Browse Articles
Discover research articles across all indexed journals
Regional-scale forest aboveground biomass mapping using temporally consistent ICESat-2, Landsat, and field inventory data
Spatially continuous and accurate estimation of forest aboveground biomass (AGB) is essential for understanding carbon storage, ecosystem health, and biodiversity. Forests of the southeastern United States (US) represent about 40% of the nation’s forest area and one of the most significant carbon sequestration and storage potentials in the US. The availability of data from more recent and long-standing Earth-observing missions, like spaceborne light detection and ranging data from NASA’s Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) and imagery from Landsat satellites, present an exemplary opportunity to characterize vegetation structure and AGB. Despite this potential, the extent to which data from these ongoing missions can be used synergistically for AGB estimation at the regional scale is not well known. This study served to better understand the combined utility of Landsat and ICESat-2 for developing a large-area AGB mapping framework. Specifically, this work served to: (1) determine the best modeling technique for estimating field-derived AGB using ICESat-2 and Landsat-derived variables, among machine learning (random forest (RF) and support vector machine (SVM)) and geostatistical approaches (random forest regression kriging (RFRK) and support vector machine regression kriging (SVMRK)), and (2) create a high-resolution (30 m) baseline AGB map for the year 2020 across ~254,266 km² of forests of the southeastern US. Canopy height information from ICESat-2, Landsat-8 imagery and imagery-derived variables, digital elevation models, and canopy cover were used to model AGB. Resulting models yielded R2 values ranging from 0.34 to 0.61, and RMSEs between 22 and 31 Mg/ha. Evidently, AGB estimated using the SVMRK model was substantially better than the other models (R2 = 0.61 and RMSE = 23.99 Mg/ha), highlighting its potential for broad-scale AGB mapping. Overall, this work highlights a feasible approach for deriving spatially comprehensive AGB information for southeastern US forests and provides a high-resolution AGB baseline product to support regional-scale monitoring.
Correction: The prognostic impact of monocyte fluorescence, immunosuppressive monocytes and peripheral blood immune cell numbers in HIV-associated Diffuse Large B-cell Lymphoma
Point-of-care creatinine-based eGFR (StatSensor) in detecting kidney dysfunction (KD) among people living with HIV in Tanzania
Introduction Kidney Dysfunction (KD) is prevalent among people living with HIV (PLHIV) in low- and middle-income countries (LMICs), but routine screening is limited due to inadequate laboratory infrastructure. The StatSensor® Point-of-Care (POC) Creatinine Test offers a rapid, cost-effective alternative for early KD detection, though its accuracy in PLHIV remains uncertain. Methods We conducted a diagnostic accuracy cross-sectional study at Temeke Regional Referral Hospital (TRRH) HIV Clinic from January to March 2025 among PLHIV aged ≥18 years. Kidney dysfunction (KD) was defined as an estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m² using the CKD-EPI 2021 equation. We compared StatSensor point-of-care eGFR results with eGFR derived from serum creatinine measured by the Jaffe method. Diagnostic performance metrics including sensitivity, specificity, predictive values, and receiver operating characteristic (ROC) curves were reported. Results Among 358 participants, the median age was 48 years, with 66.2% female and 15.6% having KD (eGFR < 60 mL/min/1.73m²). The StatSensor demonstrated 92.9% sensitivity, 94.7% specificity, and 94.4% overall diagnostic accuracy compared to the Jaffe method. The ROC curve (AUC = 0.938) indicated strong test performance, showing substantial agreement with a kappa value of 0.805. Bland-Altman analysis revealed a negative bias of 4.36 mL/min/1.73 m² with limits of agreement from −19.68 to 28.40 and a strong correlation (R² = 0.813) between the two methods. Conclusion The StatSensor POC Creatinine test demonstrated high diagnostic accuracy and strong agreement with the standard Jaffe method, indicating its potential as a reliable screening tool for kidney dysfunction in PLHIV in resource-limited settings.
