Clinical features and possible pathogenesis of multiple evanescent white dot syndrome with different retinal diseases and events: a narrative review.

Multiple evanescent white dot syndrome (MEWDS) is a rare fundus disease, characterized by acute vision loss and visual field defects. Many previous studies have explained the possible pathogenesis and clinical features of primary MEWDS. However, as the number of reported cases increases, secondary MEWDS occurs in other related retinal diseases and injuries, exhibiting some special characteristics. The associated retinal diseases include multifocal choroiditis/punctate inner choroidopathy (MFC/PIC), acute zonal occult outer retinopathy, best vitelliform macular dystrophy, pseudoxanthoma elasticum, and ocular toxoplasmosis. The related retinal injury is laser photocoagulation, surgery, and trauma. Although primary MEWDS often have a self-limiting course, secondary MEWDS may require treatment in some cases, according to the severity of concomitant diseases and complications. Notably, MEWDS secondary to MFC/PIC that is prone to forming choroidal neovascularization and focal choroidal excavation, needs positive treatment with corticosteroids. The possible underlying pathogenesis of secondary MEWDS is the exposure of choroidal antigen after the disruption of Bruch's membrane. The MEWDS-related features in secondary MEWDS are still evanescent under most circumstances. Its prognosis and treatment depend on the severity of complications. Current studies propose that the etiology is associated with immune factors, including viral infection, inflammation in choroid and Bruch's membrane, and antigen exposure caused by retinal and/or choroidal insults. More pathogenic studies should be conducted in the future. Accurate diagnosis for secondary MEWDS could benefit patients in aspects of management and prognosis.

Artificial intelligence powered advancements in upper extremity joint MRI: A review.

Magnetic resonance imaging (MRI) is an indispensable medical imaging examination technique in musculoskeletal medicine. Modern MRI techniques achieve superior high-quality multiplanar imaging of soft tissue and skeletal pathologies without the harmful effects of ionizing radiation. Some current limitations of MRI include long acquisition times, artifacts, and noise. In addition, it is often challenging to distinguish abutting or closely applied soft tissue structures with similar signal characteristics. In the past decade, Artificial Intelligence (AI) has been widely employed in musculoskeletal MRI to help reduce the image acquisition time and improve image quality. Apart from being able to reduce medical costs, AI can assist clinicians in diagnosing diseases more accurately. This will effectively help formulate appropriate treatment plans and ultimately improve patient care. This review article intends to summarize AI's current research and application in musculoskeletal MRI, particularly the advancement of DL in identifying the structure and lesions of upper extremity joints in MRI images.

The role of iron-rich hydrosaline liquids in the formation of Kiruna-type iron oxide-apatite deposits.

Kiruna-type iron oxide-apatite (IOA) deposits, an important source of iron, show close associations with andesitic subvolcanic intrusions. However, the processes of ore formation and the mechanism controlling iron concentration remain uncertain. Here, we report the widespread presence of high-temperature (>800°C) water-poor multisolid hydrosaline liquid inclusions in pre- and syn-ore minerals from IOA deposits of eastern China. These inclusions consistently homogenize to a liquid phase by vapor disappearance and mostly contain 3 to 10 wt % Fe, signifying a substantial capacity for iron transportation by such hydrosaline liquids. We propose that the hydrosaline liquids were likely immiscible from the dioritic magmas with high Cl/H2O in subvolcanic settings. Subsequent reaction with host rocks and/or decompression and cooling of the hydrosaline liquids is deemed responsible for the simultaneous formation of high-temperature alteration and magnetite ores, thereby providing important insights into the distinctive characteristics of IOA deposits in shallow magmatic-hydrothermal systems.

Comparative Analysis of Six Complete Plastomes of Tripterospermum spp.

