CT scanning parameter evaluation method and system based on body fat rate

Abstract

The invention provides a body fat rate-based CT scanning parameter evaluation method and system, and relates to the technical field of medical image processing. Comprising the steps of extracting the number of fat points in a CT image to obtain fat proportion features; objective factors of a patient corresponding to the CT image are obtained, and patient features are formed; fusing the fat proportion characteristics of the CT image and the patient characteristics, respectively inputting the fused characteristics into the two FFN network models, and respectively predicting CT scanning voltage and current to obtain predicted scanning voltage and predicted scanning current of the current CT image; and based on the actual scanning voltage, the predicted scanning voltage, the actual scanning current and the predicted scanning current, evaluating whether the actual scanning voltage and the actual scanning current conform to expectations or not. According to the method, the fat proportion characteristics of the patient corresponding to the CT image and the patient characteristics are fused, the single CT image and the multiple CT images are evaluated respectively, and the evaluation result is more accurate.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical image processing, and in particular to a method and system for evaluating CT scanning parameters based on body fat percentage. Background Art

[0002] CT (Computed Tomography) scan is a clinical imaging test. As one of the most common imaging modalities for clinical diagnosis, CT helps healthcare providers detect diseases and injuries. It uses a series of X-rays and computer software to create detailed images of bones and soft tissues and is widely used in screening and diagnosing major diseases.

[0003] Medical workers generally operate CT according to a unified standard and perform CT scans on patients according to the set CT parameters. The scanning protocol in the CT examination determines the voltage and current used by the X-ray emitter in an examination. The voltage level unit of the X-ray emitter in CT scanning is generally kilovolts (kV). Higher voltages produce higher energy X-rays, so that X-rays can better penetrate tissues of different densities and have higher tissue resolution. On the other hand, the current intensity level unit of the X-ray emitter in CT scanning is generally milliamperes (mA). Higher currents produce more X-rays, which can reduce image noise and improve image quality.

[0004] Although higher current and voltage can improve image quality, they also lead to increased radiation doses, thereby increasing radiation risks for patients and operators. In addition, most medical institutions often do not have the link to evaluate the patient's CT scan parameters later. When the patient undergoes a CT scan again in a short period of time, the patient's CT parameters cannot be optimized well. Summary of the invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a CT scanning parameter evaluation method and system based on body fat percentage, which fuses the fat proportion characteristics of the patient corresponding to the CT image and the patient characteristics, uses a neural network to obtain the predicted scanning voltage and predicted scanning current of the CT image, and sets an evaluation method to evaluate a single CT image and multiple CT images respectively, so that the evaluation results are more accurate.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0007] A first aspect of the present invention provides a method for evaluating CT scan parameters based on body fat percentage.

[0008] The CT scan parameter evaluation method based on body fat percentage includes the following steps:

[0009] Acquire multiple CT images, extract the number of fat points in each CT image, calculate the fat ratio, and obtain the fat ratio feature;

[0010] Obtain multiple objective factors corresponding to each patient in each CT image to form patient characteristics;

[0011] The fat ratio features of the corresponding CT images and the patient features are fused, and the fused features are respectively input into two FFN network models to predict the CT scanning voltage and current respectively, and obtain the predicted scanning voltage and predicted scanning current of the current CT image;

[0012] The actual scanning voltage and actual scanning current of the corresponding CT image are obtained, and based on the actual scanning voltage and predicted scanning voltage, as well as the actual scanning current and predicted scanning current, whether the actual scanning voltage and actual scanning current corresponding to a single CT image and multiple CT images meet expectations are evaluated respectively.

[0013] A second aspect of the present invention provides a CT scan parameter evaluation system based on body fat percentage.

[0014] The CT scan parameter evaluation system based on body fat percentage includes:

[0015] The fat ratio feature acquisition module is configured to: acquire multiple CT images, extract the number of fat points in the CT images respectively, calculate the fat ratio, and obtain the fat ratio feature;

[0016] The patient feature acquisition module is configured to: acquire multiple objective factors corresponding to each patient in each CT image to form patient features;

[0017] The prediction module is configured to: fuse the fat ratio features and patient features of the corresponding CT image, input the fused features into two FFN network models respectively, predict the CT scanning voltage and current respectively, and obtain the predicted scanning voltage and predicted scanning current of the current CT image;

[0018] The comparison and judgment module is configured to: obtain the actual scanning voltage and the actual scanning current corresponding to the CT image, and based on the actual scanning voltage and the predicted scanning voltage, as well as the actual scanning current and the predicted scanning current, evaluate whether the actual scanning voltage and the actual scanning current corresponding to a single CT image and multiple CT images are in line with expectations.

[0019] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in the method for evaluating CT scan parameters based on body fat percentage as described in the first aspect of the present invention.

