Measurement method of tear meniscus height based on deep learning.

Tear meniscus height (TMH) is an important reference parameter in the diagnosis of dry eye disease. However, most traditional methods of measuring TMH are manual or semi-automatic, which causes the measurement of TMH to be prone to the influence of subjective factors, time consuming, and laborious. To solve these problems, a segmentation algorithm based on deep learning and image processing was proposed to realize the automatic measurement of TMH. To accurately segment the tear meniscus region, the segmentation algorithm designed in this study is based on the DeepLabv3 architecture and combines the partial structure of the ResNet50, GoogleNet, and FCN networks for further improvements. A total of 305 ocular surface images were used in this study, which were divided into training and testing sets. The training set was used to train the network model, and the testing set was used to evaluate the model performance. In the experiment, for tear meniscus segmentation, the average intersection over union was 0.896, the dice coefficient was 0.884, and the sensitivity was 0.877. For the central ring of corneal projection ring segmentation, the average intersection over union was 0.932, the dice coefficient was 0.926, and the sensitivity was 0.947. According to the evaluation index comparison, the segmentation model used in this study was superior to the existing model. Finally, the measurement outcome of TMH of the testing set using the proposed method was compared with manual measurement results. All measurement results were directly compared via linear regression; the regression line was y0.98x-0.02, and the overall correlation coefficient was r 20.94. Thus, the proposed method for measuring TMH in this paper is highly consistent with manual measurement and can realize the automatic measurement of TMH and assist clinicians in the diagnosis of dry eye disease.

Comparisons of the short-term effectiveness and safety of surgical treatment for neovascular glaucoma: a systematic review and network meta-analysis.

OBJECTIVE: To compare the effectiveness and safety of the six interventions for neovascular glaucoma. DESIGN: A systematic review and network meta-analysis. METHODS: Randomised controlled trials and cohort studies which compared the six interventions in neovascular glaucoma were identified using the following databases searched up to 1 September 2020: PubMed, Cochrane Library, Embase and Web of Science. The quality assessment was conducted by using the Cochrane risk of bias tool and the Newcastle-Ottawa scale. The primary outcome measure was the weighted mean differences for intraocular pressure reduction. Secondary one was ORs for success rate. Outcome measures were reported with a 95% CI and p<0.05 was considered statistically significant. Network meta-analysis was performed using Stata V.15.0. RESULTS: Twenty-three studies involving a total of 1303 patients were included. The types of surgical treatments included Ahmed glaucoma valve (AGV) implant surgery, AGV combined with intravitreal anti-vascular endothelial growth factor (AGV +IVAV), cyclophotocoagulation (CPC), cyclocryotherapy (CCT), trabeculectomy with mitomycin (Trab(MMC)) and Trab(MMC) combined with IVAV (Trab(MMC)+IVAV). Network meta-analysis showed that in comparison with AGV, AGV +IVAV (MD=4.74, 95% CI 1.04 to 8.45) and Trab(MMC)+IVAV (MD=6.19, 95% CI 0.99 to 11.40) showed a favourable effect in intraocular pressure reduction (IOPR) 6 months after surgery. Compared with CCT, AGV (OR=-0.17, 95% CI -0.53 to -0.05), AGV +IVAV (OR=-0.10, 95% CI -3.48 to -1.19), CPC (OR=-0.12, 95% CI -0.53 to -0.05), Trab(MMC) (OR=3.54, 95% CI 1.15 to 10.91) and Trab(MMC)+IVAV (OR=5.78, 95% CI 2.29 to 14.61) showed a superior impact in success rate. The order of efficacy as best intervention ranked as follows: Trab(MMC)+IVAV (IOPR 6 months after surgery, surface under the cumulative ranking (SUCRA)=88.1), CPC (IOPR 12 months after surgery, SUCRA=81.9), AGV +IVAV (IOPR 12 months after surgery, SUCRA=79.9) and AGV +IVAV (success rate, SUCRA=92.7). Adverse events were also summarised in detail. CONCLUSION: In the treatment of neovascular glaucoma, AGV+IVAV and CPC were more effective in IOPR and success rate than the other four interventions. Additionally, AGV+IVAV is superior to CPC concerning the success rate in the long-term treatment. However, considering the limitations of this review, more high-quality trials, especially those surgical interventions not mentioned in this review, should be carried out in the future to further confirm the current findings.

