Objective To study the clinical characteristics and analyze the treatment of primary ocular adnexal lymphoma (POAL). Methods A retrospective review was performed based on the clinical records of 26 POAL cases, who were treated from June 1999 to June 2009. The clinical manifestations,imaging features, diagnosis approaches and treatment of them were analyzed. Results The main symptoms were a slowly growing painless mass, exophthalmia and red eyelid. Mostly, imaging finding shows casting like. 50 % of POAL was at the orbit. 22 patients w ere B-cell origin [13 patients were extranodal marginal zone B-cell lymphoma-mucosal-associated lymphoid tissue (MALT) lymphoma], 2 patients were NK/T cell lymphoma. 76.9 % patients were in stage Ⅰ -Ⅱ. Surgery, radiotherapy and chemotherapy were most common treatment modalities for POAL. Conclusion Most POAL were low-grade B-cell lymphoma, with MALT by far the most common type. POAL was easy to be misdiagnosed due to non specific clinical manifestations.Imaging finding was helpful to diagnosis, but could not reliably distinguish between benign and malignant. The invasive biopsy is the main diagnosis approach. Radiotherapy had been the standard treatment for low-grade POAL. High-grade or infiltrating peri-orbit need combine chemotherapy. lmmunotherapy and molecular targeted therapy were the direction of future research.

Objective To evaluate the efficacy and safety of XELOX regimen combined with hepatic artery chemoembolization in the treatment of gastric cancer with liver metastasis.Methods 50 cases of gastric cancer with liver metastasis were randomly divided into two groups,the experimental group (25 cases) received chemotherapy regimen of XELOX first:Xeloda tablets 1 000 mg/m2,orally,twice a day,days1-14;Oxaliplatin 130 mg/m2,intravenous drip,day 1.Hepatic artery chemoembolization was performed one a week later,and a cycle consists of 4 weeks.The control group (25 cases) received chemotherapy regimen of XELOX,3 weeks as a cycle.All patients were evaluated for efficacy and toxicity every 2 cycles.Results In the experimental group,the overall response rate was 56％,the tumor control rate was 80％,the increase rate of Karnofsky was 60％,and 10 patients got chance of tumor resection.In the control group,the overall response rate was 32％,the tumor control rate was 52％,the increase rate of Karnofsky was 48％,and 6 patients got chance of operation.The overall response rate,tumor control rate,surgical treatment rate and the increase rate of Karnofsky in the experimental group were significantly different from those in the control group (P ＜0.05).The median total survival time was 12.5 months in the experimental group and 10.5 months in the control group (P ＜0.05).There was no significant difference in toxicity and side effects between the two groups.Conclusion XELOX regimen combined with hepatic artery chemoembolization is effective and safe in the treatment of gastric cancer with liver metastasis.

Purpose To use CT-arterial phase images of pre-treatment ovarian cancer patients,combined with deep learning algorithms and machine learning to build a model to predict the efficacy of neoadjuvant chemotherapy in ovarian cancer.Materials and Methods A total of 302 consecutive patients who underwent surgery and were pathologically diagnosed with ovarian cancer from March 2013 to August 2019 in Tianjin Medical University were retrospectively collected.All patients were partitioned into training and test sets according to the ratio of 7∶3.In the python environment,VGG13 model was integrated via combining deep learning network and machine learning,and features were filtered via least absolute shrinkage and selection operator algorithm to build a prediction model for classification and prediction of CT images.The area under the curve(AUC),accuracy,sensitivity,specificity,and Fl-Score were calculated,respectively.Results The AUC,accuracy,sensitivity,specificity,and Fl-Score of the model in the training set were 0.87,0.81,0.80,0.82 and 0.79,and 0.90,0.84,0.93,0.77 and 0.83 in the test set,respectively.The AUC of five-fold cross-validation were 0.86,0.88,0.88,0.90 and 0.87,respectively.Conclusion Predictive model based on CT images combined with deep learning and machine learning methods can provide a new clinical perspective for developing chemotherapy regimens for ovarian cancer.

Objective: To investigate the value of radiomics in image quality control with low-dose CT examination of solid pulmonary nodules. Methods: Images were acquired on CT750 HD scanner, and chest pulmonary nodules phantom were scanned at different tube voltage and tube current. The radiation dose CTDI_vol under different scanning conditions were recorded, as well as CNR and SNR of each scanning sequence. The variation of radiation dose, noise, tube voltage and tube current were analyzed. All data were analyzed by radiomics analysis software. R language statistics software was adopted to analyze the extracted features by principal component analysis (PCA), and the characteristic parameters with the largest contribution rate to image quality were selected for analysis. One-way ANOVA was used to analyze all the important characteristic parameters to reveal the difference of characteristic parameters under different tube voltages. Finally, the post-test method was used to find out the differences among different tube voltage groups. Results: Radiation dose rised linearly with the increase of tube current and tube voltage. Although the overall change trend of SNR and CNR in pulmonary nodules was linearly related to the change of tube voltage and tube current, there was no clear change trend threshold at low dose, which could not accurately evaluate the image quality under low radiation. Both CNR and SNR cannot evaluate the image quality effectively, and have no practiced value for optimizing the low dose scanning parameters. There main components including Uniformity, Voxel Value Sum, and Haralick Correlation extracted by radiomics analysis software were proved to play a critical role in image quality control. The cumulative contribution rate of variance was 89.20% and the eigen values were greater than 1. Uniformity curve of characteristic parameter showed that the trend of change was correlated with the change of tube voltage and tube current, and the stability and consistency were good. Uniformity one-way ANOVA analysis showed that when the tube voltage reduced from 140 to 120 kVp, there was no difference (P=0.117) in the uniformity, while from 120 to 80 kVp, significant differences revealed (P<0.001). Considering tube current, no significant variation was observed in uniformity when current was greater than 90 mA, which indicated that tube current of 90 mA could lead to better image quality. Conclusion Radiomics analysis can effectively evaluate and control the CT image quality of low dose solid pulmonary nodules.