Objective To establish an effective model for distinguishing glandular prodromal lesions(PGL)mixed with ground-glass nodules(mGGN)from minimally invasive adenocarcinoma(MIA)on CT based on artificial intelligence.Methods A retrospective analysis was conducted on the clinical and CT image data of 180 patients with lung adenocarcinoma confirmed by surgical pathology and with CT manifestations of mGGN in the First Affiliated Hospital of Wenzhou Medical University from January 2017 to June 2023,inclu-ding 66 patients with PGL and 114 patients with MIA.Patients were divided into the training set(n=144)and the test set(n=36)in an 8∶2 ratio using a completely random method.The quantitative parameters and radiomics features of the lesions in CT images were automatically extracted using artificial intelligence soft-ware(United Imaging Research Platform uRP).By incorporating the most obvious correlation features of omics through dimensionality reduction,five machine learning classifiers were established,including logistic regression(LR),support vector machine(SVM),Random forest(RF),Gaussian process(GP),and Decision Tree(DT).The classifier with the training set highest area under the curve(AUC)was selected as the best radiomics model,and output the result as radiomics score(Rad-score).The clinical information,CT morpho-logical characteristics and quantitative data of the two groups were included in the multivariate logistic regres-sion analysis to screen the independent influencing factors for effectively differentiating PGL and MIA,and a clinical model was established.Finally,a comprehensive prediction model was constructed based on Rad-score and clinical risk factors.The diagnostic performance of the three models was evaluated by using the AUC,sen-sitivity,specificity and accuracy of receiver operating characteristic(ROC)curve.Results Eleven radiomics features for distinguishing PGL from MIA were obtained through LASSO dimensionality reduction.Among the five machine learning classifiers,GP has the best diagnostic performance,with AUC of 0.865 in the train-ing set and 0.762 in the test set,respectively.Univariate and multivariate logistic regression analyses were used for clinical feature screening.The clinical model was constructed by using the average CT value,average long and short diameter,and solid partial long diameter of mGGN,and the AUCs of the training set and the test set were 0.870 and 0.794,respectively.The comprehensive prediction model demonstrated superior diag-nostic performance,with AUC,sensitivity,specificity,and accuracy in the training set being 0.948,81.1%,91.2%and 87.5%respectively,while 0.883,76.9%,91.3%and 86.1%respectively in the test set.Conclu-sion The comprehensive prediction model established based on the quantitative and omics feature analysis of pulmonary nodules by artificial intelligence can well distinguish mGGN mixed with PGL from MIA on CT,and can be used to guide clinical treatment decisions.

Objective:To explore the CT features of inflammatory pseudotumor like follicular dendritic cell sarcoma (FDCS) of the spleen.Methods:The clinical data of 12 patients with splenic inflammatory pseudotumor like FDCS admitted to 3 central hospitals including Yongjia People's Hospital in Zhejiang Province from January 2015 to December 2022 were retrospectively analyzed, including 4 males and 8 females, with a median age of 60 years old. The number, shape, size and CT features of the lesions were analyzed based on patient's CT image data.Results:CT scans of 12 patients showed 15 lesions, including 10 single lesions and 2 multiple lesions. The lesions were circular in 5 cases, elliptical in 4 cases, and irregular in 3 cases. The median maximum diameter of the mass is 6.5 cm. On plain scan, all 12 tumors showed low density or slightly low density. The CT value is (41.3±7.2) HU; 8 cases had uneven density and 4 cases had uniform density. There were 8 cases with clear tumor boundaries and 4 cases with unclear boundaries. There were 8 cases with tumor necrosis and cystic transformation, and 5 cases showed patchy bleeding lesions in the center of the tumor. Enhancement: the arterial phase shows small patches or flocculent enhancement at the edges or parenchymal parts of the tumor, with CT value of (56.0±3.8) HU. Among them, there were 7 cases of mild enhancement, 4 cases of moderate enhancement, and 1 case of significant enhancement. During the portal phase, there was mild to moderate persistent small patchy uneven enhancement, with CT value of (62.0±4.3) HU. Among them, there were 8 cases of mild enhancement and 4 cases of moderate enhancement. The delayed phase showed a slow withdrawal of enhancement, with CT value of (45.0±8.2) HU. All 12 cases underwent complete resection and were diagnosed with FDCS through pathological examination.Conclusion:FDCS plain scan shows circular or elliptical uneven low-density masses, with small patches or flocculent light to moderate uneven enhancement in the arterial phase, continuous enhancement in the portal phase, and slow withdrawal in the delayed phase as the main characteristics.

