Video-Instrument Synergistic Network for Referring Video Instrument Segmentation in Robotic Surgery.

Surgical instrument segmentation is fundamentally important for facilitating cognitive intelligence in robot-assisted surgery. Although existing methods have achieved accurate instrument segmentation results, they simultaneously generate segmentation masks of all instruments, which lack the capability to specify a target object and allow an interactive experience. This paper focuses on a novel and essential task in robotic surgery, i.e., Referring Surgical Video Instrument Segmentation (RSVIS), which aims to automatically identify and segment the target surgical instruments from each video frame, referred by a given language expression. This interactive feature offers enhanced user engagement and customized experiences, greatly benefiting the development of the next generation of surgical education systems. To achieve this, this paper constructs two surgery video datasets to promote the RSVIS research. Then, we devise a novel Video-Instrument Synergistic Network (VIS-Net) to learn both video-level and instrument-level knowledge to boost performance, while previous work only utilized video-level information. Meanwhile, we design a Graph-based Relation-aware Module (GRM) to model the correlation between multi-modal information (i.e., textual description and video frame) to facilitate the extraction of instrument-level information. Extensive experimental results on two RSVIS datasets exhibit that the VIS-Net can significantly outperform existing state-of-the-art referring segmentation methods. We will release our code and dataset for future research (Git).

Dual-Reference Source-Free Active Domain Adaptation for Nasopharyngeal Carcinoma Tumor Segmentation across Multiple Hospitals.

Nasopharyngeal carcinoma (NPC) is a prevalent and clinically significant malignancy that predominantly impacts the head and neck area. Precise delineation of the Gross Tumor Volume (GTV) plays a pivotal role in ensuring effective radiotherapy for NPC. Despite recent methods that have achieved promising results on GTV segmentation, they are still limited by lacking carefully-annotated data and hard-to-access data from multiple hospitals in clinical practice. Although some unsupervised domain adaptation (UDA) has been proposed to alleviate this problem, unconditionally mapping the distribution distorts the underlying structural information, leading to inferior performance. To address this challenge, we devise a novel Sourece-Free Active Domain Adaptation framework to facilitate domain adaptation for the GTV segmentation task. Specifically, we design a dual reference strategy to select domain-invariant and domain-specific representative samples from a specific target domain for annotation and model fine-tuning without relying on source-domain data. Our approach not only ensures data privacy but also reduces the workload for oncologists as it just requires annotating a few representative samples from the target domain and does not need to access the source data. We collect a large-scale clinical dataset comprising 1057 NPC patients from five hospitals to validate our approach. Experimental results show that our method outperforms the previous active learning (e.g., AADA and MHPL) and UDA (e.g., Tent and CPR) methods, and achieves comparable results to the fully supervised upper bound, even with few annotations, highlighting the significant medical utility of our approach. In addition, there is no public dataset about multi-center NPC segmentation, we will release code and dataset for future research (Git).

Applications of deep learning for detecting ophthalmic diseases with ultrawide-field fundus images.

AIM: To summarize the application of deep learning in detecting ophthalmic disease with ultrawide-field fundus images and analyze the advantages, limitations, and possible solutions common to all tasks. METHODS: We searched three academic databases, including PubMed, Web of Science, and Ovid, with the date of August 2022. We matched and screened according to the target keywords and publication year and retrieved a total of 4358 research papers according to the keywords, of which 23 studies were retrieved on applying deep learning in diagnosing ophthalmic disease with ultrawide-field images. RESULTS: Deep learning in ultrawide-field images can detect various ophthalmic diseases and achieve great performance, including diabetic retinopathy, glaucoma, age-related macular degeneration, retinal vein occlusions, retinal detachment, and other peripheral retinal diseases. Compared to fundus images, the ultrawide-field fundus scanning laser ophthalmoscopy enables the capture of the ocular fundus up to 200° in a single exposure, which can observe more areas of the retina. CONCLUSION: The combination of ultrawide-field fundus images and artificial intelligence will achieve great performance in diagnosing multiple ophthalmic diseases in the future.

