4^th June 2026

Ingested Plastics Directly Linked to Chronic Kidney Disease

Ingesting common plastics is directly linked to chronic kidney disease , while other microplastics accumulate in the brain to drive inflammation . Inside the body, an imbalance in gut bacteria accelerates the spread of breast cancer , and obesity worsens Alzheimer's decline by disrupting the gut-brain connection . In technology, a new artificial intelligence model trained on over a million X-rays expertly detects bone fractures . Meanwhile, engineers created a smart fabric that tracks body posture using half the usual data , a pain-relief neck patch that operates without triggering unwanted nerve signals , and a copper-based material to efficiently generate clean energy from seawater .

Support our mission to make science easy to navigate

Top 10 topics by publication and citation volume

Advanced Sensor and Energy Harvesting Materials22

Electrocatalysts for Energy Conversion20

Artificial Intelligence in Healthcare and Education19

Ethics and Social Impacts of AI18

Microplastics and Plastic Pollution18

Perovskite Materials and Applications17

Gut microbiota and health17

Metal-Organic Frameworks: Synthesis and Applications16

Corporate Social Responsibility Reporting16

Advanced Battery Materials and Technologies15

Extended Breakdown↓

The rapid evolution of the global scientific landscape is marked by a transition from broad, observational studies to precise, molecular-level mechanisms and highly optimized, resource-efficient technologies. Across fields as diverse as wearable electronics, electrocatalysis, artificial intelligence, environmental toxicology, and energy storage, researchers are delivering breakthroughs that bridge theoretical models with practical, real-world applications.

Advanced Sensors: Communication-Efficient and Biocompatible Wearables

In wearable technology, the focus has shifted from merely increasing sensor density to optimizing communication efficiency and ensuring absolute biosafety. Researchers have developed a monolithic strain-proximity-pressure trimodal flexible sensor [P9] that utilizes physics-informed wearable perception to distinguish material softness from structural compliance, enabling noncontact respiration and muscle dynamics monitoring. Addressing the high power and bandwidth demands of conventional sensor arrays, a novel intelligent dual-mode textile sensing system achieves an ultra-broad linear range of up to 800 kPa, making it ideal for plantar pressure mapping. By utilizing a one-dimensional convolutional neural network, this system classifies 14 distinct body postures with 98.45% accuracy even when the sampling rate is halved, significantly reducing data transmission overhead. Furthermore, wearable interfaces are becoming truly imperceptible; a hydrogel electronic skin designed for neck pain management integrates cervical strain sensors and sEMG arrays. Crucially, this platform underwent the first multimodal biosafety validation using fNIRS-based cortical pain mapping, confirming zero neural activation during wear.

Electrocatalysis: Molecular Reconstruction and Machine Learning Design

Green energy conversion is benefiting from advanced surface engineering and atomistic modeling. To accelerate the sluggish oxygen evolution reaction (OER) in water splitting, a layered double perovskite heterostructure [P2] has been engineered with strong interfacial interactions that facilitate electron transfer and inhibit amorphization, achieving exceptional stability over 1,000 hours. For seawater electrolysis, where chloride corrosion is a major hurdle, a copper-doped amorphous high-entropy hydroxide triggers the lattice oxygen mechanism (LOM) and creates a chromate re-adsorption layer that repels chloride ions, delivering superior durability. Simultaneously, the design of these catalysts is being revolutionized by artificial intelligence. By integrating quantum mechanical calculations with a neural network potential (NNP), researchers simulated real-sized (5 nm) PtCo nanoparticles containing thousands of atoms under realistic conditions [P92], successfully predicting catalyst degradation and ORR activity.