An innovative synthesis approach for Na A. Zeolite: A new pathway for enhanced performance, wastewater treatment, and antibacterial applications
Wastewater treatment is essential for protecting water resources and public health. Zeolite-based adsorbents offer an effective and sustainable solution for this purpose, providing high selectivity and regeneration potential. Zeolites are inorganic, highly crystalline, micro-porous materials composed of aluminotecto-silicates (SiO₄ and AlO₄ tetrahedral). Synthetic zeolites are commercially favored over natural ones due to their higher purity, crystallinity, and uniform pore size. Na A. zeolite (NaAZ) is a type of synthetic zeolite widely used in various applications, including wastewater treatment, due to its excellent adsorption and ion-exchange properties. This study focus on synthesize zeolite A from meta-kaolinite using a wet chemical method. The synthesis involves a hydrothermal process in which chemical reagents are mixed in an aqueous medium and heated under controlled conditions. The resulting (NaAZ) was characterized using Brunauer-Emmett-Teller (BET) surface area analysis, Fourier Transform Infrared Spectroscopy (FT-IR), X-ray Diffraction (XRD), Scanning Electron Microscopy (SEM), and Energy Dispersive X-ray Spectroscopy (EDX). This study evaluates the synthesized (NaAZ) for the removal of chemical oxygen demand from synthetic wastewater. Various parameters affecting adsorption such as contact time pH, temperature, and adsorbent dosage were investigated. The optimized conditions were then applied to real wastewater, and the material was further tested for its antitoxic and antibacterial properties. The in vitro antibacterial activity of NaAZ was assessed against both Gram-positive bacteria (Bacillus subtilis ATCC 6633, Staphylococcus aureus ATCC 6538, Enterococcus faecalis ATCC 19433) and Gram-negative bacteria (Escherichia coli ATCC 25922, Enterobacter aerogenes ATCC 13048, Pseudomonas aeruginosa ATCC 15442). Under optimal conditions contact time (40 min), pH (6–7), and adsorbent dosage (0.25 g) the removal efficiencies for COD, TSS, TKN, and PO₄3⁻ were 90.69%, 90.41%, 73.75%, and 68.85%, respectively.
Mechanics of knee meniscus results from precise balance between material microstructure and synovial fluid viscosity
The meniscus plays a crucial role in the biomechanics of the knee, serving as load transmitter and reducing friction between joints. Understanding the biomechanics of the meniscus is essential to effective treatment of knee injuries and degenerative conditions. This study aims to elucidate the relationship between the porous microstructure of the human knee meniscus and its biomechanical function, specifically focusing on fluid dynamics at the pore scale. Here, we use two central-meniscus samples extracted from a human knee and reconstruct high-resolution geometry models from μ -CT scans. By eroding the channels of the original meniscus geometry, we simulate perturbed microstructures with varying porosities ( 53% to 80% ), whilst preserving the connectivity of the porous structure. We numerically solve for the fluid dynamics in the meniscus using a mesh-free particle method, considering various inlet pressure conditions, characterising the fluid flow within the microstructures. The results of the original microstructure associated with a physiological dynamic viscosity of synovial fluid are in accordance with biophysical experiments on menisci. Furthermore, the eroded microstructure with a 33% increase in porosity exhibited a remarkable 120% increase in flow velocity. This emphasises the sensitivity of meniscus physiology to the porous microstructure, showing that detailed computational models can explore physiological and pathological conditions, advancing further knee biomechanics research.
