Image cropping aims at improving the aesthetic quality of images by adjusting their composition. Most weakly supervised cropping methods (without bounding box supervision) rely on the sliding window mechanism. The sliding window mechanism requires fixed aspect ratios and limits the cropping region with arbitrary size. Moreover, the sliding window method usually produces tens of thousands of windows on the input image which is very time-consuming. Motivated by these challenges, we firstly formulate the aesthetic image cropping as a sequential decision-making process and propose a weakly supervised Aesthetics Aware Reinforcement Learning (A2-RL) framework to address this problem. Particularly, the proposed method develops an aesthetics aware reward function which especially benefits image cropping. Similar to human's decision making, we use a comprehensive state representation including both the current observation and the historical experience. We train the agent using the actor-critic architecture in an endto-end manner. The agent is evaluated on several popular unseen cropping datasets. Experiment results show that our method achieves the state-of-the-art performance with much fewer candidate windows and much less time compared with previous weakly supervised methods.

BackgroundThe beneficial effect of surgical resection for hepatic metastasis from gastric cancer (HMGC) remains elusive. This study was conducted to analyze surgical outcomes of HMGC and determine the prognostic factors associated with survival.ResultsThe in-hospital mortality rate was zero, and the overall morbidity rate was 56%. The overall 1-, 3-, and 5-year survival rate after surgery was 87.5%, 47.6%, and 21.7%, respectively, with a median survival time of 34.0 months. Multiple liver metastases (hazard ratio [HR] =1.998; 95% confidence interval [CI] = 1.248-3.198; P = 0.004) and ≥ T3 stage of the primary gastric cancer (HR = 2.065; 95% CI = 1.201–3.549; P = 0.009) were independent prognostic determinants in the multivariate analysis.Materials and MethodsData on surgical resection of 96 patients with HMGC at six institutions in China were analysed retrospectively. Prognostic factors were assessed by multiple stepwise regression analysis using the Cox model.ConclusionsSurgical resection for HMGC is feasible and beneficial to long-term survival in selected patients.

BACKGROUNDPerioperative blood transfusion may be associated with negative clinical outcomes in oncological surgery. A meta-analysis of published studies was conducted to evaluate the impact of blood transfusion on short- and long-term outcomes following liver resection of colorectal liver metastasis (CLM).MATERIALS AND METHODSA systematic search was performed to identify relevant articles. Data were pooled for meta-analysis using Review Manager version 5.3.RESULTSTwenty-five observational studies containing 10621 patients were subjected to the analysis. Compared with non-transfused patients, transfused patients experienced higher overall morbidity (odds ratio [OR], 1.98; 95% confidence intervals [CI] =1.49-2.33), more major complications (OR, 2.12; 95% CI =1.26-3.58), higher mortality (OR, 4.13; 95% CI =1.96-8.72), and longer length of hospital stay (weighted mean difference, 4.43; 95% CI =1.15-7.69). Transfusion was associated with reduced overall survival (risk ratio [RR], 1.24, 95% CI =1.11-1.38) and disease-free survival (RR, 1.38, 95% CI=1.23-1.56).CONCLUSIONPerioperative blood transfusion has a detrimental impact on the clinical outcomes of patients undergoing CLM resection.

No abstract

No abstract