Defective kernel 58 encodes an Rrp15p domain-containing protein essential to ribosome biogenesis and seed development in maize.

Ribosome biogenesis is a highly dynamic and orchestrated process facilitated by hundreds of ribosomal biogenesis factors and small nucleolar RNAs. While many of the advances are derived from studies in yeast, ribosome biogenesis remains largely unknown in plants despite its importance to plant growth and development. Through characterizing the maize (Zea mays) defective kernel and embryo-lethal mutant dek58, we show that DEK58 encodes an Rrp15p domain-containing protein with 15.3% identity to yeast Rrp15. Over-expression of DEK58 rescues the mutant phenotype. DEK58 is localized in the nucleolus. Ribosome profiling and RNA gel blot analyses show that the absence of DEK58 reduces ribosome assembly and impedes pre-rRNA processing, accompanied by the accumulation of nearly all the pre-rRNA processing intermediates and the production of an aberrant processing product P-25S*. DEK58 interacts with ZmSSF1, a maize homolog of the yeast Ssf1 in the 60S processome. DEK58 and ZmSSF1 interact with ZmCK2α, a putative component of the yeast UTP-C complex involved in the small ribosomal subunit processome. These results demonstrate that DEK58 is essential to seed development in maize. It functions in the early stage of pre-rRNA processing in ribosome biogenesis, possibly through interacting with ZmSSF1 and ZmCK2α in maize.

Machine learning identified MDK score has prognostic value for idiopathic pulmonary fibrosis based on integrated bulk and single cell expression data.

Idiopathic pulmonary fibrosis (IPF) is a progressive and fatal lung disease that poses a significant challenge to medical professionals due to its increasing incidence and prevalence coupled with the limited understanding of its underlying molecular mechanisms. In this study, we employed a novel approach by integrating five expression datasets from bulk tissue with single-cell datasets; they underwent pseudotime trajectory analysis, switch gene selection, and cell communication analysis. Utilizing the prognostic information derived from the GSE47460 dataset, we identified 22 differentially expressed switch genes that were correlated with clinical indicators as important genes. Among these genes, we found that the midkine (MDK) gene has the potential to serve as a marker of Idiopathic pulmonary fibrosis because its cellular communicating genes are differentially expressed in the epithelial cells. We then utilized midkine and its cellular communication-related genes to calculate the midkine score. Machine learning models were further constructed through midkine and related genes to predict Idiopathic pulmonary fibrosis disease through the bulk gene expression datasets. The midkine score demonstrated a correlation with clinical indexes, and the machine learning model achieved an AUC of 0.94 and 0.86 in the Idiopathic pulmonary fibrosis classification task based on lung tissue samples and peripheral blood mononuclear cell samples, respectively. Our findings offer valuable insights into the pathogenesis of Idiopathic pulmonary fibrosis, providing new therapeutic directions and target genes for further investigation.

Maize Dek407 Encodes the Nitrate Transporter 1.5 and Is Required for Kernel Development.

The kernel serves as the storage organ and harvestable component of maize, and it plays a crucial role in determining crop yield and quality. Understanding the molecular and genetic mechanisms of kernel development is of considerable importance for maize production. In this study, we obtained a mutant, which we designated defective kernel 407 (dek407), through ethyl methanesulfonate mutagenesis. The dek407 mutant exhibited reduced kernel size and kernel weight, as well as delayed grain filling compared with those of the wild type. Positional cloning and an allelism test revealed that Dek407 encodes a nitrate transporter 1/peptide transporter family (NPF) protein and is the allele of miniature 2 (mn2) that was responsible for a poorly filled defective kernel phenotype. A transcriptome analysis of the developing kernels showed that the mutation of Dek407 altered the expression of phytohormone-related genes, especially those genes associated with indole-3-acetic acid synthesis and signaling. Phytohormone measurements and analysis indicated that the endogenous indole-3-acetic acid content was significantly reduced by 66% in the dek407 kernels, which may be the primary cause of the defective phenotype. We further demonstrated that natural variation in Dek407 is associated with kernel weight and kernel size. Therefore, Dek407 is a potential target gene for improvement of maize yield.