Assessing maternal and newborn health readiness: Insights from a service availability assessment in five provinces in Laos
Background Global maternal mortality rates have declined significantly over the past two decades, including an 80% reduction in Laos since 2000. Effective management of obstetric complications – a major contributor to maternal deaths – requires well-staffed facilities equipped with essential supplies, medicines, and infrastructure. Despite progress, Laos still faces gaps in service availability and readiness limiting further reductions in preventable maternal mortality. Objective This analysis aimed to assess the service availability and readiness of public health facilities in five provinces of Laos to deliver maternal and newborn healthcare, including basic emergency obstetric and newborn care services (BEmONC). Methods A cross-sectional survey was conducted In October-November 2023 across 212 health centers and 20 district hospitals under the Laos Maternal Child Health and Nutrition project. Descriptive analysis was used to analyze the data. Service availability was measured based on the number of facilities and beds relative to the population. Service readiness was measured across three domains: guidelines and trained staff, essential equipment and supplies, and essential medicines. A composite readiness score was calculated as the mean across these domains. BEmONC availability was assessed using the presence of seven signal functions. Findings The overall service availability score was 71.3% across all provinces. The antenatal care readiness score across both facility types was 69.0%, CI95:62.7–75.3%, with district hospitals scoring higher than health centers, at 77.1%, CI95%:75.4–78.8%, compared to 68.2%, CI95%: 62.1–4.3%. The mean readiness score for delivery and newborn care was 68.3%, CI95%: 59.5–77.1%, with district hospitals again performing better at 81.3%, CI95%:79.7–82.9% compared to 67.0%, CI95%:58.2–75.8% for health centers. Conclusion Critical gaps in maternal and newborn health services remain, particularly in health centers. Investments in staffing, infrastructure, and availability of equipment and medicines is essential to address current gaps, improve service readiness, and contribute to improved quality of care and health outcomes.
BirdNET can be as good as experts for acoustic bird monitoring in a European city
BirdNET has become a leading tool for recognising bird species in audio recordings. However, its applicability in ecological research has been questioned over the sometimes large number of species falsely identified. Using species-specific confidence thresholds has been identified as a powerful approach to solving this issue. However, determining these thresholds is time and resource-consuming. While optimising the parameter setting of the algorithm could be an alternative strategy, the effect of parameter settings on the algorithm’s performance is not well understood. Here, we compared the species identification of BirdNET against expert identification using an acoustic dataset from a single site in Munich, Germany. The performance of BirdNET was evaluated using three performance metrics: precision, recall, and F1-score, using 24 combinations of the parameters: week, sensitivity, and overlap at four temporal aggregations (pooling of data across time intervals). We found that BirdNET performance varied widely depending on parameter settings (0.46–0.84). When given more data (higher temporal aggregation) and with tuned parameters, BirdNET came close to matching the expert identification (F1 score = 0.84). While BirdNET missed five species of the 23 species identified by the experts, our confirmation test revealed that BirdNET also found one species missed by the experts. To understand how each parameter affects F1 score, we trained linear mixed effects models. Our models showed that the confidence threshold had the strongest effect on the F1 score (p < 0.001) and significantly interacted with temporal aggregation, sensitivity, and overlap. Our results showed that while there are still limitations, using appropriate parameter settings, aggregating results over longer periods and undertaking some basic validation, BirdNET can yield results comparable to experts without the need for time-consuming estimation of species-specific thresholds.
Scaling up coral spawn collection: Impacts of method and timing on Acropora valida larval quality
Scaling up coral reef restoration to ecologically relevant scales presents a significant challenge during propagule collection. Mass coral spawning events are a vast source of propagules for reef restoration, but these events are typically limited to a few nights annually. Various methods of spawn collection following spawning events are available, ranging from traditional small-scale collection to industrial large-scale collection. However, comparisons between methods and potential effects on larval integrity are poorly understood. In this laboratory-based study, different methods of spawn collection – buckets, nets, and diaphragm pumping – were tested at various time points following spawning to explore potential impacts on embryo integrity, larval size, rate of deformities, and larval settlement. Results indicated that the collection method and, especially, the timing of collection, were critical. While bucket collection had minimal impact on embryo integrity, net and pump techniques caused high embryo fragmentation (>45%) at 5–11 hours post-fertilisation when embryos were >8 cells. This significantly reduced the average size of developing larvae in net and pump collections 3–11 hours post-fertilisation. When collections took place within the first hour of fertilisation before embryo cleavage, using any collection method resulted in minimal fragmentation (<4%). In general, net samples had larger larvae than pumped samples. However, larger larvae appeared to be more prone to deformities, and deformed larvae exhibited reduced settlement success (4% deformed vs 25% intact). These results highlight how large-scale spawn collections can be conducted without compromising larval quality when timed carefully, offering practical guidance for scaling coral reef restoration efforts.
