This review explores how multi-source data-driven machine learning is transforming lung cancer care, from early screening and diagnosis to personalized treatment and prognosis. It examines imaging, multi-omics, liquid biopsy, pathology, and clinical data, alongside traditional and advanced machine learning approaches. By linking data characteristics with model selection and clinical applications, the review highlights opportunities for precision medicine while addressing challenges in data standardization, multimodal integration, interpretability, and effective clinical translation.

Ocean Networks Canada (ONC) is co-engineering key infrastructure to create a deep-sea neutrino telescope that will transform vast stretches of deep ocean into a massive neutrino detector. The technology innovation will give scientists a glimpse into the universe from under the sea. 

The emergence of brain-inspired (neuromorphic) computing offers a promising route to overcome the limitations of conventional von Neumann architectures in artificial intelligence (AI). While traditional systems separate memory and computation—resulting in high energy consumption—the human brain integrates these functions efficiently within a compact structure. Addressing this gap, the research from Prof. HAN Suting, Associate Professor of Department of Chemistry at The Hong Kong Polytechnic University, focuses on memristor-based neuromorphic hardware, enabling AI systems that more closely emulate biological intelligence.

Researchers in the Riccio College of Engineering at the University of Massachusetts Amherst have demonstrated that redesigning both the hardware and the algorithm can make AI applications on edge devices more efficient. As proof of concept, their system achieved 95.24% accuracy in language identification while reducing computing resources by 90%—the highest reported accuracy from a system of its kind.

A comprehensive synthesis published in the South African Journal of Geology evaluates 25 years of research on the Karoo Supergroup, an exceptional ~120-million-year rock record spanning the Late Carboniferous to Early Jurassic. Integrating multi-proxy sedimentology, U-Pb geochronology, magnetostratigraphy, and palaeontology, the review evaluates long-standing basin models and refines the stratigraphic architecture across southern Gondwana.

The synthesised data challenge simplistic models of linear aridification, revealing non-linear palaeoclimate oscillations – including short-lived pluvial episodes in the Carnian (Late Triassic) and extensive lacustrine-erg interactions in the Pliensbachian (Early Jurassic). Furthermore, it spotlights the improved resolution of the Karoo biozonation, a core of studies on palaeo-ecosystem dynamics, and documents patterns of ecological collapse, persistence, and recovery across four major deep-time biocrises: the end-Capitanian, end-Permian, end-Triassic, and end-Pliensbachian. This synthesis establishes an updated framework to resolve intra-basin heterogeneity and optimise the correlation of Pan-Gondwanan tectonic-climatic events.

Researchers have developed a cellulose-based hydrogel sensor that maintains stable sensing performance in extreme cold using a zinc chloride and lithium bromide molten salt hydrate system, offering a sustainable and biodegradable platform for next-generation flexible wearable devices and human-machine interfaces. Published in Journal of Bioresources and Bioproducts, the study introduces a binary molten salt hydrate approach to dissolve natural cellulose efficiently within ten minutes. The resulting ionically conductive hydrogel achieves an impressive conductivity of 4.48 S/m and compressive strength of 2.48 MPa. Differential scanning calorimetry confirmed no water crystallization exothermic peaks from -80°C to 20°C, indicating exceptional anti-freezing capability. The material's shear-thinning properties further enable 3D printing of customized sensor geometries for human-machine interfaces.