous threat signal as the second information.

[0160] The present invention uses the first processor to identify the number of alarm signals and then roughly locate the diseased part. The more alarm signals output, the closer it is to the nerve center, and a central nerve threat signal is output; the higher the output danger level, the fewer alarm signals, the more marginalized it is, and the smaller the danger is, and a central nerve threat signal or a marginal nerve threat signal is output, thereby helping doctors to more quickly lock in the area of ​​the problematic blood vessels, reduce the labor intensity of radiologists, and speed up medical efficiency.

[0161] For example, after testing a patient, if a total of 8 alarm signals are output, it means that the system has found an area suspected of blockage or stenosis in the central part of the patient's brain. Therefore, the first processor outputs a central nervous threat signal as the second information to the information display module. After seeing the central nervous threat signal, the doctor indeed found the suspected blocked blood vessels in the central nervous area and contacted doctors in related fields of the central nervous system to discuss medical plans together.

[0162] The seventh threshold may be 7 to 10 times, preferably 8 times.

[0163] The present invention sets the seventh threshold to 7 to 10 times, preferably 8 times, so that when the seventh threshold is exceeded, an alarm is often triggered from the nerve center.

[0164] The eighth threshold may be 2 to 4 times, preferably 3 times.

[0165] In the present invention, the eighth threshold value may be set to 2 to 4 times, preferably 3 times, so as to reasonably divide the area where suspected blockage or stenosis occurs.

[0166] The embodiments described above are merely descriptions of preferred implementation modes of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. An automatic analysis system for MRA brain images, characterized in that: include An image acquisition module, used for acquiring a first image; A first database, which is used to store initial threshold values; An information display module, which is used to display the first signal and the received second image; The first signal includes an alarm signal and an observation signal; a first processor, configured to identify the first image to obtain a first center point, and cut out the second image from the first image with the first center point as the center and a first length as the radius; The first processor converts the initial threshold into a first threshold after receiving the first image; The first processor determines whether the number of closed figures in the second image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs an alarm signal as the first signal, and if not, outputs the first signal through a second method; The first processor outputs the second image corresponding to the alarm signal to the information display module while outputting the alarm signal; Also comprising a second processor, The information display module is also used to display the received third image; The second processor includes an instruction acquisition module, and the instruction acquisition module is used to acquire a start instruction or a stop instruction; The second processor starts running or continues running when receiving the start instruction, and stops running when receiving the stop instruction; The second method comprises the following steps: S2.1, the first processor sends the start instruction to the second processor; S2.2, the second processor increases the first length by the first variable to obtain a second length; S2.3, the second processor cuts out the third image from the first image with the first center point as the center and the second length as the radius; S2.4, the second processor determines whether the number of closed figures in the third image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs the alarm signal as the first signal; if not, determines whether the second length is greater than the first multiple of the long side of the first image, and if so, outputs the stop instruction and the observation signal as the first signal; otherwise, runs through the third method; The second processor outputs the third image corresponding to the alarm signal to the information display module after outputting the alarm signal; The second processor outputs the third image corresponding to the observation signal to the information display module after outputting the observation signal; The third method is to delete the third image, delete the first length, convert the second length to the first length, and jump to S2.2; The centroid of the first image is selected as the first center point.

2. An automatic analysis system for MRA brain images, characterized in that: include An image acquisition module, used for acquiring a first image; A first database, which is used to store initial threshold values; An information display module, which is used to display the first signal and the received second image; The first signal includes an alarm signal and an observation signal; a first processor, configured to identify the first image to obtain a first center point, and cut out the second image from the first image with the first center point as the center and a first length as the radius; The first processor converts the initial threshold into a first threshold after receiving the first image; The first processor determines whether the number of closed figures in the second image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs an alarm signal as the first signal, and if not, outputs the first signal through a second method; The first processor outputs the second image corresponding to the alarm signal to the information display module while outputting the alarm signal; Also comprising a second processor, The information display module is also used to display the received third image; The second processor includes an instruction acquisition module, and the instruction acquisition module is used to acquire a start instruction or a stop instruction; The second processor starts running or continues running when receiving the start instruction, and stops running when receiving the stop instruction; The second method comprises the following steps: S2.1, the first processor sends the start instruction to the second processor; S2.2, the second processor increases the first length by the first variable to obtain a second length; S2.3, the second processor cuts out the third image from the first image with the first center point as the center and the second length as the radius; S2.4, the second processor determines whether the number of closed figures in the third image whose grayscale is greater than the first threshold value exceeds one, and if so, outputs the alarm signal as the first signal; if not, determines whether the second length is greater than the first multiple of the long side of the first image, and if so, outputs the stop instruction and the observation signal as the first signal; otherwise, runs through the third method; The second processor outputs the third image corresponding to the alarm signal to the information display module after outputting the alarm signal; The second processor outputs the third image corresponding to the observation signal to the information display module after outputting the observation signal; The third method is to delete the third image, delete the first length, convert the second length to the first length, and jump to S2.2; The identifying the first image to obtain the first center point comprises the following steps: The first processor determines whether a circle with a radius greater than a third length can be placed in a closed figure whose grayscale value in the first image is greater than a first threshold, and if not, selects the centroid of the first image as the first center point, and if yes, draws a first number of first circles in a closed figure whose grayscale value is greater than the first threshold according to the radius from large to small, and each of the first circles does not overlap, and uses the center of the first circle as a preliminary positioning point; A first number of the preliminary positioning points of the first circle are used as first positioning points, and a figure formed by connecting the first positioning points with straight lines that do not intersect each other is formed as a first closed figure; The first processor determines whether the straight-line distance between the centroid of the first closed figure and the centroid of the first image exceeds a fourth length. If so, the centroid of the first image is selected as the first center point. Otherwise, the midpoint of the line formed by the centroid of the first image and the centroid of the first closed figure is selected as the first center point.

3. The automatic analysis system for MRA brain images according to any one of claims 1 or 2, characterized in that: The first processor screens and removes closed figures whose areas do not exceed a ninth threshold, and the ninth threshold may be 0.5% of the area of ​​the first image.

4. The automatic analysis system for MRA brain images according to any one of claims 1 or 2, characterized in that: The second processor includes an instruction acquisition module, and the instruction acquisition module is a keyboard and a button.

5. The automatic analysis system for MRA brain images according to any one of claims 1 or 2, characterized in that: The first database is electrically connected to the first processor.

6. The automatic analysis system for MRA brain images according to any one of claims 1 or 2, characterized in that: The image display module is a liquid crystal display.

7. The automatic analysis system for MRA brain images according to any one of claims 1 or 2, characterized in that: The second processor is electrically connected to the image display module.

Images