Search Results (153)

The hospital-at-home (HaH) model offers hospital-level care within patients’ homes and has proven effective for managing conditions such as pneumonia. The point-of-care ultrasonography (PoCUS) is a key diagnostic tool in this model, especially when traditional imaging modalities are unavailable. This review explores how PoCUS can be optimized to manage pneumonia in HaH settings, focusing on its diagnostic accuracy in patients with comorbidities, differentiation from mimickers, and role in assessing disease severity. Pulmonary comorbidities, such as heart failure and interstitial lung disease (ILD), can complicate lung ultrasound (LUS) interpretation. In heart failure, combining lung, cardiac, and venous assessments (e.g., IVC collapsibility, VExUS score) improves diagnostic clarity. In ILD, distinguishing chronic changes from acute infections requires attention to B-line patterns and pleural abnormalities. PoCUS must differentiate pneumonia from conditions such as atelectasis, lung contusion, cryptogenic organizing pneumonia, eosinophilic pneumonia, and neoplastic lesions—many of which present with similar sonographic features. Serial LUS scoring provides useful information on pneumonia severity and disease progression. Studies, particularly during the COVID-19 pandemic, show correlations between worsening LUS scores and poor outcomes, including increased ventilator dependency and mortality. Furthermore, LUS scores correlate with inflammatory markers and gas exchange metrics, supporting their prognostic value. In conclusion, PoCUS in HaH care requires clinicians to integrate multi-organ ultrasound findings, clinical context, and serial monitoring to enhance diagnostic accuracy and patient outcomes. Mastery of LUS interpretation in complex scenarios is crucial to delivering personalized, high-quality care in the home setting. Full article

Traditional studies on Chinese Buddhism have largely relied on printed canons from the Song, Yuan, Ming, and Goryeo dynasties. However, recent discoveries of Dunhuang and Turfan manuscripts, along with growing recognition of Nihon kosha issaikyō (Japanese Ancient Manuscript Canons), have expanded the scope of Buddhist textual research. Despite their significance, Japanese manuscript Buddhist canons remain underexplored, particularly in relation to their textual lineages and connections to Tang-dynasty texts. This study examines Nihon kosha issaikyō through a philological analysis of the Shisong lü (Ten Recitation Vinaya), assessing textual variants, structural patterns, and transmission histories. By situating Nihon kosha issaikyō within the broader East Asian Buddhist tradition, this research clarifies their role in preserving alternative textual lineages beyond standardized printed canons. The findings contribute to a deeper understanding of Buddhist textual transmission, canon formation, and the interplay between manuscript and printed traditions in China, Korea, and Japan. This study highlights the historical processes that shaped East Asian Buddhist canons and offers new insights into their adaptation and preservation across different cultural contexts. Full article

