Enhanced robust spatial feature selection and correlation filter learning for UAV tracking.

Spatial boundary effect can significantly reduce the performance of a learned discriminative correlation filter (DCF) model. A commonly used method to relieve this effect is to extract appearance features from a wider region of a target. However, this way would introduce unexpected features from background pixels and noises, which will lead to a decrease of the filter's discrimination power. To address this shortcoming, this paper proposes an innovative method called enhanced robust spatial feature selection and correlation filter Learning (EFSCF), which performs jointly sparse feature learning to handle boundary effects effectively while suppressing the influence of background pixels and noises. Unlike the ℓ2-norm-based tracking approaches that are prone to non-Gaussian noises, the proposed method imposes the ℓ2,1-norm on the loss term to enhance the robustness against the training outliers. To enhance the discrimination further, a jointly sparse feature selection scheme based on the ℓ2,1 -norm is designed to regularize the filter in rows and columns simultaneously. To the best of the authors' knowledge, this has been the first work exploring the structural sparsity in rows and columns of a learned filter simultaneously. The proposed model can be efficiently solved by an alternating direction multiplier method. The proposed EFSCF is verified by experiments on four challenging unmanned aerial vehicle datasets under severe noise and appearance changes, and the results show that the proposed method can achieve better tracking performance than the state-of-the-art trackers.

Process Evaluation of an Implementation Trial: Design, Rationale, and Early Lessons Learnt From an International Cluster Clinical Trial in Intracerebral Hemorrhage.

Background: The third INTEnsive care bundle with blood pressure Reduction in Acute Cerebral Hemorrhage Trial (INTERACT3) is an ongoing, international, multicenter, stepped-wedge cluster, prospective, randomized, open, blinded endpoint assessed trial evaluating the effectiveness of a quality improvement "care bundle" for the management of patients with acute spontaneous intracerebral hemorrhage (ICH) in low- and middle-income countries (LMICs). An embedded process evaluation aims to explore the uptake and implementation of the intervention, and understand the context and stakeholder perspectives, for interpreting the trial outcomes. Methodology: The design was informed by Normalization Process Theory and the UK Medical Research Council process evaluation guidance. Mixed methods are used to evaluate the implementation outcomes of fidelity, reach, dose, acceptability, appropriateness, adoption, sustainability, and relevant contextual factors and mechanisms affecting delivery of the care bundle. Semi-structured interviews and non-participant observations are conducted with the primary implementers (physicians and nurses) and patients/carers to explore how the care bundle was integrated into routine care. Focus group discussions are conducted with investigators and project operational staff to understand challenges and possible solutions in the organization of the trial. Data from observational records, surveys, routine monitoring data, field notes and case report forms, inform contextual factors, and adoption of the intervention. Purposive sampling of sites according to pre-specified criteria is used to achieve sample representativeness. Discussion: Implementation outcomes, and relevant barriers and facilitators to integrating the care bundle into routine practice, will be reported after completion of the process evaluation. The embedded process evaluation will aid understanding of the causal mechanisms between care bundle elements and clinical outcomes within complex health systems across diverse LMIC settings. Trial Registration: The INTERACT3 study is registered at ClinicalTrials.gov (NCT03209258).

Implementing a Goal-Directed Care Bundle after Acute Intracerebral Haemorrhage: Process Evaluation for the Third INTEnsive Care Bundle with Blood Pressure Reduction in Acute Cerebral Haemorrhage Trial Study in China.

BACKGROUND: The third INTEnsive care bundle with blood pressure Reduction in Acute Cerebral Haemorrhage Trial is an ongoing international, multicentre, stepped wedge, cluster-randomized trial to determine the effectiveness of a goal-directed care bundle (early intensive blood pressure [BP] lowering, glycaemic control, treatment of pyrexia, and reversal of anticoagulation), as compared to standard of care, on patient-centred outcomes after acute intracerebral haemorrhage (ICH). An embedded process evaluation aims to identify factors related to the uptake and implementation of the intervention. Herein, we present the process evaluation results for hospital sites in China. METHODS/DESIGN: A mixed methods approach, including surveys, focused group discussions and interviews with clinicians, routine monitoring, and recruitment logs were used to collect data across purposively sampled hospitals. Medical Research Council guidance and normalization process theory were used as theoretical frameworks for design, data analysis, and synthesis. RESULTS: Twenty quantitative surveys were completed with clinicians, and 26 interviews and 2 focus group discussions were conducted during 2019-2020. The care bundle was generally delivered as planned and acceptable by doctors and nurses, but difficulties were reported in achieving the protocol-defined target levels of BP and glycaemic control. Resistance to implementing the care bundle occurred for patients perceived to be at high risk of adverse effects. Common organizational contextual factors that impeded implementation included delayed processes and limited medication supply, while established background care procedures, expertise, and capacity influenced its integration into routine practice. Areas to facilitate implementation included optimizing workflow within available resources, having a dedicated team, and recognizing the potential benefits of the intervention. CONCLUSIONS: Varied established care protocols across sites, different levels of background expertise, and lack of staff capacity impeded the integration of goal-directed care bundle into routine practice for ICH patients in China. Ready identification, and efforts to address, these barriers could facilitate uptake of future guideline-recommended interventions for the management of patients with ICH.