Characterizing centrality: Obsidian consumption, supra-regional connectivity, and social reproduction at the Early Bronze Age sanctuary of Keros (Cyclades, Greece)
Early Bronze Age [EBA] Keros was a central place in the 3rd millennium cal BC Cycladic islands (Greece). Its material culture attests links with communities throughout the Aegean and beyond. This study uses obsidian sourcing to help reconstruct the socio-economic networks that coalesced at the site. Some 207 artifacts were elementally characterized using portable x-ray fluorescence spectroscopy [pXRF], the material coming from two ritual deposits in the Kavos area (n = 103), and the opposite islet settlement of Dhaskalio (n = 104). The results are consonant with the cosmopolitan character of Keros’ ceramic and metallurgical assemblages with not only the expected Melian sources of Dhemenegaki and Sta Nychia represented, but also handfuls of much rarer material from Giali A in the Dodecanese and East Göllü Dağ in central Anatolia. The study also provides further evidence for a Cycladic and Cretan preference for Sta Nychia raw materials in the EBA. A more complex picture of Melian obsidian consumption locally and regionally is then produced by integrating the sourcing data with the artifacts’ techno-typological and metrical attributes, which enables us to detail several EBA cultural traditions or ‘communities of practice’ across the Aegean region. The small quantities of Giali A and East Göllü Dağ obsidian are testimony to the supra-regional networks that coalesced at the site. Both raw materials likely circulated alongside the flow of Anatolian metals into the Aegean (including tin and gold), a network that introduced socially significant media and knowledge to Keros from as far east as the Indus. This congregation of people, resources, and technical know-how on Keros formed a key mode of social reproduction in Cycladic society with the mortuary and commemorative rituals on Kavos and the commensal gatherings on Dhaskalio comprising important spaces for the initiation, maintenance, and celebration of social relations.
Evaluating tree biomass estimation in trans-Atlantic mangrove species: Comparing bole diameter measurements for improved accuracy
Estimating the biomass of terrestrial forests generally, and mangrove forests in particular, is an area of considerable interest. Most approaches rely on empirically derived allometric models to predict tree biomass. The single parameter with the strongest predictive ability in most studies is diameter at breast height (DBH); however the use of DBH arose primarily out of convenience, not from an analysis of tree form. While DBH explains a lot of variability in other tree metrics such as height or above ground biomass, its utility in smaller species is uncertain. Here we used measurements from 302 destructively sampled mangrove trees of four species to test which of three bole diameter measurements, basal stem diameter (BSD), diameter at 30 cm (D30), and DBH, is the best predictor of aboveground biomass. D30 had the highest mean coefficient of determination ( R 2 ) and lowest mean root mean squared error (RMSE) across all site/species combinations. However, the improvement over DBH was modest, with a mean across all site/species combinations of 1.58 kg RMSE and R 2 of 0.948 for D30, compared to 1.63 kg RMSE and R 2 of 0.917 for DBH. Nevertheless, D30 may have utility in future studies as it allows for lower size thresholds and has better overall explanatory power than DBH.
Prevalence of asymptomatic non-falciparum and falciparum malaria in the 2014-15 Rwanda Demographic Health Survey
Background Recent molecular surveillance suggests an unexpectedly high prevalence of non-falciparum malaria in Africa. Malaria control is also challenged by undetected asymptomatic P. falciparum malaria resulting in an undetectable reservoir for potential transmission. Context-specific surveillance of asymptomatic P. falciparum and non-falciparum species is needed to properly inform malaria control programs. Methods We performed quantitative real time PCR for four malaria species in 5,050 primarily adult individuals in Rwanda using the 2014–2015 Demographic Health Survey. We assessed correlates of infection by species to explore attributes associated with each species. Asymptomatic P. ovale spp., P. malariae, and P. falciparum malaria infection had broad spatial distribution across Rwanda. P. vivax infection was rare. Results Overall infection prevalence was 22.3% (95%CI [20.3, 24.3]), with P. falciparum and non-falciparum at 16.3% [14.5, 18.1] and 8.0% [6.6, 9.3], respectively. Parasitemias tended to be low and mixed species infections were common, especially where malaria transmission and overall prevalence was the highest. P. falciparum infection was associated with lower wealth, rural residence and low elevation. Fewer factors were significantly associated with non-falciparum malaria. Conclusions Asymptomatic non-falciparum malaria and P. falciparum malaria are common and widely distributed across Rwanda in adults. Continued molecular monitoring, preferably done by the national malaria control program, of Plasmodium diversity using routine survey samples is needed to strengthen malaria control.
