Candidate Regions Of Interest Research Articles

Exploiting spatial transcriptomics to investigate the molecular mechanisms involved in cardiac metastasis development

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): IEO-Monzino Foundation Background Cardiac metastasis is considered a relatively rare malignancy, although an increasing prevalence is expected in the future due to the increasing life expectancy. Several cases of cardiac metastases remain untreatable, given the anatomical location that often hampers both surgical eradication and bioptic collection for biological characterization. Here, we use spatial transcriptomics to provide an unbiased molecular characterization of cardiac metastases, and identify novel targets for their treatment. Material and methods We performed spatial transcriptomics on samples of primary tumor, cardiac, and extra-cardiac metastasis from three different patients. First, we checked RNA integrity by means of DV200 analysis. Spatial transcriptomic analysis was performed by using the GeoMX technology (Nanostring), which provides whole-transcriptome data along with spatial distribution information of the preselected samples from the candidate regions of interest (ROIs). ROIs were selected on the formalin-fixed paraffin-embedded sections by means of geometric shapes in combination with cell markers specific for tumor cells, cardiomyocytes and endothelial cells, respectively. Differential gene expression analysis and Gene Set Enrichment Analysis (GSEA) were performed on the entire transcriptome. Cell-cell interactions between cardiac cells and cancer cells were then evaluated by means of InterCellaR. Results Differential expression analysis between cardiac metastases and extra-cardiac tumors (primary tumors + extra-cardiac metastases) identified 92 differentially expressed genes (DEGs). Of these, 53 genes were significantly overexpressed in cardiac metastases, while 39 genes were overexpressed in the extra-cardiac samples, regardless of the site of origin of the tumor. Unsupervised hierarchical clustering of protein-coding genes clearly separated the samples according to their anatomical location, in particular cardiac metastasis from extra-cardiac tumors. GSEA identified several pathways enriched in cardiac metastases compared to the extra-cardiac tumors, including mechanotransduction (e.g. keratinization and cell-matrix interaction) and triglyceride metabolism, suggesting a potential role of these signaling pathways in the development of cardiac metastasis. Interactome analysis revealed a rich plethora of receptor-ligand interactions in cardiac metastasis, especially between tumor cells and cardiomyocytes. Conclusions These results start shedding light on the mechanisms that allow cancer cell growth in the heart and pave the way to the development of novel therapies to treat this malignancy.

Optimizing strategy for the discovery of compositionally-biased or low-complexity regions in proteins

Proteins can contain tracts dominated by a subset of amino acids and that have a functional significance. These are often termed ‘low-complexity regions’ (LCRs) or ‘compositionally-biased regions’ (CBRs). However, a wide spectrum of compositional bias is possible, and program parameters used to annotate these regions are often arbitrarily chosen. Also, investigators are sometimes interested in longer regions, or sometimes very short ones. Here, two programs for annotating LCRs/CBRs, namely SEG and fLPS, are investigated in detail across the whole expanse of their parameter spaces. In doing so, boundary behaviours are resolved that are used to derive an optimized systematic strategy for annotating LCRs/CBRs. Sets of parameters that progressively annotate or ‘cover’ more of protein sequence space and are optimized for a given target length have been derived. This progressive annotation can be applied to discern the biological relevance of CBRs, e.g., in parsing domains for experimental constructs and in generating hypotheses. It is also useful for picking out candidate regions of interest of a given target length and bias signature, and for assessing the parameter dependence of annotations. This latter application is demonstrated for a set of human intrinsically-disordered proteins associated with cancer.

