Method and device for predicting responsiveness to immune checkpoint inhibitor of non-small cell lung cancer patient, and computer program

Abstract

The present invention relates to a method for predicting the responsiveness to an immune checkpoint inhibitor of a non-small cell lung cancer patient, in which both whole transcriptome sequencing and radiology are used to build a prediction model reflecting a tumor-infiltrating lymphocyte configuration favorable for an immune checkpoint inhibitor in tumor-infiltrating lymphocytes, and thus the method has the advantage of enabling a more accurate prediction of an immune checkpoint inhibitor therapeutic response and a patient prognosis of an advanced non-small cell lung cancer patient.

Description

Method, device, and computer program for predicting responsiveness to immune checkpoint inhibitors in patients with non-small cell lung cancer

[0001] This application claims priority to Republic of Korea Patent Application No. 10-2024-0031205, filed with the Korean Intellectual Property Office on March 5, 2024, the disclosure of which is incorporated herein by reference.

[0002] The present invention was made under the support of the Ministry of Science and ICT, under the project identification number 1711184924 and project number 2022R1A2C1003999. The project management organization of the project is the National Research Foundation of Korea, the research project name is "Individual Basic Research (MSIT)", the research project name is "Development and verification of a model for predicting immuno-oncology drug response and prognosis based on multi-omics data integration and artificial intelligence algorithm", the project performing organization is Samsung Medical Center, and the research period is from March 1, 2022 to February 28, 2025.

[0003] The present invention relates to a method, a device, and a computer program for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor, and more particularly, to a method, a device, and a computer program for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor by inputting a CT image into a prediction model.

[0004] The advent of immune checkpoint inhibitors (ICIs) has opened a new chapter in lung cancer treatment. Numerous clinical trials have demonstrated the clinical benefits of ICIs in non-small cell lung cancer (NSCLC), and many clinical guidelines now consider ICIs as a first-line treatment option for patients with NSCLC without targetable mutations.

[0005] The primary target of immune checkpoint inhibitors is to counteract tumor cell immune evasion by upregulating immunosuppressive pathways, such as the programmed cell death protein 1 (PD-1) / programmed cell death ligand 1 (PD-L1) pathway. Most immune checkpoint inhibitors are designed to block these pathways and subsequently stimulate an anti-tumor immune response.

[0006] A successful anti-tumor immune response requires an appropriate tumor microenvironment (TME) comprised of diverse tumor-infiltrating lymphocytes (TILs). Many researchers have reported a correlation between the quality and abundance of TILs and the clinical success of immune checkpoint inhibitors. A previous study reported a radiologically-based prediction model for TIL enrichment, which significantly correlated with the outcome of immune checkpoint inhibitors in patients with non-small cell lung cancer.

[0007] However, the clinical success of immune checkpoint inhibitors depends not only on the quantity of tumor-infiltrating lymphocytes but also on the composition of these cells within the tumor microenvironment. For example, a high proportion of cytotoxic CD8+ T cells and CD45+ type 1 helper T cells in the tumor microenvironment is associated with improved patient survival, whereas a high proportion of infiltrating M2 macrophages is associated with a poor prognosis.

[0008] The validity of this approach has been demonstrated in previous studies. Danaher et al. reported that immune cell subpopulations can be measured in the tumor microenvironment using a set of marker genes and immune cell scores. While several radiomics models exist to predict specific immune cells in the tumor microenvironment, these models are not frequently used clinically and remain sketchy. The present inventors concluded that radiomics models could be more precise and predictive if they reflected the composition of tumor-infiltrating lymphocytes in the tumor microenvironment.

[0009] Therefore, there is a need for a predictive model that reflects the tumor-infiltrating lymphocyte composition favorable for immune checkpoint inhibitors in the microtumoral environment for immune checkpoint inhibitor treatment, utilizing both radiomics and whole transcriptome sequencing (WTS), which shows more accurate performance in predicting immune checkpoint inhibitor treatment response and patient prognosis in patients with advanced non-small cell lung cancer.

[0010] [Prior Art Literature]

[0011] [Patent Document]

[0012] Korean Patent No. 10-2216725

[0013] The present invention aims to provide a method, device, and computer program for predicting the responsiveness of non-small cell lung cancer patients to immune checkpoint inhibitors by accessing the composition of tumor-infiltrating lymphocytes favorable for immune checkpoint inhibitors through whole transcriptome sequencing and constructing a predictive model for immune checkpoint inhibitor treatment through the correlation between the composition of tumor-infiltrating lymphocytes and radiological features.

[0014] In order to achieve the above object, the present invention is characterized in that it comprises the steps of: receiving whole-length transcriptome sequencing and CT images of one or more patients; calculating a cell type score that quantifies the expression value of a specific marker gene obtained from the whole-length transcriptome sequencing; identifying a tumor-infiltrating lymphocyte cell population having the cell type score equal to or higher than a specific value in a patient group responsive to an immune checkpoint inhibitor among the one or more patients, and calculating an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte in the identified cell population; extracting one or more radiological features from the CT image; selecting a significant radiological feature based on a correlation between the immune score and the one or more radiological features; and training a regression model using the immune score and the selected significant radiological feature.

[0015] Preferably, the one or more patients may be patients receiving immune checkpoint inhibitor treatment.

[0016] Preferably, the receiving step may receive a full-length transcriptome sequence acquired within one year from the time of immune checkpoint inhibitor treatment, and may receive a CT image acquired within one month from the time of immune checkpoint inhibitor treatment.

[0017] Preferably, the receiving step can receive demographic characteristics and tumor characteristics of one or more patients.

[0018] Preferably, the receiving step may receive at least one of gender, age, and smoking history as demographic characteristics, and at least one of pathological characteristics, tumor proportional score (PD-L1 TPS), and target mutation as tumor characteristics.

[0019] Preferably, the step of calculating the cell type score may be a marker gene that measures the expression level of a population of tumor-infiltrating lymphocyte cells.

[0020] Preferably, the step of calculating the cell type score may calculate the cell type score as a simple average of the log-transformed expression values ​​of specific marker genes obtained from the full-length transcriptome sequencing.

[0021] Preferably, the step of calculating the immune score can identify a tumor-infiltrating lymphocyte cell population having a significantly high cell type score by performing a chi-square test.

[0022] Preferably, the step of extracting the radiological feature may extract at least one of a primary feature, a shape feature, a gray level co-occurrence matrix feature, a gray level magnitude domain matrix feature, a cumulative distribution function feature, a physical feature, and a fractal feature as the radiological feature.

[0023] Preferably, the step of selecting the significant radiological features may calculate a Pearson correlation coefficient between the immune score and the extracted radiological features.

[0024] Preferably, the step of selecting the significant radiological feature may be performed such that the Pearson correlation coefficient is selected as a significant radiological feature when ρ<0.05 or ρ>0.2.

[0025] Preferably, the method may further include a step of inputting a CT image of a specific patient into the regression model to predict an immune score.

[0026] In addition, the present invention provides a device for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor, comprising: a processor including one or more cores; and a memory; wherein the processor receives whole-length transcriptome sequencing and CT images of one or more patients, calculates a cell type score that quantifies the expression value of a specific marker gene obtained from the whole-length transcriptome sequencing, identifies a tumor-infiltrating lymphocyte cell population having the cell type score equal to or higher than a specific value in a patient population responsive to an immune checkpoint inhibitor among the one or more patients, calculates an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte in the identified cell population, extracts one or more radiological features from the CT image, selects a significant radiological feature based on a correlation between the immune score and the one or more radiological features, and trains a regression model using the immune score and the selected significant radiological feature.