The current work records for the first time the rare-metal pegmatites with mixed NYF-LCT located at Wadi Sikait, south Eastern Desert of the Egyptian Nubian Shield. Most of the Sikait pegmatites are associated with sheared granite and are surrounded by an alteration zone cross-cutting through greisen bodies. Sikait pegmatites show zoned and complex types, where the outer wall zones are highly mineralized (Nb, Ta, Y, Th, Hf, REE, U) than the barren cores. They consist essentially of K-feldspar, quartz, micas (muscovite, lepidolite, and zinnwaldite), and less albite. They contain a wide range of accessory minerals, including garnet, columbite, fergusonite-(Y), cassiterite, allanite, monazite, bastnaesite (Y, Ce, Nd), thorite, zircon, beryl, topaz, apatite, and Fe-Ti oxides. In the present work, the discovery of Li-bearing minerals for the first time in the Wadi Sikait pegmatite is highly significant. Sikait pegmatites are highly mineralized and yield higher maximum concentrations of several metals than the associated sheared granite. They are strongly enriched in Li (900–1791 ppm), Nb (1181–1771 ppm), Ta (138–191 ppm), Y (626–998 ppm), Hf (201–303 ppm), Th (413–685 ppm), Zr (2592–4429 ppm), U (224–699 ppm), and ∑REE (830–1711 ppm). The pegmatites and associated sheared granite represent highly differentiated peraluminous rocks that are typical of post-collisional rare-metal bearing granites. They show parallel chondrite-normalized REE patterns, enriched in HREE relative to LREE [(La/Lu)_n = 0.04–0.12] and strongly negative Eu anomalies [(Eu/Eu*) = 0.03–0.10]. The REE patterns show an M-type tetrad effect, usually observed in granites that are strongly differentiated and ascribed to hydrothermal fluid exchange. The pegmatite has mineralogical and geochemical characteristics of the mixed NYF-LCT family and shows non-CHARAC behavior due to a hydrothermal effect. Late-stage metasomatism processes caused redistribution, concentrated on the primary rare metals, and drove the development of greisen and quartz veins along the fracture systems. The genetic relationship between the Sikait pegmatite and the surrounding sheared granite was demonstrated by the similarities in their geochemical properties. The source magmas were mostly derived from the juvenile continental crust of the Nubian Shield through partial melting and subsequently subjected to a high fractional crystallization degree. During the late hydrothermal stage, the exsolution of F-rich fluids transported some elements and locally increased their concentrations to the economic grades. The investigated pegmatite and sheared granite should be considered as a potential resource to warrant exploration for REEs and other rare metals. Full article

Inkjet-printed circuits on flexible substrates are rapidly emerging as a key technology in flexible electronics, driven by their minimal fabrication process, cost-effectiveness, and environmental sustainability. Recent advancements in inkjet-printed devices and circuits have broadened their applications in both sensing and computing. Building on this progress, this work has developed a nonlinear computational element coined as mTanh to serve as an activation function in neural networks. Activation functions are essential in neural networks as they introduce nonlinearity, enabling machine learning models to capture complex patterns. However, widely used functions such as Tanh and sigmoid often suffer from the vanishing gradient problem, limiting the depth of neural networks. To address this, alternative functions like ReLU and Leaky ReLU have been explored, yet these also introduce challenges such as the dying ReLU issue, bias shifting, and noise sensitivity. The proposed mTanh activation function effectively mitigates the vanishing gradient problem, allowing for the development of deeper neural network architectures without compromising training efficiency. This study demonstrates the feasibility of mTanh as an activation function by integrating it into an Echo State Network to predict the Mackey–Glass time series signal. The results show that mTanh performs comparably to Tanh, ReLU, and Leaky ReLU in this task. Additionally, the vanishing gradient resistance of the mTanh function was evaluated by implementing it in a deep multi-layer perceptron model for Fashion MNIST image classification. The study indicates that mTanh enables the addition of 3–5 extra layers compared to Tanh and sigmoid, while exhibiting vanishing gradient resistance similar to ReLU. These results highlight the potential of mTanh as a promising activation function for deep learning models, particularly in flexible electronics applications. Full article

Overextraction of groundwater, as well as rapidly changing land use patterns, climatic change, and anthropogenic activities, in the densely populated Melur of Tamil Nadu state in India, has led to aquifer degradation. This study maps the groundwater potential (GWPZ) by evaluating 678 km² of this region in the Analytical Hierarchy Processes (AHP) and by using remote sensing and GIS tools as part of SDG 6 for the sustainable management of drinking, irrigation, and industrial uses for future generations. Data information layers, such as aquifer (a), topography (t), lineaments (l), land-use/land-cover (LuLc), soil (s), rainfall (r), and drainage (d) characteristics, separated the study area between poor and excellent groundwater potential zones with 361 km² or 53% of the study area remaining as low GWP and the prospective excellent groundwater potential zone covering only 9 km² (1.3% of total area). The integrated approach of the GWPZ and Water Quality Index (WQI) can effectively identify different zones based on their suitability for extraction and consumption for better understanding. This study also evaluates the performance of three machine learning models, such as Random Forest (RF), Gradient Boosting, and Support Vector Machine (SVM), based on a classification method using the same layers that govern the groundwater potential. The results indicate that both the RF model and Gradient Boosting achieved 100% accuracy, while SVM had a lower accuracy of 50%. Performance metrics such as precision, recall, and F1-score were analyzed to assess classification effectiveness. The findings highlight the importance of model selection, dataset size, and feature importance in achieving optimal classification performance. Results of this study highlight that the aquifer system of Melur has a low groundwater reserve, and it requires adequate water resource management strategies such as artificial recharge, pumping restriction, and implementation of groundwater tariffs for sustainability. Full article

