Candidate Regions Research Articles

Automated detection of lacunar infarcts from MRI images

AbstractBackgroundWhen MRI images of patients with memory problems are assessed, the presence of lacunar infarcts is evaluated. Lacunar infarcts, fluid‐filled cavities of 3‐15 mm in diameter, are one possible result of cerebral small vessel disease and therefore relevant in dementia diagnostics. In this study, we propose an automated tool for detecting lacunar infarcts on MRI images.MethodWe adapted our previous approach [1] in three ways: 1) post‐processing steps were added to the convolutional neural network segmentation pipeline, e.g., to exclude enlarged perivascular spaces and segmentations located inside sulci or ventricles, 2) memory clinic data from the Amsterdam Dementia Cohort and PredictND cohort were included in addition to the LADIS cohort data, and 3) the ground truth was defined by three experienced neuroradiologists (A, B, C) instead of one rater. T1 and T2‐FLAIR images were available for all cases and T2 for a subset in visual rating. The dataset was composed of 393 subjects and each expert rated 1156 potential lacunar infarcts as “1 = no”, “2 = probably no”, “3 = probably yes” or “4 = yes”. A consensus (CO) was created by calculating the average and using a cutoff at 2.5. The agreement was measured between individual raters (A‐B, B‐C, A‐C), between automatic rating and individual raters (AU‐A, AU‐B, AU‐C), and between the automatic rating and the consensus rating (AU‐CO) using Kappa. Correlation coefficient for the number of lacunar infarcts per subject was defined. Cross‐validation was used in training the algorithm.ResultOut of 1156 candidate regions evaluated, 540 were considered as lacunar infarcts based on consensus. The individual raters agreed with each other in 74% of cases whether the shown lesion was lacunar infarct (kappa = 0.48). Correspondingly, automatic rating agreed with the individual raters in 74% of cases (kappa = 0.47). When automatic rating was compared with consensus, agreement/accuracy was 78% (kappa = 0.55). The correlation coefficient for the number of lacunar infarcts detected automatically from the whole image and visually (consensus) was 0.85 (Figure 1).ConclusionThis study suggests that the agreement in detecting lacunar infarcts automatically is comparable to inter‐rater variability between experienced neuroradiologists. [1] Jokinen et al. Stroke 2020 Nov 9; 50(1)

Comparing one-stage and two-stage learning strategy in object detection

Object detection plays a vital role in computer social perception and computer vision. It could be applied to computer navigation, video monitoring, industrial detection, and so on. It greatly reduces the human labours by automatically locate and identify objects. Nowadays, the mainstream methods of object detection could be separated into the one- and two-stage method. The one-stage method leverages Convolutional Neural Network (CNN) for obtaining features and directly locate the target objects and their corresponding category probabilities. Different from the two-stage solutions, its accuracy is lower and the recognition speed is higher. The two-stage method is a straight forward solution, which process is mainly completed through a complete CNN, so CNN features will be leveraged to extract the feature description of the target among candidate regions through a CNN. The accuracy of the two-step method has been greatly improved, but the running speed is much slower than the one-step method. While the one-step method is less accurate, it is much faster. In this work, representative works for object detection are conducted and compared. The results could further demonstrate their respective advantages.

Image semantic segmentation using deep learning technique

With the deepening research on image understanding in many application fields, including auto drive system, unmanned aerial vehicle (UAV) landing point judgment, virtual reality wearable devices, etc., computer vision and machine learning researchers are paying more and more attention to image semantic segmentation (ISS). In this paper, according to the different region generation algorithms, the regional classification image semantic segmentation methods are classified into the candidate region method and the segmentation mask method, according to different learning methods, the image semantic segmentation methods based on super pixels are divided into fully supervised learning method and weakly supervised learning method. The typical algorithms in these various categories are summarized and compared. In addition, this paper also systematically expounds the role of DL technology in the field of ISS, and discusses the main challenges and future development prospects in this field.

