Early detection of fruit fly (Diptera: Tephritidae) infestation in traded fruit is important for maintaining phytosanitary security, particularly for quarantine pests. This study evaluated the integration of micro-focus X-ray computed tomography (µCT) and deep learning as a non-destructive, automated, and scalable solution for postharvest pest detection. The Mediterranean fruit fly (medfly), Ceratitis capitata Wiedemann, 1824, was used as a model species and Navel oranges, Citrus sinensis (Osbeck), served as the host commodity. Controlled infestation experiments were conducted using both artificially punctured fruit to promote oviposition and intact fruit to avoid confounding effects associated with mechanical damage. Oranges were exposed to varying adult medfly densities to generate different infestation levels. Following exposure, the infested oranges were subjected to varying incubation periods to simulate different infestation stages. High-resolution µCT scans were first assessed manually and subsequently analyzed using a staged deep learning pipeline incorporating convolutional neural networks (CNNs), 3D CNNs, CNN-Long Short-Term Memory (LSTM) hybrids and Transformer-based models. Manual μCT inspection achieved mean correct classification rates of 0.84 in punctured fruit and 0.69 in intact fruit, with detection performance increasing with infestation stage. Deep learning analyses were conducted on a curated punctured-fruit dataset using a slice-level classification and voting-based aggregation strategy. Among the evaluated transfer-learning architectures, DenseNet121 achieved the best overall performance, reaching an accuracy of 0.88, a precision of 0.86, a sensitivity of 1.00, and an AUC of 0.80; because the manual and automated analyses were performed on different sets of fruit, these accuracies are complementary rather than directly comparable.Grad-CAM and attention visualizations confirmed that model predictions were driven by biologically meaningful tissue regions associated with larval damage. Although the deep learning workflow was developed and validated only on punctured fruit, the results demonstrate that μCT images contain sufficient information for automated detection of medfly infestation when infestation-associated damage is visible. These findings support the potential of combining μCT imaging and artificial intelligence (AI) for automated, reproducible and non-destructive postharvest pest detection and enhanced phytosanitary compliance in postharvest supply chains.
Kesterite Cu2ZnSn(S,Se)4 (CZTSSe) is an attractive thin-film absorber based on earth-abundant and low-toxicity elements. Nevertheless, high-performance CZTSSe solar cells still typically employ CdS buffer layers, which pose significant limitations due to the hazardous nature of Cd and introduce parasitic absorption at short wavelengths. The use of alternative buffer layers is therefore crucial for further advancing this technology. Among the possible candidates, ZnxSn1−xO (ZTO) has emerged as a promising option due to its wide, tunable bandgap and benign constituents. In this work, ZTO buffer layers were deposited by Atomic Layer Deposition (ALD) at three different temperatures (150, 120, 90 °C). However, the as-deposited ZTO films exhibited high resistivity, which was mainly attributed to the controversial role of defect-related intragap states. To overcome this limitation, complete photovoltaic (PV) devices were annealed in air at 260 °C and subsequently exposed to light soaking. These post-deposition treatments synergistically improved the electrical conductivity of ZTO and reduced trap-assisted recombination, thereby enhancing final PV performance. Optical characterisation revealed a widened bandgap compared to conventional CdS, effectively suppressing parasitic absorption in the blue spectral region and enhancing short-circuit current density (JSC) as expected. The best-performing devices were obtained with ZTO deposited at 90 °C, yielding efficiencies (η) exceeding 7% by optimising the trade-off between junction quality and ZTO conductivity. These experimental findings are corroborated by numerical SCAPS simulations and confirm the potential of ALD-grown ZTO as a Cd-free buffer layer for kesterite solar cells.
