ABSTRACT
Conventional metric learning methods usually assume that the training and test samples are captured in similar scenarios so that their distributions are assumed to be the same. This assumption does not hold in many real visual recognition applications, especially when samples are captured across different data sets. In this paper, we propose a new deep transfer metric learning (DTML) method to learn a set of hierarchical nonlinear transformations for cross-domain visual recognition by transferring discriminative knowledge from the labeled source domain to the unlabeled target domain. Specifically, our DTML learns a deep metric network by maximizing the inter-class variations and minimizing the intra-class variations, and minimizing the distribution divergence between the source domain and the target domain at the top layer of the network. To better exploit the discriminative information from the source domain, we further develop a deeply supervised transfer metric learning (DSTML) method by including an additional objective on DTML, where the output of both the hidden layers and the top layer are optimized jointly. To preserve the local manifold of input data points in the metric space, we present two new methods, DTML with autoencoder regularization and DSTML with autoencoder regularization. Experimental results on face verification, person re-identification, and handwritten digit recognition validate the effectiveness of the proposed methods.
ABSTRACT
Objective To investigate the absorption characteristics of galangin in various intestinal segments. Methods Single-pass intestinal perfusion was employed in rats, and the mass quality was used to correct the volume;Galangin in rat intestinal perfusion was determined by HPLC to investigate the effects of intestinal segments, drug concentration and P-glycoprotein ( P-gp) inhibitor on drug’ s absorption. Results Galangin could be absorbed in the whole intestine, and its Ka values in the segments of duodenum, jejunum, ileum and colon were (5. 12±1. 14)í10-2,(2. 23±1. 02)í10-2,(4. 61± 0. 75)í 10-2 and(2. 68 ± 0. 70)í10-2 ·min-1 ,respectively. Meanwhile, the values of the Ka in the segment of ileum were not affected by the drug concentration and P-gp inhibitor. Conclusion The galangin is well absorbed in rats intestinal segments. The absorption procedure is mainly controlled by passive diffusion but unaffected by P-gp efflux protein.
ABSTRACT
To evaluate the efficacy of Tuina and Chinese patent drug Shuxuetong injection in preventing patients undergoing total knee arthroplasty from deep venous thrombosis and in functional rehabilitation.
ABSTRACT
Objective This study is aimed to explore the effect of rhTNFR:Fc and methotrexate (MTX)-rhTN FR:Fc on joint destruction of collagen-induced arthritis ( CIA ) rat by establishing CIA rat model which imitates pathogenic factors of rheumatoid arthritis (RA).Methods CIA rat model were developed by subcutaneous injection of bovine type Ⅱ collagen.The rats with inflammation scores of two or above were randomly divided into four groups:the sterilized water treatment group (0.4 ml/w,intra-peritoneal injection),the MTX treatment group (1 mg/w,intra-peritoneal injection),the rhTNFR:Fc treatment group(0.8 mg Biw,intra-peritoneal injection),the MTX + rhTNFR:Fc treatment group (MTX 1 mg/w and rhTNFR:Fc 0.8 mg Biw,intraperitoneal injection).After treatment for 8 weeks,the rats were sacrificed and took the ankle radiography.Micro-CT scan of proximal tibia was performed and hard-tissue slices were made,and then the ankle's bone damage of each group was observed in order to evaluate trabecular variation and bone quantity changes of proximal tibia.Statisstical analysis was conducted with ANK-q test.Results After treatment for 8 weeks,the percentage of trabecular area and the trabecular number of the rhTNFR:Fc treatment group and the MTX-rhTNFR:Fc treatment were [(29.1±0.3)％,(26.7±0.6)％,(4.4±0.5)/mm,(4.0±0.6)/mm] (P＜0.01),which were evidently higher than the sterilized water treatment group and MTX treatment group (P＜0.01).The trabecular separation of Etanercept treatment group and MTX-rhTNFR:Fc group was obviously less than the sterilized water treatment group and MTX treatment group [(12.9±0.5)％,(13.2±0.4)％ vs (2.0±0.3)/mm,(2.2t0.2)/mm] (P＜0.01).Conclusion rhTNFR:Fc and MTX-rhTNFR:Fc can remarkably inhibit joint destruction of CIA rat.And their effect on inhibiting of inflammation and increasing peri-articular bone quantity.In addition,they are effective on inhibiting the reduction of local trabecular structure and increase of trabecular separation.
ABSTRACT
Objective To evaluate the outcome of primary total knee arthroplasty (TKA) in treatment of stiff knee, and discuss the key points of operative technique and rehabilitation. Methods From February 2005 to April 2009, 23 patients with 34 stiff knees were treated with primary TKA. The study included 3 males (4 knees) and 20 females (30 knees), with the mean age of 56.9 years. Primary disease of the patients included rheumatoid arthritis (26 knees in 15 cases) and osteoarthritis (8 knees in 8 cases). Varus deformity was found in 10 knees (5°-15°), and valgus was found in 5 knees (5°-10°). Evaluations included preoperative and postoperative range of motion (ROM) measurement, hospital for special surgery knee score (HSS), blood loss, operative time and assessment of postoperative complications. Results All patients were followed up. The mean follow-up time was 32.2 (range, 24 to 40) months. At the final follow-up visit, the HSS score increased from 42.9±5.2 preoperatively to 85.7±4.3, the range of motion increased from 42.6°±5.7° preoperatively to 89.2°±10.5°. Sixteen knees in 12 cases underwent manipulation at 3 to 8 weeks postoperatively for unsatisfied ROM, but ROM was still less than 90° in 8 knees at the last follow-up. The average blood loss were (632.4±180.2) ml in first 24 hours (450-850 ml) and the operative time were (98.1±18.6) min (80-150 min). Deep venous thrombosis was found in 3 patients. All the symptoms relieved after anticoagulant therapy. Postoperative varus deformity was seen in one patient, but the function of knee was good. No revision was needed. Conclusion Primary total knee arthroplasty is reliable method in treatment of stiff knees. Sufficient soft tissue release during the operation, postoperative muscle strength exercise and essential manipulation are key points for satisfactory outcomes.