Cork classification based on multi-scale faster-RCNN with machine vision

In traditional industrial production, the classification of cork mainly adopts manual operation. Due to round cork blocks with complex textures, manual sorting is time-consuming and often leads to inefficiency. It will lead to inefficiency and a high error rate. Therefore, an automatic cork sorting system is urgently required to replace manual operation. In this paper, an advanced deep-learning method based on machine vision is adopted to detect the surface of corks and classify them automatically. To robustly detect the texture of cork profile, a Multiple Scale Faster R-CNN based on the upper and lower layers (UMS-FRCNN) approach is proposed. The network consists of a region proposal component and the region-of-interest (RoI) detection component. There are two main contributions in our proposed network, which play a significant role in achieving state-of-the-art performance in cork classification. Firstly, the multi-scale information is grouped in two aspects: region suggestion and RoI detection, to detect on small cork texture regions. Secondly, the convolution Region Proposal Network (RPN) is used to generate the candidate regions of the UMS-FRCNN model to improve the classification accuracy. The experimental results show that our method not only can accurately classify the corks but also has better performance than the Faster RCNN. Compared with other methods, our method has the advantages of high precision, less time-consuming, and no dependence labeling. Our work has been successfully utilized in a cork production plant.

QTL Mapping for Ovary- and Fruit-Related Traits in Cucumis sativus-C. hystrix Introgression Line IL52.

IL52 is a valuable introgression line obtained from interspecific hybridization between cultivated cucumber (Cucumis sativus L., 2n = 14) and the wild relative species C. hystrix Chakr. (2n = 24). IL52 exhibits high resistance to a number of diseases, including downy mildew, powdery mildew, and angular leaf spot. However, the ovary- and fruit-related traits of IL52 have not been thoroughly investigated. Here, we conducted quantitative trait loci (QTL) mapping for 11 traits related to ovary size, fruit size, and flowering time using a previously developed 155 F7:8 RIL population derived from a cross between CCMC and IL52. In total, 27 QTL associated with the 11 traits were detected, distributed on seven chromosomes. These QTL explained 3.61% to 43.98% of the phenotypic variance. Notably, we identified a major-effect QTL (qOHN4.1) on chromosome 4 associated with the ovary hypanthium neck width and further delimited it into a 114-kb candidate region harboring 13 candidate genes. Furthermore, the QTL qOHN4.1 is co-localized with the QTL detected for ovary length, mature fruit length, and fruit neck length, all residing within the consensus QTL FS4.1, suggesting a plausible pleiotropic effect.

Clustering-based improved adaptive synthetic minority oversampling technique for imbalanced data classification

Synthetic Minority Oversampling Technique (SMOTE) and some extensions based on it are popularly used to balance imbalanced data. In this study, we concentrate on solving overfitting of the classification model caused by choosing instances to oversample that increase the occurrence of overlaps with the majority class. Our method called Clustering-based Improved Adaptive Synthetic Minority Oversampling Technique (CI-ASMOTE1) decomposes minority instances into sub-clusters according to their connectivity in the feature space and then selects minority sub-clusters which are relatively close to the decision boundary as the candidate regions to oversample. After application of CI-ASMOTE1, new minority instances are only synthesized within each connected region of the selected sub-clusters. Considering the diversity of the synthetic instances in each selected sub-cluster, CI-ASMOTE2 is put forward to extend CI-ASMOTE1 by keeping all features of those instances in the feature space as different as possible. The experimental evaluation shows that CI-ASMOTE1 and CI-ASMOTE2 improve SMOTE and its extensions, especially in the occurrence of overlaps between the minority instances and the majority instances.

Ravages: An R package for the simulation and analysis of rare variants in multicategory phenotypes.

Current software packages for the analysis and the simulations of rare variants are only available for binary and continuous traits. Ravages provides solutions in a single R package to perform rare variant association tests for multicategory, binary and continuous phenotypes, to simulate datasets under different scenarios and to compute statistical power. Association tests can be run in the whole genome thanks to C++ implementation of most of the functions, using either RAVA-FIRST, a recently developed strategy to filter and analyse genome-wide rare variants, or user-defined candidate regions. Ravages also includes a simulation module that generates genetic data for cases who can be stratified into several subgroups and for controls. Through comparisons with existing programmes, we show that Ravages complements existing tools and will be useful to study the genetic architecture of complex diseases. Ravages is available on the CRAN at https://cran.r-project.org/web/packages/Ravages/ and maintained on Github at https://github.com/genostats/Ravages.

