Accurate crop mapping is crucial for ensuring food security. Recently, many studies have developed diverse crop mapping models based on deep learning. However, these models generally rely on a large amount of labeled crop samples to investigate the intricate relationship between the crop types of the samples and the corresponding remote sensing features. Moreover, their efficacy is often compromised when applied to other areas owing to the disparities between source and target data. To address this issue, a new multi-modal deep adaptation crop classification network (MDACCN) was proposed in this study. Specifically, MDACCN synergistically exploits time series optical and SAR images using a middle fusion strategy to achieve good classification capacity. Additionally, local maximum mean discrepancy (LMMD) is embedded into the model to measure and decrease domain discrepancies between source and target domains. As a result, a well-trained model in a source domain can still maintain satisfactory accuracy when applied to a target domain. In the training process, MDACCN incorporates the labeled samples from a source domain and unlabeled samples from a target domain. When it comes to the inference process, only unlabeled samples of the target domain are required. To assess the validity of the proposed model, Arkansas State in the United States was chosen as the source domain, and Heilongjiang Province in China was selected as the target domain. Supervised deep learning and traditional machine learning models were chosen as comparison models. The results indicated that the MDACCN achieved inspiring performance in the target domain, surpassing other models with overall accuracy, Kappa, and a macro-averaged F1 score of 0.878, 0.810, and 0.746, respectively. In addition, the crop-type maps produced by the MDACCN exhibited greater consistency with the reference maps. Moreover, the integration of optical and SAR features exhibited a substantial improvement of the model in the target domain compared with using single-modal features. This study indicated the considerable potential of combining multi-modal remote sensing data and an unsupervised domain adaptive approach to provide reliable crop distribution information in areas where labeled samples are missing. Full article

This study presents a comprehensive approach for simulating the additive manufacturing process of semi-crystalline composite polymers using Fused Deposition Modeling (FDM). By combining thermomechanical Finite Element Analysis (FEA) with experimental validation, our main objective is to comprehend and model the complex behaviors of 50 wt.% carbon fiber-reinforced Polyphenylene Sulfide (CF PPS) during FDM printing. The simulations of the FDM process encompass various theoretical aspects, including heat transfer, orthotropic thermal properties, thermal dissipation mechanisms, polymer crystallization, anisotropic viscoelasticity, and material shrinkage. We utilize Abaqus user subroutines such as UMATHT for thermal orthotropic constitutive behavior, UEPACTIVATIONVOL for progressive activation of elements, and ORIENT for material orientation. Mechanical behavior is characterized using a Maxwell model for viscoelastic materials, incorporating a dual non-isothermal crystallization kinetics model within the UMAT subroutine. Our approach is validated by comparing nodal temperature distributions obtained from both the Abaqus built-in AM Modeler and our user subroutines, showing close agreement and demonstrating the effectiveness of our simulation methods. Experimental verification further confirms the accuracy of our simulation techniques. The mechanical analysis investigates residual stresses and distortions, with particular emphasis on the critical transverse in-plane stress component. This study offers valuable insights into accurately simulating thermomechanical behaviors in additive manufacturing of composite polymers. Full article

This study investigates the effect of daily consumption of wheat biscuits enriched with plant proteins in postprandial metabolic responses of women with overweight/obesity who follow an energy-restricted diet. Thirty apparently healthy women participated in a 12-week randomized controlled trial and were assigned either to a control (CB) or an intervention (PB) group. Participants consumed daily either a conventional (CB) or an isocaloric wheat biscuit enriched with plant proteins (PB) containing high amounts of amino acids with appetite-regulating properties, i.e., BCAAs and L-arg. At baseline and the end of the intervention, a mixed meal tolerance test was performed. The responses of glucose, insulin, ghrelin, GLP-1, and glicentin were evaluated over 180 min. After 12 weeks, both groups experienced significant decreases in body weight, fat mass, and waist circumference. In the PB group, a trend towards higher weight loss was observed, accompanied by lower carbohydrate, fat, and energy intakes (p < 0.05 compared to baseline and CB group), while decreases in fasting insulin and the HOMA-IR index were also observed (p < 0.05 compared to baseline). In both groups, similar postprandial glucose, ghrelin, and GLP-1 responses were detected, while iAUC for insulin was lower (p < 0.05). Interestingly, the iAUC of glicentin was greater in the PB group (p < 0.05 compared to baseline). Subjective appetite ratings were beneficially affected in both groups (p < 0.05). Consumption of wheat biscuits enriched in plant proteins contributed to greater weight loss, lower energy intake, and insulin resistance and had a positive impact on postprandial glicentin response, a peptide that can potentially predict long-term weight loss and decreased food intake. Full article

