SARS-CoV-2 replication in airway epithelia requires motile cilia and microvillar reprogramming.

How SARS-CoV-2 penetrates the airway barrier of mucus and periciliary mucins to infect nasal epithelium remains unclear. Using primary nasal epithelial organoid cultures, we found that the virus attaches to motile cilia via the ACE2 receptor. SARS-CoV-2 traverses the mucus layer, using motile cilia as tracks to access the cell body. Depleting cilia blocks infection for SARS-CoV-2 and other respiratory viruses. SARS-CoV-2 progeny attach to airway microvilli 24 h post-infection and trigger formation of apically extended and highly branched microvilli that organize viral egress from the microvilli back into the mucus layer, supporting a model of virus dispersion throughout airway tissue via mucociliary transport. Phosphoproteomics and kinase inhibition reveal that microvillar remodeling is regulated by p21-activated kinases (PAK). Importantly, Omicron variants bind with higher affinity to motile cilia and show accelerated viral entry. Our work suggests that motile cilia, microvilli, and mucociliary-dependent mucus flow are critical for efficient virus replication in nasal epithelia.

Combined protein and nucleic acid imaging reveals virus-dependent B cell and macrophage immunosuppression of tissue microenvironments.

Understanding the mechanisms of HIV tissue persistence necessitates the ability to visualize tissue microenvironments where infected cells reside; however, technological barriers limit our ability to dissect the cellular components of these HIV reservoirs. Here, we developed protein and nucleic acid in situ imaging (PANINI) to simultaneously quantify DNA, RNA, and protein levels within these tissue compartments. By coupling PANINI with multiplexed ion beam imaging (MIBI), we measured over 30 parameters simultaneously across archival lymphoid tissues from healthy or simian immunodeficiency virus (SIV)-infected nonhuman primates. PANINI enabled the spatial dissection of cellular phenotypes, functional markers, and viral events resulting from infection. SIV infection induced IL-10 expression in lymphoid B cells, which correlated with local macrophage M2 polarization. This highlights a potential viral mechanism for conditioning an immunosuppressive tissue environment for virion production. The spatial multimodal framework here can be extended to decipher tissue responses in other infectious diseases and tumor biology.

Robust single-cell matching and multimodal analysis using shared and distinct features.

The ability to align individual cellular information from multiple experimental sources is fundamental for a systems-level understanding of biological processes. However, currently available tools are mainly designed for single-cell transcriptomics matching and integration, and generally rely on a large number of shared features across datasets for cell matching. This approach underperforms when applied to single-cell proteomic datasets due to the limited number of parameters simultaneously accessed and lack of shared markers across these experiments. Here, we introduce a cell-matching algorithm, matching with partial overlap (MARIO) that accounts for both shared and distinct features, while consisting of vital filtering steps to avoid suboptimal matching. MARIO accurately matches and integrates data from different single-cell proteomic and multimodal methods, including spatial techniques and has cross-species capabilities. MARIO robustly matched tissue macrophages identified from COVID-19 lung autopsies via codetection by indexing imaging to macrophages recovered from COVID-19 bronchoalveolar lavage fluid by cellular indexing of transcriptomes and epitopes by sequencing, revealing unique immune responses within the lung microenvironment of patients with COVID.

Safety and Immunogenicity of mRNA-1010, an Investigational Seasonal Influenza Vaccine, in Healthy Adults: Final Results From a Phase 1/2 Randomized Trial.

BACKGROUND: Seasonal influenza remains a global public health concern. A messenger RNA (mRNA)-based quadrivalent seasonal influenza vaccine, mRNA-1010, was investigated in a 3-part, first-in-human, phase 1/2 clinical trial. METHODS: In Parts 1-3 of this stratified, observer-blind study, adults aged ≥18 years old were randomly assigned to receive a single dose (6.25 µg to 200 µg) of mRNA-1010 or placebo (Part 1) or an active comparator (Afluria; Parts 2-3). Primary study objectives were assessment of safety, reactogenicity, and humoral immunogenicity of mRNA-1010, placebo (Part 1), or active comparator (Parts 2-3). Exploratory endpoints included assessment of cellular immunogenicity (Part 1) and antigenic breadth against vaccine heterologous (A/H3N2) strains (Parts 1-2). RESULTS: In all study parts, solicited adverse reactions were reported more frequently for mRNA-1010 than placebo or Afluria and most were grade 1 or 2 in severity. No vaccine-related serious adverse events or deaths were reported. In Parts 1-2, a single dose of mRNA-1010 (25 µg to 200 µg) elicited robust Day 29 hemagglutination inhibition (HAI) titers that persisted through 6 months. In Part 3, lower doses of mRNA-1010 (6.25 µg to 25 µg) elicited Day 29 HAI titers that were higher or comparable to Afluria for influenza A strains. Compared with Afluria, mRNA-1010 (50 µg) elicited broader A/H3N2 antibody responses (Part 2). mRNA-1010 induced greater T-cell responses than placebo at Day 8 that were sustained or stronger at Day 29 (Part 1). CONCLUSIONS: Data support the continued development of mRNA-1010 as a seasonal influenza vaccine. CLINICALTRIALS.GOV IDENTIFIER: NCT04956575 (https://clinicaltrials.gov/study/NCT04956575).

