Search for: All records

Mapping agricultural fields using high-resolution satellite imagery and deep learning (DL) models has advanced significantly, even in regions with small, irregularly shaped fields. However, effective DL models often require large, expensive labeled datasets, which are typically limited to specific years or regions. This restricts the ability to create annual maps needed for agricultural monitoring, as changes in farming practices and environmental conditions cause domain shifts between years and locations. To address this, we focused on improving model generalization without relying on yearly labels through a holistic approach that integrates several techniques, including an area-based loss function, Tversky-focal loss (TFL), data augmentation, and the use of regularization techniques like dropout. Photometric augmentations helped encode invariance to brightness changes but also increased the incidence of false positives. The best results were achieved by combining photometric augmentation, TFL, and Monte Carlo dropout, although dropout alone led to more false negatives. Input normalization also played a key role, with the best results obtained when normalization statistics were calculated locally (per chip) across all bands. Our U-Net-based workflow successfully generated multi-year crop maps over large areas, outperforming the base model without photometric augmentation or MC-dropout by 17 IoU points. 

Abstract Three-dimensional (3D) microneedle arrays (MAs) have shown remarkable performances for a wide range of biomedical applications. Achieving advanced customizable 3D MAs for personalized research and treatment remain a formidable challenge. In this paper, we have developed a high-resolution electrohydrodynamic (EHD) 3D printing process for fabricating customizable 3D MAs with economical and biocompatible molten alloy. The critical printing parameters (i.e., voltage and pressure) on the printing process for both two-dimensional (2D) and 3D features are characterized, and an optimal set of printing parameters was obtained for printing 3D MAs. We have also studied the effect of the tip-nozzle separation speed on the final tip dimension, which will directly influence MAs' insertion performance and functions. With the optimal process parameters, we successfully EHD printed customizable 3D MAs with varying spacing distances and shank heights. A 3 × 3 customized 3D MAs configuration with various heights ranging from 0.8 mm to 1 mm and a spacing distance as small as 350 μm were successfully fabricated, in which the diameter of each individual microneedle was as small as 100 μm. A series of tests were conducted to evaluate the printed 3D MAs. The experimental results demonstrated that the printed 3D MAs exhibit good mechanical strength for implanting and good electrical properties for electrophysiological sensing and stimulation. All results show the potential applications of the EHD printing technique in fabricating cost-effective, customizable, high-performance MAs for biomedical applications. 

Abstract Organic electrochemical transistors (OECTs) are gaining significant attention due to their high sensitivity, customizability, ease of integration, and low‐cost manufacturing. In this paper, we design and develop a flexible dual‐gate OECT based on laser‐scribed graphene (LSG) with modified OECT gates for the detection of dopamine and glutamate, two critical neurotransmitters (NTs). The developed OECTs are fully carbon‐based and environmentally friendly. By modifying the gates of OECTs with biopolymer chitosan and L‐Glutamate oxidase enzyme, highly selective and sensitive measurements are successfully achieved with detection limits of 5 nmfor dopamine and 1 µmfor glutamate, respectively. The modified dual‐gate shows no interference between the detections of two neurotransmitters, making it a promising tool for customized multi‐neurotransmitter analysis. The results demonstrate the potential of LSG‐based OECTs in customizable biosensing applications, offering a flexible, cost‐effective platform for biomedical disorder diagnostics. 

Abstract Winter Arctic sea-ice concentration (SIC) decline plays an important role in Arctic amplification which, in turn, influences Arctic ecosystems, midlatitude weather and climate. SIC over the Barents-Kara Seas (BKS) shows large interannual variations, whose origin is still unclear. Here we find that interannual variations in winter BKS SIC have significantly strengthened in recent decades likely due to increased amplitudes of the El Niño-Southern Oscillation (ENSO) in a warming climate. La Niña leads to enhanced Atlantic Hadley cell and a positive phase North Atlantic Oscillation-like anomaly pattern, together with concurring Ural blocking, that transports Atlantic ocean heat and atmospheric moisture toward the BKS and promotes sea-ice melting via intensified surface warming. The reverse is seen during El Niño which leads to weakened Atlantic poleward transport and an increase in the BKS SIC. Thus, interannual variability of the BKS SIC partly originates from ENSO via the Atlantic pathway.