Functional mapping of microRNA promoters with dCas9 fused to transcriptional regulators.

MicroRNAs are small non-coding RNAs that control gene expression during development, physiology, and disease. Transcription is a key factor in microRNA abundance and tissue-specific expression. Many databases predict the location of microRNA transcription start sites and promoters. However, these candidate regions require functional validation. Here, dCas9 fused to transcriptional activators or repressors - CRISPR activation (CRISPRa) and inhibition (CRISPRi)- were targeted to the candidate promoters of two intronic microRNAs, mmu-miR-335 and hsa-miR-3662, including the promoters of their respective host genes Mest and HBS1L. We report that in mouse embryonic stem cells and brain organoids, miR-335 was downregulated upon CRISPRi of its host gene Mest. Reciprocally, CRISPRa of Mest promoter upregulated miR-335. By contrast, CRISPRa of the predicted miR-335-specific promoter (located in an intron of Mest) did not affect miR-335 levels. Thus, the expression of miR-335 only depends on the promoter activity of its host gene Mest. By contrast, miR-3662 was CRISPR activatable both by the promoter of its host gene HBS1L and an intronic sequence in HEK-293T cells. Thus, CRISPRa and CRISPRi are powerful tools to evaluate the relevance of endogenous regulatory sequences involved in microRNA transcription in defined cell types.

Juvenile Idiopathic Epilepsy in Arabian Horses is not a Single Gene Disorder.

Valued for their temperament, beauty, athletic ability, and exhibition in the show ring, Arabian horses are an important component of the horse industry. Juvenile Idiopathic Epilepsy (JIE), a seizure disorder, is most often reported in Arabian foals from birth to six months of age. Affected foals exhibit tonic-clonic seizures lasting as long as five minutes and risking secondary complications like temporary blindness and disorientation. Some foals outgrow this condition, while others die or suffer lifelong complications if not treated. Previous work suggested a strong genetic component to JIE and proposed JIE to be a single-gene trait. In this work we conducted a GWAS in 60 cases of JIE and 120 genetically matched controls, identifying loci suggesting JIE is not caused by a single locus. Coat color (chestnut, grey) phenotypes were used as positive control traits to assess the efficacy of GWAs in this population. Future work will attempt to future define candidate regions and explore a polygenic mode of inheritance.

41 A preliminary investigation of candidate genes from a GWAS identified QTL for relative body size in the American Quarter Horse

Artificial selection during domestication has influenced an array of morphological traits in the horse, many that are now important economic traits for the horse industry. Growth and skeletal phenotypes impact the desirability and function of horses in diverse disciplines and are often key characteristics in the development of breeds. Skeletal traits also impact functionality: benefitting horses for specialized work or predisposing them to musculoskeletal injury. To investigate variations in body size and shape, we utilized measurements taken at 2 years of age (+/− one week) from 91 American Quarter Horses bred and owned by the University of Florida. We summarized 32 body measures from each horse using a principal component analysis followed by a varimax rotation. Factor 1 summarizes 19.5% of the variation and describes overall body size, except for head measurements which are better captured in Factor 2. Factor 1 traits encompass body proportions, heavily loading on the measure's height at dock and croup. DNA was extracted from blood and hair through previously published methods and genotyped at GeneSeek Inc. (Lansing, MI) using the Affymetrix Axiom Equine HD 670k array. We recently reported results from a preliminary genome-wide association study with 311,085 high-quality SNPs (poly-high resolution, MAF >0.05)suggesting a quantitative trait locus for the Body Factor 1 trait on chromosome 6 (raw P = 1.356e-09, Bonferroni P = 4.51e-07) and defines a candidate region spanning 1.5Mb. The QTL is in a region unrelated to a previously characterized HMGA2 body height locus. The candidate region contains 5 protein-coding genes, influencing human height and skeletal traits including one associated with Noonan Syndrome- a disease impacting height and facial morphology, an interesting parallel to the phenotype observed in our population. None of the 16,860 variants found in the European Nucleotide Archive for our candidate region had a functional impacted on a protein coding region. Therefore, we sequenced 2 whole genomes from American Quarter Horses to discover variants specific to the haplotype bearing this new body size QTL. Using the 10x Chromium library prep with 150bp paired end reads on an Illumina HiSeq2500, the 2 genomes totaled 1.5 billion reads. Following alignment and variant calling using the LongRanger package (v2.2, 10x Genomics Inc.), we observed 4,703 SNPs in the candidate region. Future investigation aims to investigate additional loci relevant to body morphological variations and develop predictive phenotype models in the horse.