The Tripterospermum, comprising 34 species, is a genus of Gentianaceae. Members of Tripterospermum are mostly perennial, entwined herbs with high medicinal value and rich in iridoids, xanthones, flavonoids, and triterpenes. However, our inadequate understanding of the differences in the plastid genome sequences of Tripterospermum species has severely hindered the study of their evolution and phylogeny. Therefore, we first analyzed the 86 Gentianae plastid genomes to explore the phylogenetic relationships within the Gentianae subfamily where Tripterospermum is located. Then, we analyzed six plastid genomes of Tripterospermum, including two newly sequenced plastid genomes and four previously published plastid genomes, to explore the plastid genomes' evolution and phylogenetic relationships in the genus Tripterospermum. The Tripterospermum plastomes have a quadripartite structure and are between 150,929 and 151,350 bp in size. The plastomes of Tripterospermum encoding 134 genes were detected, including 86 protein-coding genes (CDS), 37 transfer RNA (tRNA) genes, eight ribosomal RNA (rRNA) genes, and three pseudogenes (infA, rps19, and ycf1). The result of the comparison shows that the Tripterospermum plastomes are very conserved, with the total plastome GC content ranging from 37.70% to 37.79%. In repeat sequence analysis, the number of single nucleotide repeats (A/T) varies among the six Tripterospermum species, and the identified main long repeat types are forward and palindromic repeats. The degree of conservation is higher at the SC/IR boundary. The regions with the highest divergence in the CDS and the intergenic region (IGS) are psaI and rrn4.5-rrn5, respectively. The average pi of the CDS and the IGS are only 0.071% and 0.232%, respectively, indicating that the Tripterospermum plastomes are highly conserved. Phylogenetic analysis indicated that Gentianinae is divided into two clades, with Tripterospermum as a sister to Sinogeniana. Phylogenetic trees based on CDS and CDS + IGS combined matrices have strong support in Tripterospermum. These findings contribute to the elucidation of the plastid genome evolution of Tripterospermum and provide a foundation for further exploration and resource utilization within this genus.

TRPV1 analgesics disturb core body temperature via a biased allosteric mechanism involving conformations distinct from that for nociception.

Efforts on developing transient receptor potential vanilloid 1 (TRPV1) drugs for pain management have been hampered by deleterious hypo- or hyperthermia caused by TRPV1 agonists/antagonists. Here, we compared the effects of four antagonists on TRPV1 polymodal gating and core body temperature (CBT) in Trpv1+/+, Trpv1-/-, and Trpv1T634A/T634A. Neither the effect on proton gating nor drug administration route, hair coverage, CBT rhythmic fluctuations, or inflammation had any influence on the differential actions of TRPV1 drugs on CBT. We identified the S4-S5 linker region exposed to the vanilloid pocket of TRPV1 to be critical for hyperthermia associated with certain TRPV1 antagonists. PSFL2874, a TRPV1 antagonist we discovered, is effective against inflammatory pain but devoid of binding to the S4-S5 linker and inducing CBT changes. These findings implicate that biased allosteric mechanisms exist for TRPV1 coupling to nociception and CBT regulation, opening avenues for the development of non-opioid analgesics without affecting CBT.

Bi-Directional Ensemble Feature Reconstruction Network for Few-Shot Fine-Grained Classification.

The main challenge for fine-grained few-shot image classification is to learn feature representations with higher inter-class and lower intra-class variations, with a mere few labelled samples. Conventional few-shot learning methods however cannot be naively adopted for this fine-grained setting - a quick pilot study reveals that they in fact push for the opposite (i.e., lower inter-class variations and higher intra-class variations). To alleviate this problem, prior works predominately use a support set to reconstruct the query image and then utilize metric learning to determine its category. Upon careful inspection, we further reveal that such unidirectional reconstruction methods only help to increase inter-class variations and are not effective in tackling intra-class variations. In this paper, we introduce a bi-reconstruction mechanism that can simultaneously accommodate for inter-class and intra-class variations. In addition to using the support set to reconstruct the query set for increasing inter-class variations, we further use the query set to reconstruct the support set for reducing intra-class variations. This design effectively helps the model to explore more subtle and discriminative features which is key for the fine-grained problem in hand. Furthermore, we also construct a self-reconstruction module to work alongside the bi-directional module to make the features even more discriminative. We introduce the snapshot ensemble method in the episodic learning strategy - a simple trick to further improve model performance without increasing training costs. Experimental results on three widely used fine-grained image classification datasets, as well as general and cross-domain few-shot image datasets, consistently show considerable improvements compared with other methods. Codes are available at https://github.com/PRIS-CV/BiEN.

CreativeSeg: Semantic Segmentation of Creative Sketches.