[0020] The fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the CT scan parameter evaluation method based on body fat percentage as described in the first aspect of the present invention are implemented.

[0021] One or more of the above technical solutions have the following beneficial effects:

[0022] The present invention provides a CT scanning parameter evaluation method and system based on body fat percentage. The method and system fuse the fat ratio characteristics of the patient corresponding to the CT image and the patient characteristics, input the fused characteristics into the FFN network, obtain the predicted scanning voltage and predicted scanning current of the current CT image, and use the predicted scanning voltage and predicted scanning current to evaluate whether the actual scanning voltage and actual scanning current of the corresponding CT image meet expectations. This provides a means for medical institutions to perform post-evaluation of the scanning parameters of CT images, facilitates medical institutions to grasp feedback information, and provides data support for the optimization of CT scanning parameters.

[0023] The present invention selects the patient's fat ratio characteristics and patient characteristics composed of objective factors of the patient to predict the parameters of CT scanning, and can realize personalized customization of CT scanning parameters for patients. The patient characteristics include five objective factors of the patient's height, weight, BMI, gender, and age. Using the above characteristics for prediction can obtain more accurate parameter prediction results.

[0024] The present invention provides an evaluation method for a single CT image and multiple CT images, which can evaluate the accuracy of the actual scanning voltage and the actual scanning current locally and globally, respectively, and determine whether the actual scanning voltage and the actual scanning current meet expectations, making the evaluation more comprehensive and accurate.

[0025] For patients who undergo multiple CT scans in a short period of time and whose body fat percentage does not change much, the predicted scanning voltage and predicted scanning current predicted by the previous CT scan can be used to perform the next scan on the patient. Compared with the existing technology of unifying CT scanning parameters for patients according to set standards, this is safer for patients and reduces radiation hazards.

[0026] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0028] Figure 1This is a flow chart of the method of the first embodiment.

[0029] Figure 2 It is a system structure diagram of the second embodiment. DETAILED DESCRIPTION

[0030] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0031] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.

[0032] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0033] Embodiment 1

[0034] This embodiment discloses a method for evaluating CT scan parameters based on body fat percentage.

[0035] like Figure 1 As shown, the CT scan parameter evaluation method based on body fat percentage includes the following steps:

[0036] Acquire multiple CT images, extract the number of fat points in each CT image, calculate the fat ratio, and obtain the fat ratio feature;

[0037] Obtain multiple objective factors corresponding to each patient in each CT image to form patient characteristics;

[0038] The fat ratio features of the corresponding CT images and the patient features are fused, and the fused features are respectively input into two FFN network models to predict the CT scanning voltage and current respectively, and obtain the predicted scanning voltage and predicted scanning current of the current CT image;

[0039] The actual scanning voltage and actual scanning current of the corresponding CT image are obtained, and based on the actual scanning voltage and predicted scanning voltage, as well as the actual scanning current and predicted scanning current, whether the actual scanning voltage and actual scanning current corresponding to a single CT image and multiple CT images meet expectations are evaluated respectively.

[0040] This embodiment mainly includes the following steps:

[0041] (a) Obtain the image data sequence set ct_series of the chest coronal CT. Each image ct in the sequence 0 ,ct 1 ,ct 2 ,......,ct N-1,N is the length of ct_series.

[0042] (b) Each image ct of the CT image data in (a) i , through the following transformation, ct i The value of each pixel in is converted to HU value. The specific HU value conversion formula is as follows:

[0043] HU=pix _ value*Rescale_slope+Rescale_intercept

[0044] Where: pix_value is ct i The value of each pixel point, Rescale_slope is ct i The value of the header information (0028,1053), Rescale_intercept is ct i The value of the header information (0028,1052).

[0045] HU value, full name Hounsfield Unit, is a term commonly used in medical imaging diagnosis, especially in computed tomography (CT), HU value is used to indicate the degree of absorption of X-rays by different tissues. HU value is a measurement standard used to describe the density of various tissues and organs in the human body. The density of water is used as the benchmark, and its HU value is 0. Different tissues have different HU values.

[0046] (c) According to official medical guidelines, the HU value range of CT fat is between (-80, -20). Based on this range, each CT image is counted. i The number of fat points fat_num i , thus we can count each ct i The fat ratio of the image fat_ratio i .

[0047] Therefore, the fat ratio feature composed of N CT images can be expressed as: f :[fat_ratio 0 ,fat_ratio 1 ,...,fat_ratio i ,...fat_ratio N-1 ].

[0048] (d) Obtain the five objective factors of the patient's height, weight, BMI, gender, and age of the current scan image to form a patient vector:

[0049] obj f :[height f,weight f ,bmi f ,sex f ,age f ].

[0050] Among them, height f Indicates the patient's height factor; weight f Indicates the patient's weight factor; bmi f Indicates the patient's BMI factor; sex f Indicates the patient's gender factor; age f Indicates the patient's age factor.