Molecular basis and evolutionary cost of a novel macrolides/lincosamides resistance phenotype in Staphylococcus haemolyticus.

Staphylococcus haemolyticus (S. haemolyticus) is a coagulase-negative Staphylococcus that has become one of the primary causes of nosocomial infection. After a long period of antibiotic use, S. haemolyticus has developed multiple resistance phenotypes for macrolides and lincosamides. Herein, we evaluated four S. haemolyticus clinical isolates, of which three had antibiotic resistance patterns reported previously. The fourth isolate was resistant to both erythromycin and clindamycin in the absence of erythromycin induction. This novel phenotype, known as constitutive macrolides-lincosamides-streptogramins resistance, has been reported in other bacteria but has not been previously reported in S. haemolyticus. Investigation of the isolate demonstrated a deletion in the methyltransferase gene ermC, upstream leader peptide. This deletion resulted in constitutive MLS resistance based on whole-genome sequencing and experimental verification. Continuous expression of ermC was shown to inhibit the growth of S. haemolyticus, which turned out to be the fitness cost with no MLS pressure. In summary, this study is the first to report constitutive MLS resistance in S. haemolyticus, which provides a better understanding of MLS resistance in clinical medicine. IMPORTANCE This study identified a novel phenotype of macrolides/lincosamides resistance in Staphylococcus haemolyticus which improved a better guidance for clinical treatment. It also clarified the mechanistic basis for this form of antibiotic resistance that supplemented the drug resistance mechanism of Staphylococcus. In addition, this study elaborated on a possibility that continuous expression of some resistance genes was shown to inhibit the growth of bacteria themselves, which turned out to be the fitness cost in the absence of antibiotic pressure.

Fine-grained attention & knowledge-based collaborative network for diabetic retinopathy grading.

Accurate diabetic retinopathy (DR) grading is crucial for making the proper treatment plan to reduce the damage caused by vision loss. This task is challenging due to the fact that the DR related lesions are often small and subtle in visual differences and intra-class variations. Moreover, relationships between the lesions and the DR levels are complicated. Although many deep learning (DL) DR grading systems have been developed with some success, there are still rooms for grading accuracy improvement. A common issue is that not much medical knowledge was used in these DL DR grading systems. As a result, the grading results are not properly interpreted by ophthalmologists, thus hinder the potential for practical applications. This paper proposes a novel fine-grained attention & knowledge-based collaborative network (FA+KC-Net) to address this concern. The fine-grained attention network dynamically divides the extracted feature maps into smaller patches and effectively captures small image features that are meaningful in the sense of its training from large amount of retinopathy fundus images. The knowledge-based collaborative network extracts a-priori medical knowledge features, i.e., lesions such as the microaneurysms (MAs), soft exudates (SEs), hard exudates (EXs), and hemorrhages (HEs). Finally, decision rules are developed to fuse the DR grading results from the fine-grained network and the knowledge-based collaborative network to make the final grading. Extensive experiments are carried out on four widely-used datasets, the DDR, Messidor, APTOS, and EyePACS to evaluate the efficacy of our method and compare with other state-of-the-art (SOTA) DL models. Simulation results show that proposed FA+KC-Net is accurate and stable, achieves the best performances on the DDR, Messidor, and APTOS datasets.

Delineation of clinical target volume and organs at risk in cervical cancer radiotherapy by deep learning networks.