Deep learning-based classification of peptide analytes from single-channel nanopore translocation events
Rapid and accurate detection of peptide biomarkers using nanopore biosensors is critical for disease diagnosis and other biomedical applications. Processing large, complex single-channel translocation data streams poses a significant challenge for peptide analyte classification. Here, we present a supervised deep learning data processing pipeline for peptide classification from translocation events. The first stage employs a convolutional and recurrent neural network, adapted from the Deep-Channel multi-channel classifier, to accurately classify raw current recordings into discrete conductance states, including partially blocked sub-conductance intermediates. The second stage, peptide classification, utilizes a novel branched input network with a temporal convolutional network for processing translocation event conductance state sequences and a dense network for incorporating computed event-level and global kinetic features. Using idealized simulated multi-state translocation data for seven peptides, we demonstrate high classification accuracy (0.9998 (±0.0006)) when global features are included alongside event-level features. For classifying mixture samples, where only event-level features are applicable, performance shows more modest accuracy (0.70 (±0.01)). Peptide mixture predictions showed reasonable accuracy (MAE 0.045–0.161), although misclassification resulted in false positives. Event stochasticity and the fact that some peptides possessed similar kinetic parameters posed challenging for event-level prediction. However, vote aggregation from translocation event streams achieves perfect 100% accuracy, when predicting pure peptide samples. This proof-of-concept study demonstrates a robust deep learning framework for nanopore peptide classification using simulated data, laying the groundwork for classifying peptides from complex mixtures using real experimental data with the anthrax toxin protective antigen nanopore.
Estimating the population size of gay, bisexual, and other men who have sex with men in four major provinces in Canada: A descriptive study using data from a population-based survey
Background Estimating the size of key populations is critical for effective research and policy development. We estimated the population size of gay, bisexual, and other men who have sex with men (GBM) based on different definitions and compared the demographic composition of the GBM and non-GBM populations in Canada. Methods This descriptive study used data from the 2015–2016 and 2019–2020 Canadian Community Health Survey (CCHS) cycles. We selected men aged 18–64 years who had valid responses to the sexual identity and sexual behaviour contents. We explored different combinations of the survey questions to estimate the size of the GBM population in Canada and conducted a separate analysis for Canada’s four most populous provinces, comparing sociodemographic characteristics. Results Using a definition of GBM combining sexual identity and behaviour (i.e., men who identify as gay or bisexual or who had sex with men in the last 12 months), the weighted proportion of GBM in the 2015–2016 cycle was 2.7% (95% Confidence Interval (CI) 1.9%–3.4%) in Alberta, 3.5% (95% CI 2.7%–4.4%) in British Columbia, 4.1% (95% CI 3.2%–4.9%) in Ontario, and 4.8% (95% CI 4.0%–5.7%) in Quebec. In the 2019–2020 cycle, the weighted proportion of GBM (i.e., men who identify as gay, bisexual or pansexual, or who had sex with men in the last 12 months) was 4.4% (95% CI 3.3%–5.4%) in British Columbia and 4.7% (95% CI 3.9%–5.4%) in Ontario. Overall, compared to non-GBM, GBM were more likely to be single/never married, have an annual household income of less than $30,000, live in medium and large population centres and have lower mean age. Conclusion Our estimates showed sexual orientation discordance in Canada. Our findings also suggested that the GBM population might be increasing over time.