Acute respiratory failure and the acute respiratory distress syndrome (ARDS) are characterized by profound spatial and temporal heterogeneity in lung aeration, mechanics, and inflammation. While mechanical ventilation is often vital to ensure gas exchange, it can also cause or worsen lung injury as a result of the superimposed mechanical stress due to the application of positive pressure on a heterogeneous parenchyma. Characterizing the lung microstructure and function at small length scales is essential for understanding the pathogenesis of ventilator-induced lung injury. This is a challenging task due to the complex architecture of the lung, to its constant motion and deformation with breathing and pulsatile blood flow, and its multiscale organization. Indeed, mechanical forces act on the components of the extracellular matrix and cells, acini, airways and blood vessels at the microscale to impact numerous biological processes. Beyond these effects, the global mechanical behavior and function of the lung emerge from this complex dynamic system. Because synchrotron radiation imaging techniques such as phase-contrast computed tomography (PC-CT), and K-edge subtraction CT (KES-CT), have a high spatial resolution, are quantitative, and are fast, they offer unique insights into the multiscale, dynamic behavior of the lung. These modalities enable mapping of regional ventilation, blood distribution, within-breath alveolar recruitment/derecruitment, airway closure, and tissue strain. This review explains synchrotron radiation imaging modalities, and discusses the findings in models of acute respiratory failure, their translational implications, as well as future areas of investigation.
Green areas in urban and peri-urban landscapes are increasingly recognized as important habitats for wild bees, but the extent to which species adjust their reproductive behavior to local conditions in these environments remains poorly studied. We studied the reproductive output, offspring’s body size, sex allocation, and larval diet (assessed using pollen morphology and DNA metabarcoding) of the solitary bee Osmia caerulescens (Linnaeus, 1758) in a Mediterranean peri-urban environment using trap nests and local-scale environmental descriptors. Based on a total of 215 nests and 648 brood cells, females increased their reproductive output with greater floral resource availability and increasing openness of the habitat, but this was accompanied by a reduction in offspring body size (ITD) and a sex allocation skewed toward males. These findings are consistent with a trade-off between offspring number and per-offspring investment and contrary to general expectations that higher resource availability should favor larger and more female-biased offspring. Together, these responses indicate a shift toward quantity-oriented reproductive strategies, in contrast to patterns reported for this species in agricultural systems, where higher resource availability is typically associated with greater per-offspring investment and more female-biased sex ratios. Pronounced among-site variation in pollen diet, reflecting differences in local floral resource composition, further highlights how use of floral resources varies at a fine spatial scale and may shape reproductive strategies and, ultimately, population persistence. Our results show that the ecological value of urban habitats for solitary bees depends not only on the amount of floral resources available, but also on how local conditions influence species-specific reproductive responses. These findings highlight the value of species-specific reproductive analyses in abundant, functionally important solitary bees as a complementary approach to evaluating habitat quality at local scales in urban and peri-urban environments. Understanding how solitary bees adjust reproductive strategies to urban conditions is essential for supporting their conservation and long-term persistence in urban green spaces.
PURPOSE:The intellectual profile of monolingual children with Developmental Language Disorder is typically characterized by below-average verbal IQ (VIQ) and average Performance IQ (hence, PIQ) scores. Our knowledge of IQ functioning in bilingual children with Developmental Language Disorder is still very limited. Furthermore, previous research in IQ functioning in children with Developmental Language Disorder has not addressed maternal education as a possible alternative exploratory variable affecting children's performance in IQ tests. Here, we aimed to investigate if intellectual functioning of children with Developmental Language Disorder is affected by bilingualism, and whether this relation is affected by the children's socioeconomic characteristics, including maternal and parental education, and family income. METHODS:We focused on IQ profile comparisons between 125 bilingual children and 109 monolingual children with Developmental Language Disorder. RESULTS:We found that the bilingual children exhibited 'normalized' performance in those VIQ tests that tapped into metalinguistic knowledge and social understanding. The bilingual children were also more likely to exhibit average skills across PIQ tests. Finally, we found that the positive effect of bilingualism on children's IQ was only observed for the children whose mothers had a low educational level. CONCLUSION:The overall findings hold implications for the broader understanding of intellectual functioning in bilingual neurodiverse populations, also highlighting the unique role of maternal education in studying children's cognitive development.
