Poulsenharder7811

Convolutional neural networks (CNNs) are emerging as powerful tools for EEG decoding these techniques, by automatically learning relevant features for class discrimination, improve EEG decoding performances without relying on handcrafted features. Nevertheless, the learned features are difficult to interpret and most of the existing CNNs introduce many trainable parameters. Here, we propose a lightweight and interpretable shallow CNN (Sinc-ShallowNet), by stacking a temporal sinc-convolutional layer (designed to learn band-pass filters, each having only the two cut-off frequencies as trainable parameters), a spatial depthwise convolutional layer (reducing channel connectivity and learning spatial filters tied to each band-pass filter), and a fully-connected layer finalizing the classification. This convolutional module limits the number of trainable parameters and allows direct interpretation of the learned spectral-spatial​ features via simple kernel visualizations. Furthermore, we designed a post-hoc gradient-based technique to enhance interpretation by identifying the more relevant and more class-specific features. Sinc-ShallowNet was evaluated on benchmark motor-execution and motor-imagery datasets and against different design choices and training strategies. Results show that (i) Sinc-ShallowNet outperformed a traditional machine learning algorithm and other CNNs for EEG decoding; (ii) The learned spectral-spatial features matched well-known EEG motor-related activity; (iii) The proposed architecture performed better with a larger number of temporal kernels still maintaining a good compromise between accuracy and parsimony, and with a trialwise rather than a cropped training strategy. In perspective, the proposed approach, with its interpretative capacity, can be exploited to investigate cognitive/motor aspects whose EEG correlates are yet scarcely known, potentially characterizing their relevant features.A guidance document for the identification of endocrine disruptors (EDs) in the regulatory assessment of plant protection products (PPP) and biocidal products (BP) has been published by the European Chemical Agency (ECHA) and the European Food Safety Authority (EFSA). The ECHA/EFSA guidance, mainly addressing EATS (estrogen, androgen, thyroid, steroidogenesis) modalities, is intended to guide applicants and assessors of the competent regulatory authorities on the implementation of the scientific criteria for the determination of ED properties pursuant to the recently implemented PPP (EU 2018/605) and BP (EU 2017/2100) EU Regulations. In this study, a search filter for targeted literature search in context of assessing if a substance can be identified as an ED relevant for human health was developed and validated. Development of the search filter was based on the search strategy presented in the ECHA/EFSA guidance and using the estrogenic chemical Bisphenol AF (BPAF) as a model substance. Information specialists from two independent institutions developed refined search filters based on the suggested original search strategy published (ECHA/EFSA guidance - Appendix F). Articles identified by a systematic literature search for BPAF were screened for relevance with inclusion and exclusion criteria by two independent reviewers obtaining positive (relevant) and negative (irrelevant) controls. The developed search filter was quantitatively evaluated in terms of sensitivity, specificity and precision based on the positive and negative controls. The developed filter was then validated for T modality by its application to the known thyroid-disruptor perchlorate. The result is a sensitive search filter with sufficient specificity, which can be applied for all chemicals where a targeted literature search is needed to assess and identify ED properties of chemicals with relevance for humans. AUY-922 concentration Future application of the filter to a broader range of chemicals may identify further points of improvement.Background Corona Virus Disease 19 (COVID-19) had a worldwide negative impact on healthcare systems, which were not used to coping with such pandemic. Adaptation strategies prioritizing COVID-19 patients included triage of patients and reduction or re-allocation of other services. The aim of our survey was to provide a real time international snapshot of modifications of breast cancer management during the COVID-19 pandemic. Methods A survey was developed by a multidisciplinary group on behalf of European Breast Cancer Research Association of Surgical Trialists and distributed via breast cancer societies. One reply per breast unit was requested. Results In ten days, 377 breast centres from 41 countries completed the questionnaire. RT-PCR testing for SARS-CoV-2 prior to treatment was reported by 44.8% of the institutions. The estimated time interval between diagnosis and treatment initiation increased for about 20% of institutions. Indications for primary systemic therapy were modified in 56% (211/377), with upfront surgery increasing from 39.8% to 50.7% (p less then 0.002) and from 33.7% to 42.2% (p less then 0.016) in T1cN0 triple-negative and ER-negative/HER2-positive cases, respectively. Sixty-seven percent considered that chemotherapy increases risks for developing COVID-19 complications. Fifty-one percent of the responders reported modifications in chemotherapy protocols. Gene-expression profile used to evaluate the need for adjuvant chemotherapy increased in 18.8%. In luminal-A tumours, a large majority (68%) recommended endocrine treatment to postpone surgery. Postoperative radiation therapy was postponed in 20% of the cases. Conclusions Breast cancer management was considerably modified during the COVID-19 pandemic. Our data provide a base to investigate whether these changes impact oncologic outcomes.Background Walking training is an essential intervention to improve the function in stroke patients. However, only a limited number of gait training strategies are available for stroke patients with relatively severe disabilities. Research question Is underwater gait training or overground gait training more effective in severe stroke patients? Methods A total of 21 patients with severe hemiplegic stroke were randomly assigned to the experimental and control groups. All participants (n = 21) received 60-minute sessions of general physical therapy, 5 times a week for a period of 12 weeks. Additionally, the experimental and control groups underwent underwater and overground walking training, respectively, for 30 min twice times a week for 12 weeks. Postural assessment for stroke score, center of pressure path length and velocity, step time and step length difference, and walking velocity were measured before and after the 12-week training. Results Both groups showed a significant decrease in the center of pressure path length and velocity after the intervention compared to the values before the intervention (p .