A system for differentiating benign and malignant lung nodules by CT and imprint gene detection

Abstract

The application discloses a system for identifying lung nodule benignity and malignancy by combining CT and imprint gene detection. The system comprises a CT scanner for acquiring lung images, a CT feature quantification classification processing system, a kit for in situ hybridization detection of imprint genes GNAS, GRB10, SNRPN and HM13, and an imprint gene grading processing system; wherein the standard for judging lung nodule benignity and malignancy by combining CT classification and imprint gene expression grading is as follows: when CT imaging is grade 2 and imprint genes are IV grade, it is judged as malignant; when CT imaging is 3S and imprint genes are III grade and IV grade, it is judged as malignant; when CT imaging is 3L or above and imprint genes are II grade, III grade and IV grade, it is judged as malignant; and in other cases, it is judged as benign. The total sensitivity, total specificity and total accuracy of the diagnosis method based on the system are high, the positive diagnosis accuracy of lung nodule diagnosis based on CT imaging is improved, and the clinical value is high.

Description

Technical Field

[0001] This invention relates to a system for identifying benign and malignant pulmonary nodules by combining CT and imprint gene detection, belonging to the field of medical diagnostic technology. Background Technology

[0002] The widespread use of CT screening has improved the early detection rate of lung cancer; however, CT imaging has a high false-positive rate, and high-risk nodules require further biopsy pathology to differentiate between benign and malignant nodules, which is crucial for determining whether surgery is necessary. Preoperative biopsy diagnosis relies on pathologists' interpretation of the morphology of biopsy cells and tissues, resulting in significant variations in diagnostic accuracy, and some samples cannot be diagnosed due to insufficient morphological characteristics. Currently, various molecular diagnostic techniques can assist in the diagnosis of biopsy samples, but their accuracy remains to be improved when used alone. There are currently no reports on the combined use of molecular diagnostics and CT features to differentiate between benign and malignant lung nodules. Summary of the Invention

[0003] The technical problem to be solved by this invention is: how to combine CT and imprint gene detection to differentiate between benign and malignant pulmonary nodules, so as to improve the accuracy of preoperative diagnosis of pulmonary nodules based on CT images.

[0004] To address the aforementioned technical problems, this invention provides a system for identifying benign and malignant pulmonary nodules through a combination of CT and imprinted gene detection. The system includes a CT scanner for acquiring lung images, a CT feature quantification and classification processing system, a kit for in situ hybridization detection of imprinted genes GNAS, GRB10, SNRPN, and HM13, and an imprinted gene grading processing system.

[0005] The CT feature quantification and classification system includes an input terminal for inputting CT images and patient clinical information. Based on CT image features, it categorizes nodules into: Category 1 - No nodules found; Category 2 - Benign nodules; Category 3S - Newly developed or stable small nodules (5-9 mm); Category 3L - Newly developed or nodules ≥10 mm found during baseline screening; Category 4A - Low risk of malignancy; Category 4B - Carcinoma in situ or microinvasive carcinoma; Category 4C - High probability of malignancy; Category 5 - Nodules strongly suggestive of malignancy on screening CT; and Category 6 - Histologically confirmed malignant nodules. Categories 1 and 6 are nodules without nodules and histologically confirmed malignant nodules, respectively, and do not undergo imprint gene testing. Category 2 CT image features include nodules with benign calcification, interlobar fissure nodules, lipid-containing hamartomas, spherical atelectasis, benign nodules on central biopsy, nodules <5 mm of any density, nodules ≥5 mm of solid density, and nodules ≥5 mm of solid density with stable follow-up for more than 2 years. The following are considered criteria for CT imaging characteristics: 1) Stable solid nodules with a density of ≥5mm and a follow-up period of more than 5 years; 2) Stable solid nodules with a follow-up period of <2 years or stable subsolid nodules with a follow-up period of <5 years; 3) Blurred or ground-glass opacities, partial ground-glass opacities, multiple lesions, satellite lesions, or tree-in-bud appearance; 4) Solid nodules of 10-25mm with benign features, but unclear; 4) Persistent subsolid nodules with a density of ≥10mm, of which the solid component is ≤5mm; 4) Solid nodules with no significant improvement in short-term follow-up, of which the solid component is ≥5mm, or malignant growth of the solid portion of a solid nodule or subsolid nodule, or the presence of nodules ≥10mm with clear borders, lobulation and spiculation, without CT signs and clinical manifestations of inflammatory lesions; 5) Infiltrating the chest wall or mediastinum.