A nanoparticle-anchored three-dimensional microsphere flower-structured layered double hydroxide (LDH) material with Fe₃O₄ particles was successfully prepared using simple hydrothermal and hot solvent methods. Micro-nanostructured Fe₃O₄@LDHs (SLF) composites balance microwave absorption, corrosion protection, and UV aging resistance. The minimum reflection loss value of SLF is −35.75 dB at 14.16 GHz, when the absorber thickness is 8 mm, and the absorption bandwidth at this frequency is up to 2.56 GHz for RL values less than −10 dB, while the LL is only 1 GHz. The SLF /EP coating has not only excellent microwave absorption performance but also excellent corrosion and UV aging resistance performance. The coating still has some anti-corrosion effect after 10 d of immersion. This work is intended as a reference for the development of new coatings with excellent microwave absorption properties as well as corrosion and UV aging resistance for wind turbine tower barrels (seaside wind power generation equipment) surfaces. Full article

Identifying the factors affecting organizational sustainability is a crucial topic in the field of social science and business research. Especially in family businesses, the most crucial issue is how to maintain corporate sustainability across generations. In this regard, succession planning plays a key role in maintaining the sustainable development of family businesses. From the perspective of socioemotional wealth, this study discusses the motivations and consequences of intrafamily succession by measuring the impact of the internal determining factors of succession planning on family business performance. Based on a sample of 281 Chinese family firms, this study uncovers the relationship between succession planning and organizational performance. The core findings of this study include succession planning is positively related to organizational performance in the matter of the successor’s training; succession planning has a positive effect on organizational performance in terms of the successor’s self-preparation; and succession planning is positively correlated with organizational performance in the aspect ofthe relationship between the successor and business. By illustrating that the formulation of succession planning is an essential pursuit for family businesses to preserve sustainability and socioemotional wealth, the results reveal ways to facilitate succession planning through internal factors in the family business. This study contributes to organizational sustainable development literature, family business sustainability studies, and succession management research by validating the positive relationship between succession planning and organizational performance, indicating that succession planning is a vital driving force for achieving organizational sustainability. Full article

Bladder cancer (BC) diagnosis is reliant on cystoscopy, an invasive procedure associated with urinary tract infections. This has sparked interest in identifying noninvasive biomarkers in body fluids such as blood and urine. A source of biomarkers in these biofluids are extracellular vesicles (EVs), nanosized vesicles that contain a wide array of molecular cargo, including small noncoding RNA such as transfer RNA-derived fragments (tRF) and microRNA. Here, we performed small-RNA next-generation sequencing from EVs from urine and serum, as well as from serum supernatant. RNA was extracted from 15 non-cancer patients (NCPs) with benign findings in cystoscopy and 41 patients with non-muscle invasive BC. Urine and serum were collected before transurethral resection of bladder tumors (TUR-b) and at routine post-surgery check-ups. We compared levels of tRFs in pre-surgery samples to samples from NCPs and post-surgery check-ups. To further verify our findings, samples from 10 patients with stage T1 disease were resequenced. When comparing tRF expression in urine EVs between T1 stage BC patients and NCPs, 14 differentially expressed tRFs (DEtRFs) were identified. In serum supernatant, six DEtRFs were identified among stage T1 patients when comparing pre-surgery to post-surgery samples and four DEtRFs were found when comparing pre-surgery samples to NCPs. By performing a blast search, we found that sequences of DEtRFs aligned with genomic sequences pertaining to processes relevant to cancer development, such as enhancers, regulatory elements and CpG islands. Our findings display a number of tRFs that may hold potential as biomarkers for the diagnosis and recurrence-free survival of BC. Full article

In this paper, we propose a method for establishing if a variable is capable of modifying the Macroscopic Fundamental Diagram (MFD) of a street network. The variables have many different configurations, and a simulation is performed for each one. Then, based on the output data of each simulation, the representative speed, density, and flow of the network are calculated. We use three metrics to establish if a variable affects the MFD: the first establishes a distance between the compared density and speed patterns, the second establishes a distance between capacities, and the third establishes a distance between critical densities. We select four variables to test our method: the precision of driving, the vehicles’ top speeds distribution, the procedure for selecting routes, and the procedure for selecting destinations; we determine whether each of these variables can modify the MFD shape. Additionally, we detect which configurations of a variable are able to reach and exceed the critical density (causing congestion) so we can establish which configurations are sustainable and which are not. The novelties of this work are twofold: (1) we introduce a method to detect if a variable modifies the MFD; (2) we establish if the selected variables modify the MFD. Full article