[0051] (e) Because of fat f and obj f These are vectors of two different modes, and the two features need to be fused.

[0052] e1).fat f After a Self-Attention network, the output dimension is (1, M, L), which is composed of fat fsa Indicates. f After the Self-attention network, the output dimension is (1, M, L), which is obj fsa express.

[0053] e2).fat fsa and obj fsa Combined together to form a vector of dimension (1, M, 2L), the two vectors are fused through the transformer-econoderm layer, and the output is the fused vector f fuse , the dimensions are (1,M,2L)

[0054] (f). Through the f of the previous step fuse The vector passes through two FFN networks, where the FFN network activation layer uses the ReLU network. The two network structures are the same, but the weights are different. One network is responsible for outputting voltage A network is responsible for output current Their dimensions are all (1, N). Since each image needs to predict voltage and current, the N here is consistent with the number of CT images N.

[0055] The specific network formula of FFN is as follows:

[0056] f1=max(0,f fuse *W 1 +b 1 )*W 2 +b 2

[0057] Among them, W 1 is a matrix of dimension 2L*D, W 2 is a matrix of dimension (D, 2L), b 2 b 1 is the offset;

[0058] f2=a*f1+b*f fuse

[0059] Among them, a and b are hyperparameters.

[0060] f3=sigmod(f2*W 3 )

[0061] W 3 is a matrix of dimension 2L*1.

[0062] The above FFN network can obtain the output predicted current and voltage. Using the MSE loss function, the optimizer uses SGD, and back-propagation training is used to obtain the model model for predicting voltage and current.

[0063] The fused vector f fuse The information is input into two FFN network models respectively, and the information is transmitted layer by layer through the hidden layer and the output layer to extract the features. The features are transformed nonlinearly through the activation function between the hidden layer and the output layer, and the predicted scanning voltage is finally obtained. and predicted scan current

[0064] (g). Get each ct i The voltage in the image header information (0018,0060)v i and current (0018,1150)mAs i The values ​​of and Make comparison.

[0065] The specific comparison method is as follows:

[0066] g1. Get the predicted scanning voltage respectively The predicted scanning voltage value of the i-th CT image in i=1,2,…n, n is the number of CT images;

[0067] Get the actual scan voltage value v;

[0068] For voltage and v i , get their ratio: The voltage ratio is obtained and a ratio threshold range T (Ts, Tm) is set. If the voltage ratio is within this range, the actual scanning voltage of the current CT image meets expectations.

[0069] g2. Step g1 is for a single photo, but the whole photo still needs to be compared.

[0070] Overall comparison method:

[0071] By v and Two vectors do cosine similarity v s , if v s >V sim , then it is considered appropriate. sim is the threshold of the similarity between two vectors.

[0072] The current comparison method is similar:

[0073] g3. Get the predicted scanning current respectively The predicted scanning current value of the i-th CT image in i=1,2,…n, n is the number of CT images;

[0074] Get the actual scanning current value mAs;

[0075] calculate Get the current ratio, where mAs i is the actual scanning current value of the i-th CT image;

[0076] The current ratio is determined to be greater than the preset threshold value, and whether the actual scanning current of a single CT image meets expectations.

[0077] g4. Calculate mAs and The cosine similarity mAs s ;

[0078] Determine the cosine similarity mAs s The actual scanning current of multiple CT images is judged as a whole based on the size of the preset threshold value to see whether it meets expectations.

[0079] Embodiment 2

[0080] This embodiment discloses a CT scan parameter evaluation system based on body fat percentage.

[0081] like Figure 2 As shown, the CT scanning parameter evaluation system based on body fat percentage includes:

[0082] The fat ratio feature acquisition module is configured to: acquire multiple CT images, extract the number of fat points in the CT images respectively, calculate the fat ratio, and obtain the fat ratio feature;

[0083] The patient feature acquisition module is configured to: acquire multiple objective factors corresponding to each patient in each CT image to form patient features;

[0084] The prediction module is configured to: fuse the fat ratio features and patient features of the corresponding CT image, input the fused features into two FFN network models respectively, predict the CT scanning voltage and current respectively, and obtain the predicted scanning voltage and predicted scanning current of the current CT image;

[0085] The comparison and judgment module is configured to: obtain the actual scanning voltage and the actual scanning current corresponding to the CT image, and based on the actual scanning voltage and the predicted scanning voltage, as well as the actual scanning current and the predicted scanning current, evaluate whether the actual scanning voltage and the actual scanning current corresponding to a single CT image and multiple CT images are in line with expectations.

[0086] Embodiment 3

[0087] The purpose of this embodiment is to provide a computer-readable storage medium.

[0088] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the method for evaluating CT scan parameters based on body fat percentage as described in Example 1 of the present disclosure.

