Objective To investigate a diabetic retinopathy ( DR ) detection algorithm based on transfer learning in small sample dataset. Methods Total of 4465 fundus color photographs taken by Gaoyao People ' s Hospital was used as the full dataset. The model training strategies using fixed pre-trained parameters and fine-tuning pre-trained parameters were used as the transfer learning group to compare with the non-transfer learning strategy that randomly initializes parameters. These three training strategies were applied to the training of three deep learning networks:ResNet50,Inception V3 and NASNet. In addition,a small dataset randomly extracted from the full dataset was used to study the impact of the reduction of training data on different strategies. The accuracy and training time of the diagnostic model were used to analyze the performance of different training strategies. Results The best results in different network architectures were chosen. The accuracy of the model obtained by fine-tuning pre-training parameters strategy was 90. 9％,which was higher than the strategy of fixed pre-training parameters (88. 1％) and the strategy of randomly initializing parameters ( 88. 4％) . The training time for fixed pre-training parameters was 10 minutes,less than the strategy of fine-tuning pre-training parameters ( 16 hours ) and the strategy of randomly initializing parameters (24 hours). After the training data was reduced,the accuracy of the model obtained by the strategy of randomly initializing parameters decreased by 8. 6％ on average,while the accuracy of the transfer learning group decreased by 2. 5％ on average. Conclusions The proposed automated and novel DR detection algorithm based on fine-tune and NASNet structure maintains high accuracy in small sample dataset,is found to be robust,and effective for the preliminary diagnosis of DR.

Objective To study the efficiency and accuracy of artificial intelligence (AI) system based on fundus photograph in diabetic retinopathy(DR)screening,and evaluate the clinical application value of AI system. Methods A diagnostic trial was adopted. Total of 13683 color fundus photos were collected in Zhaoqing Gaoyao People's Hospital from March,2017 to November,2018. The AI system for DR (ZOC-DR-V1) was established,based on transfer learning + NASNet algorithm,by training 4465 precisely labeled fundus images (2510 normal,and 1955 with any stage of DR). One thousand confirmed fundus images (300 normal and 700 with any stage of DR),diagnosed by AI ( AI group ) and doctors ( 3 ophthalmologist doctors and 3 endocrinologist doctors ) ( doctor group ) , respectively. Ophthalmologist group and endocrinologist group were both composed of primary,intermediate and senior physicians. The mean reading time of each image and the total time of 1000 images were recorded. The accuracy and efficiency of AI system and doctor groups were compared. The reading process was divided into two stages. The diagnostic coincidence rate and the average reading time of each group between the two parts were calculated and compared. This study protocol was approved by Ethic Committee of Zhongshan Ophthalmic Center, Sun Yat-sen University (No. 2017KYPJ104). Results After training,the diagnostic coincidence rate of AI system (ZOC-DR-V1) in test set was 94. 7％,AUC was 0. 994. In this "man-machine to war",the diagnostic coincidence rate of primary,intermediate and senior endocrinologist was 94. 0％,91. 4％ and 93. 4％;the diagnostic coincidence rate of primary,intermediate and senior ophthalmologist was 92. 7％,94. 4％ and 95. 6％;the diagnostic coincidence rate of AI system was 95. 2％. There was no difference in the diagnostic coincidence rate between AI system and senior ophthalmologist ( P = 0. 749 ) . The mean reading time of each image of primary, intermediate and senior endocrinologists was (4. 63±1. 87),(3. 74±3. 47) and (5. 71±3. 47) seconds,and the total time of 1000 images of primary,intermediate and senior endocrinologists was 1. 29,1. 04 and 1. 58 hours;the mean reading time of each image of primary,intermediate and senior ophthalmologists was ( 7. 25 ± 6. 58 ) , ( 5. 18 ± 5. 01 ) and ( 5. 18 ± 3. 47 ) seconds,and the total time of 1000 images of primary,intermediate and senior endocrinologists was 2. 02,1. 44 and 1. 44 hours;the mean and total time of AI system was (1. 62±0. 67) seconds and 0. 45 hours,and the reading time of AI system was significantly shorter than that of the doctor groups (all at P=0. 000). The diagnostic coincidence rates between previous and posterior part of primary endocrinologist, primary and intermediate ophthalmologist were significantly different (χ2=11. 986,6. 517,10. 896;all at P<0. 05),and the mean reading time in the posterior part was significantly shorter than that in the previous part of intermediate and senior endocrinologist and primary ophthalmologist (t=4. 175,8. 189,5. 160;all at P<0. 01). While the reading time of AI system remained stable throughout the process(χ2=3. 151,P=0. 103;t=0. 038,P=0. 970). Conclusions The ophthalmic AI system based on fundus images has a good diagnostic efficiency,and its diagnostic coincidence rate can compare with senior ophthalmologist,providing a new method and platform for large-scale DR screening.