[0006] The imprinted gene grading system includes an input terminal for inputting imprinted gene detection results. The imprinted gene grading is divided into levels 0-IV. The imprinted gene expression level (levels 0-IV) is calculated according to the model in the literature Zhou, J., Cheng, T., Li, X. et al. Clin Epigenet 2021; 13, 220.

[0007] The system uses CT classification combined with imprint gene expression grading to determine the benign or malignant nature of pulmonary nodules as follows: CT imaging grade 2 with imprint gene grade IV indicates malignancy; CT imaging grade 3S with imprint gene grade III and IV indicates malignancy; CT imaging grade 3L or higher with imprint gene grade II, III, and IV indicates malignancy; other conditions are considered benign.

[0008] Preferably, the samples used for in situ hybridization detection of imprinted genes GNAS, GRB10, SNRPN and HM13 in the kit are lung brush, bronchoalveolar lavage fluid and bronchoscopic biopsy samples, and the kit detects three parameters: biallelic expression, multi-allelic expression and total expression.

[0009] Compared with the prior art, the present invention has the following beneficial effects:

[0010] The diagnostic method based on the combined CT and imprint gene detection of this invention for differentiating benign and malignant pulmonary nodules has a total sensitivity, total specificity, and total accuracy of 98.4%, 96.1%, and 97.5%, respectively, which improves the positive diagnostic accuracy of pulmonary nodules based on CT imaging. In addition, compared with the imprint gene grading diagnostic method reported in the literature, the combined diagnostic method based on this invention has higher specificity and is divided into more and finer categories according to CT classification and imprint gene grading. In some categories, it can accurately distinguish between benign and malignant nodules with 100% accuracy, which is of great clinical significance. Attached Figure Description

[0011] Figure 1 To screen for nodules that appear strongly malignant on CT scans;

[0012] Figure 2 The classification of individual imprinted genes;

[0013] Figure 3 The combined classification of imprinted genes GNAS, GRB10, SNRPN and HM13;

[0014] Figure 4 A table for determining the benign or malignant nature of pulmonary nodules by combining CT classification with imprint gene expression grading;

[0015] Figure 5 Statistical results of CT classification combined with imprint gene expression grading to determine the benign or malignant nature of pulmonary nodules in selected clinical malignant samples;

[0016] Figure 6 Statistical results of CT classification combined with imprint gene expression grading to determine the benignity or malignancy of pulmonary nodules in selected clinical benign samples;

[0017] Figure 7 The statistical results of CT classification combined with imprint gene expression grading to determine the benign or malignant nature of pulmonary nodules in the selected total clinical sample (including benign and malignant samples);

[0018] Figure 8 A comparative statistical analysis of CT classification combined with imprint gene expression grading to determine the benign or malignant nature of pulmonary nodules in clinical malignant samples and the total sample. Detailed Implementation

[0019] To make the present invention more apparent and understandable, preferred embodiments are described in detail below with reference to the accompanying drawings.

[0020] Example

[0021] A system for identifying benign and malignant pulmonary nodules by combining CT and imprinted gene detection includes a CT scanner for acquiring lung images, a CT feature quantification and classification system, a kit for in situ hybridization detection of imprinted genes GNAS, GRB10, SNRPN and HM13, and an imprinted gene grading system.

[0022] The CT feature quantification and classification system includes an input terminal for inputting CT images and patient clinical information. Based on CT image features, it categorizes nodules into: Category 1 - No nodules found; Category 2 - Benign nodules; Category 3S - Newly developed or stable small nodules (5-9 mm); Category 3L - Newly developed or nodules ≥10 mm found during baseline screening; Category 4A - Low risk of malignancy; Category 4B - Carcinoma in situ or microinvasive carcinoma; Category 4C - High probability of malignancy; Category 5 - Nodules strongly suggestive of malignancy on screening CT; and Category 6 - Histologically confirmed malignant nodules. Categories 1 and 6 are nodules without nodules and histologically confirmed malignant nodules, respectively, and do not undergo imprint gene testing. Category 2 CT image features include nodules with benign calcification, interlobar fissure nodules, lipid-containing hamartomas, spherical atelectasis, benign nodules on central biopsy, nodules <5 mm of any density, nodules ≥5 mm of solid density, and nodules ≥5 mm of solid density with stable follow-up for more than 2 years. The following are considered criteria for CT imaging characteristics: 1) Stable solid nodules with a density of ≥5mm and a follow-up period of more than 5 years; 2) Stable solid nodules with a follow-up period of <2 years or stable subsolid nodules with a follow-up period of <5 years; 3) Blurred or ground-glass opacities, partial ground-glass opacities, multiple lesions, satellite lesions, or tree-in-bud appearance; 4) Solid nodules of 10-25mm with benign features, but unclear; 4) Persistent subsolid nodules with a density of ≥10mm, of which the solid component is ≤5mm; 4) Solid nodules with no significant improvement in short-term follow-up, of which the solid component is ≥5mm, or malignant growth of the solid portion of a solid nodule or subsolid nodule, or the presence of nodules ≥10mm with clear borders, lobulation and spiculation, without CT signs and clinical manifestations of inflammatory lesions; 5) Infiltrating the chest wall or mediastinum.