The terrestrial spatial patterns were affected by human activities, primarily on regional land use (LU) changes, with habitat quality (HQ) serving as a prerequisite for achieving regional sustainable development. Assessing and predicting the spatiotemporal evolution characteristics of regional LU changes and HQ is critical for formulating regional LU strategies and enhancing ecosystem service functions. Using the Poyang Lake Region as our research object, this research employs LU data and utilizes the ‘InVEST’ model and hot-spot analysis to quantitatively evaluate the spatiotemporal changes in HQ during 2000–2020. The PLUS model is then applied to predict LU and HQ trends from 2020 to 2050. The findings are as follows: (1). From 2000 to 2020, the areas of forestland, shrubland, sparse woodland, paddy fields, and dryland in the Poyang Lake Region showed a decreasing trend, with reductions mainly occurring in urban expansion zones such as Nanchang City and largely converted into urban construction land. (2). Since 2000, HQ in the Poyang Lake Region has shown a slight retrogressive evolution, with significant spatial heterogeneity. HQ spatially exhibits a pattern of improvement radiating outward from major cities. (3). Predictions for 2030 to 2050 indicate that HQ in the Poyang Lake Region will continue to decline, with the most significant downward trends occurring in urban built-up areas and their peripheries. The spatiotemporal characteristics reveal an expansion ring around Poyang Lake and an east–west urban expansion corridor linking Pingxiang, Yichun, Xinyu, Nanchang, Fuzhou, Yingtan, and Shangrao. This study provided a research basis for LU direction and urban planning policies in the Poyang Lake Region and its surrounding areas, while also contributing to the construction of agrarian security patterns and the enhancement of ecosystem service levels in the region. Full article

The WOX (WUSCHEL-related homeobox) gene family plays pivotal roles in plant growth, development, and responses to biotic/abiotic stresses. Flax (Linum usitatissimum L.), a globally important oilseed and fiber crop, lacks a comprehensive characterization of its WOX family. Here, 18 LuWOX genes were systematically identified in the flax genome through bioinformatics analyses. Phylogenetic classification grouped these genes into three clades: Ancient, Intermediate, and WUS Clades, with members within the same clade exhibiting conserved exon–intron structures and motif compositions. Promoter analysis revealed abundant cis-acting elements associated with hormone responses (MeJA, abscisic acid) and abiotic stress adaptation (anaerobic induction, drought, low temperature). Segmental duplication events (nine gene pairs) contributed significantly to LuWOX family expansion. Protein–protein interaction networks implicated several LuWOX proteins in stress-responsive pathways. Expression profiling demonstrated that most LuWOX genes were highly expressed in 5-day-post-anthesis (DPA) flowers and embryonic tissues. qRT-PCR validation further uncovered distinct expression patterns of LuWOX genes under cold, drought, and salt stresses. This study established a foundational framework for leveraging LuWOX genes to enhance stress tolerance in flax breeding and functional genomics. Full article