[0027] In addition, the present invention is characterized in that it is a computer program including instructions stored in a computer-readable storage medium and causing a computer to perform the following operations, wherein the operations include: receiving whole-length transcriptome sequencing and CT images of one or more patients; calculating a cell type score that quantifies the expression value of a specific marker gene obtained from the whole-length transcriptome sequencing; identifying a tumor-infiltrating lymphocyte cell population having a cell type score greater than or equal to a specific value in a patient population responsive to an immune checkpoint inhibitor among the one or more patients, and calculating an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte in the identified cell population; extracting one or more radiological features from the CT image; selecting a significant radiological feature based on a correlation between the immune score and the one or more radiological features; training a regression model using the immune score and the selected significant radiological feature; and inputting a CT image of a specific patient into the regression model to predict an immune score.

[0028] The present invention utilizes both whole transcriptome sequencing and radiomics to build a predictive model that reflects the composition of tumor-infiltrating lymphocytes favorable for immune checkpoint inhibitors in tumor-infiltrating lymphocytes, thereby enabling more accurate prediction of immune checkpoint inhibitor treatment response and patient prognosis in patients with advanced non-small cell lung cancer.

[0029] Figure 1 illustrates a flowchart of a method for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor according to an embodiment of the present invention.

[0030] Figure 2 schematically illustrates a method for predicting responsiveness of non-small cell lung cancer patients to immune checkpoint inhibitors according to an embodiment of the present invention.

[0031] Figure 3 illustrates a development and training cohort patient flow diagram according to an embodiment of the present invention.

[0032] Figure 4 shows the differences in cell type scores between immune checkpoint inhibitor responders and non-responders according to an embodiment of the present invention.

[0033] Figure 5 shows the correlation between the original immune score and the predicted score of the regression model (radTIL) in both the training and validation cohorts according to an embodiment of the present invention.

[0034] Figure 6 shows the difference in cell type scores between the radTIL high group and the radTIL low group according to an embodiment of the present invention.

[0035] Figure 7 shows the performance of radTIL in predicting the outcome of immune checkpoint inhibitors in both training and validation cohorts according to an embodiment of the present invention.

[0036] Figures 8 and 9 show the results of radTIL subgroup analysis in various situations using a validation cohort according to an embodiment of the present invention.

[0037] Figure 10 shows the performance of PD-L1 in predicting the outcome of immune checkpoint inhibitors using a validation cohort according to an embodiment of the present invention.

[0038] FIG. 11 shows a schematic diagram of a device (100) for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor according to an embodiment of the present invention.

[0039] Figure 12 illustrates a schematic diagram of a computing environment according to an embodiment of the present invention.

[0040] Hereinafter, the present invention will be described in detail with reference to the contents described in the attached drawings. However, the present invention is not limited or restricted by the exemplary embodiments. The same reference numerals in each drawing indicate components that perform substantially the same functions.

[0041] The purpose and effects of the present invention can be naturally understood or made clearer by the following description, and the purpose and effects of the present invention are not limited solely by the following description. Furthermore, in describing the present invention, if a detailed description of known technologies related to the present invention is deemed to unnecessarily obscure the gist of the present invention, such detailed description will be omitted.

[0042] The terminology used herein is merely used to describe specific embodiments and is not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, it should be understood that the terms "comprise" or "have" indicate the presence of a feature, number, step, operation, component, part, or combination thereof described in the description of the invention, but do not preclude the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0043] While terms like "first" and "second" may be used to describe various components, these components should not be limited by these terms. These terms are used solely to distinguish one component from another. For example, without departing from the scope of the present invention, a first component could be referred to as a "second component," and similarly, a second component could also be referred to as a "first component."

[0044] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Terms defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and shall not be interpreted in an idealized or overly formal sense unless explicitly defined herein.

[0045] When interpreting components, even if there is no explicit description, it is interpreted as including the margin of error. When describing temporal relationships, for example, when temporal continuity is described with phrases such as "after," "following," "next to," or "before," this also includes cases where the relationship is not continuous, unless "immediately" or "directly" is used.

[0046] Hereinafter, the technical configuration of the present invention will be described in detail with reference to the attached drawings.

[0047] FIG. 1 is a flowchart of a method for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor according to an embodiment of the present invention. Referring to FIG. 1, the method for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor may include a receiving step (S100), a calculating cell type score step (S200), an immune score calculation step (S300), a radiomics feature extraction step (S400), a selecting significant radiomics feature step (S500), and a training regression model step (S600). The method for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor may further include a predicting immune score step (S700).

[0048] Tumor-infiltrating lymphocytes (TILs) in the tumor microenvironment (TME) influence the outcome of immune checkpoint inhibitors (ICIs) in non-small cell lung cancer (NSCLC). While previous studies have focused on radiological models that correlate the abundance of TILs with the outcome of immune checkpoint inhibitors in NSCLC patients, the present invention focuses on the composition of TILs, as well as the abundance of TILs.

[0049] A method for predicting the responsiveness of non-small cell lung cancer patients to immune checkpoint inhibitors is to obtain information on the composition of tumor-infiltrating lymphocytes favorable for immune checkpoint inhibitors through whole transcriptome sequencing (WTS), and then establish a prediction model for immune checkpoint inhibitor treatment by finding a correlation between the obtained tumor-infiltrating lymphocyte composition and radiomics features (RF).

[0050] Figure 2 schematically illustrates a method for predicting responsiveness to an immune checkpoint inhibitor in a non-small cell lung cancer patient according to an embodiment of the present invention. Referring to Figure 2, the method for predicting responsiveness to an immune checkpoint inhibitor in a non-small cell lung cancer patient can identify the composition of tumor-infiltrating lymphocytes favorable to an immune checkpoint inhibitor in whole transcriptome sequencing of tissues of refractory or metastatic non-small cell lung cancer patients before immune checkpoint inhibitor treatment. Next, the method for predicting responsiveness to an immune checkpoint inhibitor in a non-small cell lung cancer patient can extract radiomic features from chest CT images of the patient taken before immune checkpoint inhibitor treatment. Next, the method for predicting responsiveness to an immune checkpoint inhibitor in a non-small cell lung cancer patient can train a prediction model using the composition of tumor-infiltrating lymphocytes and the extracted radiomic features.

[0051] The receiving step (S100) may receive whole-genome sequencing and CT images of one or more patients. One or more patients may be patients receiving immune checkpoint inhibitor treatment and may be refractory or metastatic non-small cell lung cancer patients.

[0052] The receiving step (S100) may receive a full-length transcriptome sequence obtained within one year from the time of immune checkpoint inhibitor treatment and a CT image obtained within one month from the time of immune checkpoint inhibitor treatment. The receiving step (S100) may receive a full-length transcriptome sequence and a CT image of a patient received before immune checkpoint inhibitor treatment.

[0053] The receiving step (S100) may receive demographic characteristics and tumor characteristics of one or more patients. The receiving step (S100) may receive at least one of gender, age, and smoking history as demographic characteristics, and may receive at least one of pathological characteristics, tumor proportional score (PD-L1 TPS), and target mutation (EGFR mutation, ALK translocation) as tumor characteristics. The tumor proportional score (PD-L1 TPS) may be determined through the results of the PD-L1 IHC SP263 assay, the PD-L1 IHC 22C3 pharmDx assay, and the PD-L1 IHC SP142 assay.