Multi-Scale Region-Of-Interest based Deep Learning for fruit disease identification and classification

Agricultural productivity is mainly affected by the different types of fruit diseases. To reduce such impacts, automated fruit disease identification and categorization model is vital. So, a Multi-scale Deep Learning (MDL) model was designed to identify and categorize the variety of strawberry fruit diseases. This model has an Adaptive Receptive Field Module (ARFM), a Bottleneck Block (BB), Depth-wise Residual Blocks (DRBs) and the Squeeze Excitation Residual Network (SE-ResNet) units to extract the Feature Maps (FMs) related to all disease classes. But, this model cannot partition individual fruit samples. Also, the number of fruits with multiple infections within a particular image was not easily captured. So, this article develops a Multi-scale Region-Of-Interest (ROI)-based DL model called (MRDL) to improve the accuracy of identification and categorization of a single fruit sample with a certain disease in a real-time scenario. This model is performed in 2 different phases. First, more robust features are extracted by the DRB, ARFM, and BB in the initial phase. Then, an ROI Fused (ROIF) unit, the SE-ResNet unit, a classification unit, and a mask unit are involved in the second phase. The ROIF unit obtains the FMs for all candidate ROIs using the candidate bounding boxes from the initial phase. The extracted FMs per sampled position are aggregated to get the fused FM, which is resized by integrating the attention strategy with the SE-ResNet unit. Those resized fused FMs are independently passed to the classification and mask units to classify the candidate bounding boxes and localize the infected pixels in the candidate ROIs, respectively. Finally, the test results exhibit that the MRDL model achieves 91.1%, 90.8%, and 91.6% accuracy on apple, citrus, and tomato disease databases, respectively compared to the classical models.

Age, sex and APOE-ε4 modify the balance between soluble and fibrillar β-amyloid in non-demented individuals: topographical patterns across two independent cohorts

Amyloid (Aβ) pathology is the earliest detectable pathophysiological event along the Alzheimer’s continuum, which can be measured both in the cerebrospinal fluid (CSF) and by Positron Emission Tomography (PET). Yet, these biomarkers identify two distinct Aβ pools, reflecting the clearance of soluble Aβ as opposed to the presence of Aβ fibrils in the brain. An open question is whether risk factors known to increase Alzheimer’s’ disease (AD) prevalence may promote an imbalance between soluble and deposited Aβ. Unveiling such interactions shall aid our understanding of the biological pathways underlying Aβ deposition and foster the design of effective prevention strategies. We assessed the impact of three major AD risk factors, such as age, APOE-ε4 and female sex, on the association between CSF and PET Aβ, in two independent samples of non-demented individuals (ALFA: n = 320, ADNI: n = 682). We tested our hypotheses both in candidate regions of interest and in the whole brain using voxel-wise non-parametric permutations. All of the assessed risk factors induced a higher Aβ deposition for any given level of CSF Aβ42/40, although in distinct cerebral topologies. While age and sex mapped onto neocortical areas, the effect of APOE-ε4 was prominent in the medial temporal lobe, which represents a target of early tau deposition. Further, we found that the effects of age and APOE-ε4 was stronger in women than in men. Our data indicate that specific AD risk factors affect the spatial patterns of cerebral Aβ aggregation, with APOE-ε4 possibly facilitating a co-localization between Aβ and tau along the disease continuum.

Simultaneous detection and diagnosis of mammogram mass using bilateral analysis and soft label based metric learning

In the clinics, mammogram masses appear as asymmetric structures between the left and right breasts. In this paper, we design a bilateral image analysis method based on convolutional neural network which can detect and classify breast mass regions simultaneously. It mainly consists of three parts: a feature similarity based region matching technique, mass region of interest (ROI) selection step and a deep metric learning based classifier. Firstly, discriminative score maps are calculated relied on the deep features extracted from bilateral left and right mammograms respectively in global or local spatial image domain. The contralateral correspondences are determined by minimum discriminative scores. Secondly, to select the mass candidate ROIs and further remove false positive mass-to-normal pairs, we propose a dynamic histogram weighting mechanism with three new constrains imposed on the distribution of discriminative score histogram. In addition, a novel soft label based deep metric learning regularization is designed for mass ROI classifier to tackle the large variation of masses in shape, size, texture and breast density. We apply it to the open dataset Digital Database for Screening Mammography. Compared with other state-of-the-art approaches, the proposed scheme gives competitive results in classification and localization tasks for mammographic lesions.

A human-computer collaboration for COVID-19 differentiation: combining a radiomics model with deep learning and human auditing.