Don't panic any more: MORE PANICLES 3 increases rice yield at elevated CO2 levels.

Don't panic any more: MORE PANICLES 3 increases rice yield at elevated CO2 levels.

43 A QTL for skull morphology in Arabian horses

Domestication and selective pressures alter developmental pathways of mammals. The identification of genomic loci for skull morphology in particular can illuminate pathways controlling facial shape. Previous work in the dog identified QTLs for skull morphology and highlighted relationships formed during the development of distinct breeds of dogs from lupine ancestors through the direction of 6 loci across the genome. The Arabian horse is known for its unique concave facial profile in comparison to other equine breeds. We extracted DNA from blood and hair samples for 103 Arabian horses exhibiting diverse skull morphology and genotyped these with the Affymetrix Axiom Equine HD 670k array. Phenotyping was accomplished using 7 defined face measures from a defined protocol and summarized face shape variation using a principal component analysis. Utilizing these 2 principal component scores as skull morphology phenotypes, we performed a genome wide association study using 315,553 SNPs to identify significant preliminary QTLs for the strongest variation in facial morphologywithin this population. We used a linear regression model (PLINK v1.90b6.26) including quality filters for MAF >0.05 and genotyping rate of above 95%. Given the stratification of this population and diversity of subgroups within the Arabian breed, efforts to account for inflation were implemented using the genomic relatedness as a covariate in the linear model. Strong variation in head lengths were noted in the skull principal component 1 phenotype, independent from body size and skull length. Preliminary GWAs identified a promising locus on ECA 5, candidate base pair region: 27,578,838–27,848,837, (raw P = 1.421e-11, Bonferroni corrected P = 1.584e-7) despite strong population stratification across the diverse Arabian breed. Three protein coding genes were noted in this candidate region and further work will investigate these for potentially functional variants and improve phenotype accuracy by applying photo morphometics.

ROI Scissor: Interactive Segmentation of Feature Region of Interest in a Triangular Mesh

AbstractWe present a simple and effective method for the interactive segmentation of feature regions in a triangular mesh. From the user‐specified radius and click position, the candidate region that contains the desired feature region is defined as geodesic disc on a triangle mesh. A concavity‐aware harmonic field is then computed on the candidate region using the appropriate boundary constraints. An initial isoline is chosen by evaluating the uniformly sampled ones on the harmonic field based on the gradient magnitude. A set of feature points on the initial isoline is selected and the anisotropic geodesics passing through them are then determined as the final segmentation boundary, which is smooth and locally shortest. The experimental results show several segmentation results for various 3D models, revealing the effectiveness of the proposed method.

High-resolution mapping of SrTm4, a recessive resistance gene to wheat stem rust

Key messageThe diploid wheat recessive stem rust resistance gene SrTm4 was fine-mapped to a 754-kb region on chromosome arm 2AmL and potential candidate genes were identified.Race Ug99 of Puccinia graminis f. sp. tritici (Pgt), the causal agent of wheat stem (or black) rust is one of the most serious threats to global wheat production. The identification, mapping, and deployment of effective stem rust resistance (Sr) genes are critical to reduce this threat. In this study, we generated SrTm4 monogenic lines and found that this gene confers resistance to North American and Chinese Pgt races. Using a large mapping population (9522 gametes), we mapped SrTm4 within a 0.06 cM interval flanked by marker loci CS4211 and 130K1519, which corresponds to a 1.0-Mb region in the Chinese Spring reference genome v2.1. A physical map of the SrTm4 region was constructed with 11 overlapping BACs from the resistant Triticum monococcum PI 306540. Comparison of the 754-kb physical map with the genomic sequence of Chinese Spring and a discontinuous BAC sequence of DV92 revealed a 593-kb chromosomal inversion in PI 306540. Within the candidate region, we identified an L-type lectin-domain containing receptor kinase (LLK1), which was disrupted by the proximal inversion breakpoint, as a potential candidate gene. Two diagnostic dominant markers were developed to detect the inversion breakpoints. In a survey of T. monococcum accessions, we identified 10 domesticated T. monococcum subsp. monococcum genotypes, mainly from the Balkans, carrying the inversion and showing similar mesothetic resistant infection types against Pgt races. The high-density map and tightly linked molecular markers developed in this study are useful tools to accelerate the deployment of SrTm4-mediated resistance in wheat breeding programs.