A pandemic risk index to improve supply chains decision-making between US and Mexico: A COVID-19 case study
The Social Vulnerability Index (SVI) developed by the Centers for Disease Control and Prevention (CDC), has been widely used as a benchmark to measure the state of vulnerability of counties across the United States. The SVI is integrated using a simple aggregation methodology on a set of variables reflecting the region’s socioeconomic status, household characteristics, racial & ethnic minority status, and housing type/transportation. Due to its simple construction and inclusion of significant variables publicly available, the SVI has grown exponentially in popularity among organizations and government officials as a tool for decision-making, especially for resource allocation and for regional risk assessment. Furthermore, the COVID-19 pandemic brought a set of unprecedented challenges in the bi-national health between the United States and Mexico, particularly on the state of risk of supply chains. Since the North American Free Trade Agreement (NAFTA) became effective in 1994 and then renewed in 2020 as USMCA, Mexico has grown to be the biggest trading partner of the U.S., fast approaching a trade value of more than a trillion USD a year. For which conducting regional risk assessment following the SVI formulation can be a significant impact for multiple stakeholders and organizations. In this work, the formulation of the SVI is analyzed using a risk framework as a reference, to corroborate its applicability for decision-making, and to expand it to account for variables and processes impacting supply chains during the COVID-19 pandemic. This analysis shows that vulnerability is only one of three factors required to conduct risk assessment (i.e., hazards vulnerability, and consequences), needed to produce a baseline of reference to make informed decisions. A case study is also developed based on the use of the SVI during the COVID-19 pandemic for supply chains between the U.S. and Mexico, by introducing the formulation of a risk index that is compatible with the proposed risk framework. The first step to expand the SVI into a risk index for supply chains between U.S. and Mexico, was to reproduce the CDC methodology, followed by using an Empirical Cumulative Density Function (ECDF) aggregation methodology to justify it statistically, and then to illustrate its benefits and limitations when extended into a new risk index (accounting for the three required risk components). As a result, a bi-national risk index map is produced after harmonizing publicly available variables in the U.S. and Mexico, illustrating the potential to quantify the state of regional risk for supply chains and other path-dependent systems, and setting a reference to further improve it.
This is the world’s largest ‘mosquito factory’: its goal is to stop dengue
Taphonomy of aquatic insects from the Crato Formation Lagerstätte (Aptian, Lower Cretaceous) under an actualistic look
The Crato Formation (Aptian, Lower Cretaceous) is a fossiliferous deposit of global significance, representing a lacustrine palaeoenvironment which offers insights into aquatic insect taphonomy. Despite its importance, prior studies lacked an actualistic approach. Here, we analyze the preservation of mayflies (Ephemeroptera) and dragonflies (Odonata) from this formation using experimental taphonomy on 253 extant Ephemeroptera and 236 Odonata, alongside 306 fossil specimens. Disarticulation experiments showed that the thorax of modern mayfly larvae disarticulated first, yet Crato Hexagenitidae larvae retained intact thoraces, indicating minimal disturbance and autochthonous deposition. Fossil alate specimens rarely exhibited decay-related wing damage, aligning with short decay times. Dragonfly carcasses exhibited a characteristic leg posture in death, also preserved in Crato fossils, further suggesting minimal transport. Additionally, fossil dragonflies retained labial masks, the first structure to disarticulate experimentally, consistent with parautochthonous assemblages. Mayfly larvae exposed to low salinity during experiments exhibited excessive defecation before death, hinting at possible low salinity conditions in the Crato palaeoenvironment, though preservational challenges obscure confirmation. During experimentation, we also noticed that all carcasses immediately floated under hypersaline conditions, while carcasses immersed in non-hypersaline conditions went through slower decomposition. Thus, we can safely propose with experimental data that microbial biofilms on the surface of the water were acting during carcass sinking in this deposit.