A comprehensive, open-source data model for wastewater-based epidemiology.

The recent SARS-COV-2 pandemic has sparked the adoption of wastewater-based epidemiology (WBE) as a low-cost way to monitor the health of populations. In parallel, the pandemic has encouraged researchers to openly share their data to serve the public better and accelerate science. However, environmental surveillance data are highly dependent on context and are difficult to interpret meaningfully across sites. This paper presents the second iteration of the Public Health Environmental Surveillance Open Data Model (PHES-ODM), an open-source dictionary and set of data tools to enhance the interoperability of environmental surveillance data and enable the storage of contextual (meta)data. The data model describes how to store environmental surveillance program data, metadata about measurements taken on various specimens (water, air, surfaces, sites, populations) and data about measurement protocols. The model provides software tools that support the collection and use of PHES-ODM formatted data, including performing PCR calculations and data validation, recording data into input templates, generating wide tables for analysis, and producing SQL database definitions. Fully open-source and already adopted by institutions in Canada, the European Union, and other countries, the PHES-ODM provides a path forward for creating robust, interoperable, open datasets for environmental public health surveillance for SARS-CoV-2 and beyond.

Inflammatory molecular endotypes of nasal polyps derived from White and Japanese populations.

BACKGROUND: Emerging evidence suggests that chronic rhinosinusitis with nasal polyps (CRSwNP) is a highly heterogeneous disease with disparate inflammatory characteristics between different racial groups and geographies. Currently, little is known about possible underlying distinguishing factors between these inflammatory differences. OBJECTIVE: Our aim was to interrogate differences in CRSwNP disease between White/non-Asian patients and Japanese patients by using whole transcriptome and single-cell RNA gene expression profiling of nasal polyps (NPs). METHODS: We performed whole transcriptome RNA sequencing with endotype stratification of NPs from 8 White patients (residing in the United States) and 9 Japanese patients (residing in Japan). Reproducibility was confirmed by quantitative PCR in an independent validation set of 46 White and 31 Japanese patients. Single-cell RNA sequencing (scRNAseq) was used to stratify key cell types for contributory transcriptional signatures. RESULTS: Unsupervised clustering analysis identified 2 major endotypes that were present within both cohorts of patients with NPs and had previously been reported at the cytokine level: (1) type 2 endotype and (2) non-type 2 endotype. Importantly, there was a statistically significant difference in the proportion of these endotypes between these geographically distinct subgroups with NPs (P = .03). Droplet-based single-cell RNA sequencing further identified prominent type 2 inflammatory transcript expression: C-C motif chemokine ligand 13 (CCL13) and CCL18 in M2 macrophages, as well as cystatin SN (CST1) and CCL26 in basal, suprabasal, and secretory epithelial cells. CONCLUSION: NPs from both racial groups harbor the same 2 major endotypes, which we have determined to be present in differing ratios between each cohort with CRSwNP disease. Distinct inflammatory and epithelial cells contribute to the type 2 inflammatory profiles observed.

Ocular syphilis on the rise: a 10-year analysis from 2010 to 2020.