Cultural crop-production practices are not only engineered to minimize pest incidence but also improve resource use efficiency and increase the diversity of habitat for beneficial insects that provide pollination services. With the increasing cultivation of industrial hemp and the benefits associated with the cultivation of multiple crops, its integration into a polyculture cropping system remains to be evaluated. We intercropped two pollinator-attractive crops, hemp and cowpea, with squash, a pollinator-dependent crop, to evaluate the impact of pollinator abundance and diversity on crop yield. Intercropping significantly increased the overall abundance of pollinators with 79.1% recorded from the intercropping systems compared to 21.9% in the monocropping systems. Sweat bees and bumble bees were the most abundant bees, and Squash+Cowpea was the most diverse cropping system. Intercropping significantly increased the yield of squash with higher squash yield (155%) in Hemp+Squash and (161%) in Squash+Cowpea than in squash monocrop. Also, intercropping resulted in higher hemp yield (64%) in Hemp+Cowpea and (165%) in Hemp+Squash compared to hemp monocrop. This study demonstrated that agricultural systems such as intercropping that are designed to attract pollinators are much more productive by not only improving crop yield but also growers’ returns on investments. Full article

In this paper, we propose a training method to reduce the VR sickness that occurs while viewing VR content with an HMD on. The proposed approach is a new method that involves pre-exposing users to VR sickness to enable them to adapt to VR sickness. In the proposed method, the training process was designed based on the features of existing studies related to exposure and adaptation to motion sickness and simulator sickness. The effectiveness of the proposed method was evaluated through experiments with 15 subjects (SSQ and VR sickness response were used in the analysis). As a result of the experiment, nausea was significantly decreased by 47%, and oculomotor discomfort was significantly decreased by 34% after the proposed training method. The VR sickness response decreased by 31%; however, this difference was not statistically significant. Furthermore, we analyzed the VR sickness response in two groups: those whose sickness decreased and those whose sickness increased. We confirmed that the decrease group (pre-experiment mean: 1.34 times, post-experiment mean: 0.58 times) had a larger change than the increase group (pre-experiment mean: 0.31 times, post-experiment mean: 0.42 times). Therefore, from the experimental results, it was confirmed that the proposed method is effective in reducing VR sickness. Full article

The aim of this work is research dedicated to the search for new bactericidal systems for use in cosmetic formulations, dermocosmetics, or the production of wound dressings. Over the last two decades, chitosan, due to its special biological activity, has become a highly indispensable biopolymer with very wide application possibilities. Reports in the literature on the antibacterial effects of chitosan are very diverse, but our research has shown that they can be successfully improved through chemical modification. Therefore, in this study, results on the synthesis of new chitosan-based Schiff bases, dCsSB-SFD and dCsSB-PCA, are obtained using two aldehydes: sodium 4-formylbenzene-1,3-disulfonate (SFD) and 2-pyridine carboxaldehyde (PCA), respectively. Chitosan derivatives synthesized in this way demonstrate stronger antimicrobial activity. Carrying out the procedure of grafting chitosan with a caproyl chain allowed obtaining compatible blends of chitosan derivatives with κ-carrageenan, which are stable hydrogels with a high swelling coefficient. Furthermore, the covalently bounded poly(ε-caprolactone) (PCL) chain improved the solubility of obtained polymers in organic solvents. In this respect, the Schiff base-containing polymers obtained in this study, with special hydrogel and antimicrobial properties, are very promising materials for potential use as a controlled-release formulation of both hydrophilic and hydrophobic drugs in cosmetic products for skin health. Full article

