Objective To investigate the value of the volume percentage of solid component in differential diagnosis of stage Ⅰ lung adenocarcinoma subtypes,and to predict the optimal critical CT value of solid component.Methods A total of 962 nodules with sur-gical pathological findings confirmed as adenocarcinoma in situ(AIS)or pulmonary adenocarcinoma and manifested as subsolid nod-ules on thin-section CT were analyzed retrospectively,and divided the lesions into:(1)AIS(n=350)or pulmonary adenocarcinoma(n=612)groups;and(2)minimally invasive adenocarcinoma(MIA)(n=213)or invasive adenocarcinoma(IAC)(n=399)groups based on pathological findings.The volume percentage of solid components within pulmonary nodules was measured via CT density histogram analysis at different thresholds,the diagnostic efficacy of different thresholds was analyzed and the optimal critical CT val-ues of solid component was found.Results In the AIS versus pulmonary adenocarcinoma groups,age,mass,and volume percentage of solid component were statistically significant(P＜0.001);the CT threshold of-350 HU had the highest diagnostic efficacy[area under the curve(AUC)0.859,accuracy 81.69％,sensitivity 85.60％,and specificity 74.86％,respetively],and the critical value of volume percentage of solid component determined under this threshold was 8.3％.In the MIA versus IAC groups,age(P=0.017),mass(P＜0.001),and volume percentage of solid component(P＜0.001)were statistically significant;the CT threshold of-350 HU had the highest diagnostic efficacy(AUC 0.857,accuracy 78.27％,sensitivity 82.91％,and specificity 69.48％,respectively),and the critical value of volume percentage of solid component determined under this threshold was 16.9％.Conclusion The volume percentage of solid component measured based on CT density histogram analysis may be helpful in the differential diagnosis of early stage lung ade-nocarcinoma subtypes.

Objective To study the application value of multi-parameter imaging of Revolution energy spectrum CT in identifying the tumor-lung interface between central lung cancer and accompanying atelectasis.Methods The spectral CT imaging data of 73 patients with central lung cancer and accompanying atelectasis confirmed by pathology were collected.The polychromatic image,the best monochromatic image,the best monochromatic image combined with iodine concentration map and the effective atomic number(Eff-Z)of the tumor-lung interface in the arterial,venous and delayed phases were compared,and the CT value,iodine concentration(IC)value,Eff-Z and the slope of energy spectrum curve(λHU)of tumors and atelectasis in the venous phase were also compared.Results(1)On tri-phase CT,the subjective score of the tumor-lung interface in the venous phase was the highest.(2)There were differences among the subjective scores of tumor-lung interface on polychromatic image,best monochromatic image,best monochromatic image com-bined with iodine concentration map and Eff-Z on tri-phasic enhanced CT.The subjective score of the best monochromatic image combined with iodine concentration map was the highest,followed by the best monochromatic image.(3)There were significant differences in CT values,IC values,Eff-Z and λHU between tumors and atelectasis in venous phase,with the highest difference of IC value.Conclusion Revolution energy spectrum CT can identify the tumor-lung interface between central lung cancer and accompanying ate-lectasis through multi-parameter imaging,which may provide helpful information for the diagnosis and therapeutic evaluation of the disease.

Objective To summarize and analyze the clinical and imaging features of pulmonary nodular lymphoid hyperplasia(PNLH)and its pathological features.Methods The clinical and imaging features of PNLH in 9 patients were analyzed retrospec-tively.Results Of the 9 patients with PNLH,5 were found incidentally while other 4 presented with chest symptoms.Laboratory tests showed no specificity,and the lesions were located under the pleura with adjacent pleural indentation.Most lesions were solid with spiculation sign or spiculate protuberance.Hypodense area and vocule were visible within some lesions.Enhanced scan showed the lesion with marked enhancement,however most did not had lymph node enlargement and pleural effusion.Conclusion The clin-ical manifestations of PNLH are non-specific,and the imaging characteristics overlap with those of malignant lung tumors.Thus the diagnosis of PNLH relies on the pathological examination.

Objective To investigate the correlation between plaque characteristics and perivascular fat attenuation index(FAI)in patients with coronary heart disease by coronary computed tomography angiography(CCTA)based on artificial intelligence.Methods The clinical and CCTA imaging data of 154 patients suspected to be coronary heart disease were retrospectively collected.The rela-tionship between FAI and multiple factors affecting plaque was analyzed to identify the independent risk factors affecting FAI measure-ments.Results The FAI values in calcified plaque(CP),non-calcified plaque(NCP)and mixed plaque of left anterior descending artery(LAD)were statistically significant(P＜0.05).There were significantly different of FAI values among NCP and mixed plaque in different coronary arteries(P＜0.05),while the FAI values of CP in different coronary arteries had no statistical significance(P＞0.05).The FAI values proximal coronary artery plaques were independently correlated with low density plaque and plaque location.Conclusion The location of the coronary artery plaque and the low density plaque are FAI independent factors.The analysis based on plaque around FAI helps to identify low density plaque of high-risk coronary artery plaque,which may be monitor plaque inflammation more accu-rate around the change.