Visual-Text Reference Pretraining Model for Image Captioning.

People can accurately describe an image by constantly referring to the visual information and key text information of the image. Inspired by this idea, we propose the VTR-PTM (Visual-Text Reference Pretraining Model) for image captioning. First, based on the pretraining model (BERT/UNIML), we design the dual-stream input mode of image reference and text reference and use two different mask modes (bidirectional and sequence to sequence) to realize the VTR-PTM suitable for generating tasks. Second, the target dataset is used to fine tune the VTR-PTM. To the best of our knowledge, VTR-PTM is the first reported pretraining model to use visual-text references in the learning process. To evaluate the model, we conduct several experiments on the benchmark datasets of image captioning, including MS COCO and Visual Genome, and achieve significant improvements on most metrics. The code is available at https://github.com/lpfworld/VTR-PTM.

Guided Punctures with Ultrasound Volume Navigation in Percutaneous Transforaminal Endoscopic Discectomy: A Technical Note.

OBJECTIVE: Ultrasound volume navigation (UVN) has been widely used for accurate guidance and decreased radiation exposure. However, few studies have focused on the clinical significance of UVN in guiding percutaneous puncture in percutaneous transforaminal endoscopic discectomy (PTED). We evaluated UVN to guide percutaneous puncture in PTED. METHODS: We retrospectively reviewed the medical records of 12 patients (8 men and 4 women), who had undergone PTED with the help of UVN or fluoroscopic guidance for lumbar disc herniation from November 2017 to December 2017. RESULTS: The age of these 12 patients range was 26-71 years, and the body mass index range was 18.19-26.91 kg/m2. Of the 12 patients, 6 were in UVN group and 6 were in fluoroscopy group. The mean number of punctures was 1.00 in UVN group and 3.83 in fluoroscopy group. The mean exposure time was 3.60 and 13.80 seconds in UVN and fluoroscopy groups, respectively. The mean operation time was 48.17 minutes and 61.33 minutes in UVN and fluoroscopy groups, respectively. A positive relationship was found between operation time and exposure time (P < 0.05). All patients achieved excellent or good clinical outcomes. The Oswestry Disability Index and visual analog scales for leg pain and back pain all showed significant improvement after the procedure (P < 0.05). None of patients experienced a complication. CONCLUSIONS: UVN decreased the number of puncture attempts, radiation exposure, and operation time compared with fluoroscopic guidance in PTED. Therefore, UVN is a feasible and efficient method for guiding percutaneous puncture in PTED.

Contrast-Enhanced Ultrasonography Versus Contrast-Enhanced Computed Tomography for Assessment of Residual Tumor From Hepatocellular Carcinoma Treated With Transarterial Chemoembolization: A Meta-analysis.

