Lung cancer prognosis auxiliary evaluation method and system based on CT radiomics

[0024] In order to make the objects, technical solutions, and advantages of the present invention more clearly, the technical solutions in the embodiments of the present invention will be described in contemplation in the embodiments of the present invention, and will be described, and the embodiments described in the embodiments of the present invention will be described. It is a part of the embodiments of the present invention, not all of the embodiments, based on the embodiments of the present invention, and those of ordinary skill in the art without all other embodiments obtained by creative labor premise, .

[0025] Such as figure 1 As shown, a prognostic auxiliary evaluation method based on CT imaging group, including: S101, collecting original medical imaging and clinical information using radiation therapy lung cancer, screening; S102, the proportion of lung cancer patients is randomly divided into training Set and verification set, the training set is used for training prediction models for verifying the performance of the prediction model; S103, scanning the lesser region of the original medical image by high-resolution computer tomography, and from the lesser region Extract the imaging histology characteristics, resulting in the preliminary imaging syndrome; S104, analyzing the preliminary imaging histology characteristics, resulting in target imaging syndrome; S105, the target imaging syndrome of the training concentration is used to use machine learning algorithm training forecast Model, construct imagery moderate evaluation model, and verify the centralized verification model; S106, establish a dynamic imaging syndrome label, and verify it in the verification set; S107, calculate the image of each patient on the basis of the prediction model Formula score; S108, transforming the patient's imaging method score to the probability of sensitivity of patient radiation therapy.

[0026]Specifically, the original medical imaging of patients with radiation therapy is collected, and the proportion of patients with lung cancer is randomly divided into training set and verification set, and the original medical image of the training set is 2 to 4 times the original medical image in the verification group. The resolved computer fault scans the lesion area of ​​the original medical image, and extracts the imaging histology characteristics from the lesser region. After treatment, it obtains the target imaging histology characteristics, and the target imaging method for training concentration is used to use machine learning algorithm training. The prediction model is constructed, and the imaging method evaluation model is constructed, and the model is verified, and the imaging histological characterization score of each patient is calculated on the basis of the predictive model, thereby obtaining the probability of sensitivity of patient radiation therapy, Figure 4 For the conversion table of imaging histology characterization and probability; the present invention is divided, feature extraction, feature screening and establishing an imaging method, combined with the segmentation, feature extraction, and feature screening of pulmonary cancer patients in a conventional imaging method. The patient's clinical information is qualitatively and quantitatively analyzed to the patient's radiotherapy, thereby assists the doctor to formulate a personalized treatment plan, and assists the doctor to evaluate the patient's survival and recurrence, and the performance of the resulting imaging method evaluation model Verify that the accuracy of the imaging method evaluation model is guaranteed.

[0027] The lesion region of the high-resolution computer tomography is extracted from the lesion profile, extracts the imaging group in the lesion profile to obtain a preliminary imaging syndrome, including one-stage characteristics, shape characteristics 14, grayscale symbiotic matrix characteristics, 14 grayscale correlation matrix features, 16 grayscale run matrix characteristics, 16 grayscale region matrix characteristics, 5 adjacent gray difference matrix features 5, total 107 features Specifically, the gray level region in the gray level region size matrix quantifier image is defined as the number of connection voxels having the same gray strength; the gradation stroke matrix quantifies the grayscale run, defined as a pixel The length of the number, and a continuous pixel having the same gray value; the grayscale unevenness measures the variability of the grayscale intensity value in the image, the lower value indicates more uniform intensity; the higher cluster shadow reflects the average value. Asymmetry. The cluster is a metric that the grayscale symbiotic matrix beagration and asymmetry are measured. The higher value reflects the asymmetry of the average, and the lower value indicates the peak value, and the average value is adjacent. Less, large area distribution is an indicator for measuring large area distribution. The larger value indicates a large region and a thicker texture, and the high grach-length distribution calculates a distribution of higher gray value lengths.

[0028] The preliminary imaging component is analyzed to obtain a target imaging syndrome, specifically comprises: S31, setting a variance feature value, excluding a feature smaller than the set variance threshold; S32, using a single variable feature selection method ANOVA (variance analysis) algorithm, resulting in the relationship between data concentrated imaging phenotypes and therapeutic response, evaluating the difference between different groups of different groups, excluding training groups and verification groups @P value more than 0.05 Features; S33, using the minimum absolute value of high dimensional data regression and selection operator regression model (LASSO) method, selecting a valid predictive feature from the original data set, using 10 times cross-validation method to calculate the minimum mean square error, according to The minimum mean square error is obtained by the optimal penalty parameters of the selection operator regression model; S34, the screening of the model is not zero.

[0029] The acquisition uses the original medical imaging and clinical information of patients with radiation therapy and screening, including: excluding patients with other cancers; excluding patients who have received hands-free treatment or radiotherapy but did not complete the complete treatment plan; exclude will lack Patients with high-resolution computer faults; specifically, intravitable pulmonary fibrosis, clinical symptoms, pathogenic microbial detection confirmed acute infections, tumor-related interstitial pulmonary disease, drug-related intermediate lung disease, dust lung, alveolar Protein sickness, lack of HRCT data, repeated hospitalization, etc., the corresponding patient's original medical image is not treated, the data is not adopted.

[0030] The clinical information specifically includes, but is not limited to, age, gender, basic disease, pathological type, TMN stage, first-treatment Hb, WBC average level, breathing difficult, cough, cough, and fever symptoms.