PURPOSE: Delineation of the clinical target volume (CTV) and organs-at-risk (OARs) is important in cervical cancer radiotherapy. But it is generally labor-intensive, time-consuming, and subjective. This paper proposes a parallel-path attention fusion network (PPAF-net) to overcome these disadvantages in the delineation task. METHODS: The PPAF-net utilizes both the texture and structure information of CTV and OARs by employing a U-Net network to capture the high-level texture information, and an up-sampling and down-sampling (USDS) network to capture the low-level structure information to accentuate the boundaries of CTV and OARs. Multi-level features extracted from both networks are then fused together through an attention module to generate the delineation result. RESULTS: The dataset contains 276 computed tomography (CT) scans of patients with cervical cancer of staging IB-IIA. The images are provided by the West China Hospital of Sichuan University. Simulation results demonstrate that PPAF-net performs favorably on the delineation of the CTV and OARs (e.g., rectum, bladder and etc.) and achieves the state-of-the-art delineation accuracy, respectively, for the CTV and OARs. In terms of the Dice Similarity Coefficient (DSC) and the Hausdorff Distance (HD), 88.61% and 2.25 cm for the CTV, 92.27% and 0.73 cm for the rectum, 96.74% and 0.68 cm for the bladder, 96.38% and 0.65 cm for the left kidney, 96.79% and 0.63 cm for the right kidney, 93.42% and 0.52 cm for the left femoral head, 93.69% and 0.51 cm for the right femoral head, 87.53% and 1.07 cm for the small intestine, and 91.50% and 0.84 cm for the spinal cord. CONCLUSIONS: The proposed automatic delineation network PPAF-net performs well on CTV and OARs segmentation tasks, which has great potential for reducing the burden of radiation oncologists and increasing the accuracy of delineation. In future, radiation oncologists from the West China Hospital of Sichuan University will further evaluate the results of network delineation, making this method helpful in clinical practice.

Integrated System for On-Site Rapid and Safe Screening of COVID-19.

Since the outbreak of coronavirus disease 2019 (COVID-19), the epidemic has been spreading around the world for more than 2 years. Rapid, safe, and on-site detection methods of COVID-19 are in urgent demand for the control of the epidemic. Here, we established an integrated system, which incorporates a machine-learning-based Fourier transform infrared spectroscopy technique for rapid COVID-19 screening and air-plasma-based disinfection modules to prevent potential secondary infections. A partial least-squares discrimination analysis and a convolutional neural network model were built using the collected infrared spectral dataset containing 857 training serum samples. Furthermore, the sensitivity, specificity, and prediction accuracy could all reach over 94% from the results of the field test regarding 968 blind testing samples. Additionally, the disinfection modules achieved an inactivation efficiency of 99.9% for surface and airborne tested bacteria. The proposed system is conducive and promising for point-of-care and on-site COVID-19 screening in the mass population.

Maize Shrek1 encodes a WD40 protein that regulates pre-rRNA processing in ribosome biogenesis.

Ribosome biogenesis is a fundamental and highly orchestrated process that involves hundreds of ribosome biogenesis factors. Despite advances that have been made in yeast, the molecular mechanism of ribosome biogenesis remains largely unknown in plants. We uncovered a WD40 protein, Shrunken and Embryo Defective Kernel 1 (SHREK1), and showed that it plays a crucial role in ribosome biogenesis and kernel development in maize (Zea mays). The shrek1 mutant shows an aborted embryo and underdeveloped endosperm and embryo-lethal in maize. SHREK1 localizes mainly to the nucleolus and accumulates to high levels in the seed. Depleting SHREK1 perturbs pre-rRNA processing and causes imbalanced profiles of mature rRNA and ribosome. The expression pattern of ribosomal-related genes is significantly altered in shrek1. Like its yeast (Saccharomyces cerevisiae) ortholog Periodic tryptophan protein 1 (PWP1), SHREK1 physically interacts with ribosomal protein ZmRPL7a, a transient component of the PWP1-subcomplex involved in pre-rRNA processing in yeast. Additionally, SHREK1 may assist in the A3 cleavage of the pre-rRNA in maize by interacting with the nucleolar protein ZmPOP4, a maize homolog of the yeast RNase mitochondrial RNA-processing complex subunit. Overall, our work demonstrates a vital role of SHREK1 in pre-60S ribosome maturation, and reveals that impaired ribosome function accounts for the embryo lethality in shrek1.