Objective To investigate the correlation between non-alcoholic fatty liver disease(NAFLD)and coronary artery plaque and perivascular adipose tissue(PVAT)parameters.Methods A total of 119 patients who underwent liver ultrasound and coronary computed tomography angiography(CCTA)examination were selected and divided into NAFLD group(50 cases)and non-NAFLD group(69 cases).The types of coronary artery plaque,plaque vulnerability,plaque involvement length,plaque load,vascular remode-ling,degree of vessel stenosis,and parameters of pericoronary fat[pericoronary fat attenuation index(pFAI),pericoronary fat total volume of voxels(pFV),perivascular water attenuation index(PVWI)]were evaluated.The data were analyzed statistically.Results The incidence of vulnerable plaque in NAFLD group was higher than that in non-NAFLD group,and the difference was statistically significant(P＜0.001).The incidence of positive vascular remodeling in NAFLD group was higher than that in non-NAFLD group,the incidence of no remodeling was the opposite,and the differences were statistically significant(P=0.001).The incidence of non-calcified plaque in NAFLD group was higher than that in non-NAFLD group,and the incidence of calcified plaque was the opposite(P=0.011).The incidence of severe stenosis in NAFLD group was higher than that in non-NAFLD group,and the incidence of moderate stenosis was the opposite,with statistical significance(P=0.004).The pFAI and plaque load in NAFLD group was higher than that in non-NAFLD group,and the pFV value was lower than that in non-NAFLD group,with statistical significance(P＜0.001,P=0.029,P=0.002).Conclusion The incidences of vulnerable plaque,positive remodeling,non-calcified plaque,vascular severe stenosis,and plaque load are significantly higher in NAFLD group than those in non-NAFLD group.In addition,pFAI increases and pFV decreases in NAFLD group.

Objective To analyze the correlation between digital mammography features,lymphocyte to monocyte ratio(LMR)and prognosis in patient with invasive breast cancer,and to explore the prognostic value of digital mammography features and LMR in invasive breast cancer.Methods Ninety-six patients with invasive breast cancer were selected.The data of digital mammography features and serum LMR levels were analyzed.The patients were followed up after discharge.According to different prognosis,all patients were divided into good prognosis group(68 cases)and poor prognosis group(28 cases).Spearman was used to analyze the correlation of digital mammography features,LMR level and prognosis of patients.Receiver operating characteristic(ROC)curve was drawn to analyze the prognostic value of digital mammography features and LMR level.Results The spicule sign and proportion of calcifica-tion in the poor prognosis group was significantly higher than that in the good prognosis group,the maximum diameter of tumor was significantly larger than that in the good prognosis group,and the serum LMR level was significantly lower than that in the good prognosis group(P＜0.05).Spearman correlation analysis showed that spicule sign,proportion of calcification and maximum diame-ter of tumor were positively correlated with poor prognosis of invasive breast cancer(P＜0.05).Serum LMR level was negatively correlated with poor prognosis of invasive breast cancer(P＜0.05).ROC curve analysis showed that the area under the curve(AUC)of spicule sign,calcification and tumor maximum diameter combined with LMR level to evaluate the prognosis of invasive breast cancer was 0.927,the sensitivity was 89.71％,and the specificity was 82.14％,which were significantly higher than those of each index alone.Conclusion Digital mammography and serum LMR levels may be closely related to the prognosis of patients with invasive breast cancer,and spicule sign,calcification,tumor maximum diameter combined with serum LMR may be of significant prognostic value in invasive breast cancer.

Objective To explore the diagnostic value of MRI quantitative enhancement parameters in breast cancer and their rela-tionships with human epidermal growth factor receptor-2(HER-2)and angiopoietin-2(Ang-2).Methods A total of 80 patients with breast cancer were selected as the study objects.The MRI quantitative enhancement parameters[rate constant(Kep),extravascular extracellular space volume fraction(Ve),volume transfer constant(Ktrans)]of all patients with different clinical staging and lymph node metastasis were compared,respectively.The receiver operating characteristic(ROC)curve was used to analyze the diagnostic value of MRI quantitative enhancement parameters on the expression of HER-2 and Ang-2.Results Kep and Ktrans in stage Ⅱ group were significantly lower than those in stage Ⅲ and Ⅳ groups(F=6.322,F=6.676;P＜0.05).Kep and Ktrans in lymph node metas-tasis group were significantly higher than those in lymph node non-metastasis group(t=4.265,t=4.557;P＜0.05).Kep and Ktrans in HER-2 positive group were significantly higher than those in HER-2 negative group(t=3.988,t=4.616;P＜0.05).ROC curve analysis showed that the area under the curve(AUC)of the combination of Kep and Ktrans for the diagnosis of HER-2 expression in breast cancer patients was 0.887,with specificity of 0.811 and sensitivity of 0.889,respectively.Kep and Ktrans in Ang-2 positive group were signifi-cantly higher than those in Ang-2 negative group(t=2.534,t=3.189;P＜0.05).ROC curve analysis showed that the AUC of the com-bination of Kep and Ktrans for the diagnosis of Ang-2 expression in breast cancer patients was 0.767,with specificity of 0.867 and sensitivity of 0.569,respectively.Conclusion MRI quantitative enhancement parameters can improve the diagnostic value of breast cancer and the expression of HER-2 and Ang-2,which are beneficial to guide the selections of clinical therapeutic regimens.