[0031] Such as figure 2 As shown, a lung cancer prognostic assessment system based on CT imaging group, including: pretreatment module 101: Collecting original medical imaging and clinical information using radiopathic lung cancer; classification module 102: Patients with lung cancer The proportion is randomly divided into training set and verification set, the training set for training prediction models, the test set is used to verify the performance of the prediction model; the extraction module 103: Scan the outline of the original medical image by high-resolution computer tomography Region, and extracting the imaging histology characteristics from the lesser region to obtain preliminary imaging histology; screening module 104: Screening of preliminary imaging histology characteristics, resulting in target imaging syndrome; prediction module 105: Target for training Imaging Method Training Prediction Model, Building a Machine Evaluation Model, and constructing an imaging method assessment model, and verifying a centralized verification model; verification module 106: Establishing a radiotherapy effective imaging system characteristic tag, and verifying in verification; rating; Module 107: Calculate the imaging syndrome score of each patient on the basis of a predictive model; converting the translation module 108: transforming the patient's imaging histology feature score to the probability of sensitivity of patient radiation therapy.

[0032] In the extraction module 103, the imaging component comprises 18 characteristics, 14 shape characteristics, 24 grayscale symbiotic matrix characteristics, 14 grayscale correlation matrix characteristics, 16 grayscale run matrix features 16, ash There are 16 matrix characteristics of the degree area, and 5 adjacent gray difference matrices, a total of 107 features.

[0033] Such as image 3 As shown, the screening module 104, specifically includes: a first screening unit 1041: sets the variance feature value, excluding features smaller than the set variance threshold; second screening unit 1042: ANOVA in a single variable feature selection method ( Variance analysis) Algorithm, obtaining the relationship between data concentrated imaging phenotypes and therapeutic response, evaluating the difference between different groups of different groups in the treatment reaction, excluding training group and verification group @P value exceeding 0.05 characteristics; Third screening unit 1043: Using the minimum absolute value of the high dimensional data regression and selection operator regression model (LASSO) method, select the effective prediction feature from the original data set, and use 10 times cross-validation method to calculate the minimum mean square error. The optimal penalty parameters of the selection operator regression model are obtained according to the minimum mean square error; the fourth screening unit 1044: Screening the characteristic set of the coefficient is not zero.

[0034] A computer device comprising a memory, a processor, and a computer program stored in the memory and can operate on the processor, the processor performs the computer program to implement CT imaging-based groups as described above. Lung cancer prognostic assistance evaluation method.

[0035] The processor can be a central processing unit (CPU) or other form of processing unit having data processing capabilities and / or instruction execution capabilities, and other components in the electronic device can control the desired function.

[0036] Memory can include one or more computer program products, computer program products can include various forms of computer readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory, for example, can include a random access memory (RAM) and / or a cache, and the like. Non-volatile memory, for example, can include read only memory (ROM), hard disk, flash memory, and the like. One or more computer program instructions can be stored on the computer readable storage medium, and the processor can run the program command to implement the steps in the storage authorization change method of the various embodiments of the present application above, and / or other desired functions. . Information such as light intensity, positioning of light intensity, filter, etc. can also be stored in a computer readable storage medium.

[0037] A computer readable storage medium, the storage medium stores a computer program, and when the computer program is executed by the processor, a prognosis method based on CT imaging group-based lung cancer prognostic auxiliary evaluation method is implemented.

[0038] The computer readable storage medium can be any combination of one or more readable media, and the readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can, for example, include, but are not limited to, systems, devices or devices, or devices, or devices, or components, or semiconductors, or semiconductors. More specific examples of readable storage media (non-exhaustive list) include: electrical connection, portable disc, hard disk, random access memory ((RAM), read-only memory (ROM) having one or more wires Wipe-type programmable read-only memory (EPROM or flash), fiber, portable compact disk read only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.

[0039] The basic principles of the present application are described above, but it is necessary to point out that the advantages, advantages, effects, etc. mentioned in the present application are only example rather than limitations, and these advantages, advantages, effects, etc. are not considered. The various embodiments of the present application must have. Further, the specific details of the above disclosure are merely for example and to facilitate understanding, and not limitation, the details are not limited to the present invention must be implemented in the specific details.

[0040] The block diagram of the device, device, device, and system according to the present application, as an example of exemplary examples and is not intended to require or imply, must be connected, arranged, configured in a manner shown in block diagram. As will be recognized in any way, these devices, devices, devices, and systems can be configured in any way. Words such as "including", "include", "having", etc., are open vocabulary, refers to "including but not limited to", and can be used interchangeably. The vocabulary "or" and "referred to herein" and / or ", and can be used interchangeably, unless the context explicitly indicates the case. The vocabulary used herein ",", ",", ",",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

[0041] It will also be noted that in the apparatus, apparatus, and methods of the present application, each component or each step is to decompose and / or recombine. These decomposition and / or re-combinations should be considered an equivalent solution of the present application.

[0042] The above description of the disclosed aspects is provided to enable any skilled in the art to make or use the present application. Various modifications to these aspects will be apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of the present application. Accordingly, the present application is not intended to be limited to the aspects shown here, but in accordance with the broadest range consistent with the principles and novel features disclosed herein.

[0043] The above description has been presented for purposes of illustration and description. Moreover, this description is not intended to limit the embodiments of the present application to the form disclosed herein. Although multiple examples and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, add, and sub-combinations.

[0044] It will be noted in that the above embodiments are intended to illustrate the technical solutions of the present invention, not limiting the invention, although the present invention will be described in detail, and those skilled in the art will understand: The technical scheme described in the foregoing embodiments can still be modified, or partially or all of the techniques is equivalent to alternative; and these modifications or replacements do not allow the present invention to take advantage of the present invention. range.