The problem of sketch semantic segmentation is far from being solved. Despite existing methods exhibiting near-saturating performances on simple sketches with high recognisability, they suffer serious setbacks when the target sketches are products of an imaginative process with high degree of creativity. We hypothesise that human creativity, being highly individualistic, induces a significant shift in distribution of sketches, leading to poor model generalisation. Such hypothesis, backed by empirical evidences, opens the door for a solution that explicitly disentangles creativity while learning sketch representations. We materialise this by crafting a learnable creativity estimator that assigns a scalar score of creativity to each sketch. It follows that we introduce CreativeSeg, a learning-to-learn framework that leverages the estimator in order to learn creativity-agnostic representation, and eventually the downstream semantic segmentation task. We empirically verify the superiority of CreativeSeg on the recent "Creative Birds" and "Creative Creatures" creative sketch datasets. Through a human study, we further strengthen the case that the learned creativity score does indeed have a positive correlation with the subjective creativity of human. Codes are available at https://github.com/PRIS-CV/Sketch-CS.

Arthroscopic Bone Grafting and Robot-assisted Fixation for Scaphoid Nonunion.

OBJECTIVES: Scaphoid nonunion remains a challenging problem to manage with no general consensus on its treatment recommendations. We propose a novel minimally invasive (MIS) technique of arthroscopic bone grafting (ABG) with robot-assisted fixation for the treatment of scaphoid nonunions. METHODS: Patients with radiographically proven scaphoid nonunion treated by this novel surgical technique were included. Following arthroscopic debridement and iliac crest bone grafting, the scaphoid was fixed percutaneously using either multiple Kirschner (K)-wires or a headless compression screw using a robotic navigation system. RESULTS: Six male patients with an average age of 29.2 years were enrolled. Four patients had scaphoid waist fractures, and the other two were proximal pole fractures. During wrist arthroscopy, punctate bleeding of the proximal scaphoid fragment was observed in four out of the six patients. Half of the patients were fixed using a headless compression screw and the other half using multiple K-wires. All the guidewires were placed with a single-attempt using the robotic navigation system. Postoperatively, all the scaphoid fractures had complete radiographic union by 16 weeks. At a mean follow-up of 18.3 months, there were significant improvements in wrist range of motion, grip strength, and patient-rated outcomes. No intraoperative or early postoperative complications were encountered in any of our patients. CONCLUSION: Arthroscopic bone grafting with robot-assisted fixation is a feasible and promising therapeutic option for scaphoid nonunions, regardless of the vascularity of the proximal pole fragment. This novel technique allows for anatomic restoration of the scaphoid alignment and accurate, targeted placement of implants into the scaphoid nonunion site within a single-attempt using a robotic navigation system.

A comparative cadaveric study for percutaneous scaphoid fixation: robotic vs freehand.

PURPOSES: To compare the robotic-assisted and the traditional freehand percutaneous scaphoid fixation in number of guidewire attempts, duration of fluoroscopy time, amount of radiation dose, and screw centrality. METHODS: Twenty cadaveric specimens were randomized into either the robotic or freehand group. The scaphoids in both groups were fixed by either the same attending or resident from our hand surgery department. The operation duration, amount of radiation from intraoperative fluoroscopy, total fluoroscopy time, and the number of guidewire attempts were documented and compared. Postoperatively, all the specimens had a computed tomography (CT) scan performed, and the difference in the final position of the screw and the central axis of the scaphoid was examined. RESULTS: In the robotic group, all the guide wires were satisfactorily positioned within a single attempt, while the median number of attempts in the traditional freehand group was 18 (quaternion 14-65). This also meant that the surgeon in the robotic group experienced significantly lower radiation exposure dose and time as compared to the freehand group. There were no significant differences in the final screw position as compared to the central axis of the scaphoid in both groups. Although there was no difference in surgeon performance in the robotic group, the operative time for the attending was significantly lower as compared to the resident in the freehand group. CONCLUSION: Robotic-assisted surgery for scaphoid fracture fixation is superior to the traditional freehand method as it facilitates accurate screw placement with lower radiation exposure and fewer guide wire attempts.

Semi-Supervised Learning for FGVC With Out-of-Category Data.

