ABSTRACT
In optical remote sensing image object detection, discontinuous boundaries often limit detection accuracy, particularly at high Intersection over Union (IoU) thresholds. This paper addresses this issue by proposing the Spatial Adaptive Angle-Aware (SA3) Network. The SA3 Network employs a hierarchical refinement approach, consisting of coarse regression, fine regression, and precise tuning, to optimize the angle parameters of rotated bounding boxes. It adapts to specific task scenarios using either class-aware or class-agnostic strategies. Experimental results demonstrate its effectiveness in significantly improving detection accuracy at high IoU thresholds. Additionally, we introduce a Gaussian transform-based IoU factor during angle regression loss calculation, leading to the development of Edge-aware Skewed Bounding Box Loss (EAS Loss). The EAS loss enhances the loss gradient at the final stage of angle regression for bounding boxes, addressing the challenge of further learning when the predicted box angle closely aligns with the real target box angle. This results in increased training efficiency and better alignment between training and evaluation metrics. Experimental results show that the proposed method substantially enhances the detection accuracy of ReDet and ReBiDet models. The SA3 Network and EAS loss not only elevate the mAP of the ReBiDet model on DOTA-v1.5 to 78.85% but also effectively improve the model's mAP under high IoU threshold conditions.
ABSTRACT
Signal transduction processes in living organisms are mainly transmitted through conformational changes in transmembrane protein receptors. So far, the development of signal transduction models induced by artificial simulation of conformational changes remains limited. We herein report a new artificial receptor that achieves controllable "ON/OFF" signal transduction through conformational changes between the folding and unfolding of a transmembrane foldamer moiety. The receptor contains three functional modules: a lipid-anchored cholic acid headgroup, a foldamer transmembrane moiety, and a precatalyst tailgroup. After inserting in the lipid membrane, the addition of Zn2+ induces unfolding of the foldamer, which changes the molecular conformation and activates the tailgroup to enter the cavity to perform its catalytic task, resulting in signal transduction in an "ON" state. By further adding a competitive ligand to bind Zn2+, the transduction can be turned "OFF". External signals can be used to reversibly switch intravesicular catalysis on and off, which provides a new model for constructing artificial signal transduction systems.
Subject(s)
Receptors, Artificial , Signal Transduction , Molecular Conformation , Membrane Proteins , Lipids , Protein ConformationABSTRACT
The design of artificial ion channels with high activity, selectivity and gating function is challenging. Herein, we designed the light-driven motor molecule MC2, which provides new design criteria to overcome these challenges. MC2 forms a selective K+ channel through a single molecular transmembrane mechanism, and the light-driven rotary motion significantly accelerates ion transport, which endows the irradiated motor molecule with excellent cytotoxicity and cancer cell selectivity. Mechanistic studies reveal that the rotary motion of MC2 promotes K+ efflux, generates reactive oxygen species and eventually activates caspase-3-dependent apoptosis in cancer cells. Combined with the spatiotemporally controllable advantages of light, we believe this strategy can be exploited in the structural design and application of next-generation synthetic cation transporters for the treatment of cancer and other diseases.
Subject(s)
Apoptosis , Neoplasms , Ion TransportABSTRACT
Peptide-based supramolecular hydrogels have attracted great attention due to their good biocompatibility and biodegradability and have become promising candidates for biomedical applications. The bottom-up self-assembly endows the peptides with a highly ordered secondary structure, which has proven to be an effective strategy to improve the mechanical properties of hydrogels through strong physical interactions and energy dissipation. Inspired by the excellent mechanical properties of spider-silk, which can be attributed to the rich ß-sheet crystal formation by the hydrophobic peptide fragment, a hydrophobic peptide (HP) that can form a ß-sheet assembly was designed and introduced into a poly(vinyl alcohol) (PVA) scaffold to improve mechanical properties of hydrogels by the cooperative intermolecular physical interactions. Compared with hydrogels without peptide grafting (P-HP0), the strong ß-sheet self-assembly domain endows the hybrid hydrogels (P-HP20, P-HP29, and P-HP37) with high strength and toughness. The fracture tensile strength increased from 0.3 to 2.1 MPa (7 times), the toughness increased from 0.4 to 21.6 MJ m-3 (54 times), and the compressive strength increased from 0.33 to 10.43 MPa (31 times) at 75% strain. Moreover, the hybrid hydrogels are enzymatically degradable due to the dominant contribution of the ß-sheet assembly for network cross-linking. Combining the good biocompatibility and sustained drug release of the constructed hydrogels, this hydrophobic ß-sheet peptide represents a promising candidate for the rational design of hydrogels for biomedical applications.
Subject(s)
Hydrogels , Silk , Hydrophobic and Hydrophilic Interactions , Peptides , Protein Conformation, beta-StrandABSTRACT
Inspired by the design criteria of heteroditopic receptors for ion-pair binding, we herein describe a new strategy to construct a rotaxane transporter (RR[2]) for K+/Cl- co-transport. The use of a rigid axle improves the transport activity with an EC50 value of 0.58 µM, presenting a significant step toward developing rotaxane artificial channels.
Subject(s)
Rotaxanes , Rotaxanes/chemistry , Membrane Transport Proteins , LipidsABSTRACT
Cardiovascular disease is one of the leading threats to human lives and its fatality rate still rises gradually year by year. Driven by the development of advanced information technologies, such as big data, cloud computing, and artificial intelligence, remote/distributed cardiac healthcare is presenting a promising future. The traditional dynamic cardiac health monitoring method based on electrocardiogram (ECG) signals only has obvious deficiencies in comfortableness, informativeness, and accuracy under motion state. Therefore, a non-contact, compact, wearable, synchronous ECG and seismocardiogram (SCG) measuring system, based on a pair of capacitance coupling electrodes with ultra-high input impedance, and a high-resolution accelerometer were developed in this work, which can collect the ECG and SCG signals at the same point simultaneously through the multi-layer cloth. Meanwhile, the driven right leg electrode for ECG measurement is replaced by the AgCl fabric sewn to the outside of the cloth for realizing the total gel-free ECG measurement. Besides, synchronous ECG and SCG signals at multiple points on the chest surface were measured, and the recommended measuring points were given by their amplitude characteristics and the timing sequence correspondence analysis. Finally, the empirical mode decomposition algorithm was used to adaptively filter the motion artifacts within the ECG and SCG signals for measuring performance enhancement under motion states. The results demonstrate that the proposed non-contact, wearable cardiac health monitoring system can effectively collect ECG and SCG synchronously under various measuring situations.
Subject(s)
Artificial Intelligence , Wearable Electronic Devices , Humans , Signal Processing, Computer-Assisted , Electrocardiography/methods , HeartABSTRACT
Light signal transduction pathways are the central components of mechanisms that regulate plant development, in which photoreceptors receive light and participate in light signal transduction. Chemical systems can be designed to mimic these biological processes that have potential applications in smart sensing, drug delivery and synthetic biology. Here, we synthesized a series of simple photoresponsive molecules for use as photoreceptors in artificial light signal transduction. The hydrophobic structures of these molecules facilitate their insertion into vesicular lipid bilayers, and reversible photoisomerization initiates the reciprocating translocation of molecules in the membrane, thus activating or deactivating the hydrolysis reaction of a precatalyst in the transducer for an encapsulated substrate, resulting in a light controllable output signal. This study represents the first example of using simplified synthetic molecules to simulate light signal transduction performed by complex biomolecules.