PURPOSE: The purpose of the study was to evaluate the incidence of ocular syphilis as well as diagnostic parameters, comorbidities, and visual outcomes over a 10-year time period in West Virginia. METHODS: A retrospective chart review included 25 eyes of 17 patients with ocular syphilis between 2010 and 2020. RESULTS: The incidence of systemic syphilis at a large tertiary referral center has increased from 27 cases in 2010 to 105 cases in 2020. Seventeen patients were identified with ocular syphilis. Bilaterality was present in 47.1% of cases. In this study, 70.6% of patients were male and 29.4% were female. The median age of presentation was 40.2 years (range 21-63). Panuveitis was the most common (60.0%) followed by isolated anterior uveitis (16.0%), chorioretinitis (12.0%), inner retinitis (4.0%), and papillitis (8.0%). Forty percent of patients had visual acuity worse than 20/400 on presentation. Post-treatment visual acuity improved in all patients. Rapid plasma reagin (RPR) and Treponema pallidum particle agglutination (TP-PA) tests were positive in 84.6% and 100% of cases, respectively. CSF venereal disease research laboratory (VDRL) was positive in 36.4%, CSF pleocytosis was present in 72.7%, and elevated CSF protein was observed in 81.8%. Human immunodeficiency virus (HIV) co-infection was present in 31.3%. A majority of patients experienced maculopapular rash and/or history of genital chancre. The anatomic classification of presenting uveitis (anterior, intermediate, posterior, and panuveitis) did not correlate with clinical variables including age, gender, HIV status, serologic test, presence of rash, or year of diagnosis (p > 0.05). CONCLUSION: Ocular syphilis is becoming increasingly prevalent and can present with a variety of ocular findings; therefore, it should be considered in the differential diagnosis for patients with ocular inflammation. Visual prognosis is excellent with timely diagnosis and treatment.

NEAR: Named entity and attribute recognition of clinical concepts.

Named Entity Recognition (NER) or the extraction of concepts from clinical text is the task of identifying entities in text and slotting them into categories such as problems, treatments, tests, clinical departments, occurrences (such as admission and discharge) and others. NER forms a critical component of processing and leveraging unstructured data from Electronic Health Records (EHR). While identifying the spans and categories of concepts is itself a challenging task, these entities could also have attributes such as negation that pivot their meanings implied to the consumers of the named entities. There has been little research dedicated to identifying the entities and their qualifying attributes together. This research hopes to contribute to the area of detecting entities and their corresponding attributes by modelling the NER task as a supervised, multi-label tagging problem with each of the attributes assigned tagging sequence labels. In this paper, we propose 3 architectures to achieve this multi-label entity tagging: BiLSTM n-CRF, BiLSTM-CRF-Smax-TF and BiLSTM n-CRF-TF. We evaluate these methods on the 2010 i2b2/VA and the i2b2 2012 shared task datasets. Our different models obtain best NER scores of 0.903 and 0.808 on the i2b2 2010/VA and i2b2 2012 respectively. The highest span based micro-averaged F1 polarity scores obtained were 0.832 and 0.836 on the i2b2 2010/VA and i2b2 2012 datasets respectively, and the highest macro-averaged F1 polarity scores obtained were 0.924 and 0.888 respectively. The modality studies conducted on i2b2 2012 dataset revealed high scores of 0.818 and 0.501 for span based micro-averaged F1 and macro-averaged F1 respectively.

Research advancements in optical imaging and spectroscopic techniques for nondestructive detection of mold infection and mycotoxins in cereal grains and nuts.

Cereal grains and nuts are represented as the economic backbone of many developed and developing countries. Kernels of cereal grains and nuts are prone to mold infection under high relative humidity and suitable temperature conditions in the field as well as storage conditions. Health risks caused by molds and their molecular metabolite mycotoxins are, therefore, important topics to investigate. Strict regulations have been developed by international trade regulatory bodies for the detection of mold growth and mycotoxin contamination across the food chain starting from the harvest to storage and consumption. Molds and aflatoxins are not evenly distributed over the bulk of grains, thus appropriate sampling for detection and quantification is crucial. Existing reference methods for mold and mycotoxin detection are destructive in nature as well as involve skilled labor and hazardous chemicals. Also, these methods cannot be used for inline sorting of the infected kernels. Thus, analytical methods have been extensively researched to develop the one that is more practical to be used in commercial detection and sorting processes. Among various analytical techniques, optical imaging and spectroscopic techniques are attracting growers' attention for their potential of nondestructive and rapid inline identification and quantification of molds and mycotoxins in various food products. This review summarizes the recent application of rapid and nondestructive optical imaging and spectroscopic techniques, including digital color imaging, X-ray imaging, near-infrared spectroscopy, fluorescent, multispectral, and hyperspectral imaging. Advance chemometric techniques to identify very low-level mold growth and mycotoxin contamination are also discussed. Benefits, limitations, and challenges of deploying these techniques in practice are also presented in this paper.

