Making waves: Knowledge and data fusion in urban water modelling.

Mathematical modeling plays a crucial role in understanding and managing urban water systems (UWS), with mechanistic models often serving as the foundation for their design and operations. Despite the wide adoptions, mechanistic models are challenged by the complexity of dynamic processes and high computational demands. Data-driven models bring opportunities to capture system complexities and reduce computational cost, by leveraging the abundant data made available by recent advance in sensor technologies. However, the interpretability and data availability hinder their wider adoption. This paper advocates for a paradigm shift in the application of data-driven models within the context of UWS. Integrating existing mechanistic knowledge into data-driven modeling offers a unique solution that reduces data requirements and enhances model interpretability. The knowledge-informed approach balances model complexity with dataset size, enabling more efficient and interpretable modeling in UWS. Furthermore, the integration of mechanistic and data-driven models offers a more accurate representation of UWS dynamics, addressing lingering uncertainties and advancing modelling capabilities. This paper presents perspectives and conceptual framework on developing and implementing knowledge-informed data-driven modeling, highlighting their potential to improve UWS management in the digital era.

A Semantic-Consistent Few-Shot Modulation Recognition Framework for IoT Applications.

The rapid growth of the Internet of Things (IoT) has led to the widespread adoption of the IoT networks in numerous digital applications. To counter physical threats in these systems, automatic modulation classification (AMC) has emerged as an effective approach for identifying the modulation format of signals in noisy environments. However, identifying those threats can be particularly challenging due to the scarcity of labeled data, which is a common issue in various IoT applications, such as anomaly detection for unmanned aerial vehicles (UAVs) and intrusion detection in the IoT networks. Few-shot learning (FSL) offers a promising solution by enabling models to grasp the concepts of new classes using only a limited number of labeled samples. However, prevalent FSL techniques are primarily tailored for tasks in the computer vision domain and are not suitable for the wireless signal domain. Instead of designing a new FSL model, this work suggests a novel approach that enhances wireless signals to be more efficiently processed by the existing state-of-the-art (SOTA) FSL models. We present the semantic-consistent signal pretransformation (ScSP), a parameterized transformation architecture that ensures signals with identical semantics exhibit similar representations. ScSP is designed to integrate seamlessly with various SOTA FSL models for signal modulation recognition and supports commonly used deep learning backbones. Our evaluation indicates that ScSP boosts the performance of numerous SOTA FSL models, while preserving flexibility.

Lactobacillus inoculation mediated carboxylates and alcohols production from waste activated sludge fermentation system: Insight into process outcomes and metabolic network.

Producing medium chain fatty acids (MCFAs) from waste activated sludge (WAS) is crucial for sustainable chemical industries. This study addressed the electron donor requirement for MCFAs production by inoculating Lactobacillus at varying concentrations (7.94 × 1010, 3.18 × 1011, and 6.35 × 1011 cell/L) to supply lactate internally. Interestingly, the highest MCFAs yield (∼2000 mg COD/L) occurred at the lowest Lactobacillus inoculation. Higher inoculation concentrations redirected more carbon from WAS towards alcohols production rather than MCFAs generation, with up to 2852 mg COD/L alcohols obtained under 6.35 × 1011 cell/L inoculation. Clostridium dominance and increased genes abundance for substrate hydrolysis, lactate conversion, and MCFAs/alcohol production collectively enhanced WAS-derived MCFAs and alcohols synthesis after Lactobacillus inoculation. Overall, the strategy of Lactobacillus inoculation regulated fermentation outcomes and subsequent carbon recovery in WAS, presenting a sustainable technology to achieve liquid bio-energy production from underutilized wet wastes.

Promoting patient-centered care in CAR-T therapy for hematologic malignancy: a qualitative meta-synthesis.