Objective To investigate the CT features of early infection stage of thoracic and pulmonary paragonimiasis. Methods Medical records of 56 patients with thoracic and pulmonary paragonimiasis from January 2010 to June 2017 were collected, and the patients were diagnosed and treated at Yongjia County People's Hospital, and the results of laboratory examination and CT imaging features were analyzed retrospectively. Results The absolute value of eosinophils in peripheral blood of 56 patients was (5.61 ± 3.18) × 109/L, and the percentage of eosinophils was (35.90 ± 19.16)％, all of which increased to varying degrees. Forty-two patients had different degrees of pleural effusion and 52 cases with lung lesions. Lung lesions demonstrated one or several kinds of foci at the same time, randomly distributed in the lung field, mostly located in the sub-pleural lung tissue. There were 12 cases with pulmonary ground glass shadow, 4 cases with peribronchitis, 31 cases with pulmonary invasive lesions and 28 cases with pulmonary nodular/strip shadow. The size of most nodules were 0.5 - 1.0 cm, accompanied with halo sign. Conclusions The CT features of early infection stage of thoracic and pulmonary paragonimiasis are diverse. The size of 0.5 - 1.0 cm lung nodules with halo sign has certain characteristics in the diagnosis of paragonimiasis. Peribronchitis, infiltrative lesions, pleural effusion and increased peripheral blood eosinophil percentage can suggest diagnosis.

Objective To evaluate the reliability of CT perfusion parameter values of the normal hepatic segments and neoplasms, obtained with deconvolution (DC) and maximum slope (MS) algorithms. Methods Perfusion parameter values of 111 ROIs in 62 normal hepatic segments and 49 neoplasms derived from 62 CT perfusion studies performed with 320 multi-slice CT, were retrospectively analyzed by two experienced radiologists. BF,BV and PI according to DC and MS algorithms were compared with t paired test, Pearson correlation and Bland-Altman agreement analysis. Interobserver agreement for all perfusion parameters was calculated using intraclass correlation coefficients (ICC). Results Interobserver agreement measured with ICC was very good for all perfusion parameters (≥0.95). BFdc and BVdc exceeded the BFms and BVms in normal hepatic segments and neoplasms (P<0.05); PIdc significantly exceeded the PIms in normal hepatic segments(P<0.05) ,while no difference were found in hepatic neoplasms(P>0.05). Both pairs of perfusion measurements significantly correlated with each other(r>0.9, P<0.01),but the agreement of BF, BV and PI according to DC and MS algorithms was not good. Conclusions CT perfusion values such as BF,BV and PI obtained by DC and MS algorithms correlated significantly with each other, but with poor agreement.

Objective To evaluate the clinical value of 320-detector CT in interventional treatment of bronchial artery hemoptysis. Methods CTA and DSA images of 30 patients with bronchial artery hemoptysis were retrospectively analyzed.Spatial anatomical characters of the bronchial arteries,such as the type of branches,origin and opening positions of the bronchial arteries were observed and recorded.Results In 30 patients,6 bronchial arteries distribution patterns were found,and the most common type was R1 L1 (43.3%).83 bronchial arteries were identified using CTA,including 38 on the right and 45 on the left.The right bronchial arteries mainly originated from the intercostal artery (52.6%),while the left bronchial arteries mainly from the descending aorta and aortic arch (82.2%).The opening positions of right and left bronchial arteries were mainly located at the right wall of the descending aorta (78.9%),and anterior wall of the descending aorta (62.2%),respectively.When the cacarina of trachea was used as the reference position,the left and right bronchial arteries were mainly located in the range of above 2 cm to below 1 cm from tracheal bifurcation, accounting for 80% and 89.5%,respectively.Compared with DSA,the sensitivity and specificity of CTA were 97.5% and 100%, respectively.Conclusion 320-detector CT can be used to clearly display the distribution patterns,origin and opening positions of bronchial arteries,and especially to find bronchial arteries with ectopic origin.It is possible to apply 320-detector CT in preoperative routine examination and postoperative evaluation of massive hemoptysis.