Impact of clinicopathologic factors on survival in patients with sebaceous carcinoma of the eyelid - a population-based analysis.

Purpose: To evaluate prognostic factors and survival of patients with sebaceous carcinoma of the eyelid through a population-based analysis. Methods: A total of 940 patients with primary sebaceous carcinoma of the eyelid were derived from the Surveillance, Epidemiology, and End Results (SEER) database in the United States from 1973 to 2013. Kaplan-Meier univariate analysis and Cox Regression multivariate analysis were performed to examine prognostic factors in overall survival (OS). 5- and 10-year survival rates, median survival, and prognostic variables with statistical significance were measured. Results: Kaplan-Meier analysis showed that OS is 66% and 44% at 5 years and 10 years respectively. Median OS is 9.4 years. Multivariate Cox regression analysis demonstrated that independent prognostic factors for OS are age at diagnosis (HR = 4.61 [95% CI 1.93-11.0], P = 0.001), surgical treatment (HR = 0.196 [95% CI 0.07-0.55], P = 0.002), combined surgical and radiation treatment (HR = 0.227 [95% CI 0.06-0.81], P = 0.023), and greater tumor size at diagnosis (HR = 3.381, [95% CI 1.77-6.45], P < 0.001). Conclusion: We report the largest population study to date to evaluate prognostic factors of patients with sebaceous carcinoma of the eyelid. Multivariate analysis shows that older age, and greater tumor size correlate with decreased overall survival, whereas surgical treatment or combined surgical and radiation treatment correlate with increased overall survival. Interestingly, tumor grade, lymph node involvement, and distant extent of tumor have not demonstrated to be independent prognostic factors for overall survival.

Combining anti-IgE with oral immunotherapy.

Food allergy is a significant medical problem that affects up to 8% of children in developed countries. At present, there are no curative therapies available in routine practice and management of food allergy involves strict allergen avoidance, education, and prompt treatment upon accidental exposure. Oral immunotherapy (OIT) is an efficacious experimental approach to food allergy and has been shown to provide a substantial benefit in terms of allergen desensitization. However, OIT is associated with high rates of allergic reactions, and the period of protection offered by OIT appears to be limited and highly variable. Recurrence of allergen sensitivity after a period of treatment discontinuation is commonly observed. With the aim of overcoming these limitations of OIT, several trials have studied omalizumab (anti-IgE monoclonal antibody) as an adjuvant treatment for patients undergoing OIT. Results from these trials have shown that the addition of omalizumab to OIT leads to a significant decrease in the frequency and severity of reactions, which allows for an increase in the threshold of tolerance to food allergens. This review provides a summary of the current literature and addresses some of the key questions that remain regarding the use of omalizumab in conjunction with OIT.

Extensive diversity in circadian regulation of plasma lipids and evidence for different circadian metabolic phenotypes in humans.

The circadian system regulates daily rhythms in lipid metabolism and adipose tissue function. Although disruption of circadian clock function is associated with negative cardiometabolic end points, very little is known about interindividual variation in circadian-regulated metabolic pathways. Here, we used targeted lipidomics-based approaches to profile the time course of 263 lipids in blood plasma in 20 healthy individuals. Over a span of 28 h, blood was collected every 4 h and plasma lipids were analyzed by HPLC/MS. Across subjects, about 13% of lipid metabolites showed circadian variation. Rhythmicity spanned all metabolite classes examined, suggesting widespread circadian control of lipid-mediated energy storage, transport, and signaling. Intersubject agreement for lipids identified as rhythmic was only about 20%, however, and the timing of lipid rhythms ranged up to 12 h apart between individuals. Healthy subjects therefore showed substantial variation in the timing and strength of rhythms across different lipid species. Strong interindividual differences were also observed for rhythms of blood glucose and insulin, but not cortisol. Using consensus clustering with iterative feature selection, subjects clustered into different groups based on strength of rhythmicity for a subset of triglycerides and phosphatidylcholines, suggesting that there are different circadian metabolic phenotypes in the general population. These results have potential implications for lipid metabolism disorders linked to circadian clock disruption.