Various published COVID-19 models have been used in epidemiological studies and healthcare planning to model and predict the spread of the disease and appropriately realign health measures and priorities given the resource limitations in the field of healthcare. However, a significant issue arises when these models need help identifying the distribution of the constituent variants of COVID-19 infections. The emergence of such a challenge means that, given limited healthcare resources, health planning would be ineffective and cost lives. This work presents a universal neural network (NN) computational instrument for predicting the mainstream symptomatic infection rate of COVID-19 and models of the distribution of its associated variants. The NN is based on a mixture density network (MDN) with a Gaussian mixture model (GMM) object as a backbone. Twelve use cases were used to demonstrate the validity and reliability of the proposed MDN. The use cases included COVID-19 data for Canada and Saudi Arabia, two date ranges (300 and 500 days), two input data modes, and three activation functions, each with different implementations of the batch size and epoch value. This array of scenarios provided an opportunity to investigate the impacts of epistemic uncertainty (EU) and aleatoric uncertainty (AU) on the prediction model’s fitting. The model accuracy readings were in the high nineties based on a tolerance margin of 0.0125. The primary outcome of this work indicates that this easy-to-use universal MDN helps provide reliable predictions of COVID-19 variant distributions and the corresponding synthesized profile of the mainstream infection rate. Full article

The correlation femtoscopy technique makes it possible to estimate the geometric dimensions and lifetime of the particle emission region after the collision of ions. Measurements of the emission region characteristics not only at midrapidity but also at backward (forward) rapidity can provide new information about the source and make it possible to impose constraints on the heavy-ion collision models. This work is devoted to revealing the dependence of the spatial and temporal parameters of the emission region of identical pions in Au+Au collisions at $\sqrt{s_{NN}}$ = 3 GeV from the fixed-target program of the STAR experiment. The extracted femtoscopic radii, $R_{out}$, $R_{side}$, $R_{long}$, $R_{out-long}^2$, and the correlation strength, $\lambda$, are presented as a function of collision centrality, pair rapidity, and transverse momentum. Physics implications will be discussed. Full article

The advancement of cryptographic systems presents both opportunities and challenges in the realm of digital forensics. In an era where the security of digital information is crucial, the ability to non-invasively detect and analyze cryptographic configurations has become significant. As cryptographic algorithms become more robust with longer key lengths, they provide higher levels of security. However, non-invasive side channels, specifically through electromagnetic (EM) emanations, can expose confidential cryptographic details, thus presenting a novel solution to the pressing forensic challenge. This research delves into the capabilities of EM side-channel analysis (EM-SCA), specifically focusing on detecting both cryptographic key lengths and the algorithms employed utilizing a machine-learning-based approach, which can be instrumental for digital forensic experts during their investigations. Through meticulous data processing and analysis, the Support Vector Machine (SVM) model, among others, demonstrated a notable accuracy of 94.55% in distinguishing between AES and ECC cryptographic operations. This capability significantly enhances digital forensic methodologies, offering a novel avenue for noninvasively uncovering encrypted data’s cryptographic settings. By identifying key lengths and algorithms without invasive procedures, this research contributes substantially to the advancement of forensic investigations in encrypted environments. Full article

The morphology and crystal structure of Pt films grown by pulsed laser deposition (PLD) on yttria-stabilized zirconia (YSZ)at high temperatures Tg = 900 °C was studied for four different film thicknesses varying between 10 and 70 nm. During the subsequent growth of the capping layer, the thermal stability of the Pt was strongly influenced by the Pt film’s thickness. Furthermore, these later affected the film morphology, the crystal structure and hillocks size, and distribution during subsequent growth at Tg = 900 °C for a long duration. The modifications in the morphology as well as in the structure of the Pt film without a capping layer, named also as the as-grown and encapsulated layers in the bilayer system, were examined by a combination of microscopic and scattering methods. The increase in the thickness of the deposited Pt film brought three competitive phenomena into occurrence, such as 3D–2D morphological transition, dewetting, and hillock formation. The degree of coverage, film continuity, and the crystal quality of the Pt film were significantly improved by increasing the deposition time. An optimum Pt film thickness of 70 nm was found to be suitable for obtaining a hillock-free Pt bottom electrode which also withstood the dewetting phenomena revealed during the subsequent growth of capping layers. This achievement is crucial for the deposition of functional bottom electrodes in ferroelectric and multiferroic heterostructure systems. Full article

Since methods for reducing macromolecule entanglements have been developed, it has become possible to better understand the impact of polymer chain entanglement on the crystallization process. The article presents basic information about the disentangling of macromolecules and the characterization of the degree of entanglement. The basic knowledge of polymer crystallization was also presented. Then, it was discussed how polymers crystallize during their disentangling. Non-isothermal and isothermal crystallization experiments using disentangled polymers, and for comparison using entangled polymers, are described in more detail. The influence of disentangling on both nucleation and crystal growth is highlighted. It is also shown how the crystallization of polymers changes when macromolecules re-entangle. Full article