This study aimed to build a radiomics model with deep learning (DL) and human auditing and examine its diagnostic value in differentiating between coronavirus disease 2019 (COVID-19) and community-acquired pneumonia (CAP). Forty-three COVID-19 patients, whose diagnoses had been confirmed with reverse-transcriptase polymerase-chain-reaction (RT-PCR) tests, and 60 CAP patients, whose diagnoses had been confirmed with sputum cultures, were enrolled in this retrospective study. The candidate regions of interest (ROIs) on the computed tomography (CT) images of the 103 patients were determined using a DL-based segmentation model powered by transfer learning. These ROIs were manually audited and corrected by 3 radiologists (with an average of 12 years of experience; range 6-17 years) to check the segmentation acceptance for the radiomics analysis. ROI-derived radiomics features were subsequently extracted to build the classification model and processed using 4 different algorithms (L1 regularization, Lasso, Ridge, and Z test) and 4 classifiers, including the logistic regression (LR), multi-layer perceptron (MLP), support vector machine (SVM), and extreme Gradient Boosting (XGboost). A receiver operating characteristic curve (ROC) analysis was conducted to evaluate the performance of the model. Quantitative CT measurements derived from human-audited segmentation results showed that COVID-19 patients had significantly decreased numbers of infected lobes compared to patients in the CAP group {median [interquartile range (IQR)]: 4 [3, 4] and 4 [4, 5]; P=0.031}. The infected percentage (%) of the whole lung was significantly more elevated in the CAP group [6.40 (2.77, 11.11)] than the COVID-19 group [1.83 (0.65, 4.42); P<0.001], and the same trend applied to each lobe, except for the superior lobe of the right lung [1.81 (0.09, 5.28) for COVID-19 vs. 1.32 (0.14, 7.02) for CAP; P=0.649]. Additionally, the highest proportion of infected lesions were observed in the CT value range of (-470, -370) Hounsfield units (HU) in the COVID-19 group. Conversely, the CAP group had a value range of (30, 60) HU. Radiomic model using corrected ROIs exhibited the highest area under ROC (AUC) of 0.990 [95% confidence interval (CI): 0.962-1.000] using Lasso for feature selection and MLP for classification. The proposed radiomics model based on human-audited segmentation made accurate differential diagnoses of COVID-19 and CAP. The quantification of CT measurements derived from DL could potentially be used as effective biomarkers in current clinical practice.

Remote Photoplethysmography With an EEMD-MCCA Method Robust Against Spatially Uneven Illuminations

Remote photoplethysmography (rPPG) is an attractive video-based technique to monitor heart rate (HR) information in telehealth screening. The subjects are required to stay stationary in the remote screening scenario and ambient lights become the main interference source of measuring HR with rPPG. In this paper, we introduce a novel approach robust against spatially uneven illumination interference in rPPG by combining ensemble empirical mode decomposition (EEMD) with multiset canonical correlation analysis (MCCA). We adopt the following procedures to ensure that the pulses dominate the correlations across multiple signal sets while the illumination noises are as diverse as possible. Specifically, a group of optimal regions of interest (ROIs) are selected according to the quality indicators defined from the green channel in each candidate ROI. A multi-channel signal set is then constructed by decomposing the green signal with EEMD. Only those intrinsic mode functions (IMFs) with the dominant frequencies falling into the interested HR range are utilized as the input of MCCA. The canonical variables (CVs) with the highest cross correlations are derived as the underlying candidate pulses. Finally, fast Fourier transform (FFT) is employed to calculate the dominant frequency and the target HR is determined based on both the quasi-periodic property and the continuity of HR. The proposed EEMD-MCCA method is validated on both the in-house BSIPL-RPPG and the public COHFACE databases, which achieves superior performance over several typical rPPG methods. This study will provide a promising tool for realistic rPPG applications in telehealth screening.

Association between alexithymia and impaired reward valuation in patients with fronto-insular damage.