Prefrontal cortex function and gait alterations during single- and dual-task walking in knee osteoarthritis
Over-recruitment of the prefrontal cortex (PFC) during complex walking conditions may reflect altered motor and cognitive performance in people with knee osteoarthritis (OA). Our objectives were (1) to assess PFC activation, and motor and cognitive performance, during single- and dual-task walking in people with knee OA and (2) to examine the association of PFC activation with the performance. Forty-eight people with symptomatic knee OA completed three tasks, (1) single-task walking (STW) (2) subtraction by 7 from a 3-digit number (S7), and (3) dual-task walking (DTW), a combination of STW and S7. Oxygenated hemoglobin concentration changes (ΔHbO2) in bilateral prefrontal cortex (PFC) were assessed using functional Near-Infrared Spectroscopy. Motor performance outcomes included gait speed, step duration variability, and stride length variability. Cognitive performance was assessed as the correct response rate during S7. We used repeated measures ANCOVA to compare the outcomes by tasks. Correlation and multiple linear regression analyses were used to determine the association between PFC activation and performance outcomes. PFC activation was higher during STW and DTW compared to S7 but not significantly different between STW and DTW. People with knee OA walked slower (d = 0.63) and had higher variability in step duration (d = 0.45) and stride length (d = 0.37) during DTW compared to STW. Greater activation in right ventrolateral PFC (R2 = 0.15) and left dorsomedial PFC (R2 = 0.12) were associated with lower step duration variability. When walking is challenged with a cognitive task, people with knee OA show deterioration of gait performance and no change in PFC activation.
Analog optical computer for AI inference and combinatorial optimization
Abstract Artificial intelligence (AI) and combinatorial optimization drive applications across science and industry, but their increasing energy demands challenge the sustainability of digital computing. Most unconventional computing systems1–7 target either AI or optimization workloads and rely on frequent, energy-intensive digital conversions, limiting efficiency. These systems also face application-hardware mismatches, whether handling memory-bottlenecked neural models, mapping real-world optimization problems or contending with inherent analog noise. Here we introduce an analog optical computer (AOC) that combines analog electronics and three-dimensional optics to accelerate AI inference and combinatorial optimization in a single platform. This dual-domain capability is enabled by a rapid fixed-point search, which avoids digital conversions and enhances noise robustness. With this fixed-point abstraction, the AOC implements emerging compute-bound neural models with recursive reasoning potential and realizes an advanced gradient-descent approach for expressive optimization. We demonstrate the benefits of co-designing the hardware and abstraction, echoing the co-evolution of digital accelerators and deep learning models, through four case studies: image classification, nonlinear regression, medical image reconstruction and financial transaction settlement. Built with scalable, consumer-grade technologies, the AOC paves a promising path for faster and sustainable computing. Its native support for iterative, compute-intensive models offers a scalable analog platform for fostering future innovation in AI and optimization.
Bronze Age make-up recipes from Sudanese Lower Nubia point to a greater diversity across cultural borders in ancient Northeast Africa
Previous scientific explorations of kohl and other make-up substances from ancient Egypt have revealed a considerable diversity of materials and recipes used in different regions and time periods. However, samples from Sudanese Nubia have never been included in scientific investigations of make-up substances used along the Nile valley. For the first time, 24 samples of kohl and other cosmetics from Bronze Age Sudanese Lower Nubia (c. 2055–1070 BCE) were analysed using optical microscopy, GC-MS, SEM-EDS, ATR-FTIR and XRD. Beyond expanding our knowledge of make-up usage in the ancient Nile valley by including samples from Sudan, this study adds further depth to our understanding of make-up substances in ancient Northeast Africa by exploring samples from well-defined archaeological contexts. The multi-analytical approach presented here sheds light on the diversity of recipes used by various communities in the Middle Nile valley during the Bronze Age. Most samples are dominated by lead sulphides, but these occur in various mixtures with quartz, clay, calcite, gypsum and zinc compounds, in addition to plant gums and animal fats. We also report for the first time the use of synthetic calcium antimonate in ancient cosmetic mixtures. Besides expanding our knowledge of make-up mixtures in ancient Northeast Africa, our study suggests that the considerable variation detected across the cultural borders of Bronze Age Egypt and Nubia reflects distinctive bodily ideals.