Can we detect common objects in a variety of image domains without instance-level annotations? In this paper, we present a framework for a novel task, cross-domain weakly supervised object detection, which addresses this question. For this paper, we have access to images with instance-level annotations in a source domain (e.g., natural image) and images with image-level annotations in a target domain (e.g., watercolor). In addition, the classes to be detected in the target domain are all or a subset of those in the source domain. Starting from a fully supervised object detector, which is pre-trained on the source domain, we propose a two-step progressive domain adaptation technique by fine-tuning the detector on two types of artificially and automatically generated samples. We test our methods on our newly collected datasets 1 containing three image domains, and achieve an improvement of approximately 5 to 20 percentage points in terms of mean average precision (mAP) compared to the best-performing baselines.

Background-The influences of antiplatelet therapy discontinuation on the risk of stent thrombosis and long-term clinical outcomes after drug-eluting stent implantation have not yet been addressed adequately. Methods and Results-In an observational study in Japan, 2-year outcomes were assessed in 10 778 patients undergoing sirolimus-eluting stent implantation. Data on status of antiplatelet therapy during follow-up were collected prospectively. Incidences of definite stent thrombosis were 0.34% at 30 days, 0.54% at 1 year, and 0.77% at 2 years. Thienopyridine use was maintained in 97%, 62%, and 50% of patients at 30 days, 1 year, and 2 years, respectively. Patients who discontinued both thienopyridine and aspirin had a significantly higher rate of stent thrombosis than those who continued both in the intervals of 31 to 180 days, 181 to 365 days, and 366 to 548 days after stent implantation (1.76% versus 0.1%, PϽ0.001; 0.72% versus 0.07%, Pϭ0.02; and 2.1% versus 0.14%, Pϭ0.004, respectively). When discontinuation of aspirin was taken into account, patients who discontinued thienopyridine only did not have an excess of stent thrombosis in any of the time intervals studied. Adjusted rates of death or myocardial infarction at 24 months were 4.1% for patients taking thienopyridine and 4.1% for patients not taking thienopyridine (Pϭ0.99) in the 6-month landmark analysis. Conclusions-Discontinuation of both thienopyridine and aspirin, but not discontinuation of thienopyridine therapy only, was associated with an increased risk of stent thrombosis. Landmark analysis did not suggest an apparent clinical benefit of thienopyridine use beyond 6 months after sirolimus-eluting stent implantation. (Circulation. 2009;119:987-995.)

Unacceptably high viscosity is observed in high protein concentration formulations due to extremely large therapeutic dose of antibodies and volume restriction of subcutaneous route of administration. Here, we show that a protein aggregation suppressor, arginine hydrochloride (ArgHCl), specifically decreases viscosity of antibody formulations. The viscosities of bovine gamma globulin (BGG) solution at 250 mg/mL and human gamma globulin (HGG) solution at 292 mg/mL at a physiological pH were too high for subcutaneous injections, but decreased to an acceptable level (below 50 cP) in the presence of 1,000 mM ArgHCl. ArgHCl also decreased the viscosity of BGG solution at acidic and alkaline pHs. Interestingly, ArgHCl decreased the viscosity of antibody solutions (BGG, HGG, and human immunoglobulin G) but not globular protein solutions (α-amylase and α-chymotrypsin). These results indicate not only high potency of ArgHCl as an excipient to decrease the solution viscosity of high concentration antibodies formulations but also specific interactions between ArgHCl and antibodies.