Molecular Marker-Assisted Selection for Frost Tolerance in a Diallel Population of Potato.

A multi-parental population is an innovative tool for mapping large numbers of loci and genetic modifications, particularly where they have been used for breeding and pre-breeding in crops. Frost injury is an environmental stress factor that greatly affects the growth, development, production efficiency, and geographical distribution of crops. No reported study has focused on genetic mapping and molecular marker development using diallel populations of potatoes. In this study, 23 successful cross combinations, obtained by a half diallel cross among 16 parents, including eight frost-tolerant advanced breeding lines and eight cultivars, were used to map the genetic loci for frost tolerance and to create a molecular marker-assisted selection (MAS) system. Three candidate regions related to frost tolerance on chromosomes II, V, and IX were mapped by bulked segregant analysis (BSA). Furthermore, six SNP markers associated with frost tolerance from candidate regions were developed and validated. Above all, a MAS system for the frost tolerance screening of early breeding offspring was established. This study highlights the practical advantages of applying diallel populations to broaden and improve frost-tolerant germplasm resources.

Fine mapping and identifying candidate gene of Y underlying yellow peel in Cucurbita pepo.

As a conspicuous trait, peel color is one of the most important characteristics that affects commodity quality and consumer preferences. The locus Y underlying yellow peel in Cucurbita pepo (zucchini) was first reported in 1922; however, its molecular mechanism is still unknown. In this study, a genetic analysis revealed that yellow peel is controlled by a single dominant genetic factor. Furthermore, Y was mapped in a ~170 kb region on chromosome 10 by bulked segregated analysis (BSA) and fine mapping in F2 and BC1 segregating populations. The candidate region harbors fifteen annotated genes, among which Cp4.1LG10g11560 (CpCHLH) is regarded as a promising candidate gene. CpCHLH encodes a magnesium chelatase H subunit involved in chlorophyll biosynthesis, and its mutation can result in a reduction in chlorophyll content and yellow phenotype. Interestingly, a large fragment (~15 kb) duplication containing incomplete CpCHLH was inserted in the candidate interval, resulting in two reformed CpCHLH proteins in the yellow parental line. It is most likely that the reformed CpCHLH proteins act as a malfunctional competitor of the normal CpCHLH protein to interrupt the formation of chlorophyll. Overall, the isolation of Y will shed light on the molecular mechanism of the peel color regulation of zucchini and lay a foundation for breeding.

Genetic fine-mapping reveals single nucleotide polymorphism mutations in the MC1R regulatory region associated with duck melanism.

Birds are among the most colourful terrestrial vertebrates, with various plumage colours and patterns. We conducted a genome-wide association study (GWAS) on an intercross F2 population of Pekin ducks and mallards (n=722) and identified a 1.57-Mb genetic region (Chr11: 20,176,480-21,750,101 bp) related to duck melanism. Fine mapping by linkage disequilibrium (LD) and FST analysis narrowed the final candidate region to a region of 22,500 bp (Chr11: 20,677,500-20,700,000 bp) including three coding genes, TCF25, MC1R and TUBB3. Combined with transcriptome and qRT-PCR analysis, MC1R was identified as the unique genetic locus responsible for black plumage in ducks, and it was significantly more highly expressed in the feather bulbs of black ducks. We also identified 52G > A (Chr11: 20,696,354G > A) and 376G > A (Chr11: 20,696,678G > A) mutations in the MC1R coding region that have been widely studied in ducks. In addition, structural variations (SVs) were screened by nanopore sequencing, and no significant SV was found to be associated with the duck black plumage trait. However, we identified four novel single nucleotide polymorphisms in the MC1R regulator region (Chr11: 20,678,412G > A, Chr11: 20,679,236G > A, Chr11: 20,692,496 A > G and Chr11: 20,692,791 A > G) that had a strong association with the black plumage phenotype of ducks and combined with potential changes in transcription binding affinities. The luciferase reporter gene assay demonstrated that Chr11: 20,678,412G > A and Chr11: 20,679,236G > A led to significant promoter activity changes. Our research emphasizes the importance of MC1R regulatory region mutation in determining the duck black plumage phenotype, and these results expand our understanding of the genetic mechanism underlying duck plumage colour.