[0054] The step of calculating a cell type score (S200) can calculate a cell type score that quantifies the expression value of a specific marker gene obtained from full-length transcriptome sequencing.

[0055] The step of calculating a cell type score (S200) may be a marker gene measuring the expression level of a specific marker gene in a tumor-infiltrating lymphocyte population. The step of calculating a cell type score (S200) may employ marker gene information measuring the expression level of a tumor-infiltrating lymphocyte population as shown in Table 1 below to identify a tumor-infiltrating lymphocyte composition favorable for immune checkpoint inhibitors.

[0056] GeneCell typeBLKB-cellsCD19B-cellsMS4A1B-cellsFCRL2B-cellsPTPRCCD45CD8ACD8 T cellsCTSWCytotoxic cellsGNLYCytotoxic cellsGZMACytotoxic cellsGZMBCytotoxic cellsGZMHCytotoxic cellsKLRB1Cytotoxic cellsKLRD1Cytotoxic cellsKLRK1Cytotoxic cellsPRF1Cytotoxic cellsNKG7Cytotoxic cellsCCL13DCCD209DCHSD11B1DCCD244Exhausted CD8EOMESExhausted CD8LAG3Exhausted CD8PTGER4Exhausted CD8CD163MacrophagesCD68MacrophagesCD84MacrophagesMS4A4AMacrophagesMS4A2Mast cellsTPSAB1Mast cellsCPA3Mast cellsHDCMast cellsTPSB2Mast cellsCSF3RNeutrophilsFCGR3BNeutrophilsFPR1NeutrophilsIL21RNK CD56dim cellsCD3DT-cellsCD3ET-cellsCD3GT-cellsCD6T-cellsSH2D1AT-cellsTRAT1T-cellsTBX21Th1 cellsFOXP3Treg

[0057] Assuming that each marker gene is present at a fixed number per cell, the average log-transformed expression level of the marker gene is equal to the log-transformed abundance of the cell type plus an unknown constant. Therefore, the step of calculating the cell type score (S200) can calculate the cell type score as a simple average of the log-transformed expression values ​​of specific marker genes obtained from whole-length transcriptome sequencing.

[0058] The step of calculating an immune score (S300) may identify a tumor-infiltrating lymphocyte cell population having a cell type score higher than a specific value in a patient group responsive to an immune checkpoint inhibitor among one or more patients, and calculate an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte in the identified cell population. The step of calculating an immune score (S300) may identify a tumor-infiltrating lymphocyte cell population having a significantly high cell type score in a group of immune checkpoint inhibitor responders by performing a chi-square test.

[0059] The step of extracting radiological features (S400) can extract one or more radiological features from a CT image. The target lesion can be segmented by drawing a volume of interest (VOI) using commercial software, AVIEW Research, and a slice-by-slice approach. The step of extracting radiological features (S400) can modify the lesion boundary to avoid adjacent air, fat, blood vessels, and surrounding organs. The step of extracting radiological features (S400) can extract radiometric features from the raw image for a given region of interest (ROI) using a combination of open source and MATLAB code.

[0060] The step (S400) of extracting radiological features may extract at least one of a first order feature (intensity), a shape feature, a gray level co-occurrence matrix feature (GLCM), a gray level size zone matrix feature (GLSZM), a cumulative distribution function feature (CDF), a physical feature, and a fractal feature as the radiological features.

[0061] Primary features are based on statistics on the distribution of voxel intensities within an image region defined by a ROI. Gray-level co-occurrence matrix features are texture-based features that describe a second-order joint probability function of an image region and consider the connectivity between adjacent voxels. Gray-level magnitude area matrix features quantify gray-level areas, defined as the number of connected voxels sharing the same gray-level intensity. Shape features describe the morphological characteristics of a tumor. Physical features can be density and mass. Density values ​​are calculated using a predefined mass density value, specifically, mass density * the average of intensity values ​​for a given ROI + a constant. Here, mass density is 0.0011 and the constant is 1.1. Mass values ​​can be calculated as the volume of a voxel * mass density. Fractal features provide a means to assess the structural shape of tumor vessels.

[0062] Primary features can be extracted through the formulas described in Table 2 below, gray level co-occurrence matrix features can be extracted through the formulas described in Table 3 below, gray level size domain matrix features can be extracted through the formulas described in Table 4 below, cumulative distribution function features can be extracted through the formulas described in Table 5 below, shape features can be extracted through the formulas described in Table 6 below, physical features can be extracted through the formulas described in Table 7 below, and fractal features can be extracted through the formulas described in Table 8 below.

[0063] First order featuresParameterFormulaDescriptionMaximum Where denote the 3d image matrixMeasures maximum intensity value of a histogramMinimum Where denote the 3d image matrixMeasures minimum intensity value of a histogramMedian Where denote the 3d image matrixMeasures median intensity value of a histogramMean Where denote the 3d image matrix with voxel.Measures mean intensity value of a histogramVariance Measures squared distances of each value of a histogram from the meanEnergy Where denote the 3d image matrix with voxel.Measures squared magnitude value of a histogramTotal Energy Where denote the 3d image matrix with voxel and V voxelis the volume of voxel.Measures squared magnitude value of a histogram scaled by the volume of the voxelStandard deviation Where denote the 3d image matrix with voxel.Measures amount of variation of a histogram.Skewness Where is the mean of , is the standard deviation of , is the expectation operator.Measures asymmetry of a histogram.Kurtosis Where is the mean of , is the standard deviation of , is the expectation operator.Measures 'peakedeness' of a histogram (flatness of histogram)Root mean square (RMS) Where denote the 3d image matrix with voxel.Measures the square-root of the mean of the squares of the values of the histogram. This feature is another measure of the magnitude of a histogramMean Absolute Deviation Where denote mean value from histogram with voxelMeasure the mean distance of all intensity values from histogramRobust Mean Absolute Deviation (rMAD) Where denote the number of values on histogram between 10 th and 90 th percentile and denote the mean value on histogram between 10 th and 90 th percentileMeasures the mean distance of all intensity values from the mean value calculated on the histogram between the 10 th and 90 th percentileInter quartile range Where denote the 3 rd quartile of histogram, denote the 1 st quartile of histogramMeasures of variability, based on dividing a histogram into quartilesRange Measures difference between the highest and lowest voxel values of a histogramEntropy Where denote the first order histogram with discrete intensity levels.Measures irregularity of a histogram.Uniformity Where denote the first order histogram with discrete intensity levels.Measures uniformity of a histogram.Percentile Measures intensity value at the 2.5 th , 25 th , 50 th , 75 th , and 97.5 th percentile on histogram