OBJECTIVES: This study reviewed the literature to directly evaluate the diagnostic performance of contrast-enhanced ultrasonography (CEUS) versus contrast-enhanced computed tomography (CECT) for assessing residual tumors of hepatocellular carcinoma treated with transarterial chemoembolization. METHODS: PubMed, Embase, the Cochrane Library, and the China National Knowledge Infrastructure were searched through April 30, 2017. The pooled sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, negative likelihood ratio, diagnostic odds ratio, and summary receiver operating characteristic curve were calculated and compared to examine the diagnostic performance of CEUS versus CECT. RESULTS: A total of 11 studies, including 421 patients and 491 nodules were analyzed. The pooled diagnostic performances of CEUS versus CECT were as follows: (1) sensitivity (95% confidence interval), 0.97 (0.95-0.99) versus 0.72 (0.67-0.76); (2) specificity, 0.86 (0.74-0.94) versus 0.99 (0.95-1.00); (3) positive predictive value, 0.97 (0.95-0.99) versus 1.00 (0.98-1.00); (4) negative predictive value, 0.90 (0.83-0.95) versus 0.51 (0.44-0.58); (5) positive likelihood ratio, 7.79 (4.73-12.82) versus 12.50 (5.74-27.20); (6) negative likelihood ratio, 0.05 (0.03-0.09) versus 0.35 (0.26-0.48); (7) diagnostic odds ratio, 150.56 (57.03-397.49) versus 35.54 (14.89-84.83); and (8) area under the summary receiver operating characteristic curve, 0.9875 versus 0.9239. The sensitivity and negative predictive value of CEUS were significantly higher than those of CECT (both P < .001). The specificity and positive predictive value of CECT were significantly higher than those of CEUS (both P < .05). CONCLUSIONS: Contrast-enhanced US, with better sensitivity and negative predictive value versus CECT, was an effective method for exclusion of residual tumors after transarterial chemoembolization. Contrast-enhanced CT, with higher specificity than CEUS, is a valid approach for identifying residual tumors.

Discovery of cell-type specific DNA motif grammar in cis-regulatory elements using random Forest.

BACKGROUND: It has been observed that many transcription factors (TFs) can bind to different genomic loci depending on the cell type in which a TF is expressed in, even though the individual TF usually binds to the same core motif in different cell types. How a TF can bind to the genome in such a highly cell-type specific manner, is a critical research question. One hypothesis is that a TF requires co-binding of different TFs in different cell types. If this is the case, it may be possible to observe different combinations of TF motifs - a motif grammar - located at the TF binding sites in different cell types. In this study, we develop a bioinformatics method to systematically identify DNA motifs in TF binding sites across multiple cell types based on published ChIP-seq data, and address two questions: (1) can we build a machine learning classifier to predict cell-type specificity based on motif combinations alone, and (2) can we extract meaningful cell-type specific motif grammars from this classifier model. RESULTS: We present a Random Forest (RF) based approach to build a multi-class classifier to predict the cell-type specificity of a TF binding site given its motif content. We applied this RF classifier to two published ChIP-seq datasets of TF (TCF7L2 and MAX) across multiple cell types. Using cross-validation, we show that motif combinations alone are indeed predictive of cell types. Furthermore, we present a rule mining approach to extract the most discriminatory rules in the RF classifier, thus allowing us to discover the underlying cell-type specific motif grammar. CONCLUSIONS: Our bioinformatics analysis supports the hypothesis that combinatorial TF motif patterns are cell-type specific.

Impact of sequencing depth and read length on single cell RNA sequencing data of T cells.

Single cell RNA sequencing (scRNA-seq) provides great potential in measuring the gene expression profiles of heterogeneous cell populations. In immunology, scRNA-seq allowed the characterisation of transcript sequence diversity of functionally relevant T cell subsets, and the identification of the full length T cell receptor (TCRαß), which defines the specificity against cognate antigens. Several factors, e.g. RNA library capture, cell quality, and sequencing output affect the quality of scRNA-seq data. We studied the effects of read length and sequencing depth on the quality of gene expression profiles, cell type identification, and TCRαß reconstruction, utilising 1,305 single cells from 8 publically available scRNA-seq datasets, and simulation-based analyses. Gene expression was characterised by an increased number of unique genes identified with short read lengths (<50 bp), but these featured higher technical variability compared to profiles from longer reads. Successful TCRαß reconstruction was achieved for 6 datasets (81% - 100%) with at least 0.25 millions (PE) reads of length >50 bp, while it failed for datasets with <30 bp reads. Sufficient read length and sequencing depth can control technical noise to enable accurate identification of TCRαß and gene expression profiles from scRNA-seq data of T cells.

CIDR: Ultrafast and accurate clustering through imputation for single-cell RNA-seq data.