UPG: 3D vision-based prediction framework for robotic grasping in multi-object scenes

Robotic grasping has the challenge of accurately extracting the graspable target from a complicated scenario. To address the issue, we propose a 3D vision prediction framework including visual observation and pose estimation. Firstly, we exploit the continuity characteristics in the U-disparity map to identify the isolated objects and occluded objects which can quickly partition the grasping scene and produce valid candidate regions for grasping. Secondly, an end-to-end approach based on PointNet＋＋ is improved to obtain the topmost target if there is a pile of stacked objects. We also provide a robust labeling method for generating the datasets comprising the multi-object scenes. Moreover, a designed evaluation criterion is presented to assist with estimating the 6-DOF (degree of freedom) pose. Our method UPG (U-disparity and PointNet＋＋ grasping) simplifies the segmentation task and makes the training model lightweight in order to apply in practical bin-picking and assembly. To validate the feasibility, UPG is evaluated on simulation and real-world scenes, respectively. The extensive results indicate that UPG can achieve better segmentation accuracy and grasping success rates against other state-of-the-arts.

Two Path Gland Segmentation Algorithm of Colon Pathological Image Based on Local Semantic Guidance.

Colonic adenocarcinoma is a disease severely endangering human life caused by mucosal epidermal carcinogenesis. The segmentation of potentially cancerous glands is the key in the detection and diagnosis of colonic adenocarcinoma. The appearance of cancerous tissue is different in gland segmentation in colon pathological images, and it is impossible to accurately segment the changes of glands from benign to malignant using a single network. Given these issues, a two-path gland segmentation algorithm of colon pathological image based on local semantic guidance is proposed in this paper. The improved candidate region search algorithm is adopted to expand the original image data set and generate sub-datasets sensitive to specific features. Then, the semantic feature-guided model is employed to extract the local adenocarcinoma features and acts on the backbone network together with context feature extraction based on the attention mechanism. In this way, a larger receptive field and more local feature information are obtained, the learning ability of the network to the morphological features of glands is enhanced, and the performance of automatic gland segmentation is finally improved. The algorithm is verified on Warwick Qu-Dataset. Compared with the current popular segmentation algorithms, our algorithm has good performance in Dice coefficient, F1 score, and Hausdorff distance on different types of test sets.

Reliability Updating with a System Reliability Method Based on Adaptive Kriging

Reliability updating can improve the failure probability estimates of structural systems when new observations are obtained. At present, an efficient reliability updating method is one that transforms the equality observation information into the inequality type by introducing an auxiliary variable and then employs structural reliability methods to perform reliability updating. For this method, the computation of posterior failure probability Pr(F|Z), where Z is an observation event, requires two reliability analysis processes to estimate Pr(Z) and Pr(F∩Z), respectively. To improve the efficiency, this paper develops a reliability updating method with output observations, where the computation of Pr(Z) and Pr(F∩Z) is integrated into one system reliability problem, which is then solved by adaptive kriging with truncated candidate region (AKTCR). A composite adaptive learning process is developed to construct the kriging model, which can provide accurate estimates for both Pr(Z) and Pr(F∩Z). In this way, the proposed method can also deal with the problem of multiple failure modes and multiple observation events. In addition, a combined global and local sampling method is embedded into the proposed algorithm to reduce the size of candidate sample pool and improve the modeling efficiency. Finally, three examples can prove the efficiency and accuracy of the proposed reliability updating method.