BACKGROUND: CAR-T therapy has emerged as a potentially effective treatment for individuals diagnosed with hematologic malignancies. Understanding patients' unique experiences with this therapeutic approach is essential. This knowledge will enable the development of tailored nursing interventions that align with the increasing importance of patient-centered care. OBJECTIVE: To examine and synthesize qualitative data on patients and their family caregivers' experiences during the treatment journey. DESIGN: We conducted a systematic review and qualitative meta-synthesis. Eligible studies contained adult patient or family caregiver quotes about experiences of CAR-T therapy, published in English or Chinese in a peer-reviewed journal since 2015. Data sources included MEDLINE, CINAHL, Embase, PsycINFO, Web of Science, Scopus, Cochrane Library, CNKI, and WanFang. METHODS: Systematic search yielded 6373 identified articles. Of these, 12 reports were included in the analysis, which covered 11 separate studies. Two reviewers independently extracted data into NVIVO 12.0. Qualitative meta-synthesis was performed through line-by-line coding of full text, organization of codes into descriptive themes, and development themes. RESULTS: The qualitative meta-synthesis yielded eight primary themes. Noteworthy revelations from patients and their family caregivers regarding the CAR-T therapy journey encompassed various aspects. Prior to CAR-T therapy, patients experienced a lack of actual choice, struggled with expectations for treatment outcomes, and encountered intricate emotional experiences. During or immediately after CAR-T therapy, patients reported both comfortable and uncomfortable experiences. Additionally, patients emphasized that concerns regarding treatment efficacy and adverse reactions intensified treatment-related distress. After CAR-T therapy, significant changes were observed, and the burden of home-based rehabilitation. Additionally, we found factors contributed to the high CAR-T therapy cost. CONCLUSIONS: To ensure the safety and sustainability of CAR-T therapy, it is crucial to address the physical and psychological aspects of the patient's experience. Effective communication and comprehensive management are highly valued by patients and their caregivers. Further research should investigate ways to reduce burdens and develop self-management education programs for patients and their families.

UniHead: Unifying Multi-Perception for Detection Heads.

The detection head constitutes a pivotal component within object detectors, tasked with executing both classification and localization functions. Regrettably, the commonly used parallel head often lacks omni perceptual capabilities, such as deformation perception (DP), global perception (GP), and cross-task perception (CTP). Despite numerous methods attempting to enhance these abilities from a single aspect, achieving a comprehensive and unified solution remains a significant challenge. In response to this challenge, we develop an innovative detection head, termed UniHead, to unify three perceptual abilities simultaneously. More precisely, our approach: 1) introduces DP, enabling the model to adaptively sample object features; 2) proposes a dual-axial aggregation transformer (DAT) to adeptly model long-range dependencies, thereby achieving GP; and 3) devises a cross-task interaction transformer (CIT) that facilitates interaction between the classification and localization branches, thus aligning the two tasks. As a plug-and-play method, the proposed UniHead can be conveniently integrated with existing detectors. Extensive experiments on the COCO dataset demonstrate that our UniHead can bring significant improvements to many detectors. For instance, the UniHead can obtain + 2.7 AP gains in RetinaNet, + 2.9 AP gains in FreeAnchor, and + 2.1 AP gains in GFL. The code is available at https://github.com/zht8506/UniHead.

Prototype Correlation Matching and Class-Relation Reasoning for Few-Shot Medical Image Segmentation.

Few-shot medical image segmentation has achieved great progress in improving accuracy and efficiency of medical analysis in the biomedical imaging field. However, most existing methods cannot explore inter-class relations among base and novel medical classes to reason unseen novel classes. Moreover, the same kind of medical class has large intra-class variations brought by diverse appearances, shapes and scales, thus causing ambiguous visual characterization to degrade generalization performance of these existing methods on unseen novel classes. To address the above challenges, in this paper, we propose a Prototype correlation Matching and Class-relation Reasoning (i.e., PMCR) model. The proposed model can effectively mitigate false pixel correlation matches caused by large intra-class variations while reasoning inter-class relations among different medical classes. Specifically, in order to address false pixel correlation match brought by large intra-class variations, we propose a prototype correlation matching module to mine representative prototypes that can characterize diverse visual information of different appearances well. We aim to explore prototypelevel rather than pixel-level correlation matching between support and query features via optimal transport algorithm to tackle false matches caused by intra-class variations. Meanwhile, in order to explore inter-class relations, we design a class-relation reasoning module to segment unseen novel medical objects via reasoning inter-class relations between base and novel classes. Such inter-class relations can be well propagated to semantic encoding of local query features to improve few-shot segmentation performance. Quantitative comparisons illustrates the large performance improvement of our model over other baseline methods.

Impact of nutrient deficiency on biological sewage treatment - Perspectives towards urine source segregation.