Despite great strides made on fine-grained visual classification (FGVC), current methods are still heavily reliant on fully-supervised paradigms where ample expert labels are called for. Semi-supervised learning (SSL) techniques, acquiring knowledge from unlabeled data, provide a considerable means forward and have shown great promise for coarse-grained problems. However, exiting SSL paradigms mostly assume in-category (i.e., category-aligned) unlabeled data, which hinders their effectiveness when re-proposed on FGVC. In this paper, we put forward a novel design specifically aimed at making out-of-category data work for semi-supervised FGVC. We work off an important assumption that all fine-grained categories naturally follow a hierarchical structure (e.g., the phylogenetic tree of "Aves" that covers all bird species). It follows that, instead of operating on individual samples, we can instead predict sample relations within this tree structure as the optimization goal of SSL. Beyond this, we further introduced two strategies uniquely brought by these tree structures to achieve inter-sample consistency regularization and reliable pseudo-relation. Our experimental results reveal that (i) the proposed method yields good robustness against out-of-category data, and (ii) it can be equipped with prior arts, boosting their performance thus yielding state-of-the-art results.

[Therapeutic efficacy of volume-guaranteed high frequency oscillation ventilation on respiratory failure in preterm infants with a gestational age of 28-34 weeks: a prospective randomized controlled study]. / 容量保证的高频振荡通气模式对胎龄28~34å‘¨æ—©äº§å„¿å‘¼å¸è¡°ç«­ç–—æ•ˆçš„å‰çž»æ€§éšæœºå¯¹ç §ç ”ç©¶.

OBJECTIVES: To investigate the therapeutic efficacy of volume-guaranteed high frequency oscillation ventilation (HFOV-VG) versus conventional mechanical ventilation (CMV) in the treatment of preterm infants with respiratory failure. METHODS: A prospective study was conducted on 112 preterm infants with respiratory failure (a gestational age of 28-34 weeks) who were admitted to the Department of Neonatology, Jiangyin Hospital Affiliated to Medical School of Southeast University, from October 2018 to December 2022. The infants were randomly divided into an HFOV-VG group (44 infants) and a CMV group (68 infants) using the coin tossing method based on the mode of mechanical ventilation. The therapeutic efficacy was compared between the two groups. RESULTS: After 24 hours of treatment, both the HFOV-VG and CMV groups showed significant improvements in arterial blood pH, partial pressure of oxygen, partial pressure of carbon dioxide, and partial pressure of oxygen/fractional concentration of inspired oxygen ratio (P<0.05), and the HFOV-VG group had better improvements than the CMV group (P<0.05). There were no significant differences between the two groups in the incidence rate of complications, 28-day mortality rate, and length of hospital stay (P>0.05), but the HFOV-VG group had a significantly shorter duration of invasive mechanical ventilation than the CMV group (P<0.05). The follow-up at the corrected age of 6 months showed that there were no significant differences between the two groups in the scores of developmental quotient, gross motor function, fine motor function, adaptive ability, language, and social behavior in the Pediatric Neuropsychological Development Scale (P>0.05). CONCLUSIONS: Compared with CMV mode, HFOV-VG mode improves partial pressure of oxygen and promotes carbon dioxide elimination, thereby enhancing oxygenation and shortening the duration of mechanical ventilation in preterm infants with respiratory failure, while it has no significant impact on short-term neurobehavioral development in these infants.

Horizontally Transferred Salivary Protein Promotes Insect Feeding by Suppressing Ferredoxin-Mediated Plant Defenses.

Herbivorous insects such as whiteflies, planthoppers, and aphids secrete abundant orphan proteins to facilitate feeding. Yet, how these genes are recruited and evolve to mediate plant-insect interaction remains unknown. In this study, we report a horizontal gene transfer (HGT) event from fungi to an ancestor of Aleyrodidae insects approximately 42 to 190 million years ago. BtFTSP1 is a salivary protein that is secreted into host plants during Bemisia tabaci feeding. It targets a defensive ferredoxin 1 in Nicotiana tabacum (NtFD1) and disrupts the NtFD1-NtFD1 interaction in plant cytosol, leading to the degradation of NtFD1 in a ubiquitin-dependent manner. Silencing BtFTSP1 has negative effects on B. tabaci feeding while overexpressing BtFTSP1 in N. tabacum benefits insects and rescues the adverse effect caused by NtFD1 overexpression. The association between BtFTSP1 and NtFD1 is newly evolved after HGT, with the homologous FTSP in its fungal donor failing to interact and destabilize NtFD1. Our study illustrates the important roles of horizontally transferred genes in plant-insect interactions and suggests the potential origin of orphan salivary genes.

Interleukin-22 promotes proliferation and reverses LPS-induced apoptosis and steroidogenesis attenuation in human ovarian granulosa cells: implications for polycystic ovary syndrome pathogenesis.