[0064] Gray-level Co-Occurrence Matrix (GLCM) based featuresAutocorrelation Measures of the magnitude of the fineness and coarseness of textureCluster Prominence Measures the skewness and asymmetry of the GLCMCluster Shade Measures the skewness and uniformity of the GLCMCluster tendency Measures of the homogeneity of GLCMMaximum probability Measures maximum value of GLCM matrixContrast Measures of the local intensity variation of GLCMCorrelation Measures the linear dependency of gray level values to their respective voxels in the GLCMDifference Average Measure the relationship between occurrences of pairs with similar intensity values and occurrences of pairs with differing intensity valuesDifference entropy Measures entropy of processed GLCM matrix Px-yDifference Variance Measures the heterogeneity that places higher weights on differing intensity level pairs that deviate more from the meanInverse Difference (ID) Measures the local homogeneity of an imageInverse Difference Moment (IDM) Measures the normalized local homogeneity of an image.Inverse Difference Moment Normalized (IDMN) Measures the local homogeneity of an imageInverse Difference Normalized (IDN) Measures the normalized local homogeneity of an image.Informational measure of correlation 1 (IMC1) Secondary measure of Homogeneity1Informational measure of correlation 2 (IMC2) Inverse Variance Measures the inverse variance of the GLCMJoint Average Measures the mean gray level intensity of the distribution on GLCMJoint Energy Measures the homogeneous patterns in the imageJoint Entropy Measures the randomness and variability in neighborhood intensity valuesMaximum Correlation Coefficient (MCC) Measures the complexity of the textureSum average Measures the relationship between occurrences of pairs with lower and higher intensity valuesSum entropy Sum of neighborhood intensity value differencesSum Square Value in the distribution of neighboring intensity level pairs about the mean intensity level in the GLCM

[0065] Gray Level Size Zone Matrix (GLSZM) FeaturesGray Level Non-Uniformity (GLN) Measures the variability of gray-level intensity values in the imageGray Level Non-uniformity Normalized (GLNN) Measures the variability of gray-level intensity values in the image and then normalizedGray Level Variance (GLV) Measures the variance in gray level intensities for the zonesHigh Gray Level Zone Emphasis (HGLZE) Measures the distribution of the higher gray-level valuesLarge Area Emphasis (LAE) Measures the distribution of large area size zonesLarge Area High Gray Level Emphasis (LAHGLE) Measures the proportion in the image of the joint distribution of larger size zones with higher gray-level valuesLarge Area Low Gray Level Emphasis (LALGLE) Measures the proportion in the image of the joint distribution of larger size zones with lower gray-level valuesLow Gray Level Zone Emphasis (LGLZE) Measures the distribution of lower gray-level size zonesSize Zone Non-Uniformity (SZM) Measures the variability of size zone volumes in the imageSize Zone Non-Uniformity Normalized (SZNN) Measures the variability of size zone volumes in the image and then normalizedSmall Area Emphasis (SAE) Measures the distribution of large area size zonesSmall Area High Gray Level Emphasis (SAHGLE) Measures the proportion in the image of the joint distribution of smaller size zones with higher gray-level valuesSmall Area Low Gray Level Emphasis (SALGLE) Measures the proportion in the image of the joint distribution of smaller size zones with higher gray-level valuesZone Entropy (ZE) Measures the uncertainty and randomness in the distribution of zone size and gray levelsZone Percentage (ZP) Measures the coarseness of the texture by taking the ratio of number of zones and number of voxels in the ROIZone Variance (ZV) Measures the variance in zone size volumes for the zones

[0066] Cumulative Distribution Function (CDF) based Features (Ref)Standard Deviation of SlopeSee description in the next columnMeasures the standard deviation of slope values extracted form CDFMean of SlopeSee description in the next columnMeasures the mean of slope values extracted form CDF75 th Percentile of SlopeSee description in the next columnMeasures the slope value at the 75 th percentile on histogram for slope valueSkewness of SlopeSee description in the next columnMeasures asymmetry of a histogram for slope value.Kurtosis of SlopeSee description in the next columnMeasures "peakedness"of a histogram for slope value (flatness of histogram)

[0067] Morphological featuresShape and Size based featuresMesh Volume Volume calculated from the triangle mesh of the ROIElongation Where and are the lengths of the largest and second largest principal component axesMeasure the relationship between the two largest principal components in the ROI shapeFlatness Where and are the lengths of the largest and smallest principal component axesMeasures the relationship between the largest and smallest principal components in the ROI shapeLeast Axis Length Where denote the smallest principal componentMeasures the smallest axis length of the ROI-enclosing ellipsoidMajor Axis Length Where denote the largest principal componentMeasures the largest axis length of the ROI-enclosing ellipsoidMinor Axis Length Where denote the second largest principal componentMeasures the second-largest axis length of the ROI-enclosing ellipsoidCompactness Where denote the volume and denote the surface area of the volume of interest (VOI)Quantifies how close an object to the smoothest shape, the circleSurface area Where is the total number triangle (coved surface area) and are edge vectorsThe surface area of the ROIConvexity Where denote tumor volume and denote convex hull volumeMeasures ratio of the ROI volume contained within the tumor to the calculated convex hull volumeSphericity Where denote area and denote tumor volumeMeasures of the roundness of the ROIMaximum 2D Diameter ColumnSee description in the next columnMeasures the largest pairwise Euclidean distance between tumor surface mesh vertices in the row-slice (coronal) planeMaximum 2D DiameterSee description in the next columnMeasures the largest pairwise Euclidean distance between tumor surface mesh verticesMaximum 2D Diameter SliceSee description in the next columnMeasures the largest pairwise Euclidean distance between tumor surface mesh vertices in the row-column (axial) planeMaximum 3D diameterSee description in the next columnMeasures of the maximum 3D ROI diameter. It is measured as the largest pairwise Euclidean distance, between surface voxels of the ROISpherical disproportion Where is the radius of a sphere with the same volume as the ROIThe ratio of the surface area of the ROI to the surface area of a sphere with the same volume as the ROISurface to volume ratio (SVR) Where is area and is volumeSurface to volume ratioVolume Where denote the 3d image resolutionVolume of tumor (ROI)

[0068] Physical (Density) featuresDensityMass density * Mean of intensity values + ConstantDensity of tumor (ROI)MassVoxel volume * Mass densityTumor mass (ROI)

[0069] Fractal FeaturesFractal Dimension (Box-counting method)Fractal dimension and fractal abundance are calculated by using box-counting method on the basis of the equation as follow (Ref): Where is the box size, is the number of boxes of size needed to cover the object, and is the fractal dimensionBy plotting a log-log plot of versus fractal abundance, or log , can be obtained from the y-intercept of the straight portion of the curveFractal dimension quantifies morphological complexity and provides information on the self-similarity propertiesFractal Abundance (Box-counting method)Lacunarity (Box-counting method) Where is lacunarity at box size , is mass or pixel of interest, and is probability of M in box size Measures of the texture or distribution of gaps within an imageFractal Signature Dissimilarity (Ref) (Blanket method) Where is calculated by using nth slice of the tumor volume, is the sub-volume of the tumor's nth slice and is the distance of the nth slice to the tumor's central sliceMeasure of tumor heterogeneity information

[0070]

[0071] The step of selecting meaningful radiological features (S500) may select meaningful radiological features based on the correlation between the immunological score and one or more radiological features. The step of selecting meaningful radiological features (S500) may calculate a Pearson correlation coefficient (ρ) between the immunological score and the extracted radiological features. The step of selecting meaningful radiological features (S500) may select radiological features as significant if the Pearson correlation coefficient is ρ <0.05 or ρ >0.2. The step of selecting meaningful radiological features (S500) may exclude features exhibiting multicollinearity by calculating a variance inflation factor, and may exclude features with a variance inflation factor value of 20 or more.

[0072] The regression model training step (S600) can train a regression model using the immune score and selected significant radiological features. The regression model training step (S600) can utilize Elastic Net regression.

[0073] The step of training the regression model (S600) can obtain the optimal cutoff value that best predicts the patient's progression-free survival through ROC (Receiver operating characteristic) analysis.

[0074] The step of predicting an immune score (S700) can predict an immune score by inputting a CT image of a specific patient into the regression model.