Human urine contains 9 g/L of nitrogen (N) and 0.7 g/L of phosphorus (P). The recovery of N and P from urine helps close the nutrient loop and increase resource circularity in the sewage treatment sector. Urine contributes an average of 80 % N and 50 % P in sewage, whereby urine source segregation could reduce the burden of nutrient removal in sewage treatment plants (STPs) but result in N and P deficiency and unintended negative consequences. This review examines the potential impacts of N and P deficiency on the removal of organic carbon and nutrients, sludge characteristics and greenhouse gas emissions in activated sludge processes. The details of how these impacts affect the operation of STPs were also included. This review helps foresee operational challenges that established STPs may face when dealing with nutrient-deficient sewage in a future where source separation of urine is the norm. The findings indicate that the requirement of nitrification-denitrification and biological P removal processes could shrink at urine segregation above 80 % and 100 %, respectively. Organic carbon, N and biological P removal processes can be severely affected under full urine segregation. The decrease in solid retention time due to urine segregation increases treatment capacity up to 48 %. Sludge flocculation and settleability would deteriorate due to changes in extracellular polymeric substances and induce various forms of bulking. Beneficially, N deficiency reduces nitrous oxide emissions. These findings emphasise the importance of considering and preparing for impacts caused by urine source segregation-induced nutrient deficiency in sewage treatment processes.

Developing a conceptual model for understanding caregiving experience and their impacts on quality of life for Chinese breast cancer family caregivers: A qualitative study.

AIM: The purpose of this study was to understand the caregiving experiences of breast cancer family caregivers and explore the profound impacts of those experiences on their quality of life. DESIGN: A qualitative research method was used. METHODS: We extended invitations to 23 family caregivers of outpatients and inpatients receiving breast surgery and oncology treatments in Taiyuan, China, to participate in semi-structured interviews. The interviews were audio-recorded and transcribed verbatim. Thematic analysis was employed to analyse the interview data. RESULTS: Four themes and associated categories were identified: (1) changes in family dynamics, (2) the socio-medical context, (3) interactions between family and society, (4) self-efficacy and nine subthemes and their related categories, where virtually all participants expressed future uncertainty, emotional contagion, and personal challenges, and self-efficacy had a moderating influence on the first three themes. PATIENT OR PUBLIC CONTRIBUTION: This study did not involve direct participation of patients or the public. However, their experiences and perspectives on caregiving were indirectly reflected through the insights provided by the family caregivers who participated in the interviews. Their valuable input contributed to a deeper understanding of the caregiving experience and its impact on the quality of life for Chinese breast cancer family caregivers.

Occurrence, spatiotemporal trends, fate, and treatment technologies for microplastics and organic contaminants in biosolids: A review.

This review provides a comprehensive overview of the occurrence, fate, treatment and multi-criteria analysis of microplastics (MPs) and organic contaminants (OCs) in biosolids. A meta-analysis was complementarily analysed through the literature to map out the occurrence and fate of MPs and 10 different groups of OCs. The data demonstrate that MPs (54.7% occurrence rate) and linear alkylbenzene sulfonate surfactants (44.2% occurrence rate) account for the highest prevalence of contaminants in biosolids. In turn, dioxin, polychlorinated biphenyls (PCBs) and phosphorus flame retardants (PFRs) have the lowest rates (<0.01%). The occurrence of several OCs (e.g., dioxin, per- and polyfluoroalkyl substances, polycyclic aromatic hydrocarbons, pharmaceutical and personal care products, ultraviolet filters, phosphate flame retardants) in Europe appear at higher rates than in Asia and the Americas. However, MP concentrations in biosolids from Australia are reported to be 10 times higher than in America and Europe, which required more measurement data for in-depth analysis. Amongst the OC groups, brominated flame retardants exhibited exceptional sorption to biosolids with partitioning coefficients (log Kd) higher than 4. To remove these contaminants from biosolids, a wide range of technologies have been developed. Our multicriteria analysis shows that anaerobic digestion is the most mature and practical. Thermal treatment is a viable option; however, it still requires additional improvements in infrastructure, legislation, and public acceptance.

MRL-Seg: Overcoming Imbalance in Medical Image Segmentation With Multi-Step Reinforcement Learning.