OBJECTIVE: Interleukin 22 (IL-22) plays a role in inflammatory diseases. However, whether IL-22 affects the function of ovarian granulosa cells (GCs) and its relationship with Polycystic Ovary Syndrome (PCOS)remains unclear. METHODS: We investigated the level of IL-22 in human follicular fluid using ELISA. The expression and localization of the IL-22 receptor 1 (IL-22R1) in GCs were investigated by RT-PCR and immunofluorescence staining, respectively. The proliferation of KGN cells (human GCs line) was assessed by CCK-8 assay and EdU assay after treatment with recombinant human IL-22 (rhIL-22) and lipopolysaccharide (LPS). Apoptosis was assessed using flow cytometry. Apoptotic proteins and steroidogenic genes were detected by western blotting. RESULTS: ELISA's results showed that compared with the control group, PCOS patients showed lower expression of IL-22 in follicular fluid. Immunofluorescence showed that IL-22R1 is expressed and localized in human granulosa cell membranes. IL-22 promoted cell proliferation and reversed LPS-induced inhibition of cell proliferation. IL-22 alone did not affect apoptotic or steroidogenic protein expression, however, it reversed LPS-induced apoptosis via downregulation of Bcl-2, upregulation of Bax and cleaved caspase-3, and restoration of LPS-downregulated StAR, CYP11A1, and CYP19A1 expression. Western blotting confirmed that IL-22 activated the JAK2/STAT3 signaling. CONCLUSION: IL-22 promotes cell proliferation, inhibits apoptosis, and regulates KGN cell steroidogenesis confronted with LPS, and decreased IL-22 may be involved in the development of PCOS.

Transcriptomic Analysis of Tobacco Plants in Response to Whitefly Infection.

The whitefly Bemisia tabaci is one of the most destructive pests worldwide, and causes tremendous economic losses. Tobacco Nicotiana tabacum serves as a model organism for studying fundamental biological processes and is severely damaged by whiteflies. Hitherto, our knowledge of how tobacco perceives and defends itself against whiteflies has been scare. In this study, we analyze the gene expression patterns of tobacco in response to whitefly infestation. A total of 244 and 2417 differentially expressed genes (DEGs) were identified at 12 h and 24 h post whitefly infestation, respectively. Enrichment analysis demonstrates that whitefly infestation activates plant defense at both time points, with genes involved in plant pattern recognition, transcription factors, and hormonal regulation significantly upregulated. Notably, defense genes are more intensely upregulated at 24 h post infestation than at 12 h, indicating an increased immunity induced by whitefly infestation. In contrast, genes associated with energy metabolism, carbohydrate metabolism, ribosomes, and photosynthesis are suppressed, suggesting impaired plant development. Taken together, our study provides comprehensive insights into how plants respond to phloem-feeding insects, and offers a theoretical basis for better research on plant-insect interactions.

Current development of pyrazole-azole hybrids with anticancer potential.

Chemotherapy is a critical treatment modality for cancer patients, but multidrug resistance remains one of the major challenges in cancer therapy, creating an urgent need for the development of novel potent chemical entities. Azoles, particularly pyrazole, could interact with different biological targets and exhibit diverse biological properties including anticancer activity. Many clinically used anticancer agents own an azole moiety, demonstrating that azoles are privileged and pivotal templates in the discovery of novel anticancer chemotherapeutics. The present article is an attempt to highlight the recent advances in pyrazole-azole hybrids with anticancer potential and discuss the structure-activity relationships, covering articles published from 2018 to present, to facilitate the rational design of more effective anticancer candidates.

Sketch-Segformer: Transformer-Based Segmentation for Figurative and Creative Sketches.

