234 papers
Bioengineering sits at the vibrant intersection of biology and engineering, where scientists design new tools to understand life and solve real-world problems. From creating artificial organs to developing smart materials that mimic natural tissues, this field turns complex biological questions into tangible innovations that improve human health and our environment.
At Gist.Science, we track every new preprint in this category directly from bioRxiv. Our team processes each submission to provide both plain-language explanations for curious minds and detailed technical summaries for researchers, ensuring these cutting-edge discoveries are accessible to everyone. Below are the latest papers in bioengineering that have recently appeared on bioRxiv.
This paper demonstrates that leveraging large language models, particularly Gemma 3 with SentencePiece tokenization and few-shot prompting, as a post-processing step can effectively correct character-level errors in P300 speller output, thereby enabling faster communication with fewer stimulus repetitions while maintaining high accuracy for users with motor and speech disabilities.
This study demonstrates that catalytic rewiring of the RuvC-II site in Fanzor2 activates robust trans-cleavage activity, enabling the development of versatile diagnostic platforms for sensitive pathogen detection and precise single-nucleotide variant genotyping.
This paper presents a semi-quantitative HIV-1 detection method that utilizes a ResNet convolutional neural network to analyze spatiotemporal fluorescence patterns from isothermal amplification in microfluidic chips, achieving high accuracy across a five-order-of-magnitude concentration range and enabling rapid, decentralized molecular diagnostics.
This paper demonstrates that engineered T cells can be programmed to produce and transfer genome-editing machinery to target cells via a customizable genetic circuit, offering a novel cell-based delivery system that leverages immune cell properties for enhanced specificity and efficiency.
This study characterizes the shear-thinning, non-Newtonian rheological properties of venoms from thirteen snake species and demonstrates that safe, inexpensive aqueous solutions of bovine serum albumin and xanthan gum can effectively serve as scalable surrogates to model venom flow dynamics for biomedical and biomechanical research.
This paper introduces ZT Scan DIA, a dual-dimensional deconvolution strategy integrated into MS-DIAL that significantly enhances metabolite annotation rates and enables precise isomer separation and quantification in complex hydrophilic and lipidomic samples.
This study demonstrates that integrating a U-Net deep learning architecture into variable-density GRAPPA reconstruction significantly outperforms conventional and MultiNet-based methods by enabling higher acceleration factors for 7T 1H MRSI while effectively suppressing lipid artifacts and preserving metabolite quantification fidelity.
By combining ultra-high-speed imaging with digital image correlation, this study demonstrates that a cell's viscoelastic properties and cytoskeletal organization govern the propagation of microbubble-induced deformation waves at ultrasound frequencies, thereby determining the efficiency of therapeutic permeabilization.
This study presents and systematically validates a novel multiscale, fibre-type-specific Hill-type neuromuscular actuator with refined mechanistic excitation-activation dynamics against comprehensive experimental benchmarks spanning motor unit and whole-muscle scales, demonstrating its ability to accurately reproduce muscle forces across diverse physiological conditions.
This study combines feline electrophysiology and computational modeling to demonstrate that while reducing electrode contact diameter can improve the focal selectivity of lateral spinal cord stimulation, gains plateau below 1000 µm and further reductions may not enhance selectivity due to inherent physiological constraints.