Objective To investigate and analyze the reasons for the omission of adjuvant radiotherapy after breast-conserving surgery (BCS) in patients with breast cancer.Methods The clinicopathologial characteristics and socioeconomic data of 55 breast cancer patients undergoing BCS without postoperative adjuvant radiotherapy in our hospital from 2012 to 2016 were retrospectively analyzed.Results Among the 55 patients who did not receive radiotherapy,25 patients were due to low local recurrence risk,12 patients were due to economic or family reasons,12 patients were due to fear of adverse reactions of radiotherapy,and 5 patients were not recommended by primary physicians for radiotherapy.In addition,3 cases with multiple distant metastases and 3 cases with concomitant thyroid cancer didn't received radiotherapy.Conclnsions Low risk local recurrence is the main reason for the omission of adjuvant radiotherapy,followed by the fear of radiation-induced toxicity and poor financial support.Patient education and medical insurance may improve the adjuvant radiotherapy compliance.

Objective: To improve the understanding of 17α-hydroxylase/17, 20-lyase deficiency(17OHD)disease. Methods: The clinical data of six patients suffering from 17OHD were analyzed retrospectively. Results: Two patients with complete combined defect had typical clinical presentations, including absence of secondary sexual characteristics, primary amenorrhea, hypertension, hypokalamia, lower gonadal hormone levels, as well as elevated corticotropin and progesterone levels. TAC329AA homozygous mutation, IVS1+ 2T>C, and c. 775_776delAT complex heterozygous mutation were found in 2 cases. Four cases of partial combined defect showed high progesterone, lower gonadal hormones and dehydroepiandrosterone-sulfate levels. Three females(46, XX)showed spontaneous menstrual and primary infertility, and two of them got successful pregnancy with assisted reproductive technology. TAC329AA heterozygous mutation was found in those 4 cases. Conclusions TAC329AA mutation is common in 17OHD, and heterozygous or homozygous mutation of TAC329AA may be the genetic and molecular basis for determining whether these patients present as partial or complete defect. The elevated plasma progesterone in non-pregnancy combined with lower gonadal hormones and dehydroepiandrosterone-sulfate is the main character of patients with partial 17OHD. Less severe patients may be able to get successful pregnancy with assisted reproductive technology.

Energy intake excess in modern society contributes to the abrupt increase of type 2 diabetes. The sodium-glucose cotransport 2 (SGLT-2) inhibitors, a new class of hypoglycemic agents inhibit reabsorption of glucose and excrete excessive energy by the kidney and therefore ameliorate the metabolic stress. The relief of metabolic stress further improves blood glucose, lipid levels, so as blood pressure, insulin resistance and pancreatic β cell function, which would be associated with the protection of cardiovascular diseases and amelioration of renal function.