Background-It remains unclear whether cilostazol, which has been shown to improve the clinical outcomes of endovascular therapy for femoropopliteal lesions, also reduces angiographic restenosis. Methods and Results-The Sufficient Treatment of Peripheral Intervention by Cilostazol (STOP-IC) study investigated whether cilostazol reduces the 12-month angiographic restenosis rate after percutaneous transluminal angioplasty with provisional nitinol stenting for femoropopliteal lesions. Two hundred patients with femoropopliteal lesions treated from March 2009 to April 2011 at 13 cardiovascular centers were randomly assigned 1:1 to receive oral aspirin with or without cilostazol. The primary end point was 12-month angiographic restenosis rate. Secondary end points were the restenosis rate on duplex ultrasound, the rate of major adverse cardiac events, and target lesion event-free survival. Researchers evaluated all follow-up data and assessed the end points in a blinded fashion. The mean lesion length and reference vessel diameter at the treated segment were 128±86 mm and 5.4±1.4 mm, respectively. The frequency of stent used was similar between groups (88% versus 90% in the cilostazol and noncilostazol group, respectively, P=0.82). During the 12-month follow-up period, 11 patients died and 152 patients (80%) had evaluable angiographic data at 12 months. The angiographic restenosis rate at 12 months was 20% (15/75) in the cilostazol group versus 49% (38/77) in the noncilostazol group (P=0.0001) by intention-to-treat analysis. The cilostazol group also had a significantly higher event-free survival at 12 months (83% versus 71%, P=0.02), although cardiovascular event rates were similar in both groups. Conclusions-Cilostazol reduced angiographic restenosis after percutaneous transluminal angioplasty with provisional nitinol stenting for femoropopliteal lesions. Clinical Trial Registration-URL: http://www.clinicaltrials.gov. Unique identifier: NCT00912756; and URL: https://www. umin.ac.jp. Unique identifier: UMIN000002091.

No abstract

Background:The use of a stent in the treatment of lesions of the superficial femoral artery (SFA) remains controversial. Although some reports have suggested that use of a nitinol stent in conjunction with aggressive medical management is effective for long SFA lesions, few long-term, large-scale studies have been done. Methods and Results:A retrospective analysis was conducted of data from a multicenter study in which the S.M.A.R.T. Control™ stent was used for treatment of de novo SFA lesions. A total of 528 lesions in 432 patients were included. Mean patient age was 72.5±9.1 years; mean stent length was 15.7±8.1 cm; 259 lesions (49%) were classified as C/D according to the TransAtlantic Inter-Society Consensus (TASC) II classification. Primary and secondary patency at 4 years was 66% and 87%, respectively. No cilostazol administration (41% re-stenosis group vs. 29% no-restenosis group, P<0.01), female gender (42% vs. 26%, P<0.01), younger age (70.7±9.3 years vs. 72.9±9.0 years, P<0.05), and chronic total occlusion (CTO; 72% vs. 52%, P<0.01) were independent predictors of re-stenosis.Conclusions: The S.M.A.R.T. Control™ stent provided good long-term durability in the treatment of SFA lesions, and no cilostazol administration, female gender, younger age and CTO were associated with re-stenosis. (Circ J 2011; 75: 939 - 944)

Renal denervation is a promising new non-pharmacological treatment for resistant hypertension. However, there is a lack of data from Asian patients. The REQUIRE trial investigated the blood pressure-lowering efficacy of renal denervation in treated patients with resistant hypertension from Japan and South Korea. Adults with resistant hypertension (seated office blood pressure ≥150/90 mmHg and 24-hour ambulatory systolic blood pressure ≥140 mmHg) with suitable renal artery anatomy were randomized to ultrasound renal denervation or a sham procedure. The primary endpoint was change from baseline in 24-hour ambulatory systolic blood pressure at 3 months. A total of 143 patients were included (72 renal denervation, 71 sham control). Reduction from baseline in 24-hour ambulatory systolic blood pressure at 3 months was not significantly different between the renal denervation (−6.6 mmHg) and sham control (−6.5 mmHg) groups (difference: −0.1, 95% confidence interval −5.5, 5.3; p = 0.971). Reductions from baseline in home and office systolic blood pressure (differences: –1.8 mmHg [p = 0.488] and −2.0 mmHg [p = 0.511], respectively), and medication load, did not differ significantly between the two groups. The procedure-/device-related major adverse events was not seen. This study did not show a significant difference in ambulatory blood pressure reductions between renal denervation and a sham procedure in treated patients with resistant hypertension. Although blood pressure reduction after renal denervation was similar to other sham-controlled studies, the sham group in this study showed much greater reduction. This unexpected blood pressure reduction in the sham control group highlights study design issues that will be addressed in a new trial. Clinical trial registration NCT02918305 (http://www.clinicaltrials.gov).

Endovascular therapy using nitinol stents for FP lesions yielded acceptable outcomes up to 6 years. Risk stratification for patency can play an important role in estimating future occurrence of restenosis after nitinol stent implantation in FP lesions.