[0075]

[0076] Below, the simulation results of the present invention are described.

[0077]

[0078] Patient demographics

[0079] Figure 3 illustrates a patient flow diagram of the development and training cohorts according to an embodiment of the present invention. Referring to Figure 3, the development cohort included a total of 120 non-small cell lung cancer patients, including 98 men and 22 women. Most patients in the development cohort had metastatic disease (87.5%) and received immune checkpoint inhibitor treatment, while the remaining patients had disease that was refractory to or relapsed with conventional chemotherapy (12.5%) and received immune checkpoint inhibitor treatment. Most patients did not have mutations amenable to target therapy (EGFR negative, 73.3%, ALK negative, 88.3%). In the development cohort, 107 patients (89.2%) experienced disease progression, with a median progression-free survival (PFS) of 10.2 months. The validation cohort included 319 patients and had similar demographic characteristics to the development cohort, although there were significant differences in smoking status, ALK mutation status, and PD-L1 expression. Detailed clinical characteristics of patients in the development and validation cohorts are presented in Table 9 below.

[0080] Development cohortTest cohortP value*Sex0.153Male98 (81.7%)238 (74.6%)Female22 (18.3%)81 (25.4%)Age0.812Below 6056 (48.5%)143 (45.7%)Above 6054 (51.5%)176 (54.3%)Smoking<.001Current46 (38.3%)75 (23.5%)Former47 (39.2%)96 (30.1%)Never27 (22.5%)148 (46.4%)ECOG status0.13104 (3.3%)20 (6.3%)1110 (91.7%)264 (82.8%)26 (5.0%)34 (10.7%)30 (0%)1 (0.3%)Pathology0.405SqCC39 (32.5%)84 (26.3%)ADC74 (61.7%)211 (66.1%)Others7 (5.8%)24 (7.5%)Stage (TNM 8 th)0.962315 (12.5%)38 (11.9%)4A57 (47.5%)149 (46.7%)4B48 (40.0%)132 (41.4%)EGFR status0.728Positive19 (15.8%)59 (18.5%)Negative88 (73.3%)231 (72.4%)N / A13 (10.8%)29 (9.1%)ALK status0.031Positive5 (4.9%)2 (0.6%)Negative106 (88.3%)292 (91.5%)N / A9 (7.5%)25 (7.8%)PD-L1 (TPS ≥50%)<.001Positive55 (45.8%)150 (47.0%)Negative57 (47.5%)169 (53.0%)N / A8 (6.7%)0 (0%)Type of treatment1.000ICI monotherapy117 (97.5%)310 (97.2%)ICI + chemotherapy3 (2.5%)10 (2.8%)Progression1.000Yes107 (89.2%)286 (89.7%)No13 (10.8%)33 (10.3%)Mean PFS # 10.2 ± 15.910.3 ± 16.50.972Total120319

[0081] Differences in immune cell type scores between immune checkpoint inhibitor responders and non-responders

[0082] Figure 4 shows the differences in cell type scores between immune checkpoint inhibitor responders and non-responders according to an embodiment of the present invention. Referring to Figure 4, the immune checkpoint inhibitor responder group had significantly higher cell type scores of CD45 cells, CD8 T cells, cytotoxic cells, exhausted CD8 cells, macrophages, and NK CD56dim cells compared to the non-responder group. Among them, the log2 fold changes between the groups for CD45 cells, CD8 T cells, and macrophages were 3.67, 1.34, and 1.70, respectively, which were greater than 1. On the other hand, the log2 fold changes for exhausted CD8 cells, cytotoxic cells, and NK CD56dim cells were less than 1, specifically, 0.35, 0.46, and 0.23, respectively.

[0083] The immune score was defined as the sum of the six cell type scores and was significantly higher in the immune checkpoint inhibitor responder group.

[0084] Radiological model related to immune score (radTIL)

[0085] To build and validate the radiomics model, the development and validation cohorts were randomly divided into training and validation cohorts in a ratio of 6:4. Given an expected correlation coefficient (ρ) of 0.45, the minimum sample size for the validation cohort was calculated to be 36 to achieve a power of at least 80%. Therefore, the development and validation cohorts were divided in a ratio of 6:4.

[0086] A total of 88 radiomic features were extracted from 120 ROIs in the development cohort. Of these, 46 radiomic features with high correlations (ρ > 0.9) were removed, retaining only one feature for further analysis. After filtering, 27 features with significantly low correlations with the immune score were also removed, and three features exhibiting multicollinearity were excluded.

[0087] The final radiomics model (radTIL) was built with the radiomics data of the training cohort and consists of seven radiomics features (one shape feature, three GLCM features, two GLSZM features, and one fractal feature) selected by Elastic Net regression, as shown in Table 10 below.

[0088] Figure 5 illustrates the correlation between the original immune scores and the predicted scores of the regression model (radTIL) in both the training and validation cohorts according to an embodiment of the present invention. Referring to Figure 5 , in both the training and validation cohorts, the predicted scores of the radTIL model were significantly correlated with the original immune scores, with an estimated power of 88.1% assuming an actual correlation coefficient of 0.46.

[0089] TypeName of featureCoefficientShapeSurfaceVolumeRaito0.3778249GLCMCorrelation-4.8943792InverseVariance-1.8394695Imc2 (Informational Measure of Correlation 2)2.4990431GLSZMSmallAreaLowGrayLevelEmphasis (SALGLE)15.8666986ZoneEntropy2.1272768FractalLacunarity_b5-3.5139851

[0090] Using the radTIL model, we analyzed the outcomes of patients receiving immune checkpoint inhibitors in the training and validation cohorts, demonstrating that the radTIL model accurately reflects the immune aspect of the immune score. Using the optimal cutoff value from the training cohort, patients in the training and validation cohorts were divided into radTIL-high and radTIL-low groups, respectively.

[0091] Figure 6 shows the differences in cell type scores between the radTIL high group and the radTIL low group according to an embodiment of the present invention. Referring to Figure 6, in the radTIL high group, the cell type scores of CD45 cells, CD8 T cells, and macrophages were significantly higher than those in the radTIL low group, and these scores were also significantly higher in the immune checkpoint inhibitor responder group. In contrast, the cell type scores of cytotoxic cells, exhausted CD8 cells, and NK CD56dim cells were significantly higher in the immune checkpoint inhibitor responder group, but showed low fold changes (log2 fold changes < 1) and did not show significant differences.

[0092] Performance of a regression model (radTIL) for predicting the outcome of immune checkpoint inhibitors

[0093] Figure 7 illustrates the performance of radTIL in predicting the outcome of immune checkpoint inhibitors in both the training and validation cohorts according to an embodiment of the present invention. Figures 8 and 9 illustrate the results of radTIL subgroup analyses in various situations using the validation cohort according to an embodiment of the present invention. Figure 10 illustrates the performance of PD-L1 in predicting the outcome of immune checkpoint inhibitors using the validation cohort according to an embodiment of the present invention.