Regression is a fundamental task in machine learning, and neural networks have been successfully employed in many applications to identify underlying regression patterns. However, they are often criticised for their lack of interpretability and commonly referred to as black-box models. Feature selection approaches address this challenge by simplifying datasets through the removal of unimportant features, while improving explainability by revealing feature importance. In this work, we leverage mathematical programming to identify the most important features in a trained deep neural network with a ReLU activation function, providing greater insight into its decision-making process. Unlike traditional feature selection methods, our approach adjusts the weights and biases of the trained neural network via a Mixed-Integer Linear Programming (MILP) model to identify the most important features and thereby uncover underlying relationships. The mathematical formulation is reported, which determines the subset of selected features, and clustering is applied to reduce the complexity of the model. Our results illustrate improved performance in the neural network when feature selection is implemented by the proposed approach, as compared to other feature selection approaches. Finally, analysis of feature selection frequency across each dataset reveals feature contribution in model predictions, thereby addressing the black-box nature of the neural network. Full article

Aiming at the overfitting problem caused by the limited sample size in the spectral classification of aero-engine hot jets, this paper proposed a synthetic spectral enhancement classification network FTIR-SpectralGAN for the FT-IR of aeroengine hot jets. Firstly, passive telemetry FTIR spectrometers were used to measure the hot jet spectrum data of six types of aero-engines, and a spectral classification dataset was created. Then, a spectral classification network FTIR-SpectralGAN was designed, which consists of a generator and a discriminator. The generator architecture comprises six Conv1DTranspose layers, with five of these layers integrated with BN and LeakyReLU layers to introduce noise injection. This design enhances the generation capability for complex patterns and facilitates the transformation from noise to high-dimensional data. The discriminator employs a multi-task dual-output structure, consisting of three Conv1D layers combined with LeakyReLU and Dropout techniques. This configuration progressively reduces feature dimensions and mitigates overfitting. During training, the generator learns the underlying distribution of spectral data, while the discriminator distinguishes between real and synthetic data and performs spectral classification. The dataset was randomly partitioned into training, validation, and test sets in an 8:1:1 ratio. For training strategy, an unbalanced alternating training approach was adopted, where the generator is trained first, followed by the discriminator and then the generator again. Additionally, weighted mixed loss and label smoothing strategies were introduced to enhance network training performance. Experimental results demonstrate that the spectral classification accuracy reaches up to 99%, effectively addressing the overfitting issue commonly encountered in CNN-based classification tasks with limited samples. Comparative experiments show that FTIR-SpectralGAN outperforms classical data augmentation methods and CVAE-based synthetic data enhancement approaches. It also achieves higher robustness and classification accuracy compared to other spectral classification methods. Full article

