A method and system for classifying infarction based on CTP cerebral perfusion image data

Abstract

The invention discloses an infarction classification method based on CTP cerebral perfusion image data, comprising the following steps: acquiring cerebral perfusion image data, and extracting a candidate infarction region; calculating an average CT value M1 of each sequence of a brain region where the candidate infarction region is located and acquiring a peak sequence ID, acquiring an average CT value M2 of each sequence of a brain region surrounding the candidate infarction region and acquiring a peak sequence ID, and acquiring an average CT value M0 of each sequence of a ventricular region; judging the proximity between M1 and M0, and judging that the candidate infarction region is an old infarction if M1-M0 is less than a first empirical threshold, and inputting sequence data of the brain region where the candidate infarction region is located and the brain region surrounding the candidate infarction region into a discrimination network to judge whether it is a chronic infarction or an acute infarction, and obtaining a pre-discrimination result; and performing secondary correction on the pre-discrimination result using a second empirical threshold to obtain a final classification result of chronic infarction and acute infarction.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image analysis, and in particular to an infarction classification method and system based on CTP cerebral perfusion image data. Background Art

[0002] As a very important imaging tool in the stroke diagnosis and treatment process, CTP cerebral perfusion imaging can accurately reflect the cerebral blood perfusion situation. It is an effective means to diagnose cerebral infarction at an early stage, guide emergency thrombolytic therapy and judge prognosis. CTP cerebral perfusion imaging data analysis generally requires automatic processing to save time in the emergency process. In the process of automatic calculation and processing, the emergency infarction core can easily include chronic infarction, old infarction, etc. Since acute infarction is the final emergency surgical treatment, such misjudgment will greatly affect the final surgery. Summary of the invention

[0003] To solve the above problems, the present invention provides an infarction classification method and system based on CTP cerebral perfusion image data.

[0004] The present invention adopts the following technical solutions:

[0005] A method for classifying infarction based on CTP cerebral perfusion image data comprises the following steps:

[0006] S1. Acquire cerebral perfusion imaging data and extract candidate infarction areas;

[0007] S2, calculating the average CT value M1 of each sequence in the brain region where the candidate infarction region is located and obtaining the peak sequence ID, obtaining the average CT value M2 of each sequence in the brain region surrounding the candidate infarction region and obtaining the peak sequence ID, and obtaining the average CT value M0 of each sequence in the ventricular region;

[0008] S3, judging the proximity between M1 and M0, if M1-M0 is less than a preset first empirical threshold, judging the candidate infarction region as an old infarction, if M1-M0 is greater than the preset first empirical threshold, executing step S4;

[0009] S4, inputting the sequence data of the brain region where the candidate infarction region is located and the brain region surrounding the candidate infarction region into the discriminant network, judging whether it is chronic infarction or acute infarction, and obtaining a pre-discrimination result;

[0010] S5. Use a preset second empirical threshold to perform secondary correction on the pre-discrimination result to obtain the final classification result of chronic infarction and acute infarction.

[0011] Preferably, the extraction of candidate infarction regions in step S1 is achieved by a deep learning neural network.

[0012] Preferably, the extraction of the candidate infarction region in step S1 is achieved by a parameter map threshold method.

[0013] A system for classifying infarction based on CTP cerebral perfusion image data comprises an acquisition module, a target extraction module, a CT value calculation module, a first threshold analysis module, a pre-discrimination module and a second threshold analysis module, wherein the acquisition module is used to acquire cerebral perfusion image data, the target extraction module is used to extract candidate infarction regions, the CT value calculation module is used to calculate and obtain the average CT value of each sequence of the brain region where the candidate infarction region is located, the average CT value of each sequence of the brain region surrounding the candidate infarction region and the average CT value of each sequence of the ventricle region, the first threshold analysis module is used to judge whether it is an old infarction by judging the degree of proximity between the average CT value of the sequence of the brain region where the candidate infarction region is located and the average CT value of the sequence of the ventricle region through a first empirical threshold, the pre-discrimination module is used to input the sequence data of the brain region where the candidate infarction region is located and the brain region surrounding the candidate infarction region into a discrimination network to judge whether it is a chronic infarction or an acute infarction to obtain a pre-discrimination result, and the second threshold analysis module is used to perform secondary correction on the pre-discrimination result through a second empirical threshold to obtain a final classification result of chronic infarction and acute infarction.

[0014] Preferably, the extraction of candidate infarction regions is achieved through a deep learning neural network.

[0015] Preferably, the extraction of candidate infarction regions is achieved by a parameter map threshold method.

[0016] After adopting the above technical solution, the present invention has the following advantages compared with the background technology:

[0017] The present invention can accurately classify acute infarction, chronic infarction and old infarction, determine the accuracy of diagnosis, and provide a reliable basis for surgical treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] Embodiment 1

[0021] refer to Figure 1 As shown, the present invention discloses an infarction classification method based on CTP cerebral perfusion image data, comprising the following steps:

[0022] S1. Acquire cerebral perfusion image data and extract candidate infarction regions. Extraction of candidate infarction regions is achieved by deep learning neural network or parameter map threshold method.