[0094] Patients with higher radTIL counts showed significantly longer PFS compared to patients with lower radTIL counts in both the training (median 4.5 months [95% CI 2.7 - ] vs. 2.8 months [95% CI 2.0 - 6.8], p = 0.006; Fig. 7A ) and validation cohorts (median 4.0 months [95% CI 2.3 - 6.9] vs. 2.4 months [95% CI 2.0 - 3.5], p = 0.003; Fig. 7B ). Subgroup analyses generally showed longer PFS in the radTIL high group compared to the low group ( Figs. 8A-C ). Notably, the radTIL model demonstrated efficacy in the setting of PD-L1-positive metastatic NSCLC ( Figs. 9D-F ). In PD-L1-positive NSCLC, patients with PD-L1 TPS ≥50% had significantly longer progression-free survival, but PD-L1 positivity alone did not predict outcome of immune checkpoint inhibitors (PD-L1 TPS >50%, median 3.5 months [95% CI 2.3 - 5.0] vs. 2.4 months [95% CI 2.0 - 4.0], p = 0.053; PD-L1 positivity, median 3.0 months [95% CI 2.3 - 4.1] vs. 2.2 months [95% CI 1.9 - 4.9], p = 0.840; Fig. 10A-B). However, applying the radTIL model to patients with PD-L1-positive tumors could better predict the outcome of immune checkpoint inhibitors (PD-L1+ radTIL, median 4.1 months [95% CI 3.0-8.3] vs. 2.6 months [95% CI 2.0-3.6], p <0.001; PD-L1 TPS > 50%+ radTIL, median 7.2 months [95% CI 3.9-12.4] vs. 2.6 months [95% CI 1.9-4.2], p = 0.001; Fig. 9E-F).

[0095] We performed a Cox proportional hazards model analysis to assess whether radTILs were independent of potential confounders. As shown in Table 11 below, the group with higher radTIL counts was significantly associated with prolonged PFS (HR 0.66, 95% CI 0.51–0.84, p = 0.001), independent of PD-L1 status and other clinical variables.

[0096] Univariable analysisMultivariable analysisHR (95% CI)P valueHR (95% CI)P valueAge0.336Above 600.88 (0.68 - 1.14)Below 60ReferenceSex0.554Female0.90 (0.64 - 1.27)MaleReferenceSmoking0.127Never0.73 (0.52 - 1.03)0.069Former0.77 (0.55 - 1.06)0.109CurrentReferencePathology0.314ADC0.83 (0.59- 1.15)0.255Others0.68 (0.40 - 1.16)0.158SqCCReferenceTNM 8 thStage0.0390.06630.73 (0.56 - 0.95)0.0190.75 (0.58 - 0.98)0.0324A0.68 (0.44 - 1.06)0.0870.72 (0.48 - 1.10)0.1264BReferenceReferenceEGFR0.0870.039Positive1.34 (0.96 - 1.86)1.38 (1.02 - 1.87)NegativeReferenceReferenceALK0.895Positive1.10 (0.27 - 4.56)NegativeReferencePD-L10.819Positive1.04 (0.73 - 1.48)NegativeReferencePD-L1 (≥50%)0.0280.011Positive0.72 (0.53 - 0.97)0.72 (0.57 - 0.93)NegativeReferenceReferenceradTIL model0.0010.001High group0.65 (0.50 - 0.84)0.66 (0.51 - 0.84)Low groupReferenceReference

[0097] FIG. 11 is a block diagram of a device (100) for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor according to an embodiment of the present invention. Referring to FIG. 11, the configuration of the device (100) for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor illustrated is merely a simplified example. In one embodiment of the present invention, the device (100) for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor may include other components for performing the computing environment of the device (100), and only some of the disclosed components may constitute the device (100). The device (100) for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor may include a processor (110) including one or more cores, a memory (120), and a network (130).

[0098] The processor (110) may be configured with one or more cores, and may include a processor for data analysis and deep learning, such as a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU) of a computing device. The processor (110) may read a computer program stored in the memory (120) and perform data processing for machine learning according to an embodiment of the present disclosure. According to an embodiment of the present disclosure, the processor (110) may perform operations for learning a neural network. The processor (110) may perform calculations for learning a neural network, such as processing input data for learning in deep learning (DL), extracting features from input data, calculating errors, and updating weights of a neural network using backpropagation. At least one of the CPU, GPGPU, and TPU of the processor (110) may process learning of a network function. For example, a CPU and a GPGPU can jointly process network function learning and data classification using network functions. Furthermore, in one embodiment of the present disclosure, processors of multiple computing devices can be jointly used to process network function learning and data classification using network functions. Furthermore, a computer program executed on a computing device according to one embodiment of the present disclosure may be a CPU, GPGPU, or TPU executable program.

[0099] The processor (110) can receive whole-body transcriptome sequencing and CT images of one or more patients. The processor (110) can perform the receiving step (S100) described above.

[0100] The processor (110) can calculate a cell type score that quantifies the expression value of a specific marker gene obtained from full-length transcriptome sequencing. The processor (110) can perform the step (S200) of calculating the cell type score described above.

[0101] The processor (110) can identify a tumor-infiltrating lymphocyte cell population having a cell type score greater than or equal to a specific value in a patient population responsive to an immune checkpoint inhibitor among one or more patients, and calculate an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte within the identified cell population. The processor (110) can perform the step (S300) of calculating the aforementioned immune score.

[0102] The processor (110) can extract one or more radiological features from a CT image. The processor (110) can perform the step (S400) of extracting the radiological features described above.

[0103] The processor (110) may select a significant radiological feature based on the correlation between the immune score and one or more of the radiological features. The processor (110) may perform the step of selecting the significant radiological feature (S500) described above.

[0104] The processor (110) can train a regression model using the immune score and selected significant radiological features. The processor (110) can perform the step (S600) of training the regression model described above.

[0105] The processor (110) can input a CT image of a specific patient into the regression model to predict an immune score. The processor (110) can perform the step (S700) of predicting the immune score described above.

[0106] The memory (120) can store any form of information generated or determined by the processor (110) and any form of information received by the network (130).

[0107] The memory (120) may include at least one type of storage medium among a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (e.g., SD or XD memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, and an optical disk. The computing device (100) may also operate in relation to web storage that performs the storage function of the memory (120) on the internet. The description of the above-described memory is merely an example, and the present disclosure is not limited thereto.

[0108] The network (130) may use any known wired or wireless communication system. The network (130) may receive CT images, whole-body sequencing, and the like from related devices or systems.

[0109] The network (130) can transmit and receive information, user interfaces, etc. processed by the processor (110) through communication with other terminals. For example, the network (130) can provide a user interface generated by the processor (100) to a client (e.g., a user terminal). In addition, the network (130) can receive external input from a user authorized as a client and transmit it to the processor (110). At this time, the processor (110) can process operations such as outputting, modifying, changing, and adding information provided through the user interface based on the external input of the user received from the network (130).

[0110] Meanwhile, a device (100) for predicting the responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor according to one embodiment of the present disclosure may include a server as a computing system that transmits and receives information through communication with a client. In this case, the client may be any type of terminal capable of accessing the server.

[0111] In a further embodiment, a device (100) for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor may include any form of terminal that receives data resources generated from any server and performs additional information processing.

[0112] Another embodiment of the present invention provides a computer program for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor, which may include the steps of: receiving a signal; calculating a cell type score; generating an immune score; extracting radiomics features; selecting significant radiomics features; training a regression model; and predicting an immune score. The computer program for predicting responsiveness of a non-small cell lung cancer patient to an immune checkpoint inhibitor may be stored in a computer-readable storage medium and include instructions that cause a computer to perform the following operations.

[0113] The receiving operation may receive whole-body transcriptome sequences and CT images of one or more patients. The receiving operation refers to the operation performed in the receiving step (S100) described above.