The Kongco area of Nima in the northern part of the Lhasa terrane has a suite of alkaline granitic porphyry dykes associated with Early Cretaceous granites and accompanied by Cu/Mo mineralization. LA-ICP-MS ²⁰⁶Pb/²³⁸U zircon geochronology performed on the dykes produced an age of 104.15 ± 0.94 Ma (MSWD = 0.98), indicating the Early Cretaceous emplacement of the dykes. The dykes exhibit high silica (SiO₂ = 76.22~77.90 wt.%), high potassium (K₂O = 4.97~6.21 wt.%), high alkalinity (K₂O + Na₂O = 8.07~8.98 wt.%), low calcium (CaO = 0.24~0.83 wt.%), low magnesium (MgO = 0.06~0.20 wt.%), and moderate aluminum content (Al₂O₃ = 11.93~12.45 wt.%). The Rieterman index (σ) ranges from 1.93 to 2.34. A/NK (molar ratio Al₂O₃/(Na₂O + K₂O)) and A/CNK (molar ratio Al₂O₃/(CaO + Na₂O + K₂O)) values of the dykes range from 1.06 to 1.18 and 0.98 to 1.09, respectively. The dykes are relatively enriched in Rb, Th, U, K, Ta, Ce, Nd, Zr, Hf, Sm, Y, Yb, and Lu, and they show a noticeable relative depletion in Ba, Nb, Sr, P, Eu, and Ti, as well as an average differentiation index (DI) of 96.42. The dykes also exhibit high FeO_T/MgO ratios (3.60~10.41), Ga/Al ratios (2.22 × 10⁻⁴~3.01 × 10⁻⁴), Y/Nb ratios (1.75~2.40), and Rb/Nb ratios (8.36~20.76). Additionally, they have high whole-rock Zr saturation temperatures (884~914 °C), a pronounced Eu negative anomaly (δEu = 0.04~0.23), and a rightward-sloping “V-shaped” rare earth element pattern. These characteristics suggest that the granitic porphyry dykes can be classified as A2-type granites formed in a post-collisional tectonic environment and that they are weakly peraluminous, high-potassium, and Calc-alkaline basaltic rocks. Positive ε_Hf(t) values = 0.43~3.63 and a relatively young Hf crustal model age (T_DM2 = 826~1005 Ma, ⁸⁷Sr/⁸⁶Sr ratios = 0.7043~0.7064, and ε_Nd(t) = −8.60~−2.95 all indicate lower crust and mantle mixing. The lower crust and mantle mixing model is also supported by (²⁰⁶Pb/²⁰⁴Pb)t = 18.627~18.788, (²⁰⁷Pb/²⁰⁴Pb)t = 15.707~15.719, (²⁰⁸Pb/²⁰⁴Pb)t = 39.038~39.110). Together, the Hf, Sr and Pb isotopic ratios indicate that the Kongco granitic porphyry dykes where derived from juvenile crust formed by the addition of mantle material to the lower crust. From this, we infer that the Kongco granitic porphyry dykes are related to a partial melting of the lower crust induced by subduction slab break-off and asthenospheric upwelling during the collision between the Qiangtang and Lhasa terranes and that they experienced significant fractional crystallization dominated by potassium feldspar and amphibole. These dykes are also accompanied by significant copper mineralization (five samples, copper content 0.2%), suggesting a close relationship between the magmatism associated with these dykes and regional metallogenesis, indicating a high potential for mineral exploration. Full article

Background/Objectives: Even today, interstitial lung disease (ILD) is diagnosed by chest high-resolution computed tomography (lung HR-CT). Large amounts of data are available about the usefulness of transthoracic lung ultrasound (LUS) in ILD. This study aimed to evaluate the transthoracic LUS capacity to discriminate different ILD patterns in systemic sclerosis (SSc) patients, such as usual interstitial pneumonia (UIP), non-specific interstitial pneumonia (NSIP) with ground glass opacification/opacity (GGO), and NSIP with GGO and reticulations, as well as the possibility of identifying progressive fibrosing ILD. Methods: We enrolled SSc-patients attending the outpatient Clinic of the Rheumatology Unit of Policlinico of Foggia and the Rheumatology Unit of Policlinico of Bari who satisfied these inclusion criteria: age older than 18 years; the satisfaction of ACR/EULAR 2013 classification criteria for SSc; chest HR-CT scan within three months before or three months after transthoracic LUS evaluation; and availability of recent and complete pulmonary function test. The exclusion criteria were as follows: history or recent reactivation of chronic obstructive pulmonary disease, lung cancer, lung infection, heart failure, pulmonary oedema, pulmonary arterial hypertension, acute respiratory distress syndrome and diffuse alveolar haemorrhage and thoracic surgery. All enrolled SSc-patients underwent transthoracic LUS, performed by an experienced sonographer. The ILD diagnosis and the respective patterns were assessed by chest HR-CT, which still represents the best diagnostic tool. Results: ILD was observed in 99 (63.5%) patients. Of these, 25% had the UIP pattern and 75% the NSIP pattern (46 with GGO, 28 with GGO and reticulations). By receiver operating characteristic (ROC) curve analysis, higher values of accuracy, sensitivity, specificity, and negative clinical utility index (CUI) were found for pleural line irregularity (0.84 (95% CI: 0.75–0.91), 96%, and 73.6%, p = 0.0001; 0.72), and pleural line thickness (0.84 (95% CI: 0.74–0.91), 72%, and 96.4%, p = 0.0001; 0.85) for detecting the UIP pattern. The best performance among transthoracic LUS signs for NSIP with the GGO pattern was observed for B-lines (accuracy: 0.88 (95% CI: 0.80–0.93), sensitivity: 93.4% and specificity: 82.4, p = 0.0001; CUI+: 0.75, CUI−: 0.77). LUS signs with higher accuracy, sensitivity, and specificity for NSIP with GGO and reticulations were pleural line irregularity (0.89 (95% CI: 0.80–0.95), 96.4%, and 82.4%, p = 0.0001) with CUI−: 0.72, and B-lines (0.89 (95% CI: 0.80–0.95), 96.4%, 82.4%, p = 0.0001), with CUI+: 0.80 and CUI−: 0.70. Furthermore, a total number of B-lines > 10 maximises LUS performance with 92.3% sensitivity, and an accuracy of 0.83 (p = 0.0001) for detecting the NSIP pattern, particularly GGO. A sample-restricted analysis (66 SSc patients) evidenced the presence of progressive fibrosing ILD in 77% of these patients. By binary regression analysis, the unique LUS sign associated with progressive fibrosing ILD was the presence of pleural line irregularity (OR: 3.6; 95% CI 1.08–11.9; p = 0.036). Conclusions: Our study demonstrated that transthoracic LUS presented a high capacity to discriminate the different patterns of SSc-ILD. Therefore, the hypothesis that transthoracic LUS is an effective screening method for the evaluation of the presence of SSc-ILD and establishing the correct timing of chest HR-CT, in order to avoid patients receiving excessive exposure to ionising radiation, is supported. Full article