Humans compute the anticipated reward value of stimuli in their environment in order to behave in an adaptive, goal-directed manner. This reward valuation ability is vital, and its disruption in a range of clinical populations has profound personal and social consequences. However, research has often failed to consider the reward-related functions of a central component of human emotion: conscious emotional experience. Alexithymia-a condition characterized by diminished conscious awareness of one's emotions-offers a unique opportunity to examine the link between emotional awareness and reward valuation. In the present study, we measured both acquired alexithymia and reward valuation ability in a large sample of patients with traumatic brain injuries (N = 112). Behavioral analyses provided evidence for a negative association between alexithymia and reward valuation ability. This association remained significant after controlling for several covariates in the model (anxiety, depression, posttraumatic stress disorder, and IQ). Voxel-based lesion-symptom mapping was carried out to identify brain regions-of-interest (ROIs) that, when damaged, lead to increased alexithymia and impaired reward valuation. Importantly, mediation models computed using the ROIs identified through the voxel-based lesion-symptom mapping revealed a specific indirect effect of left frontoinsular damage on impaired valuation that was mediated by increased levels of alexithymia. This indirect effect was not observed for any of the other candidate ROIs. The present study identifies a network of brain regions likely to be involved in the integration of subjective feelings and reward processes critical for the adaptive control of goal-directed behavior. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

A Coarse-to-Fine Framework for Multiple Pedestrian Crossing Detection.

When providing route guidance to pedestrians, one of the major safety considerations is to ensure that streets are crossed at places with pedestrian crossings. As a result, map service providers are keen to gather the location information about pedestrian crossings in the road network. Most, if not all, literature in this field focuses on detecting the pedestrian crossing immediately in front of the camera, while leaving the other pedestrian crossings in the same image undetected. This causes an under-utilization of the information in the video images, because not all pedestrian crossings captured by the camera are detected. In this research, we propose a coarse-to-fine framework to detect pedestrian crossings from probe vehicle videos, which can then be combined with the GPS traces of the corresponding vehicles to determine the exact locations of pedestrian crossings. At the coarse stage of our approach, we identify vanishing points and straight lines associated with the stripes of pedestrian crossings, and partition the edges to obtain rough candidate regions of interest (ROIs). At the fine stage, we determine whether these candidate ROIs are indeed pedestrian crossings by exploring their prior constraint information. Field experiments in Beijing and Shanghai cities show that the proposed approach can produce satisfactory results under a wide variety of situations.

Genome-wide analysis of high-risk primary brain cancer pedigrees identifies PDXDC1 as a candidate brain cancer predisposition gene.

There is evidence for an inherited contribution to primary brain cancer. Linkage analysis of high-risk brain cancer pedigrees has identified candidate regions of interest in which brain cancer predisposition genes are likely to reside. Genome-wide linkage analysis was performed in a unique set of 11 informative, extended, high-risk primary brain cancer pedigrees identified in a population genealogy database, which include from 2 to 6 sampled, related primary brain cancer cases. Access to formalin-fixed paraffin embedded tissue samples archived in a biorepository allowed analysis of extended pedigrees. Individual high-risk pedigrees were singly informative for linkage at multiple regions. Suggestive evidence for linkage was observed on chromosomes 2, 3, 14, and 16. The chromosome 16 region in particular contains a promising candidate gene, pyridoxal-dependent decarboxylase domain-containing 1 (PDXDC1), with prior evidence for involvement with glioblastoma from other previously reported experimental settings, and contains the lead single nucleotide polymorphism (rs3198697) from the linkage analysis of the chromosome 16 region. Pedigrees with a statistical excess of primary brain cancers have been identified in a unique genealogy resource representing the homogeneous Utah population. Genome-wide linkage analysis of these pedigrees has identified a potential candidate predisposition gene, as well as multiple candidate regions that could harbor predisposition loci, and for which further analysis is suggested.

