Efficient nonlinear function approximation in analog resistive crossbars for recurrent neural networks.

Analog In-memory Computing (IMC) has demonstrated energy-efficient and low latency implementation of convolution and fully-connected layers in deep neural networks (DNN) by using physics for computing in parallel resistive memory arrays. However, recurrent neural networks (RNN) that are widely used for speech-recognition and natural language processing have tasted limited success with this approach. This can be attributed to the significant time and energy penalties incurred in implementing nonlinear activation functions that are abundant in such models.

Risk of residual/recurrent cervical diseases in HPV-positive women post-conization depends on HPV integration status.

Background: It is crucial to identify post-operative patients with HPV infection who are at high risk for residual/recurrent disease. This study aimed to evaluate the association between HPV integration and clinical outcomes in HPV-positive women after cervical conization, as well as to identify HPV integration breakpoints.

Methods: This retrospective study analyzed data of 791 women who underwent cervical conization for cervical intraepithelial neoplasia grades 2-3 (CIN2-3) between September 2019 and September 2023, sourced from the Fujian and Hubei cervical lesion screening cohorts.

Selenium-Enriched Lactiplantibacillus plantarum ZZU 8-12 Regulates Intestinal Microbiota and Inhibits Acute Liver Injury.

Intake of certain Lactiplantibacillus strains was recognized as a potential strategy for acute liver injury (ALI) prevention. This study is aimed at developing a selenium-enriched Lactiplantibacillus strain-based ALI prevention strategy. L.

Effect of ultrasound combined with microbubbles therapy on tumor hypoxic microenvironment.

Introduction: Tumor tissues exhibit significantly lower oxygen partial pressure compared to normal tissues, leading to hypoxia in the tumor microenvironment and result in resistance to tumor treatments. Strategies to mitigate hypoxia include enhancing blood perfusion and oxygen supply, for example,by decomposing hydrogen peroxide within the tumor. Improving hypoxia in the tumor microenvironment could potentially improve the efficacy of cancer treatments.

REDT: a specialized transformer model for the respiratory phase and adventitious sound detection.

Background And Objective: In contrast to respiratory sound classification, respiratory phase and adventitious sound event detection provides more detailed and accurate respiratory information, which is clinically important for respiratory disorders. However, current respiratory sound event detection models mainly use convolutional neural networks to generate frame-level predictions. A significant drawback of the frame-based model lies in its pursuit of optimal frame-level predictions rather than the best event-level ones.