Traditional algorithms and single predictive models often face challenges such as limited prediction accuracy and insufficient modeling capabilities for complex time-series data in fault prediction tasks. To address these issues, this paper proposes a combined prediction model based on an improved temporal convolutional network (TCN) and bidirectional long short-term memory (BiLSTM), referred to as the TCN-BiLSTM model. This model aims to enhance the reliability and accuracy of time-series fault prediction. It is designed to handle continuous processes but can also be applied to batch and hybrid processes due to its flexible architecture. First, preprocessed industrial operation data are fed into the model, and hyperparameter optimization is conducted using the Optuna framework to improve training efficiency and generalization capability. Then, the model employs an improved TCN layer and a BiLSTM layer for feature extraction and learning. The TCN layer incorporates batch normalization, an optimized activation function (Leaky ReLU), and a dropout mechanism to enhance its ability to capture multi-scale temporal features. The BiLSTM layer further leverages its bidirectional learning mechanism to model the long-term dependencies in the data, enabling effective predictions of complex fault patterns. Finally, the model outputs the prediction results after iterative optimization. To evaluate the performance of the proposed model, simulation experiments were conducted to compare the TCN-BiLSTM model with mainstream prediction methods such as CNN, RNN, BiLSTM, and A-BiLSTM. The experimental results indicate that the TCN-BiLSTM model outperforms the comparison models in terms of prediction accuracy during both the modeling and forecasting stages, providing a feasible solution for time-series fault prediction. Full article