Healthcare professionals are known to suffer from workplace stress and burnout, which can negatively affect their empathy for patients and quality of care. While existing research has identified factors associated with wellbeing and empathy in healthcare professionals, these efforts are typically focused on the group level, ignoring potentially important individual differences and implications for individualized intervention approaches. In the current study, we implemented N-of-1 personalized machine learning (PML) to predict wellbeing and empathy in healthcare professionals at the individual level, leveraging ecological momentary assessments (EMAs) and smartwatch wearable data. A total of 47 mood and lifestyle feature variables (relating to sleep, diet, exercise, and social connections) were collected daily for up to three months followed by applying eight supervised machine learning (ML) models in a PML pipeline to predict wellbeing and empathy separately. Predictive insight into the model architecture was obtained using Shapley statistics for each of the best-fit personalized models, ranking the importance of each feature for each participant. The best-fit model and top features varied across participants, with anxious mood (13/19) and depressed mood (10/19) being the top predictors in most models. Social connection was a top predictor for wellbeing in 9/12 participants but not for empathy models (1/7). Additionally, empathy and wellbeing were the top predictors of each other in 64% of cases. These findings highlight shared and individual features of wellbeing and empathy in healthcare professionals and suggest that a one-size-fits-all approach to addressing modifiable factors to improve wellbeing and empathy will likely be suboptimal. In the future, such personalized models may serve as actionable insights for healthcare professionals that lead to increased wellness and quality of patient care. Full article

Nearly zero-emission buildings (nZEBs) are increasingly being constructed in Europe. There are also incentives to refurbish older buildings and transform them into nZEBs. However, permission is not always granted for their connection to the grid to infuse surplus photovoltaic electricity due to the grid being overloaded with a large number of renewables. In this study, the case of a refurbished school building in Central Greece is examined. After refurbishing it, a significant amount of photovoltaic electricity surplus is observed during the summer and neutral months, which cannot be exported to the grid. The absence of an adequate battery storage capacity resulted in the rejection of an application for exporting the school’s surplus to the network and the photovoltaic installation staying idle. An alternative approach is proposed in this work, involving a shift in the export of the photovoltaic electricity surplus to the evening hours, in order for the school to be granted permission to export it to the network. To this end, an optimal battery storage size is sought by employing a building energy system simulation. The mode of operation of the battery designed for this application is set to discharge daily, in order to export the electricity surplus in the afternoon hours to the evening hours, when it is favorable for the network. Additionally, the optimal size of the thermal energy storage of the heating system is studied to further improve its energy efficiency. Our battery and storage tank size optimization study shows that a significant battery capacity is required, with 12 kWh/kWp photovoltaic panels being recommended for installation. The ever-decreasing cost of battery installations results in the net present value (NPV) of the additional investment for the battery installation becoming positive. The solution proposed forms an alternative path to further increase the penetration of renewables in saturated networks in Greece by optimizing battery storage capacity. Full article

Membrane proteins constitute about 20% of the human proteome and play crucial roles in cellular functions. However, a complete understanding of their structure and function is limited by their hydrophobic nature, which poses significant challenges in purification and stabilization. Detergents, essential in the isolation process, risk destabilizing or altering the proteins’ native conformations, thus affecting stability and functionality. This study leverages single-particle cryo-electron microscopy to elucidate the structural nuances of membrane proteins, focusing on the SLAC1 bacterial homolog from Haemophilus influenzae (HiTehA) purified with diverse detergents, including n-dodecyl β-D-maltopyranoside (DDM), glycodiosgenin (GDN), β-D-octyl-glucoside (OG), and lauryl maltose neopentyl glycol (LMNG). This research not only contributes to the understanding of membrane protein structures but also addresses detergent effects on protein purification. By showcasing that the overall structural integrity of the channel is preserved, our study underscores the intricate interplay between proteins and detergents, offering insightful implications for drug design and membrane biology. Full article