[0023] S2. Calculate the average CT value M1 of each sequence in the brain region where the candidate infarction region is located and obtain the peak sequence ID, obtain the average CT value M2 of each sequence in the brain region surrounding the candidate infarction region and obtain the peak sequence ID, and obtain the average CT value M0 of each sequence in the ventricular region.

[0024] S3. Determine the proximity between M1 and M0. If M1-M0 is less than a preset first empirical threshold, determine that the candidate infarction area is an old infarction. If M1-M0 is greater than the preset first empirical threshold, execute step S4.

[0025] S4. Inputting the sequence data of the brain region where the candidate infarction area is located and the brain region surrounding the candidate infarction area into the discriminant network, determining whether it is a chronic infarction or an acute infarction, and obtaining a pre-discrimination result.

[0026] S5. Use a preset second empirical threshold to perform secondary correction on the pre-discrimination result to obtain the final classification result of chronic infarction and acute infarction.

[0027] The present invention also discloses an infarction classification system based on CTP cerebral perfusion image data, including an acquisition module, a target extraction module, a CT value calculation module, a first threshold analysis module, a pre-discrimination module and a second threshold analysis module, wherein:

[0028] The acquisition module is used to acquire cerebral perfusion image data; the target extraction module is used to extract candidate infarction areas, and the extraction of candidate infarction areas is achieved through deep learning neural networks or parameter map threshold methods.

[0029] The CT value calculation module is used to calculate the average CT value of each sequence of the brain region where the candidate infarction region is located, the average CT value of each sequence of the brain region surrounding the candidate infarction region, and the average CT value of each sequence of the ventricular region.

[0030] The first threshold analysis module is used to determine whether the infarction is old by judging the proximity between the average CT value of the brain region sequence where the candidate infarction region is located and the average CT value of the ventricle region sequence through the first empirical threshold.

[0031] The pre-discrimination module is used to input the sequence data of the brain region where the candidate infarction area is located and the brain region surrounding the candidate infarction area into the discrimination network to determine whether it is a chronic infarction or an acute infarction, so as to obtain a pre-discrimination result.

[0032] The second threshold analysis module is used to perform secondary correction on the pre-discrimination result through the second empirical threshold to obtain the final classification result of chronic infarction and acute infarction.

[0033] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for classifying infarction based on CTP cerebral perfusion image data, characterized in that: The following steps are involved: S1. Acquire cerebral perfusion imaging data and extract candidate infarction areas; S2, calculating the average CT value M1 of each sequence in the brain region where the candidate infarction region is located and obtaining the peak sequence ID, obtaining the average CT value M2 of each sequence in the brain region surrounding the candidate infarction region and obtaining the peak sequence ID, and obtaining the average CT value M0 of each sequence in the ventricular region; S3, judging the proximity between M1 and M0, if M1-M0 is less than a preset first empirical threshold, judging the candidate infarction region as an old infarction, if M1-M0 is greater than the preset first empirical threshold, executing step S4; S4, inputting the sequence data of the brain region where the candidate infarction region is located and the brain region surrounding the candidate infarction region into the discriminant network, judging whether it is chronic infarction or acute infarction, and obtaining a pre-discrimination result; S5. Use a preset second empirical threshold to perform secondary correction on the pre-discrimination result to obtain the final classification result of chronic infarction and acute infarction.

2. The infarction classification method based on CTP cerebral perfusion image data according to claim 1, characterized in that: The extraction of the candidate infarction region in step S1 is achieved by a deep learning neural network.

3. The infarction classification method based on CTP cerebral perfusion image data according to claim 1, characterized in that: The extraction of the candidate infarction region in step S1 is achieved by a parameter map threshold method.

4. An infarction classification system based on CTP cerebral perfusion image data, characterized in that: The method comprises an acquisition module, a target extraction module, a CT value calculation module, a first threshold analysis module, a pre-discrimination module and a second threshold analysis module, wherein the acquisition module is used to acquire cerebral perfusion image data, the target extraction module is used to extract candidate infarction regions, the CT value calculation module is used to calculate and obtain the average CT value of each sequence of the brain region where the candidate infarction region is located, the average CT value of each sequence of the brain region surrounding the candidate infarction region, and the average CT value of each sequence of the ventricle region, the first threshold analysis module is used to judge whether it is an old infarction by judging the degree of proximity between the average CT value of the sequence of the brain region where the candidate infarction region is located and the average CT value of the sequence of the ventricle region through a first empirical threshold, the pre-discrimination module is used to input the sequence data of the brain region where the candidate infarction region is located and the brain region surrounding the candidate infarction region into a discrimination network to judge whether it is a chronic infarction or an acute infarction to obtain a pre-discrimination result, and the second threshold analysis module is used to perform secondary correction on the pre-discrimination result through a second empirical threshold to obtain the final classification result of chronic infarction and acute infarction.

5. The infarction classification system based on CTP cerebral perfusion image data according to claim 4, characterized in that: The extraction of candidate infarction areas is achieved through a deep learning neural network.

6. The infarction classification system based on CTP cerebral perfusion image data according to claim 4, characterized in that: The extraction of the candidate infarction area is achieved by a parameter map threshold method.

Images