Medical image segmentation is a critical task for clinical diagnosis and research. However, dealing with highly imbalanced data remains a significant challenge in this domain, where the region of interest (ROI) may exhibit substantial variations across different slices. This presents a significant hurdle to medical image segmentation, as conventional segmentation methods may either overlook the minority class or overly emphasize the majority class, ultimately leading to a decrease in the overall generalization ability of the segmentation results. To overcome this, we propose a novel approach based on multi-step reinforcement learning, which integrates prior knowledge of medical images and pixel-wise segmentation difficulty into the reward function. Our method treats each pixel as an individual agent, utilizing diverse actions to evaluate its relevance for segmentation. To validate the effectiveness of our approach, we conduct experiments on four imbalanced medical datasets, and the results show that our approach surpasses other state-of-the-art methods in highly imbalanced scenarios. These findings hold substantial implications for clinical diagnosis and research.

Multifaceted benefits of magnesium hydroxide dosing in sewer systems: Impacts on downstream wastewater treatment processes.

Magnesium hydroxide [Mg(OH)2] is a non-hazardous chemical widely applied in sewer systems for managing odour and corrosion. Despite its proven effectiveness in mitigating these issues, the impacts of dosing Mg(OH)2 in sewers on downstream wastewater treatment plants have not been comprehensively investigated. Through a one-year operation of laboratory-scale urban wastewater systems, including sewer reactors, sequencing batch reactors, and anaerobic sludge digesters, the findings indicated that Mg(OH)2 dosing in sewer systems had multifaceted benefits on downstream treatment processes. Compared to the control, the Mg(OH)2-dosed experimental system displayed elevated sewage pH (8.8±0.1vs 7.1±0.1), reduced sulfide concentration by 35.1%±4.9% (6.7±0.9mgSL-1), and lower methane concentration by 58.0%±4.9% (19.1±3.6mgCODL-1). Additionally, it increased alkalinity by 16.3%±2.2% (51.9±5.4mgCaCO3L-1), and volatile fatty acids concentration by 207.4%±22.2% (56.6±9.0mgCODL-1) in sewer effluent. While these changes offered limited advantages for downstream nitrogen removal in systems with sufficient alkalinity and carbon sources, significant improvements in ammonium oxidation rate and NOx reduction rate were observed in cases with limited alkalinity and carbon sources availability. Moreover, Mg(OH)2 dosing in upstream did not have any detrimental effects on anaerobic sludge digesters. Magnesium-phosphate precipitation led to a 31.7%±4.1% reduction in phosphate concertation in anaerobic digester sludge supernatant (56.1±10.4mgPL-1). The retention of magnesium in sludge increased settleability by 13.9%±1.6% and improved digested sludge dewaterability by 10.7%±5.3%. Consequently, the use of Mg(OH)2 dosing in sewers could potentially reduce downstream chemical demand and costs for carbon sources (e.g., acetate), pH adjustment and sludge dewatering.

Challenges of suppressing nitrite-oxidizing bacteria in membrane aerated biofilm reactors by low dissolved oxygen control.

Membrane aerated biofilm reactor (MABR) and shortcut nitrogen removal are two types of solutions to reduce energy consumption in wastewater treatment, with the former improving the aeration efficiency and the latter reducing the oxygen demand. However, integrating these two solutions, i.e., achieving shortcut nitrogen removal in MABR, is challenging due to the difficulty in suppressing nitrite-oxidizing bacteria (NOB). In this study, four MABRs were established to demonstrate the feasibility of initiating, maintaining, and restoring NOB suppression using low dissolved oxygen (DO) control, in the presence and absence of anammox bacteria, respectively. Long-term results revealed that the strict low DO (< 0.1 mg/L) in MABR could initiate and maintain stable NOB suppression for more than five months with nitrite accumulation ratio above 90 %, but it was unable to re-suppress NOB once they prevailed. Moreover, the presence of anammox bacteria increased the threshold of DO level to maintain NOB suppression in MABRs, but it was still incapable to restore the deteriorated NOB suppression in conjunction with low DO control. Mathematical modelling confirmed the experimental results and further explored the differences of NOB suppression in conventional biofilms and MABR biofilms. Simulation results showed that it is more challenging to maintain stable NOB suppression in MABRs compared to conventional biofilms, regardless of biofilm thickness or influent nitrogen concentration. Kinetic mechanisms for NOB suppression in different types of biofilms were proposed, suggesting that it is difficult to wash out NOB developed in the innermost layer of MABR biofilms because of the high oxygen level and low sludge wasting rate. In summary, this study systematically demonstrated the challenges of NOB suppression in MABRs through both experiments and mathematical modelling. These findings provide valuable insights into the applications of MABRs and call for more studies in developing effective strategies to achieve stable shortcut nitrogen removal in this energy-efficient configuration.