Most existing dimensionality reduction and clustering packages for single-cell RNA-seq (scRNA-seq) data deal with dropouts by heavy modeling and computational machinery. Here, we introduce CIDR (Clustering through Imputation and Dimensionality Reduction), an ultrafast algorithm that uses a novel yet very simple implicit imputation approach to alleviate the impact of dropouts in scRNA-seq data in a principled manner. Using a range of simulated and real data, we show that CIDR improves the standard principal component analysis and outperforms the state-of-the-art methods, namely t-SNE, ZIFA, and RaceID, in terms of clustering accuracy. CIDR typically completes within seconds when processing a data set of hundreds of cells and minutes for a data set of thousands of cells. CIDR can be downloaded at https://github.com/VCCRI/CIDR .

Falco: a quick and flexible single-cell RNA-seq processing framework on the cloud.

Summary: Single-cell RNA-seq (scRNA-seq) is increasingly used in a range of biomedical studies. Nonetheless, current RNA-seq analysis tools are not specifically designed to efficiently process scRNA-seq data due to their limited scalability. Here we introduce Falco, a cloud-based framework to enable paralellization of existing RNA-seq processing pipelines using big data technologies of Apache Hadoop and Apache Spark for performing massively parallel analysis of large scale transcriptomic data. Using two public scRNA-seq datasets and two popular RNA-seq alignment/feature quantification pipelines, we show that the same processing pipeline runs 2.6-145.4 times faster using Falco than running on a highly optimized standalone computer. Falco also allows users to utilize low-cost spot instances of Amazon Web Services, providing a ∼65% reduction in cost of analysis. Availability and Implementation: Falco is available via a GNU General Public License at https://github.com/VCCRI/Falco/. Contact: j.ho@victorchang.edu.au. Supplementary information: Supplementary data are available at Bioinformatics online.

Transcriptional response to mitochondrial protease IMMP2L knockdown in human primary astrocytes.

IMMP2L encodes the inner membrane peptidase subunit 2, a mitochondrial protease involved in cleaving the space-sorting signals of mitochondrial membrane proteins. IMMP2L has been implicated in Tourette syndrome, but how its dysfunction contributes to the neurodevelopmental phenotype remains unclear. Here we show that IMMP2L transcription requires Topoisomerase I in human primary astrocytes, and characterize the downstream effects of IMMP2L knockdown on gene expression. We demonstrate that IMMP2L knockdown leads to dysregulation of genes involved in central nervous system development. We also find that the transcriptional response to IMMP2L knockdown partially overlaps the one induced by mitochondrial complex III inhibition. Overall, these data bring further insight into the molecular consequences of IMMP2L dysfunction in the brain.

Transcriptome analysis of human brain tissue identifies reduced expression of complement complex C1Q Genes in Rett syndrome.

BACKGROUND: MECP2, the gene mutated in the majority of Rett syndrome cases, is a transcriptional regulator that can activate or repress transcription. Although the transcription regulatory function of MECP2 has been known for over a decade, it remains unclear how transcriptional dysregulation leads to the neurodevelopmental disorder. Notably, little convergence was previously observed between the genes abnormally expressed in the brain of Rett syndrome mouse models and those identified in human studies. METHODS: Here we carried out a comprehensive transcriptome analysis of human brain tissue from Rett syndrome brain using both RNA-seq and microarrays. RESULTS: We identified over two hundred differentially expressed genes, and identified the complement C1Q complex genes (C1QA, C1QB and C1QC) as a point of convergence between gene expression changes in human and mouse Rett syndrome brain. CONCLUSIONS: The results of our study support a role for alterations in the expression level of C1Q complex genes in RTT pathogenesis.

A clinical validation study for the feasibility and reliability of three-dimensional ultrasound-ultrasound automatic image registration.