Background: Prostate-specific membrane antigen (PSMA)-directed radioligand therapy (PRLT) with Lutetium-177 ([¹⁷⁷Lu]Lu-PSMA) is a safe and effective treatment for metastatic castration-resistant prostate cancer (mCRPC). The aim of our study was to evaluate clinical variables of patients with extreme response to PRLT and to assess its immunomodulatory potential. Methods: This retrospective study included 36 patients from two centers achieving extreme response after [¹⁷⁷Lu]Lu-PSMA PRLT. The primary outcomes were the duration of maintained response in months (MR) and improvement post-therapy—clinically, serologically, and on molecular (PET/CT) imaging. We examined several variables, including pathology, gene sequencing, baseline PSA, Gleason score, prior therapies, number of PRLT cycles, and pattern of disease, to identify potential factors that may influence the extreme response. Results: Between 2018 and mid-September 2024, 36 men with mCRPC received a mean of three cycles of [¹⁷⁷Lu]Lu-PSMA PRLT. Patients were subgrouped according to clinical variables versus MR. A total of 17 patients had ≥12 months MR (17/36, 47%). The longest duration of MR was 99 months and a mean of 17.44 months (95% CI 10.05–24.84). Previous lines of treatment were evaluated for MR, p = 0.172. Pattern of disease (bone, lymph node, liver, and peritoneal) was evaluated for MR, p = 0.721. The Gleason score was evaluated for MR, p = 0.871. Patients with known BRCA sequencing status (n = 12) were analyzed with mean MR: BRCA1/2 wild-type, 6/12 (50%), 6.67 months; BRCA 1/2 negative, 1/12 (8.33%), 7 months; BRCA germline negative and somatic positive, 1/12 (8.33%), 36 months; BRCA germline negative, somatic negative, 2/12 (16.67%), 27 months; and BRCA 2 positive, 2/12 (16.67%), 43 months. Conclusions: We propose there may be intrinsic mechanisms suggesting the immunomodulatory enhancement of ionizing radiation, primarily driving extreme responses. Full article

This study investigated the genetic diversity and genomic variation in wild Cordyceps chanhua populations from four regions in China—Dazhou, Sichuan (ICD); Lu’an, Anhui (ICL); Taizhou, Zhejiang (ICT); and Yixing, Jiangsu (ICY)—to elucidate genetic differentiation patterns and provide a scientific foundation for resource conservation and sustainable utilization. Whole-genome resequencing was performed, yielding high-quality sequencing data (Q20 > 98%, Q30 > 94%, coverage: 93.62–95.79%) and enabling the detection of 82,428 single-nucleotide polymorphisms (SNPs) and 12,517 insertion–deletion markers (InDels). Genomic variations were unevenly distributed across chromosomes, with chromosome chrU05 exhibiting the highest SNP density (5187.86), suggesting a potential hotspot of genetic diversity. Phylogenetic analysis confirmed that all samples belonged to the C. chanhua lineage but revealed significant genetic differentiation among regions. Population structure analysis, supported by structure analysis and PCA, identified two distinct subgroups (G1 and G2) closely associated with geographic origins, reflecting the influence of both environmental and geographic factors on genetic differentiation. These findings underscore the substantial interregional genetic diversity in C. chanhua populations, highlighting the importance of tailored conservation strategies and region-specific germplasm utilization. The study provides critical genomic insights to support marker-assisted breeding, regional cultivation optimization, and the sustainable development of C. chanhua resources. Full article

This article presents an investigation of the use of machine learning methodologies for the prediction of surface roughness in milling operations, using sensor data as the primary source of information. The sensors, which included current transformers, a microphone, and displacement sensors, captured comprehensive machining signals at a frequency of 10 kHz. The signals were subjected to preprocessing using the Savitzky–Golay filter, with the objective of isolating relevant moments of active material machining and reducing noise. Two machine learning models, namely Elastic Net and neural networks, were employed for the prediction purposes. The Elastic Net model demonstrated effective handling of multicollinearity and reduction in the data dimensionality, while the neural networks, utilizing the ReLU activation function, exhibited the capacity to capture complex, nonlinear patterns. The models were evaluated using the coefficient of determination (R²), which yielded values of 0.94 for Elastic Net and 0.95 for neural networks, indicating a high degree of predictive accuracy. These findings illustrate the potential of machine learning to optimize manufacturing processes by facilitating precise predictions of surface roughness. Elastic Net demonstrated its utility in reducing dimensionality and selecting features, while neural networks proved effective in modeling complex data. Consequently, these methods exemplify the efficacy of integrating data-driven approaches with robust predictive models to improve the quality and efficiency of surface. Full article