Decoupling the simultaneous effects of NO2-, pH and free nitrous acid on N2O and NO production from enriched nitrifying activated sludge.

In the pursuit of energy and carbon neutrality, nitrogen removal technologies have been developed featuring nitrite (NO2-) accumulation. However, high NO2- accumulations are often associated with stimulated greenhouse gas (i.e., nitrous oxide, N2O) emissions. Furthermore, the coexistence of free nitrous acid (FNA) formed by NO2- and proton (pH) makes the consequence of NO2- accumulation on N2O emissions complicated. The concurrent three factors, NO2-, pH and FNA may play different roles on N2O and nitric oxide (NO) emissions simultaneously, which has not been systematically studied. This study aims to decouple the effects of NO2- (0-200 mg N/L), pH (6.5-8) and FNA (0-0.15 mg N/L) on the N2O and NO production rates and the production pathways by ammonia oxidizing bacteria (AOB), with the use of a series of precisely executed batch tests and isotope site-preference analysis. Results suggested the dominant factors affecting the N2O production rate were NO2- and FNA concentrations, while pH alone played a relatively insignificant role. The most influential factor shifted from NO2- to FNA as FNA concentrations increased from 0 to 0.15 mg N/L. At concentrations below 0.0045 mg HNO2-N/L, nitrite rather than FNA played a significant role stimulating N2O production at elevated nitrite concentrations. The inhibition effect of FNA emerged with further increase of FNA between 0.0045-0.015 mg HNO2-N/L, weakening the promoting effect of increased nitrite. While at concentrations above 0.015 mg HNO2-N/L, FNA inhibited N2O production especially from nitrifier denitrification pathway with the level of inhibition linearly correlated with the FNA concentration. pH and the nitrite concentration regulated the production pathways, with elevated pH promoting the nitrifier nitrification pathway, while elevated NO2- concentrations promoting the nitrifier denitrification pathway. In contrast to N2O, NO emission was less susceptible to FNA at concentrations up to 0.015 mg N/L but was stimulated by increasing NO2- concentrations. This study, for the first time, distinguished the effects of pH, NO2- and FNA on N2O and NO production, thereby providing support to the design and operation of novel nitrogen removal systems with NO2- accumulation.

Self-supervised contrastive graph representation with node and graph augmentation.

Graph representation is a critical technology in the field of knowledge engineering and knowledge-based applications since most knowledge bases are represented in the graph structure. Nowadays, contrastive learning has become a prominent way for graph representation by contrasting positive-positive and positive-negative node pairs between two augmentation graphs. It has achieved new state-of-the-art in the field of self-supervised graph representation. However, existing contrastive graph representation methods mainly focus on modifying (normally removing some edges/nodes) the original graph structure to generate the augmentation graph for the contrastive. It inevitably changes the original graph structures, meaning the generated augmentation graph is no longer equivalent to the original graph. This harms the performance of the representation in many structure-sensitive graphs such as protein graphs, chemical graphs, molecular graphs, etc. Moreover, there is only one positive-positive node pair but relatively massive positive-negative node pairs in the self-supervised graph contrastive learning. This can lead to the same class, or very similar samples are considered negative samples. To this end, in this work, we propose a Virtual Masking Augmentation (VMA) to generate an augmentation graph without changing any structures from the original graph. Meanwhile, a node augmentation method is proposed to augment the positive node pairs by discovering the most similar nodes in the same graph. Then, two different augmentation graphs are generated and put into a contrastive learning model to learn the graph representation. Extensive experiments on massive datasets demonstrate that our method achieves new state-of-the-art results on self-supervised graph representation. The source code of the proposed method is available at https://github.com/DuanhaoranCC/CGRA.

Multi-view graph representation with similarity diffusion for general zero-shot learning.