Wastewater-based surveillance of SARS-CoV-2: Short-term projection (forecasting), smoothing and outlier identification using Bayesian smoothing.

BACKGROUND: Day-to-day variation in the measurement of SARS-CoV-2 in wastewater can challenge public health interpretation. We assessed a Bayesian smoothing and forecasting method previously used for surveillance and short-term projection of COVID-19 cases, hospitalizations, and deaths. METHODS: SARS-CoV-2 viral measurement from the sewershed in Ottawa, Canada, sampled at the municipal wastewater treatment plant from July 1, 2020, to February 15, 2022, was used to assess and internally validate measurement averaging and prediction. External validation was performed using viral measurement data from influent wastewater samples from 15 wastewater treatment plants and municipalities across Ontario. RESULTS: Plots of SARS-CoV-2 viral measurement over time using Bayesian smoothing visually represented distinct COVID-19 "waves" described by case and hospitalization data in both initial (Ottawa) and external validation in 15 Ontario communities. The time-varying growth rate of viral measurement in wastewater samples approximated the growth rate observed for cases and hospitalization. One-week predicted viral measurement approximated the observed viral measurement throughout the assessment period from December 23, 2020, to August 8, 2022. An uncalibrated model showed underprediction during rapid increases in viral measurement (positive growth) and overprediction during rapid decreases. After recalibration, the model showed a close approximation between observed and predicted estimates. CONCLUSION: Bayesian smoothing of wastewater surveillance data of SARS-CoV-2 allows for accurate estimates of COVID-19 growth rates and one- and two-week forecasting of SARS-CoV-2 in wastewater for 16 municipalities in Ontario, Canada. Further assessment is warranted in other communities representing different sewersheds and environmental conditions.

A review of the immunogenicity and safety of booster doses of omicron variant-containing mRNA-1273 COVID-19 vaccines in adults and children.

INTRODUCTION: Vaccination against SARS-CoV-2 is an integral pillar of the public health approach to COVID-19. With the emergence of variants of concern that increase transmissibility and escape from vaccine- or infection-induced protection, vaccines have been developed to more closely match the newly circulating SARS-CoV-2 strains to improve protection. The safety and immunogenicity of multiple authorized messenger RNA (mRNA)-based COVID-19 vaccines targeting the omicron sublineage (BA.1, BA.4/BA.5, and XBB.1.5) have been demonstrated in several clinical trials among adults and children. AREAS COVERED: This review will comprehensively detail the available evidence (published through July 2024) from ongoing clinical trials on omicron variant-containing mRNA-1273 vaccines administered as additional doses in previously vaccinated target demographics. EXPERT OPINION: Across three clinical trials, omicron variant-containing mRNA-1273 vaccines induced immune responses to vaccine-matched omicron strains as well as ancestral SARS-CoV-2, with a safety and reactogenicity profile comparable to the original mRNA-1273 vaccine. Combined with pivotal data demonstrating the safety, efficacy, and effectiveness of the original mRNA-1273 vaccine, these findings support the use of variant-containing mRNA-1273 vaccines and provide confidence that expeditious development of updated vaccines using this established mRNA platform can maintain protection against COVID-19.

Application of specialized word embeddings and named entity and attribute recognition to the problem of unsupervised automated clinical coding.

Notes documented by clinicians, such as patient histories, hospital courses, lab reports and others are often annotated with standardized clinical codes by medical coders to facilitate a variety of secondary processing applications such as billing and statistical analyses. Clinical coding, traditionally manual and labor-intensive, has seen a surge in research interest by deep learning researchers pursuing to automate it. However, deep learning methods require large volumes of annotated clinical data for training and offer little to explain why codes were assigned to pieces of text. In this paper, we propose an unsupervised method which does not need annotated clinical text and is fully interpretable, by using Named Entity and Attribute Recognition and word embeddings specialized for the clinical domain. These methods successfully glean important information from large volumes of clinical notes and encode them effectively in order to perform automatic clinical coding.

Incidence of Epiretinal Membrane Formation After Pars Plana Vitrectomy for Giant Retinal Tear-Associated Retinal Detachment.