[0089] Embodiment 4

[0090] The purpose of this embodiment is to provide an electronic device.

[0091] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps in the CT scan parameter evaluation method based on body fat percentage as described in Example 1 of the present disclosure are implemented.

[0092] The steps involved in the apparatuses of the above embodiments 2, 3 and 4 correspond to the method embodiment 1, and the specific implementation methods can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0093] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0094] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A CT scan parameter evaluation method based on body fat percentage, characterized in that: The following steps are involved: Acquire multiple CT images, extract the number of fat points in each CT image, calculate the fat ratio, and obtain the fat ratio feature; Obtain multiple objective factors corresponding to each patient in each CT image to form patient characteristics; The fat ratio features of the corresponding CT images and the patient features are fused, and the fused features are respectively input into two FFN network models to predict the CT scanning voltage and current respectively, and obtain the predicted scanning voltage and predicted scanning current of the current CT image; The actual scanning voltage and actual scanning current of the corresponding CT image are obtained, and based on the actual scanning voltage and predicted scanning voltage, as well as the actual scanning current and predicted scanning current, whether the actual scanning voltage and actual scanning current corresponding to a single CT image and multiple CT images meet expectations are evaluated respectively.

2. The method for evaluating CT scan parameters based on body fat percentage according to claim 1, wherein: The pixel values ​​in the CT image data are converted into HU values, and the number of fat points in the CT image data is extracted based on the HU value range of fat; The ratio of the number of fat points to the total number of pixels in the current CT image is calculated to obtain the fat ratio.

3. The CT scan parameter evaluation method based on body fat percentage according to claim 1, characterized in that: The five objective factors of the patient's height, weight, BMI, gender, and age in the current scan image are obtained to form the patient's characteristics.

4. The method for evaluating CT scan parameters based on body fat percentage according to claim 1, wherein: The fat ratio feature and patient feature are input into the Self-Attention network respectively to obtain vectors fat with dimensions (1, M, L) fsa and vector obj fsa ; Fat fsa and obj fsa Input to the transformer-ecnoder layer for fusion, and obtain the fused vector f with dimension (1, M, 2L) fuse .

5. The method for evaluating CT scan parameters based on body fat percentage according to claim 4, characterized in that: The fused vector f fuse The information is input into two FFN network models respectively, and the information is transmitted layer by layer through the hidden layer and the output layer to extract the features. The features are transformed nonlinearly through the activation function between the hidden layer and the output layer, and the predicted scanning voltage is finally obtained. and predicted scan current 6. The method for evaluating CT scan parameters based on body fat percentage according to claim 1, wherein: For a single CT image: Get the predicted scan voltage respectively And the predicted scan current The predicted scanning voltage value of the i-th CT image in and each predicted scan current value i=1,2,…n, n is the number of CT images; Get the actual scan voltage value v and the actual scan current value mAs, and calculate as well as Get the voltage ratio and current ratio, where v i and mAs i are the actual scanning voltage value and actual scanning current value of the i-th CT image respectively; The voltage ratio and current ratio are respectively judged to the preset threshold values, and whether the actual scanning voltage and actual scanning current of a single CT image are in line with expectations are determined.

7. The method for evaluating CT scan parameters based on body fat percentage according to claim 6, wherein: Calculate v and The cosine similarity b s , and mAs and The cosine similarity mAs s ; Determine the cosine similarity v s and mAs s The actual scanning voltage and actual scanning current of multiple CT images are judged as a whole to be in line with expectations by comparing the actual scanning voltage and actual scanning current of the multiple CT images with the preset threshold.

8. A CT scanning parameter evaluation system based on body fat percentage, characterized in that: include: The fat ratio feature acquisition module is configured to: acquire multiple CT images, extract the number of fat points in the CT images respectively, calculate the fat ratio, and obtain the fat ratio feature; The patient feature acquisition module is configured to: acquire multiple objective factors corresponding to each patient in each CT image to form patient features; The prediction module is configured to: fuse the fat ratio features and patient features of the corresponding CT image, input the fused features into two FFN network models respectively, predict the CT scanning voltage and current respectively, and obtain the predicted scanning voltage and predicted scanning current of the current CT image; The comparison and judgment module is configured to: obtain the actual scanning voltage and the actual scanning current corresponding to the CT image, and based on the actual scanning voltage and the predicted scanning voltage, as well as the actual scanning current and the predicted scanning current, evaluate whether the actual scanning voltage and the actual scanning current corresponding to a single CT image and multiple CT images are in line with expectations.

9. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, the steps in the CT scan parameter evaluation method based on body fat percentage as described in any one of claims 1 to 7 are implemented.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the CT scan parameter evaluation method based on body fat percentage as described in any one of claims 1-7 are implemented.

Images