Zero-shot learning (ZSL) aims to predict unseen classes without using samples of these classes in model training. The ZSL has been widely used in many knowledge-based models and applications to predict various parameters, including categories, subjects, and anomalies, in different domains. Nonetheless, most existing ZSL methods require the pre-defined semantics or attributes of particular data environments. Therefore, these methods are difficult to be applied to general data environments, such as ImageNet and other real-world datasets and applications. Recent research has tried to use open knowledge to enhance the ZSL methods to adapt it to an open data environment. However, the performance of these methods is relatively low, namely the accuracy is normally below 10%, which is due to the inadequate semantics that can be used from open knowledge. Moreover, the latest methods suffer from a significant "semantic gap" problem between the generated features of unseen classes and the real features of seen classes. To this end, this paper proposes a multi-view graph representation with a similarity diffusion model, applying the ZSL tasks to general data environments. This model applies a multi-view graph to enhance the semantics fully and proposes an innovative diffusion method to augment the graph representation. In addition, a feature diffusion method is proposed to augment the multi-view graph representation and bridge the semantic gap to realize zero-shot predicting. The results of numerous experiments in general data environments and on benchmark datasets show that the proposed method can achieve new state-of-the-art results in the field of general zero-shot learning. Furthermore, seven ablation studies analyze the effects of the settings and different modules of the proposed method on its performance in detail and prove the effectiveness of each module.

Insights on biological phosphorus removal with partial nitrification in single sludge system via sidestream free ammonia and free nitrous acid dosing.

The sidestream sludge treatment by free ammonium (FA)/free nitrous acid (FNA) dosing was frequently demonstrated to maintain the nitrite pathway for the partial nitrification (PN) process. Nevertheless, the inhibitory effect of FA and FNA would severely influence polyphosphate accumulating organisms (PAOs), destroying the microbe-based phosphorus (P) removal. Therefore, a strategic evaluation was proposed to successfully achieve biological P removal with a partial nitrification process in a single sludge system by sidestream FA and FNA dosing. Through the long-term operation of 500 days, excellent phosphorus, ammonium and total nitrogen removal performance were achieved at 97.5 ± 2.6 %, 99.1 ± 1.0 % and 75.5 ± 0.4 %, respectively. Stable partial nitrification with a nitrite accumulation ratio (NAR) of 94.1 ± 3.4 was attained. The batch tests also reported the robust aerobic phosphorus uptake based on FA and FNA adapted sludge after exposure of FA and FNA, respectively, suggesting the FA and FNA treatment strategy could potentially offer the opportunity for the selection of PAOs, which synchronously have the tolerance to FA and FNA. Microbial community analysis suggested that Accumulibacter, Tetrasphaera, and Comamonadaceae collectively contributed to the phosphorus removal in this system. Summarily, the proposed work presents a novel and feasible strategy to integrate enhanced biological phosphorus removal (EBPR) and short-cut nitrogen cycling and bring the combined mainstream phosphorus removal and partial nitrification process closer to practical application.

eIF3 mRNA selectivity profiling reveals eIF3k as a cancer-relevant regulator of ribosome content.

eIF3, whose subunits are frequently overexpressed in cancer, regulates mRNA translation from initiation to termination, but mRNA-selective functions of individual subunits remain poorly defined. Using multiomic profiling upon acute depletion of eIF3 subunits, we observed that while eIF3a, b, e, and f markedly differed in their impact on eIF3 holo-complex formation and translation, they were each required for cancer cell proliferation and tumor growth. Remarkably, eIF3k showed the opposite pattern with depletion promoting global translation, cell proliferation, tumor growth, and stress resistance through repressing the synthesis of ribosomal proteins, especially RPS15A. Whereas ectopic expression of RPS15A mimicked the anabolic effects of eIF3k depletion, disruption of eIF3 binding to the 5'-UTR of RSP15A mRNA negated them. eIF3k and eIF3l are selectively downregulated in response to endoplasmic reticulum and oxidative stress. Supported by mathematical modeling, our data uncover eIF3k-l as a mRNA-specific module which, through controlling RPS15A translation, serves as a rheostat of ribosome content, possibly to secure spare translational capacity that can be mobilized during stress.

Dynamic Unary Convolution in Transformers.