Objective: To report the incidence of postoperative epiretinal membrane (ERM) formation after primary pars plana vitrectomy (PPV) for giant retinal tear associated retinal detachment (GRT-RD) repair as well as its clinical characteristics and visual outcomes at a level one trauma and tertiary referral academic center. Patients and Methods: Patients with primary RD repair for GRT-RD at West Virginia University from September 2010 to July 2021 were identified using the ICD-10 codes (H33.031, H33.032, H33.033 and H33.039). Imaging studies including optical coherence tomography (OCT) were manually reviewed pre- and post-operatively for ERM formation after PPV for GRT-RD repair in patients who underwent PPV or combined PPV and scleral buckle (SB). Univariate analysis was performed to analyze clinical factors for ERM formation. Results: The study included 17 eyes of 16 patients who underwent PPV for GRT-RD. Postoperative ERM was observed in 70.6% (13 of 17 eyes) of the patients. Anatomic success was achieved in all patients. The mean (range) preoperative and final best corrected visual acuity (BCVA) in logMAR units by macula status was 0.19 (0-0.5) and 0.28 (0-0.5) for macula-on and 1.7 (0.5-2.3) and 0.7 (0.2-1.9) for macular-off GRT-RDs. Clinical variables including use of medium-term tamponade with perfluorocarbon liquid (PFCL), cryopexy, endodiathermy, number of tears or total clock hours of tears did not correlate with an increased risk of ERM formation. Conclusion: Post-vitrectomized eyes for GRT-RD repair have a significantly higher incidence of ERM formation, nearing 70% in our study. Surgeons may consider prophylactic ILM peel at the time of removal of tamponade agents or weigh in ILM peel at the time of primary repair, a more challenging surgical technique in our opinion.

Five-Year Follow-Up of Repeated Intravitreal Bevacizumab for Macular Edema in a Pediatric Patient with Retinal Arteriovenous Malformation and Excellent Vision.

We report a case of a 9-year-old girl presenting with unilateral retinal arteriovenous malformation (AVM) with symptomatic macular edema. Over 5 years of follow-up includes optical coherence tomography (OCT), fundus photographs, and fluorescein angiography at baseline and at follow-up. Systemic and neurologic workup was completed and negative for intracranial AVM. Vision has correlated with macular edema, ranging from 20/20 to 20/80. The patient has received nine injections of intravitreal bevacizumab and has not required an injection for the last couple of years. Follow-up is ongoing.

Explainable Knowledge Distillation for On-Device Chest X-Ray Classification.

Automated multi-label chest X-rays (CXR) image classification has achieved substantial progress in clinical diagnosis via utilizing sophisticated deep learning approaches. However, most deep models have high computational demands, which makes them less feasible for compact devices with low computational requirements. To overcome this problem, we propose a knowledge distillation (KD) strategy to create the compact deep learning model for the real-time multi-label CXR image classification. We study different alternatives of CNNs and Transforms as the teacher to distill the knowledge to a smaller student. Then, we employed explainable artificial intelligence (XAI) to provide the visual explanation for the model decision improved by the KD. Our results on three benchmark CXR datasets show that our KD strategy provides the improved performance on the compact student model, thus being the feasible choice for many limited hardware platforms. For instance, when using DenseNet161 as the teacher network, EEEA-Net-C2 achieved an AUC of 83.7%, 87.1%, and 88.7% on the ChestX-ray14, CheXpert, and PadChest datasets, respectively, with fewer parameters of 4.7 million and computational cost of 0.3 billion FLOPS.

Unpaired Artistic Portrait Style Transfer via Asymmetric Double-Stream GAN.

With the development of image style transfer technologies, portrait style transfer has attracted growing attention in this research community. In this article, we present an asymmetric double-stream generative adversarial network (ADS-GAN) to solve the problems that caused by cartoonization and other style transfer techniques when they are applied to portrait photos, such as facial deformation, contours missing, and stiff lines. By observing the characteristics between source and target images, we propose an edge contour retention (ECR) regularized loss to constrain the local and global contours of generated portrait images to avoid the portrait deformation. In addition, a content-style feature fusion module is introduced for further learning of the target image style, which uses a style attention mechanism to integrate features and embeds style features into content features of portrait photos according to the attention weights. Finally, a guided filter is introduced in content encoder to smooth the textures and specific details of source image, thereby eliminating its negative impact on style transfer. We conducted overall unified optimization training on all components and got an ADS-GAN for unpaired artistic portrait style transfer. Qualitative comparisons and quantitative analyses demonstrate that the proposed method generates superior results than benchmark work in preserving the overall structure and contours of portrait; ablation and parameter study demonstrate the effectiveness of each component in our framework.