ZmEHD1 Is Required for Kernel Development and Vegetative Growth through Regulating Auxin Homeostasis.

The roles of C-terminal Eps15 homology domain (EHD) proteins in clathrin-mediated endocytosis in plants are poorly understood. Here, we isolated a maize (Zea mays) mutant, designated ehd1, which showed defects in kernel development and vegetative growth. Positional cloning and transgenic analysis revealed that ehd1 encodes an EHD protein. Internalization of the endocytic tracer FM4-64 was substantially reduced in the ehd1 mutant and ZmEHD1 knockout mutants. We further demonstrated that ZmEHD1 and the ZmAP2 σ subunit physically interact at the plasma membrane. Auxin distribution and ZmPIN1a-YFP localization were altered in the ehd1 mutant. Kernel indole-3-acetic acid levels were substantially lower in the ehd1 mutant than in wild-type maize. Exogenous application of 1-naphthaleneacetic acid, but not GA3 or 2-naphthaleneacetic acid, rescued the seed germination and seedling emergency phenotypic defects of ehd1 mutants. Taken together, these results indicate that ZmEHD1 regulates auxin homeostasis by mediating clathrin-mediated endocytosis through its interaction with the ZmAP2 σ subunit, which is crucial for kernel development and vegetative growth of maize.

Pre-rRNA processing and its response to temperature stress in maize.

Ribosome biogenesis is a fundamental process in all eukaryotic cells and is coupled with the processing and maturation of pre-rRNAs. Maize is a primary staple crop across the world, but little is known about the exact pre-rRNA processing sites and pathways in this species. In this study, we present a detailed model of the pathway by identifying the critical endonucleolytic cleavage sites and determining the pre-rRNA intermediates by circular reverse-transcription PCR and northern blot analysis. We demonstrate that two pathways coexist in maize to promote the processing of 35S pre-rRNA, and that the processing of 27SA pre-rRNA can proceed via two different pathways, which are distinguished based on the order of ITS1 removal and ITS2 cleavage. Compared with yeast and mammals, this new 27SA pre-rRNA processing mechanism is unique to maize and other higher plants. In addition, we demonstrate that maize can modulate pre-rRNA processing levels in response to chilling and heat stress, as indicated by a significant reduction of the P-A3 intermediate. Our study provides information that will facilitate future research on ribosome biogenesis and pre-rRNA processing in maize.

MicroRNA528 Affects Lodging Resistance of Maize by Regulating Lignin Biosynthesis under Nitrogen-Luxury Conditions.

Lodging under nitrogen (N)-luxury conditions substantially reduces crop yield and seed quality. However, the molecular mechanisms of plant lodging resistance remain largely unclear, especially in maize. We report here that the expression of ZmmiR528, a monocot-specific microRNA, is induced by N luxury but reduced by N deficiency. We show by the thioacidolysis and acetyl bromide analysis that N luxury significantly reduces the generation of H, G, and S monomers of the lignin as well as its total content in maize shoots. We further demonstrate that ZmLACCASE3 (ZmLAC3) and ZmLACCASE5 (ZmLAC5), which encode the copper-containing laccases, are the targets of ZmmiR528. In situ hybridization showed that ZmmiR528 is mainly expressed in maize vascular tissues. Knockdown of ZmmiR528 or overexpression of ZmLAC3 significantly increased the lignin content and rind penetrometer resistance of maize stems. In contrast, transgenic maize plants overexpressing ZmmiR528 had reduced lignin content and rind penetrometer resistance and were prone to lodging under N-luxury conditions. RNA-sequencing analysis revealed that ZmPAL7 and ZmPAL8 are upregulated in transgenic maize lines downregulating ZmmiR528. Under N-luxury conditions, the expression levels of ZmPALs were much higher in ZmmiR528-knockdown lines than in the wild type and transgenic maize lines overexpressing ZmmiR528. Taken together, these results indicate that, by regulating the expression of ZmLAC3 and ZmLAC5, ZmmiR528 affects maize lodging resistance under N-luxury conditions.