PURPOSE: The purpose of our study was to assess the feasibility and reliability of 3D ultrasound-ultrasound (US-US) automatic registration-based analysis of the hepatic vessel tree (VT) (3D VT-based automatic registration) in clinical applications. MATERIALS AND METHODS: A total of 70 pairs of 3D ultrasound data were acquired from the livers of 10 healthy volunteers enrolled in the study. An automatic registration method was applied to the acquired volumetric data pairs, and anatomic landmarks were picked by an experienced sonographer as 'ground truth'. The influences of respiration phase, subject posture, and liver lobe on data acquisition and scan volumetric angle on the registration accuracy and robustness were investigated. The registration time, success rate, median registration error distance, and sonographer's subjective feedback were assessed. RESULTS: The time required for the 3D VT-based automatic registration was approximately 15∼20 s. Overall, the success rate for the hepatic vessel-based registration was 71% (50/70), and the median registration error distance was 1.72 mm (0.57∼4.71 mm). When the influential factors were well controlled, the optimal registration accuracy (median registration error distance = 1.22 mm) could be obtained with an excellent success rate of 100% (10/10). According to the subjective assessment of the sonographer, over 90% (45/50) of the automatic registration results were not inferior to the ground truth. Among them, 42% (21/50) were superior to the fusion results from the ground truth. CONCLUSIONS: The results suggest that the 3D VT-based automatic registration is feasible and reliable and has potential for guidance and evaluation of intraoperative ablation of hepatocellular carcinoma.

Coexpression networks identify brain region-specific enhancer RNAs in the human brain.

Despite major progress in identifying enhancer regions on a genome-wide scale, the majority of available data are limited to model organisms and human transformed cell lines. We have identified a robust set of enhancer RNAs (eRNAs) expressed in the human brain and constructed networks assessing eRNA-gene coexpression interactions across human fetal brain and multiple adult brain regions. Our data identify brain region-specific eRNAs and show that enhancer regions expressing eRNAs are enriched for genetic variants associated with autism spectrum disorders.

Estimation of the methylation pattern distribution from deep sequencing data.

BACKGROUND: Bisulphite sequencing enables the detection of cytosine methylation. The sequence of the methylation states of cytosines on any given read forms a methylation pattern that carries substantially more information than merely studying the average methylation level at individual positions. In order to understand better the complexity of DNA methylation landscapes in biological samples, it is important to study the diversity of these methylation patterns. However, the accurate quantification of methylation patterns is subject to sequencing errors and spurious signals due to incomplete bisulphite conversion of cytosines. RESULTS: A statistical model is developed which accounts for the distribution of DNA methylation patterns at any given locus. The model incorporates the effects of sequencing errors and spurious reads, and enables estimation of the true underlying distribution of methylation patterns. CONCLUSIONS: Calculation of the estimated distribution over methylation patterns is implemented in the R Bioconductor package MPFE. Source code and documentation of the package are also available for download at http://bioconductor.org/packages/3.0/bioc/html/MPFE.html .

CD1d induction in solid tumor cells by histone deacetylase inhibitors through inhibition of HDAC1/2 and activation of Sp1.

CD1d is a MHC class-like molecule that presents glycolipids to natural killer T (NKT) cells, then regulates innate and adaptive immunity. The regulation of CD1d gene expression in solid tumors is still largely unknown. Gene expression can be epigenetically regulated by DNA methylation and histone acetylation. We found that histone deacetylase inhibitors, trichostatin A (TSA) and suberoylanilide hydroxamic acid (SAHA), induced CD1d gene expression in human (A549 and NCI-H292) and mouse (TC-1 and B16/F0) cancer cells. Simultaneous knockdown of HDAC1 and 2 induced CD1d gene expression. Sp1 inhibitor mitramycin A (MTM) blocked TSA- and SAHA-induced CD1d mRNA expression and Sp1 luciferase activity. Co-transfection of GAL4-Sp1 and Fc-luciferase reporters demonstrated that TSA and SAHA induced Sp1 luciferase reporter activity by enhancing Sp1 transactivation activity. The binding of Sp1 to CD1d promoter and histone H3 acetylation on Sp1 sites were increased by TSA and SAHA. These results indicate that TSA and SAHA could up-regulate CD1d expression in tumor cells through inhibition of HDAC1/2 and activation of Sp1.