Objective To investigate the application of low-dose computed tomography virtual colonoscopy(LDCTVC)in colorec-tal tumor.Methods Forty-seven colorectal tumor were given low-dose CT abdominal scan(low-dose group),15 patients with normal body mass index(BMI)who received routine-dose CT abdominal scan at the same period(routine-dose group).Volume CT dose index(CTDIvol),dose length product(DLP),virtual colonoscopy and optical colonoscopy results were recorded.Results The effective dose with normal BMI was(2.86±0.47)mSv and(4.87±1.15)mSv in the low-dose and routine-dose groups,respectively.The CTDIvol and DLP between the two groups were statistically significant(P＜0.05).There were 14 cases of true positive,4 cases of false positive,5 cases of false negative and 24 cases of true negative in the low-dose group.The sensitivity,specificity and Kappa value of LDCTVC in the diagnosis of colorectal mucosal lesions were 73.7％,85.7％,and 0.6.Conclusion LDCTVC can reduce the effective dose by 50％and has a good diagnostic value for colorectal mucosal lesions,which can make up for the deficiency of colonoscopy and make accurate judgment of extra-mucosal lesions of the bowel wall.

Objective To evaluate the CT features and differential value of gastric leiomyomas(GLs)and gastric stromal tumors(GSTs)after propensity score matching.Methods Twenty-six GLs were 1∶1 propensity score matched to GSTs based on sex,age,tumor site and size.Tumor shape and contour,mucosal ulcer,tumor growth patterns,enhancement pattern and degree,the lon-gest diameter(LD)of the tumor,and the ratio of the LD to the vertical diameter(VD)were analyzed.CT signs included hemor-rhage,calcification,peripheral invasion,and distant metastasis,etc.Regression analysis was used to determine the best quantitative evaluation for differentiation of them.Results The presence of mucosal ulcer was significantly more frequent in GSTs than in GLs(P=0.032).Both tumors showed progressive enhancement;however,the enhancement degree of GSTs was significantly higher than GLs in the arterial and portal venous phases(P=0.004,P=0.002,respectively).The above influential factors were included in a regression analysis using enhancement degree of 18 HU and 23 HU in the arterial and portal venous phases as cutoff values respectively.An enhancement degree≤18 HU in the arterial phase was identified as an independent influential factor in the diagnosis of GLs[odds ratio(OR)=12.776,95％confidence interval(CI)1.270-128.535].No significant difference was found in other morphological characteristics(P＞0.05).Conclusion Less ulceration on the tumor surface and mild enhancement in arterial phase are characteristic features of GLs compared with GSTs.

Objective To explore the feasibility of radiomics features combined with different machine learning methods based on CT scans to predict lymphovascular and perineural invasion in patient with gastric cancer.Methods A total of 142 patients with gas-tric cancer lymphovascular confirmed by operative pathological examination were retrospectively selected.Among all patients,there were 96 positive cases and 46 negative cases.Among 137 patients with perineural invasion,there were 76 positive cases and 61 nega-tive cases.The 3D-Slicer package was used for delineation,and the Pyradiomics package was used to extract radiomics features.All data were randomly divided into training set and test set in an 8∶2 ratio.Intraclass correlation coefficient(ICC),Pearson correla-tion analysis,least absolute shrinkage and selection operator(LASSO)algorithm were used for feature selection.Support vector machine(SVM),K-nearest neighbor(KNN),decision tree(DT),random forest(RF),extreme tree(ET),extreme gradient boosting(XGBoost),and LightGBM were used to compare the models of lymphovascular and perineural invasion,respectively.Receiver operating characteris-tic(ROC)curve and area under the curve(AUC)were used to evaluate the predictive performance of these models.Results The lymphovascular group AUC of SVM,KNN,DT,RF,ET,XGBoost,and LightGBM in the training set were 0.926,0.753,1.000,0.999,1.000,1.000,and 0.917,and the AUC in the test set were 0.894,0.692,0.456,0.678,0.753,0.650,and 0.650,respectively.The perineural invasion group AUC of SVM,KNN,DT,RF,ET,XGBoost,and LightGBM in the training set were 0.864,0.794,1.000,1.000,1.000,1.000,and 0.866,and the AUC in the test set were 0.861,0.706,0.700,0.672,0.731,0.667,and 0.678,respectively.Conclusion Based on venous phase CT radiomics features combined with machine learning methods,it is feasible to predict lymphovascu-lar and perineural invasion of gastric cancer preoperatively.Among the variousmachine learning methods,SVM shows the best predictive performance for lymphovascular and perineural invasion in patient with gastric cancer.