Maize Urb2 protein is required for kernel development and vegetative growth by affecting pre-ribosomal RNA processing.

Ribosome biogenesis is a fundamental process in eukaryotic cells. Although Urb2 protein has been implicated in ribosome biogenesis in yeast, the Urb2 domain is loosely conserved between plants and yeast, and the function of Urb2 protein in plants remains unknown. Here, we isolated a maize mutant, designated as urb2, with defects in kernel development and vegetative growth. Positional cloning and transgenic analysis revealed that urb2 encodes an Urb2 domain-containing protein. Compared with the wild-type (WT), the urb2 mutant showed decreased ratios of 60S/40S and 80S/40S and increased ratios of polyribosomes. The pre-rRNA intermediates of 35/33S rRNA, P-A3 and 18S-A3 were significantly accumulated in the urb2 mutant. Transcriptome profiling of the urb2 mutant indicated that ZmUrb2 affects the expression of a number of ribosome-related genes. We further demonstrated that natural variations in ZmUrb2 are significantly associated with maize kernel length. The overall results indicate that, by affecting pre-rRNA processing, the Urb2 protein is required for ribosome biogenesis in maize.

Mechanically strong and pH-responsive carboxymethyl chitosan/graphene oxide/polyacrylamide nanocomposite hydrogels with fast recoverability.

The biocompatible, quickly recoverable, mechanically strong and tough polymer hydrogels have great potential for biomedical applications. A novel carboxymethyl chitosan (CMCTs)/graphene oxide (GO)/polyacrylamide (PAM) nanocomposite (NC) hydrogel with excellent mechanical performance was synthesized through a facile, one-pot free radical polymerization, using GO nanosheets as the crosslink centers instead of toxic organic molecules. The content of GO nanosheets and CMCTs has great influence on the mechanical properties of CMCTs/GO/PAM NC hydrogels. When the amount of GO nanosheets is 0.118 wt % and CMCTs is 1.5 wt % (to the total weight of CMCTs and acrylamide), CMCTs/GO/PAM hydrogel displays a compressive stress as high as 5.8 MPa at the breaking deformation of 87.5% and a tensile strength of 223.6 kPa at the failure strain of 631%. Besides, with nearly complete recovery in 1 min at the room temperature after unloading, the CMCTs/GO/PAM hydrogels exhibit excellent fast recoverability. Additionally, the CMCTs/GO/PAM hydrogels also have an obvious pH-responsive property.

Identification of heterotic loci associated with grain yield and its components using two CSSL test populations in maize.

Heterosis has widely been used to increase grain yield and quality. In this study, the genetic basis of heterosis on grain yield and its main components in maize were examined over 2 years in two locations in two test populations constructed from a set of 184 chromosome segment substitution lines (CSSLs) and two inbred lines (Zheng58 and Xun9058). Of the 169 heterotic loci (HL) associated with grain yield and its five components identified in CSSL × Zheng58 and CSSL × Xun9058 test populations, only 25 HL were detected in both populations. The comparison of quantitative trait loci (QTLs) detected in the CSSL population with HL detected in the two test populations revealed that only 15.46% and 17.35% of the HL in the given populations respectively, shared the same chromosomal regions as that of the corresponding QTLs and showed dominant effects as well as pleiotropism with additive and dominant effects. In addition, most of the HL (74.23% and 74.49%) had overdominant effects. These results suggest that overdominance is the main contributor to the effects of heterosis on grain yield and its components in maize, and different HL are associated with heterosis for different traits in different hybrids.