Movement and muscle control are crucial for the survival of all free-living organisms. This study aimed to explore differential patterns of cortical and subcortical activation across different stages of muscle control using functional magnetic resonance imaging (fMRI). An event-related design was employed. In each trial, participants (n = 10) were instructed to gently press a button with their right index finger, hold it naturally for several seconds, and then relax the finger. Neural activation in these temporally separated stages was analyzed using a General Linear Model. Our findings revealed that a widely distributed cortical network, including the supplementary motor area and insula, was implicated not only in the pressing stage, but also in the relaxation stage, while only parts of the network were involved in the steady holding stage. Moreover, supporting the direct/indirect pathway model of the subcortical basal ganglia, their substructures played distinct roles in different stages of muscle control. The caudate nucleus exhibited greater involvement in muscle contraction, whereas the putamen demonstrated a stronger association with muscle relaxation; both structures were implicated in the pressing stage. Furthermore, the subthalamic nucleus was exclusively engaged during the muscle relaxation stage. We conclude that even the control of simple muscle movements involves intricate automatic higher sensory–motor integration at a neural level, particularly when coordinating relative muscle movements, including both muscle contraction and muscle relaxation; the cortical and subcortical regions assume distinct yet coordinated roles across different stages of muscle control. Full article

Coatomer Protein Complex-II (COPII) mediates anterograde vesicle transport from the endoplasmic reticulum (ER) to the Golgi apparatus. Here, we report that the COPII coatomer complex is constructed dependent on a small GTPase, Sar1, in spermatocytes before and during Drosophila male meiosis. COPII-containing foci co-localized with transitional endoplasmic reticulum (tER)-Golgi units. They showed dynamic distribution along astral microtubules and accumulated around the spindle pole, but they were not localized on the cleavage furrow (CF) sites. The depletion of the four COPII coatomer subunits, Sec16, or Sar1 that regulate COPII assembly resulted in multinucleated cell production after meiosis, suggesting that cytokinesis failed in both or either of the meiotic divisions. Although contractile actomyosin and anilloseptin rings were formed once plasma membrane ingression was initiated, they were frequently removed from the plasma membrane during furrowing. We explored the factors conveyed toward the CF sites in the membrane via COPII-mediated vesicles. DE-cadherin-containing vesicles were formed depending on Sar1 and were accumulated in the cleavage sites. Furthermore, COPII depletion inhibited de novo plasma membrane insertion. These findings suggest that COPII vesicles supply the factors essential for the anchoring and/or constriction of the contractile rings at cleavage sites during male meiosis in Drosophila. Full article

Deep learning has profoundly influenced various domains, particularly medical image analysis. Traditional transfer learning approaches in this field rely on models pretrained on domain-specific medical datasets, which limits their generalizability and accessibility. In this study, we propose a novel framework called real-world feature transfer learning, which utilizes backbone models initially trained on large-scale general-purpose datasets such as ImageNet. We evaluate the effectiveness and robustness of this approach compared to models trained from scratch, focusing on the task of classifying pneumonia in X-ray images. Our experiments, which included converting grayscale images to RGB format, demonstrate that real-world-feature transfer learning consistently outperforms conventional training approaches across various performance metrics. This advancement has the potential to accelerate deep learning applications in medical imaging by leveraging the rich feature representations learned from general-purpose pretrained models. The proposed methodology overcomes the limitations of domain-specific pretrained models, thereby enabling accelerated innovation in medical diagnostics and healthcare. From a mathematical perspective, we formalize the concept of real-world feature transfer learning and provide a rigorous mathematical formulation of the problem. Our experimental results provide empirical evidence supporting the effectiveness of this approach, laying the foundation for further theoretical analysis and exploration. This work contributes to the broader understanding of feature transferability across domains and has significant implications for the development of accurate and efficient models for medical image analysis, even in resource-constrained settings. Full article

While digitalization offers new opportunities for small- and medium-sized enterprises (SMEs), it also introduces the phenomenon of the “digitalization paradox”. This paper develops a theoretical model comprising digitalization, digital technology–business alignment, external social capital, and SMEs’ performance, rooted in strategic alignment theory (SAT) and social capital theory (SCT). The necessary data for the study were obtained by distributing questionnaires to 352 small and medium-sized enterprises engaged in digital practices in China, and hierarchical regression analysis was employed to investigate the impact of digitalization on the performance of SMEs and its boundaries of influence. The results indicate an inverted U-shaped relationship between digitalization and SME performance, with both digital technology–business alignment and external social capital serving as positive moderators. Specifically, digital technology–business alignment and external social capital both enhance the positive impact of digitalization on the performance of SMEs and mitigate its negative effects. The findings enhance comprehension of the “digitalization paradox” and offer new insights and solutions for SMEs to navigate the opportunities and challenges of digitalization. Full article