[0114] The operation of calculating a cell type score can calculate a cell type score that quantifies the expression value of a specific marker gene obtained from full-length transcriptome sequencing. The operation of calculating a cell type score refers to the operation performed in the step (S200) of calculating a cell type score described above.

[0115] The operation of calculating an immune score may include identifying a tumor-infiltrating lymphocyte cell population having a cell type score greater than or equal to a specific value in a patient population responsive to an immune checkpoint inhibitor among one or more patients, and calculating an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte within the identified cell population. The operation of calculating an immune score refers to the operation performed in the aforementioned step of calculating an immune score (S300).

[0116] The operation of extracting radiological features may extract one or more radiological features from a CT image. The operation of extracting radiological features refers to the operation performed in the aforementioned radiological feature extraction step (S400).

[0117] The operation of selecting meaningful radiological features may select meaningful radiological features based on the correlation between the immune score and one or more of the radiological features. The operation of selecting meaningful radiological features refers to the operation performed in the step (S500) of selecting meaningful radiological features described above.

[0118] The operation of training a regression model can train a regression model using an immune score and selected meaningful radiological features. The operation of training a regression model refers to the operation performed in the aforementioned regression model training step (S600).

[0119] The operation of predicting an immune score can predict an immune score by inputting a CT image of a specific patient into the above regression model. The operation of predicting an immune score refers to the operation performed in the aforementioned immune score prediction step (S700).

[0120] Figure 12 illustrates a schematic diagram of a computing environment according to an embodiment of the present invention.

[0121] Although the present disclosure has been described above as being generally implemented by a computing device, those skilled in the art will appreciate that the present disclosure may be implemented in combination with computer-executable instructions and / or other program modules that may be executed on one or more computers and / or as a combination of hardware and software.

[0122] Generally, program modules include routines, programs, components, data structures, and the like that perform particular tasks or implement particular abstract data types. Furthermore, those skilled in the art will appreciate that the methods of the present disclosure can be implemented with other computer system configurations, including single-processor or multiprocessor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor-based or programmable consumer electronics, and the like, each of which may be operatively connected to one or more associated devices.

[0123] The described embodiments of the present disclosure can also be practiced in distributed computing environments, where certain tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

[0124] Computers typically include a variety of computer-readable media. Computer-readable media can be any media that can be accessed by a computer, and includes both volatile and nonvolatile media, transitory and non-transitory media, removable and non-removable media. By way of example, and not limitation, computer-readable media can include computer-readable storage media and computer-readable transmission media. Computer-readable storage media includes both volatile and nonvolatile media, transitory and non-transitory media, removable and non-removable media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer-readable storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital video disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be accessed by a computer and used to store the desired information.

[0125] Computer-readable transmission media typically includes any information delivery media that embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transport mechanism. The term modulated data signal means a signal that has one or more of its characteristics set or changed so as to encode information in the signal. By way of example, and not limitation, computer-readable transmission media includes wired media, such as a wired network or direct-wired connection, and wireless media, such as acoustic, RF, infrared, or other wireless media. Combinations of any of the above are also intended to be included within the scope of computer-readable transmission media.

[0126] An exemplary environment for implementing various aspects of the present disclosure is illustrated, including a computer (1000), which includes a processing unit (1020), a system memory (1030), and a system bus (1010). The system bus (1010) connects system components, including but not limited to the system memory (1030), to the processing unit (1020). The processing unit (1020) may be any of a variety of commercially available processors. Dual processors and other multiprocessor architectures may also be utilized as the processing unit (1020).

[0127] The system bus (1010) may be any of several types of bus structures that may be additionally interconnected to a memory bus, a peripheral bus, and a local bus using any of a variety of commercial bus architectures. The system memory (1030) includes read-only memory (ROM) (1034) and random access memory (RAM) (1032). A basic input / output system (BIOS) is stored in non-volatile memory (1034), such as ROM, EPROM, or EEPROM, and includes basic routines that help transfer information between components within the computer (1000), such as during start-up. The RAM (1032) may also include high-speed RAM, such as static RAM, for caching data.

[0128] The computer (1000) also includes an internal hard disk drive (HDD) (1050) (e.g., EIDE, SATA) - which may also be configured for external use within a suitable chassis (not shown), a magnetic floppy disk drive (FDD) (1060) (e.g., for reading from or writing to removable diskettes), and an optical disk drive (1070) (e.g., for reading from or writing to CD-ROM disks or other high-capacity optical media such as DVDs). The hard disk drive (1050), the magnetic disk drive (1060), and the optical disk drive (1070) may be connected to the system bus (1010) by a hard disk drive interface, a magnetic disk drive interface, and an optical drive interface, respectively. Interfaces for implementing external drives include at least one or both of Universal Serial Bus (USB) and IEEE 1394 interface technologies.

[0129] These drives and their associated computer-readable media provide non-volatile storage of data, data structures, computer-executable instructions, and the like. In the case of the computer (1000), the drives and media correspond to storing any data in a suitable digital format. While the description of computer-readable media above refers to HDDs, removable magnetic disks, and removable optical media such as CDs or DVDs, those skilled in the art will appreciate that other types of computer-readable media, such as zip drives, magnetic cassettes, flash memory cards, cartridges, and the like, may also be used in the exemplary operating environment, and that any such media may contain computer-executable instructions for performing the methods of the present disclosure.

[0130] A number of program modules, including an operating system (1092), one or more application programs (1094), other program modules (1096), and a database (1098), may be stored in the drive and RAM (1032). All or portions of the operating system, applications, modules, and / or data may also be cached in RAM (1032). It will be appreciated that the present disclosure may be implemented in various commercially available operating systems or combinations of operating systems.

[0131] A user may enter commands and information into the computer (1000) via one or more wired / wireless input devices (1042), such as a keyboard and a pointing device such as a mouse. Other input devices (not shown) may include a microphone, an IR remote control, a joystick, a game pad, a stylus pen, a touch screen, and the like. These and other input devices are often connected to the processing unit (1020) via an input / output interface (1040) that is connected to the system bus (1010), but may be connected by other interfaces such as a parallel port, an IEEE 1394 serial port, a game port, a USB port, an IR interface, and the like.

[0132] A monitor or other type of display device is also connected to the system bus (1010) via an interface such as a video adapter. In addition to the monitor, the computer typically includes other peripheral output devices (not shown) such as speakers, a printer, and so on.

[0133] The computer (1000) may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) (1082), via wired and / or wireless communications. The remote computer(s) (1082) may be a workstation, a computing device computer, a router, a personal computer, a portable computer, a microprocessor-based entertainment device, a peer device, or other conventional network node, and may generally include many or all of the components described for the computer (1000). The logical connections include wired / wireless connections to a local area network (LAN) and / or a larger network, such as a wide area network (WAN). Such LAN and WAN networking environments are common in offices and companies and facilitate enterprise-wide computer networks, such as intranets, all of which may be connected to a worldwide computer network, such as the Internet.

[0134] When used in a LAN networking environment, the computer (1000) is connected to a local network (not shown) via a wired and / or wireless communication network interface or adapter (not shown). The adapter (not shown) may facilitate wired or wireless communication to the LAN (not shown), which may also include a wireless access point installed therein for communicating with the wireless adapter (not shown). When used in a WAN networking environment, the computer (1000) may include a modem (not shown), be connected to a communication computing device on the WAN (not shown), or have other means for establishing communications over the WAN (not shown), such as via the Internet. The modem (not shown), which may be internal or external and wired or wireless, is connected to the system bus (1010) via a serial port interface (not shown). In a networked environment, program modules described for the computer (1000), or portions thereof, may be stored in a remote memory / storage device (not shown). It will be appreciated that the network connections shown are exemplary and that other means of establishing a communications link between computers may be used.

