Systems and methods for evaluating medical image

1. A system for evaluating a medical image, comprising:

at least one storage device including a set of instructions; and

at least one processor configured to communicate with the at least one storage device, wherein when executing the set of instructions, the at least one processor is configured to direct the system to perform operations including:

obtaining the medical image;

extracting a feature of an artifact in the medical image, wherein the feature comprises a histogram of oriented gradients (HOG) feature of the medical image; and

determining an image quality of the medical image by inputting the feature of the artifact in the medical image to a determination model,

wherein extracting the feature of the artifact in the medical image includes:

calculating a gradient of each pixel in the medical image;

segmenting the medical image into a plurality of cell units, each of the plurality of cell units including a same count of pixels;

counting a HOG feature of each of the plurality of cell units based on the gradient of each pixel in each of the plurality of cell units;

combining the plurality of cell units into one or more blocks according to a predetermined number, and counting a HOG feature of each of the one or more blocks based on one or more HOG features of one or more cell units included in each of the one or more blocks; and

obtaining a HOG feature of the medical image by stitching the HOG features of the one or more blocks.

2. The system of claim 1 , wherein the operations further comprise:

before stitching the HOG features of the one or more blocks, normalizing the HOG feature of each of the one or more blocks.

3. The system of claim 1 , wherein the operations further comprise a process for obtaining the determination model, and the process comprises:

obtaining a plurality of sample medical images and a label of each of the plurality of sample medical images, wherein the label represents a degree to which an artifact in each of the plurality of sample medical images affects the tissue feature recognition;

extracting a feature of each of the plurality of sample medical images; and

obtaining the determination model by training a preliminary determination model based on the extracted feature and the label of each of the plurality of sample medical images.

4. The system of claim 3 , wherein the process further comprises:

preprocessing each of the plurality of sample medical images, wherein the preprocessing includes: regularizing a size of each of the plurality of sample medical images and/or standardizing a color of each of the plurality of sample medical images.

5. The system of claim 3 , wherein the operations further comprise:

obtaining additional determination models corresponding to different tissue types, wherein each of the determination models is obtained based on a plurality of sample medical images including a tissue type corresponding to the each of the determination models.

6. The system of claim 5 , wherein the operations further comprise:

determining a tissue type included in the medical image;

obtaining a determination model corresponding to the tissue type; and

determining the degree to which the artifact in the medical image affects recognition of the tissue feature by inputting the feature of the medical image to the determination model corresponding to the tissue type.

7. The system of claim 1 , wherein the feature further comprises a gender and an age of a user from whom the medical image is acquired, and/or a tissue type of the medical image.

8. The system of claim 1 , wherein the determination model is a support vector machine model, a logistic regression model, a naive Bayes classification model, a decision tree model, or a deep learning model.

9. A system for medical image acquisition, comprising:

at least one storage device including a set of instructions; and

at least one processor configured to communicate with the at least one storage device, wherein when executing the set of instructions, the at least one processor is configured to direct the system to perform operations including:

acquire a first medical image and conduct a determination process in real time, wherein the determination process comprises:

extracting a feature of an artifact in the first medical image, wherein the feature comprises a histogram of oriented gradients (HOG) feature of the first medical image;

determining an image quality of the medical image by inputting the feature of the artifact in the first medical image to a determination model; and

determining whether the degree satisfies a preset condition, and acquiring a second medical image in response to a determination that the degree does not satisfy the preset condition,

wherein extracting the feature of the artifact in the medical image includes:

calculating a gradient of each pixel in the medical image;

segmenting the medical image into a plurality of cell units, each of the plurality of cell units including a same count of pixels;

counting a HOG feature of each of the plurality of cell units based on the gradient of each pixel in each of the plurality of cell units;

combining the plurality of cell units into one or more blocks according to a predetermined number, and counting a HOG feature of each of the one or more blocks based on one or more HOG features of one or more cell units included in each of the one or more blocks; and

obtaining a HOG feature of the medical image by stitching the HOG features of the one or more blocks.

10. The system of claim 9 , wherein the determination model is a shallow learning model including a support vector machine model, a logistic regression model, a naive Bayes classification model, or a decision tree model.

11. A method for evaluating a medical image, comprising:

obtaining the medical image;

extracting a feature of an artifact in the medical image, wherein the feature comprises a histogram of oriented gradients (HOG) feature of the medical image; and

determining an image quality of the medical image by inputting the feature of the artifact in the medical image to a determination model,

wherein extracting the feature of the artifact in the medical image includes:

calculating a gradient of each pixel in the medical image;

segmenting the medical image into a plurality of cell units, each of the plurality of cell units including a same count of pixels;

counting a HOG feature of each of the plurality of cell units based on the gradient of each pixel in each of the plurality of cell units;

combining the plurality of cell units into one or more blocks according to a predetermined number, and counting a HOG feature of each of the one or more blocks based on one or more HOG features of one or more cell units included in each of the one or more blocks; and

obtaining a HOG feature of the medical image by stitching the HOG features of the one or more blocks.

12. The method of claim 11 , further comprising:

before stitching the HOG features of the one or more blocks, normalizing the HOG feature of each of the one or more blocks.

13. The method of claim 11 , wherein a process for obtaining the determination model comprises:

obtaining a plurality of sample medical images and a label of each of the plurality of sample medical images, wherein the label represents a degree to which an artifact in each of the plurality of sample medical images affects the tissue feature recognition;

extracting a feature of each of the plurality of sample medical images; and

obtaining the determination model by training a preliminary determination model based on the extracted feature and the label of each of the plurality of sample medical images.

14. The method of claim 13 , further comprising:

preprocessing each of the plurality of sample medical images, wherein the preprocessing includes: regularizing a size of each of the plurality of sample medical images and/or standardizing a color of each of the plurality of sample medical images.

15. The method of claim 13 , further comprising:

obtaining additional determination models corresponding to different tissue types, wherein each of the determination models is obtained based on a plurality of sample medical images including a tissue type corresponding to the each of the determination models.

16. The method of claim 15 , further comprising:

determining a tissue type included in the medical image;

obtaining a determination model corresponding to the tissue type; and

determining the degree to which the artifact in the medical image affects recognition of the tissue feature by inputting the feature of the medical image to the determination model corresponding to the tissue type.

17. The method of claim 11 , wherein the feature further comprises a gender and an age of a user from whom the medical image is acquired, and/or a tissue type of the medical image.

18. The method of claim 11 , wherein the determination model is a support vector machine model, a logistic regression model, a naive Bayes classification model, a decision tree model, or a deep learning model.