Sketch is a well-researched topic in the vision community by now. Sketch semantic segmentation in particular, serves as a fundamental step towards finer-level sketch interpretation. Recent works use various means of extracting discriminative features from sketches and have achieved considerable improvements on segmentation accuracy. Common approaches for this include attending to the sketch-image as a whole, its stroke-level representation or the sequence information embedded in it. However, they mostly focus on only a part of such multi-facet information. In this paper, we for the first time demonstrate that there is complementary information to be explored across all these three facets of sketch data, and that segmentation performance consequently benefits as a result of such exploration of sketch-specific information. Specifically, we propose the Sketch-Segformer, a transformer-based framework for sketch semantic segmentation that inherently treats sketches as stroke sequences other than pixel-maps. In particular, Sketch-Segformer introduces two types of self-attention modules having similar structures that work with different receptive fields (i.e., whole sketch or individual stroke). The order embedding is then further synergized with spatial embeddings learned from the entire sketch as well as localized stroke-level information. Extensive experiments show that our sketch-specific design is not only able to obtain state-of-the-art performance on traditional figurative sketches (such as SPG, SketchSeg-150K datasets), but also performs well on creative sketches that do not conform to conventional object semantics (CreativeSketch dataset) thanks for our usage of multi-facet sketch information. Ablation studies, visualizations, and invariance tests further justifies our design choice and the effectiveness of Sketch-Segformer. Codes are available at https://github.com/PRIS-CV/Sketch-SF.

Biological effects of IL-33/ST2 axis on oral diseases: autoimmune diseases and periodontal diseases.

IL-33 is a relatively new member of the IL-1 cytokine family, which plays a unique role in autoimmune diseases, particularly some oral diseases dominated by immune factors. The IL-33/ST2 axis is the main pathway by which IL-33 signals affect downstream cells to produce an inflammatory response or tissue repair. As a newly discovered pro-inflammatory cytokine, IL-33 can participate in the pathogenesis of autoimmune oral diseases such as Sjogren's syndrome and Behcet's disease. Moreover, the IL-33/ST2 axis also recruits and activates mast cells in periodontitis, producing inflammatory chemokines and mediating gingival inflammation and alveolar bone destruction. Interestingly, the high expression of IL-33 in the alveolar bone, which exhibits anti-osteoclast effects under appropriate mechanical loading, also confirms its dual role of destruction and repair in an immune-mediated periodontal environment. This study reviewed the biological effects of IL-33 in autoimmune oral diseases, periodontitis and periodontal bone metabolism, and elaborated its potential role and impact as a disease enhancer or a repair factor.

Uncertainty-Aware Source-Free Domain Adaptive Semantic Segmentation.

Source-Free Domain Adaptation (SFDA) is becoming topical to address the challenge of distribution shift between training and deployment data, while also relaxing the requirement of source data availability during target domain adaptation. In this paper, we focus on SFDA for semantic segmentation, in which pseudo labeling based target domain self-training is a common solution. However, pseudo labels generated by the source models are particularly unreliable on the target domain data due to the domain shift issue. Therefore, we propose to use Bayesian Neural Network (BNN) to improve the target self-training by better estimating and exploiting pseudo-label uncertainty. With the uncertainty estimation of BNNs, we introduce two novel self-training based components: Uncertainty-aware Online Teacher-Student Learning (UOTSL) and Uncertainty-aware FeatureMix (UFM). Extensive experiments on two popular benchmarks, GTA 5 → Cityscapes and SYNTHIA → Cityscapes, show the superiority of our proposed method with mIoU gains of 3.6% and 5.7% over the state-of-the-art respectively.

Prediction Calibration for Generalized Few-Shot Semantic Segmentation.

Generalized Few-shot Semantic Segmentation (GFSS) aims to segment each image pixel into either base classes with abundant training examples or novel classes with only a handful of (e. g., 1-5) training images per class. Compared to the widely studied Few-shot Semantic Segmentation (FSS), which is limited to segmenting novel classes only, GFSS is much under-studied despite being more practical. Existing approach to GFSS is based on classifier parameter fusion whereby a newly trained novel class classifier and a pre-trained base class classifier are combined to form a new classifier. As the training data is dominated by base classes, this approach is inevitably biased towards the base classes. In this work, we propose a novel Prediction Calibration Network (PCN) to address this problem. Instead of fusing the classifier parameters, we fuse the scores produced separately by the base and novel classifiers. To ensure that the fused scores are not biased to either the base or novel classes, a new Transformer-based calibration module is introduced. It is known that the lower-level features are useful of detecting edge information in an input image than higher-level features. Thus, we build a cross-attention module that guides the classifier's final prediction using the fused multi-level features. However, transformers are computationally demanding. Crucially, to make the proposed cross-attention module training tractable at the pixel level, this module is designed based on feature-score cross-covariance and episodically trained to be generalizable at inference time. Extensive experiments on PASCAL- 5i and COCO- 20i show that our PCN outperforms the state-the-the-art alternatives by large margins.