[0135] The computer (1000) operates to communicate with any wireless device or object that is arranged and operates via wireless communication, such as a printer, a scanner, a desktop and / or portable computer, a portable data assistant (PDA), a communication satellite, any equipment or location associated with a radio-detectable tag, and a telephone. This includes at least Wi-Fi and Bluetooth wireless technologies. Accordingly, the communication may be a predefined structure, as in a conventional network, or simply an ad hoc communication between at least two devices.

[0136] Wi-Fi (Wireless Fidelity) enables connections to the Internet and other devices without wires. Wi-Fi is a wireless technology that allows devices, such as computers, to send and receive data anywhere within the coverage area of ​​a base station, both indoors and outdoors, similar to cell phones. Wi-Fi networks use wireless technologies called IEEE 802.11 (a, b, g, etc.) to provide secure, reliable, and high-speed wireless connections. Wi-Fi can be used to connect computers to each other, to the Internet, and to wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in the unlicensed 2.4 and 5 GHz radio bands, at data rates of, for example, 11 Mbps (802.11a) or 54 Mbps (802.11b), or in products that include both bands (dual-band).

[0137] Those skilled in the art will appreciate that information and signals may be represented using any of a variety of different technologies and techniques. For example, the data, instructions, commands, information, signals, bits, symbols, and chips referenced in the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof.

[0138] Those skilled in the art will appreciate that the various illustrative logical blocks, modules, processors, means, circuits, and model steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, various forms of programs or design code (referred to herein, for convenience, as software), or a combination of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

[0139] The various embodiments presented herein can be implemented as a method, apparatus, or article of manufacture using standard programming and / or engineering techniques. The term article of manufacture includes a computer program, carrier, or media accessible from any computer-readable storage device. For example, computer-readable storage media include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CDs, DVDs, etc.), smart cards, and flash memory devices (e.g., EEPROMs, cards, sticks, key drives, etc.). Furthermore, various storage media presented herein include one or more devices and / or other machine-readable media for storing information.

[0140] It should be understood that the specific order or hierarchy of steps in the presented processes is merely an example of exemplary approaches. It should be understood that the specific order or hierarchy of steps in the processes may be rearranged within the scope of the present disclosure based on design priorities. The appended method claims provide elements of various steps in a sample order, but are not intended to be limited to the specific order or hierarchy presented.

[0141] The description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments disclosed herein, but is to be construed in the broadest scope consistent with the principles and novel features disclosed herein.

[0142] The embodiments of the present invention described above are not implemented solely through devices and methods. They may also be implemented through programs that implement functions corresponding to the configurations of the embodiments of the present invention, or through recording media containing such programs. Such recording media may be executed not only on servers but also on user terminals.

[0143] Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art using the basic concept of the present invention defined in the following claims also fall within the scope of the present invention.

Claims

1. A method for predicting the responsiveness of non-small cell lung cancer patients to immune checkpoint inhibitors, A step of receiving whole-body transcriptome sequencing and CT images of one or more patients; A step of calculating a cell type score by quantifying the expression value of a specific marker gene obtained from the above full-length transcriptome sequencing; A step of identifying a tumor-infiltrating lymphocyte cell population having a cell type score greater than or equal to a specific value in a patient population responsive to an immune checkpoint inhibitor among the one or more patients, and calculating an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte in the identified cell population; A step of extracting one or more radiological features from the CT image; selecting significant radiological features based on the correlation between the immune score and the one or more radiological features; and A step of training a regression model with the above immune score and selected significant radiological features; A method comprising:

2. In paragraph 1, One or more of the above patients, A method in which a patient is receiving immune checkpoint inhibitor therapy.

3. In paragraph 1, The above receiving step is, A method comprising receiving a full-length transcriptome sequence acquired within one year from the time of immune checkpoint inhibitor treatment and receiving a CT image acquired within one month from the time of immune checkpoint inhibitor treatment.

4. In paragraph 1, The above receiving step is, A method for receiving demographic characteristics and tumor characteristics of one or more patients.

5. In paragraph 4, The above receiving step is, A method for receiving at least one of sex, age, and smoking history as demographic characteristics, and at least one of pathological characteristics, tumor proportional score (PD-L1 TPS), and target mutation as tumor characteristics.

6. In paragraph 1, The step of calculating the above cell type score is: A method wherein a specific marker gene is a marker gene for measuring the expression level of a population of tumor-infiltrating lymphocyte cells.

7. In paragraph 1, The step of calculating the above cell type score is: A method for calculating a cell type score as a simple average of log-transformed expression values ​​of specific marker genes obtained from the above full-length transcriptome sequencing.

8. In paragraph 1, The step of calculating the above immune score is: A method for identifying a tumor infiltrating lymphocyte cell population having a significantly higher cell type score by performing a chi-square test.

9. In paragraph 1, The step of extracting the above radiological features is: A method for extracting at least one of a primary feature, a shape feature, a gray level co-occurrence matrix feature, a gray level magnitude domain matrix feature, a cumulative distribution function feature, a physical feature, and a fractal feature as a radiological feature.

10. In paragraph 1, The step of selecting the above significant radiological features is: A method for calculating a Pearson correlation coefficient between the above immune score and the extracted radiological features.

11. In paragraph 10, The step of selecting the above significant radiological features is: A method for selecting a significant radiological feature when the above Pearson correlation coefficient is ρ <0.05 or ρ>0.

2.

12. In paragraph 1, A method further comprising the step of inputting a CT image of a specific patient into the regression model to predict an immune score.

13. A device for predicting the responsiveness of non-small cell lung cancer patients to immune checkpoint inhibitors. a processor comprising one or more cores; and memory; Including, The above processor, Receiving whole-genome sequencing and CT images of one or more patients, Calculate a cell type score by quantifying the expression value of a specific marker gene obtained from the above full-length transcriptome sequencing, Identifying a tumor-infiltrating lymphocyte cell population having a cell type score greater than or equal to a specific value in a patient population responsive to an immune checkpoint inhibitor among one or more of the above patients, and calculating an immune score as the sum of the cell type scores for each tumor-infiltrating lymphocyte within the identified cell population, Extracting one or more radiological features from the above CT image, Selecting significant radiological features based on the association between the above immune score and one or more radiological features, and A device for training a regression model using the above immune score and selected significant radiological features.

14. A computer program stored in a computer-readable storage medium and including commands that cause a computer to perform the following operations, wherein the operations are: An operation of receiving whole-body transcriptome sequencing and CT images of one or more patients; An operation of calculating a cell type score by quantifying the expression value of a specific marker gene obtained from the above full-length transcriptome sequencing; An operation of identifying a tumor infiltrating lymphocyte cell population having a cell type score greater than or equal to a specific value in a patient population among said one or more patients responsive to an immune checkpoint inhibitor, and calculating an immune score as the sum of the cell type scores for each tumor infiltrating lymphocyte within the identified cell population; An operation of extracting one or more radiological features from the CT image; An operation of selecting a significant radiological feature based on the correlation between the above immune score and the one or more radiological features; The operation of training a regression model with the above immune score and selected significant radiological features; and A computer program stored in a computer-readable storage medium, comprising: an operation of inputting a CT image of a specific patient into the regression model to predict an immune score;