Workshop Name：1st Large Language Models for Spatial-rich Data Management (LLM+Spatial)

This study introduces a deep-learning system for rapid, automated detection and classification of tiny calcium deposits (microcalcifications) in mammograms to aid early breast cancer diagnosis. Leveraging a multi-center dataset of 4,810 biopsy-confirmed mammograms, our pipeline uses a Faster RCNN model with a feature-pyramid backbone to detect and classify microcalcifications—the pipeline requires no hand-tuned rules and provides both the overall cancer risk and highlighted lesion regions in seconds per image. On unseen test data, it achieved overall classification accuracy of 72% for discriminating between benign and malignant breasts and 78% sensitivity of malignant breast cancer prediction, marking a significant step toward AI-assisted, cost-effective breast-cancer screening that can run on standard radiology workstations.

Copper is an economic and strategic metal consumed and produced in many nations. Previous studies of future copper demand and supply have revealed some deficiencies at the macro level. To minimize these gaps, we present here a comprehensive recycling scenario to examine the future copper sustainability for more than 50 countries by 2100. End-of-life metal recycling optimization can be a transparent, flexible and broadly applicable solution, coupled with circular economy strategies to reduce the supply of such metals and related environmental risks. Substantial recycling (estimated at 40 to 65 percent) would be more conducive to reducing primary demand in regions with large populations and rapid industrial growth. Although Latin America and Caribbean, Africa and Oceania can be identified as regions with low demand and recycling intentions by 2100, carbon emissions in Latin America will peak at 70 million tons by 2050 owing to concentrated primary copper production. In order to fully realize recycling targets, more efforts should be made to formulate policies and regulations for the copper industry, systematically scrap collection, and innovate effective technologies.

Depression imposes a staggering global socioeconomic burden. Current pharmacotherapies face major limitations, including slow efficacy, adverse effects, and non-response rates of up to 55%, necessitating novel therapeutic modalities. This study introduces terahertz (THz) photoneuromodulation as an innovative physical intervention for depression, offering several advantages over conventional pharmacological or optogenetic approaches. Mild THz photoneuromodulation circumvents the need for exogenous agents or genetic modifications, mitigating potential risks while precisely modulating neurotransmitter levels and neuronal excitability to alleviate depression-like behaviors. In a chronic restraint stress (CRS) mouse model, THz photostimulation rapidly attenuated hyperactivity and increased serotonin levels by 107.5% ± 45.3% in lateral orbitofrontal cortex glutamatergic neurons (OFC^Glu) compared to those treated with antidepressants. This led to marked improvements in depressive-like behaviors and cognitive function. Furthermore, THz modulation of OFC activity recapitulated the effects of chemogenetic inhibition, underscoring the OFC's pivotal role in regulating depressive states. This research unveils THz photoneuromodulation as a promising, safe, rapid-acting, and durable neurotherapeutic strategy addressing persistent unmet needs in depression treatment.