Objective To analyze the influencing factors of venous thromboembolism (VTE) in intensive care units, and provide evidence for corresponding clinical decisions. Methods A case-control study was carried out:224 VTE patients admitted to the ICUs of tertiary hospitals in Kunshan from Nov. 2011 to Nov. 2016 were included in the case group, and 224 non-VTE patients admitted during the same period were randomly selected as the control. The patients' medical history, laboratory tests, prophylaxis and treatment of VTE, and other relevant data were retrieved. Logistic aggression analysis was utilized to identify the influencing factors of VTE in hospitalized critically ill patients. Results The univariate analysis showed that gender, age, hypertension, smoking, D-dimer, Caprini scaling, prophylaxis and treatment of VTE were associated with VTE. The multivariate analysis indicated that except hypertension, the other varieties were independent influencing factors of VTE in hospitalized critically ill patients:female ( OR=1.68, 95％CI:1.09-2.61, P=0.02), higher age(OR=1.03, 95％CI:1.01-1.04, P<0.01), smoking(OR=6.82, 95％CI:1.70-27.46, P=0.007), smoking(OR=6.82, 95％CI:1.70-27.46, P=0.007), D-dimer(OR=0.94, 95％CI:0.92-0.96, P<0.01), higher Caprini scaling(OR=1.80, 95％CI:1.33-2.44, P<0.01), prophylaxis and treatment of VTE(OR=0.34, 95％CI:0.15-0.79, P=0.01). Conclusions Those ICU patients who are female, elder, with smoking history, have higher test value of D-dimer should be screened and assessed for VTE, and those with higher Caprini scaling should be given closer attention, and receive corresponding prophylaxis and treatment to reduce the formation of VTE and its damage.

Objective: To study new ways and tools for assessing the inpatient disease management and improving refined management of the hospital. Methods: By using homepages of medical records of the patients discharged from 21 tertiary general hospitals in a city in 2016, we completed the modeling and predicted value calculation within each DRGs with the Disease Management Intelligent Analytic & Evaluation System (DMIAES System). Results: 2 192 predication models were built, to compute the theoretic values of the mortality rate, length of stay, medical fee, medical service fee, and drug cost of each inpatient. Such values were compared with the observed results to gain the O/E index. If O/E is less than 1, it indicates that the inpatient′s disease management is good and better than expected. On the other hand, O/E index greater than 1 indicates poorer disease management than expected and rooms of further improvement. With the help of O/E index, we made multidimensional comparisons assessment and analysis of different hospitals, clinical disciplines, diseases and doctors. Conclusions The DMIAES System can take risk factors of inpatients′ outcomes into account, assessing the major indicators of inpatient outcomes by means of big data and modelling. This approach proves effective in enabling administrators and doctors to rapidly analyze problems for identifying solutions and enhancing management, thus having great potential in hospital management, supervision and assessment.

The paper describes the organization framework,representation pattern,relation model and expression rule of SNOMED CT based on concept,and then explores the application of SNOMED CT in the expression of medical data and semantic retrieval.Thus,it can provide reference for the research and development of the terminology standard of clinical diagnosis and treatment,and promote the study of processing,mining and analysis of clinical medical data in China.

Objective To evaluate the value of CT enterography (CTE)imaging scores in assessing the clinical activity of Crohn's disease (CD).Methods Forty-eight patients with pathologically or clinically confirmed CD were collected retrospectively.The CTE findings associated with disease activity were analyzed and scored,and the well-known Crohn's disease activity index (CDAI)was also performed.The correlation between CTE score and CDAI score was analyzed by Pearson correlation method.The diagnostic efficacy and optimal diagnostic threshold of CTE score for clinical activity of CD were evaluated by ROC curve analysis.Results There was a good correlation between the CTE score and CDAI (r =0.790,P <0.001).ROC curve analysis showed the area under the curve (AUC)was 0.913 (P <0.001).The optimal diagnostic threshold was 5.5.A CTE score of 5.5 or above indicated the disease was active,with a sensitivity of 84.2％,specificity of 80％,positive predictive value of 94.1％,negative predictive value of 57.1％ and the Youden index of 0.642.Conclusion The CTE comprehensive score has high sensitivity and specificity for estimating clinical activity of CD.

Objective To study the role of quality control circle(QCC) in reducing operation-related acute pressure ulcer.Methods The QCC activity group composed by 13 nursing staff analyzed the current situation and causality of operation-related acute pressure ulcer,and made the pointed counter measures to reducing the incidence rate.Results Following the development of QCC activity,the incidence rate of operation-related acute pressure ulcer decreased from 1.83% to 0.73%. The sense of responsibility,communication,teamwork,motivation,quality control ability and harmony of all the group members were improved significantly after the activity.Conclusions The QCC activity could not only effectively reduced the incidence rate of operation-related acute pressure ulcer,but also consolidated our team spirit and enhanced the quality management capability of all team members.