Browse Articles

Evaluation of cationic peptide-based nanogels as delivery systems for negatively charged molecules: a formulative study

Photobiocatalytic Radical Hydroalkylation with C(sp ³ )–H Bonds Enabled by Engineered Imine Reductase and Redox Buffering

Mass spectrometry combined with machine learning identifies novel protein signatures as demonstrated with multisystem inflammatory syndrome in children

Efficient deep neural networks for cancer detection on histopathology combining attention and image downsampling

Abstract Pathology diagnosis of colorectal cancer is time-consuming and requires a high level of expertise. However, it is an essential step towards establishing the adequate treatment. The need to analyse a large number of these histopathological images calls for automatic tools capable of aiding pathologists in this arduous task. Deep learning techniques, together with the wealth of data available nowadays, provide a promising candidate for such job. Adopting state-of-the-art artificial intelligence algorithms, we developed a model to accurately detect colorectal cancer in digitalised histopathological whole-slide images. Our end-to-end approach uses the principles of multiple-instance learning combined with deep convolutional neural networks in order to fully leverage the information contained within each image and make robust predictions at the patient’s level. The model also allows to highlight the areas in the slide most likely to harbour tumour tissue. Given the finite computational resources available, working at maximum resolution can be detrimental. Therefore, we explored the impact of lowering the working image resolution. The algorithms were trained and validated on a subset of more than 1300 patients of the Molecular Epidemiology of Colorectal Cancer study with histopathology images available. These images gave rise to $$>10^5$$ tiles of $$256\times 256$$ pixels each. Once we identified the best-performing model we put it to the test on images from The Cancer Genome Atlas. We obtained the best outcomes working at 4 μm/pix, achieving the following metrics on the test dataset: F1-Score of 0.96, a Matthews correlation coefficient of 0.92 and an area under the receiver operating characteristic curve of 0.99. These results are exceptional and prove that computational costs can be reduced while keeping the performance up to standard.

People with some cancers live longer after a COVID vaccine

A-Cation-Dependent Structure–Optical Property Relationships of Halide Perovskite Heterostructures with Complex Interfaces

Development and validation of a self-management intervention for adolescents living with HIV in Namibia

Establishing the Field-Flow Competition Model to Decipher the Nonmonotonic Interfacial Li ⁺ Dynamic Process for Stabilizing the High-Voltage Cathode–Electrolyte Interface

Integrated transcriptomic analysis identifies lysosomal autophagy-related genes in sarcopenia

Tailored Electronic Metal–Support Interaction Boosts Hydrogen Release from Organic Carriers

Detection of pre-seismic magnetic field anomalies using Swarm satellite data: a case study of the 2025 Mw7.7 Myanmar earthquake

Granzyme B-Targeting Quenched Activity-Based Probes for Assessing Tumor Response to Immunotherapy

Multi-objective optimization of electromagnetic vibration parameters for corn seed phenotype prediction based on deep learning

Abstract This study presents a novel framework for adaptive optimization of electromagnetic vibration parameters in corn seed treatment using multi-objective deep learning approaches. A hybrid CNN-LSTM network architecture was developed to process heterogeneous sensor data and predict multiple seed phenotype characteristics simultaneously. The framework integrates genetic algorithms with particle swarm optimization for real-time parameter adjustment, addressing the complex relationships between electromagnetic treatment conditions and seed quality outcomes. Experimental validation using three corn varieties (Zhengdan 958, Xianyu 335, and Jingke 968) demonstrates significant performance improvements, with optimized treatment protocols achieving 12.8% enhancement in germination rates and 17.7% improvement in vigor indices compared to untreated controls. The multi-objective deep learning model achieved 93.7% prediction accuracy with 91.2% recall rate, outperforming conventional optimization approaches. The adaptive parameter optimization strategy successfully balanced competing objectives including treatment effectiveness, energy efficiency, and processing time while maintaining robust performance across different seed batches. This research provides a comprehensive solution for intelligent seed treatment systems, offering substantial potential for advancing precision agriculture and sustainable crop production technologies.

Honey, I ate the kids: how hunger and hormones make mice aggressive

Integrated Solvation Chemistry Enables High Energy Li Metal Batteries

A novel portfolio construction strategy based on the core- periphery profile of stocks

Magnitude and determinants of adverse perinatal outcomes of pregnancies complicated with preeclampsia and eclampsia at a teaching hospital in the Tigray region of Ethiopia

Macroscopic Uniform 2D Moiré Superlattices with Controllable Angles

EMG hysteresis patterns in human elbow muscles under simultaneous force and length changes

Photoelectrochemical Catalyzed Site-Selective Tryptophan β-Position Functionalization