Artificial Intelligence: Musculoskeletal Foundations, Cognitive Biases, and Governance

In healthcare AI, the transition toward specialized foundation models is accelerating. The introduction of SKELEX , a large-scale vision foundation model trained on 1.2 million musculoskeletal radiographs using self-supervised learning, marks a major milestone. SKELEX outperforms traditional task-specific models in fracture detection, osteoarthritis grading, and bone tumor classification, while enabling unsupervised anomaly localization. However, as AI becomes more integrated into clinical and high-stakes decision-making, its cognitive limitations are being exposed. A systematic quantitative assessment of eight state-of-the-art LLMs [P28] revealed that framing and primacy/recency biases are near-universal, with prompt-induced uncertainty playing a dominant role in model behavior.

This has spurred a parallel movement to reclaim data privacy and formalize AI governance. In mental health, researchers are calling for explicit opt-in consent and patient-led governance to prevent the unauthorized training of AI models on therapy transcripts [P130]. To govern the rising trend of human-directed AI development (where software is authored via natural language instructions), the Human-AI Co-Authorship Protocol (HAICA) [P75] has been proposed as an operational standard to manage prompt lineage and ensure century-scale durability. Additionally, researchers have identified a "Computational Legitimacy Gap" [P150], arguing that existing infrastructure categories cannot authoritatively verify computational decisions, and proposing a new category dedicated to "legitimacy replay verification."

Environmental Health: Microplastics and Molecular Pathogenesis

While previous weeks highlighted the general distribution of microplastics, recent studies have uncovered the precise molecular pathways of their toxicity. A systematic review of emerging evidence confirms that micro- and nanoplastics (MNPs), particularly polyethylene and polypropylene, accumulate in the human central nervous system, especially in individuals with compromised blood-brain barrier integrity, linking exposure to neuroinflammation. In the kidneys, polystyrene microplastics have been shown to activate noncanonical TGF-β signaling and induce metabolomic reprogramming . This pathway promotes epithelial-mesenchymal transition (EMT) and early-onset renal fibrosis, demonstrating a direct link between plastic ingestion and chronic kidney disease.

Perovskites and the Gut-Brain Axis

In materials science, lead-free double perovskites are achieving unprecedented stability. By utilizing a "dual-anchor molecular lock" [P45], researchers quenched ion migration in a 3D lead-free perovskite, achieving an ultralow X-ray detection limit that is 2,500 times below the medical diagnostic threshold. Meanwhile, hybrid-processed wide-bandgap perovskite solar cells are overcoming their long-standing open-circuit voltage deficit [P119] through crystallization control and bifacial passivation, opening new pathways for high-efficiency tandem photovoltaics.

In biology, the gut-brain-organ axes continue to reveal complex systemic feedback loops. Commensal dysbiosis has been shown to alter primary bile acid signaling, driving mammary gland inflammation and promoting breast tumor dissemination via a PGE2-dependent pathway . Furthermore, a study on the gut-brain axis in Alzheimer's patients with high BMIs demonstrated that obesity accelerates gut dysbiosis, which in turn disrupts the blood-brain barrier and enhances neuroinflammation, accelerating cognitive decline.

Advanced Batteries and Porous Materials

Finally, energy storage technologies are tackling interfacial instability. A comprehensive review of anode-free all-solid-state batteries [P13] outlines the chemical and mechanical vulnerabilities of sulfide electrolytes, emphasizing the need for active pressure management. Addressing these challenges, researchers utilized the thermal fusion of a lithium-philic fluorinated compound [P161] to construct an artificial solid electrolyte interface, suppressing dendrite growth and enabling the first Ah-level fluoride conversion pouch cells. In porous materials, a bio-inspired zirconium supramolecular framework [P22] demonstrates that hierarchical pore engineering and extrinsic packing pores are indispensable for high-performance ethane/ethylene separation. Simultaneously, in corporate governance, game-theoretic modeling of the "ESG Rating Game" [P152] explains how rating agencies strategically deviate from environmental fundamentals to capture market share, highlighting the persistent challenge of greenwashing.

Selected Papers
* 5
* 6
* 3
* 8
* P45
* 4
* P75
* 7
* P92
* 1
* P150
* P161