[0023] The imprinted gene grading system includes an input terminal for inputting imprinted gene detection results. The imprinted gene grading is divided into levels 0-IV. The imprinted gene expression level (levels 0-IV) is calculated according to the model in the literature Zhou, J., Cheng, T., Li, X. et al. Clin Epigenet 2021; 13, 220.

[0024] The criteria for determining the benign or malignant nature of pulmonary nodules based on CT classification combined with imprint gene expression grading are as follows: CT imaging grade 2 with imprint gene grade IV indicates malignancy; CT imaging grade 3S with imprint gene grade III and IV indicates malignancy; CT imaging grade 3L or higher with imprint gene grade II, III, and IV indicates malignancy; and other conditions indicate benignity.

[0025] The specific classification, grading, and judgment methods are as follows:

[0026] 1. CT Feature Quantification:

[0027] (1) Based on CT images, lung nodules are classified according to their size, ground-glass opacity, solidity, spiculated margins, and vascular growth, as shown in Table 1:

[0028] Table 1 CT Classification

[0029]

[0030]

[0031] 2. Imprinted gene in situ hybridization detection:

[0032] (1) Place samples such as lung brush, bronchoalveolar lavage fluid, and bronchoscopic biopsy into 10% formalin and fix at room temperature for 24 hours.

[0033] (2) Remove the cells and place them on a positively charged glass slide, then dry them.

[0034] (3) Using probes designed for RNA sequences of GNAS, SNRPN, and HM13 genes, in situ hybridization of SNRPN and HM13 genes was performed according to the operation method of the RNAscope kit.

[0035] (4) After staining, count 2000 cell nuclei under a microscope. Classify the cell nuclei according to the number of signal points in the nuclei, and calculate three parameters: biallelic expression (BAE), multiallelic expression (MAE), and total expression (TE). The formulas are as follows:

[0036] BAE = (Number of cell nuclei at two signal points) / (Number of cell nuclei at one signal point + Number of cell nuclei at two signal points + Number of cell nuclei at three or more signal points);

[0037] MAE = (Number of cell nuclei at three or more signal points) / (Number of cell nuclei at one signal point + Number of cell nuclei at two signal points + Number of cell nuclei at three or more signal points);

[0038] TE = (Number of nuclei at one signal point + Number of nuclei at two signal points + Number of nuclei at three or more signal points) / Total number of nuclei;

[0039] (5) Based on the model in the literature Zhou, J., Cheng, T., Li, X. et al. Clin Epigenet 2021; 13, 220, the imprinted gene expression levels (0-IV) were calculated, such as... Figure 2 , Figure 3 As shown.

[0040] 3. Determining the benign or malignant nature of pulmonary nodules by combining CT classification with imprint gene expression grading.

[0041] The criteria for determining the benign or malignant nature of pulmonary nodules based on CT classification combined with imprint gene expression grading are as follows: When the CT imaging is grade 2, imprint gene grade IV indicates malignancy; when the CT imaging is grade 3S, imprint gene grades III and IV indicate malignancy; when the CT imaging is grade 3L or higher, imprint gene grades II, III, and IV indicate malignancy. A malignancy risk table is then created based on these criteria. Figure 4 As shown, the risk of malignancy can be determined by referring to the table. A risk of less than 50% is considered benign, while a risk of 50% or more is considered malignant.