It is uncertain whether the power of transformer architectures can complement existing convolutional neural networks. A few recent attempts have combined convolution with transformer design through a range of structures in series, where the main contribution of this paper is to explore a parallel design approach. While previous transformed-based approaches need to segment the image into patch-wise tokens, we observe that the multi-head self-attention conducted on convolutional features is mainly sensitive to global correlations and that the performance degrades when these correlations are not exhibited. We propose two parallel modules along with multi-head self-attention to enhance the transformer. For local information, a dynamic local enhancement module leverages convolution to dynamically and explicitly enhance positive local patches and suppress the response to less informative ones. For mid-level structure, a novel unary co-occurrence excitation module utilizes convolution to actively search the local co-occurrence between patches. The parallel-designed Dynamic Unary Convolution in Transformer (DUCT) blocks are aggregated into a deep architecture, which is comprehensively evaluated across essential computer vision tasks in image-based classification, segmentation, retrieval and density estimation. Both qualitative and quantitative results show our parallel convolutional-transformer approach with dynamic and unary convolution outperforms existing series-designed structures.

Novel Use of a Ferric Salt to Enhance Mainstream Nitrogen Removal from Anaerobically Pretreated Wastewater.

This study aims to demonstrate a new technology roadmap to support the ongoing paradigm shift in wastewater management from pollutant removal to resource recovery. This is achieved by developing a novel use of an iron salt (i.e., FeCl3) in an integrated anaerobic wastewater treatment and mainstream anammox process. FeCl3 was chosen to be dosed in a proposed sidestream unit rather than in a primary settler or a mainstream reactor. This causes acidification of returned activated sludge and enables stable suppression of nitrite-oxidizing bacterial activity and excess sludge reduction. A laboratory-scale system, which comprised an anaerobic baffled reactor, a continuous-flow anoxic-aerobic (A/O) reactor, and a secondary settler, was designed to treat real domestic wastewater, with the performance of the system comprehensively monitored under a steady-state condition. The experimental assessments showed that the system had good effluent quality, with total nitrogen and phosphorus concentrations of 12.6 ± 1.3 mg N/L and 0.34 ± 0.05 mg P/L, respectively. It efficiently retained phosphorus in excess sludge (0.18 ± 0.03 g P/g dry sludge), suggesting its potential for further recovery. About half of influent organic carbon was recovered in the form of bioenergy (i.e., methane). This together with low energy consumption revealed that the system could produce a net energy of about 0.11 kWh/m3-wastewater, assessed by an energy balance analysis.

In a quest for high-efficiency mainstream partial nitritation-anammox (PN/A) implementation: One-stage or two-stage?

Partial nitritation-anammox (PN/A) process is known as an energy-efficient technology for wastewater nitrogen removal, which possesses a great potential to bring wastewater treatment plants close to energy neutrality with reduced carbon footprint. To achieve this goal, various PN/A processes implemented in a single reactor configuration (one-stage system) or two separately dedicated reactors configurations (two-stage system) were explored over the past decades. Nevertheless, large-scale implementation of these PN/A processes for low-strength municipal wastewater treatment has a long way to go owing to the low efficiency and effectiveness in nitrogen removal. In this work, we provided a comprehensive analysis of one-stage and two-stage PN/A processes with a focus on evaluating their engineering application potential towards mainstream implementation. The difficulty for nitrite-oxidizing bacteria (NOB) out-selection was revealed as the critical operational challenge to achieve the desired effluent quality. Additionally, the operational strategies of low oxygen commonly adopted in one-stage systems for NOB suppression and facilitating anammox bacteria growth results in a low nitrogen removal rate (NRR). Introducing denitrification into anammox system was found to be necessary to improve the nitrogen removal efficiency (NRE) by reducing the produced nitrate with in-situ utilizing the organics from wastewater itself. However, this may lead to part of organics oxidized with additional oxygen consumed in one-stage system, further compromising the NRR. By applying a relatively high dissolved oxygen in PN reactor with residual ammonium control, and followed by a granules-based anammox reactor feeding with a small portion of raw municipal wastewater, it appeared that two-stage system could achieve a good effluent quality as well as a high NRR. In contrast to the widely studied one-stage system, this work provided a unique perspective that more effort should be devoted to developing a two-stage PN/A process to evaluate its application potential of high efficiency and economic benefits towards mainstream implementation.