[0042] 4. Statistical analysis of clinical samples

[0043] Statistical analysis was performed on clinical samples selected from hospitals such as Zhongshan Hospital (with a small number of samples from other hospitals). The statistical results of malignant samples are as follows: Figure 5 As shown, the statistical results of the benign sample are as follows: Figure 6 As shown, the total sample statistics are as follows: Figure 7 As shown.

[0044] according to Figures 5-7 The statistical results showed that the overall sensitivity, overall specificity, and overall accuracy of the system for identifying benign and malignant pulmonary nodules using the combined CT and imprint gene detection of this invention were as follows:

[0045] Total sensitivity = 242 / 246 = 98.4%;

[0046] Overall specificity = 149 / 155 = 96.1%;

[0047] Overall accuracy = 391 / 401 = 97.5%.

[0048] Compared to the imprinted gene expression grading method reported by Zhou, J., Cheng, T., Li, X. et al. Clin Epigenet 2021; 13, 220, the overall sensitivity is basically the same as the 99.1% reported in the article (due to some error in clinical samples), while the overall specificity is significantly improved compared to the 92.1% reported in the article, and the overall accuracy is not much different from the 97.1% reported in the article. However, the imprinted gene grading method allows for more detailed classifications based on CT classification, and in some categories, it can accurately distinguish between benign and malignant tumors with 100% accuracy, which has greater clinical guidance significance.

Claims

1. A system for identifying benign and malignant pulmonary nodules by combining CT scans and imprint gene detection, characterized in that, This includes a CT scanner for acquiring lung images, a CT feature quantification and classification system, kits for in situ hybridization detection of imprinted genes GNAS, GRB10, SNRPN and HM13, and an imprinted gene grading system. The CT feature quantification and classification system includes an input terminal for inputting CT images and patient clinical information. Based on CT image features, it categorizes nodules into: Category 1 - No nodules found; Category 2 - Benign nodules; Category 3S - Newly developed or stable small nodules (5-9 mm); Category 3L - Newly developed or nodules ≥10 mm found during baseline screening; Category 4A - Low risk of malignancy; Category 4B - Carcinoma in situ or microinvasive carcinoma; Category 4C - High probability of malignancy; Category 5 - Nodules strongly suggestive of malignancy on screening CT; and Category 6 - Histologically confirmed malignant nodules. Categories 1 and 6 are nodules without nodules and histologically confirmed malignant nodules, respectively, and do not undergo imprint gene testing. Category 2 CT image features include nodules with benign calcification, interlobar fissure nodules, lipid-containing hamartomas, spherical atelectasis, benign nodules on central biopsy, nodules <5 mm of any density, nodules ≥5 mm of solid density, and nodules ≥5 mm of solid density with stable follow-up for more than 2 years. The following are considered criteria for CT imaging characteristics: 1) Stable solid nodules with a density of ≥5mm and a follow-up period of more than 5 years; 2) Stable solid nodules with a follow-up period of <2 years or stable subsolid nodules with a follow-up period of <5 years; 3) Blurred or ground-glass opacities, partial ground-glass opacities, multiple lesions, satellite lesions, or tree-in-bud appearance; 4) Solid nodules of 10-25mm with benign features, but unclear; 4) Persistent subsolid nodules with a density of ≥10mm, of which the solid component is ≤5mm; 4) Solid nodules with no significant improvement in short-term follow-up, of which the solid component is ≥5mm, or malignant growth of the solid portion of a solid nodule or subsolid nodule, or the presence of nodules ≥10mm with clear borders, lobulation and spiculation, without CT signs and clinical manifestations of inflammatory lesions; 5) Infiltrating the chest wall or mediastinum. The imprinted gene grading system includes an input terminal for inputting imprinted gene detection results, and the imprinted gene grading is divided into levels 0-IV. The system uses CT classification combined with imprint gene expression grading to determine the benign or malignant nature of pulmonary nodules as follows: CT imaging grade 2 with imprint gene grade IV indicates malignancy; CT imaging grade 3S with imprint gene grade III and IV indicates malignancy; CT imaging grade 3L or higher with imprint gene grade II, III, and IV indicates malignancy; other conditions are considered benign.

2. The system for identifying benign and malignant pulmonary nodules by combining CT and imprint gene detection as described in claim 1, characterized in that, The kit is used to detect the imprinted genes GNAS, GRB10, SNRPN and HM13 in situ hybridization. The samples used are lung brush, bronchoalveolar lavage fluid and bronchoscopic biopsy samples. The kit detects three parameters: biallelic expression, multi-allelic expression and total expression.

Images