Vehicle Detection In Remote Sensing Images

Traffic monitoring and management is one of the most crucial tasks of governing bodies in modern big cities. With each passing day the traffic problem grows in complexity due to the continuous increase of participating vehicles and the hard expansion of the road network and parking places. In this article we introduce a new method for vehicle detection and localization in parking lots using high resolution UAV images. In order to end up with practical and yet effective approach, which could be implemented on low computing hardware resources and integrated with the camera in the UAV, we considered simple steps in the proposed algorithm for optimization. It follows the machine learning pipeline such as preprocessing, sensing, feature extraction, training and classification. In preprocessing the images are thresholded iteratively in multiple color spaces to extract the candidate regions of interest (ROI). The algorithm relies on point and shape features using fast techniques in the feature extraction. The features are then clustered by the K-means algorithm and represented by the resulted clusters’ centers. Region based linear classification is finally applied using SVM to classify if the object is a vehicle or else. The proposed approach proved high detection and classification accuracy more than 86% and still running under the low complexity constraint..

Vehicle Detection In Remote Sensing Images

Traffic monitoring and management is one of the most crucial tasks of governing bodies in modern big cities. With each passing day the traffic problem grows in complexity due to the continuous increase of participating vehicles and the hard expansion of the road network and parking places. In this article we introduce a new method for vehicle detection and localization in parking lots using high resolution UAV images. In order to end up with practical and yet effective approach, which could be implemented on low computing hardware resources and integrated with the camera in the UAV, we considered simple steps in the proposed algorithm for optimization. It follows the machine learning pipeline such as preprocessing, sensing, feature extraction, training and classification. In preprocessing the images are thresholded iteratively in multiple color spaces to extract the candidate regions of interest (ROI). The algorithm relies on point and shape features using fast techniques in the feature extraction. The features are then clustered by the K-means algorithm and represented by the resulted clusters’ centers. Region based linear classification is finally applied using SVM to classify if the object is a vehicle or else. The proposed approach proved high detection and classification accuracy more than 86% and still running under the low complexity constraint..

Spatial Keyword Query of Region-Of-Interest Based on the Distributed Representation of Point-Of-Interest

The tremendous advance in information technology has promoted the rapid development of location-based services (LBSs), which play an indispensable role in people’s daily lives. Compared with a traditional LBS based on Point-Of-Interest (POI), which is an isolated location point, an increasing number of demands have concentrated on Region-Of-Interest (ROI) exploration, i.e., geographic regions that contain many POIs and express rich environmental information. The intention behind the POI is to search the geographical regions related to the user’s requirements, which contain some spatial objects, such as POIs and have certain environmental characteristics. In order to achieve effective ROI exploration, we propose an ROI top-k keyword query method that considers the environmental information of the regions. Specifically, the Word2Vec model has been introduced to achieve the distributed representation of POIs and capture their environmental semantics, which are then leveraged to describe the environmental characteristic information of the candidate ROI. Given a keyword query, different query patterns are designed to measure the similarities between the query keyword and the candidate ROIs to find the k candidate ROIs that are most relevant to the query. In the verification step, an evaluation criterion has been developed to test the effectiveness of the distributed representations of POIs. Finally, after generating the POI vectors in high quality, we validated the performance of the proposed ROI top-k query on a large-scale real-life dataset where the experimental results demonstrated the effectiveness of our proposals.

Automated localization and quality control of the aorta in cine CMR can significantly accelerate processing of the UK Biobank population data.

IntroductionAortic distensibility can be calculated using semi-automated methods to segment the aortic lumen on cine CMR (Cardiovascular Magnetic Resonance) images. However, these methods require visual quality control and manual localization of the region of interest (ROI) of ascending (AA) and proximal descending (PDA) aorta, which limit the analysis in large-scale population-based studies. Using 5100 scans from UK Biobank, this study sought to develop and validate a fully automated method to 1) detect and locate the ROIs of AA and PDA, and 2) provide a quality control mechanism.MethodsThe automated AA and PDA detection-localization algorithm followed these steps: 1) foreground segmentation; 2) detection of candidate ROIs by Circular Hough Transform (CHT); 3) spatial, histogram and shape feature extraction for candidate ROIs; 4) AA and PDA detection using Random Forest (RF); 5) quality control based on RF detection probability. To provide the ground truth, overall image quality (IQ = 0–3 from poor to good) and aortic locations were visually assessed by 13 observers. The automated algorithm was trained on 1200 scans and Dice Similarity Coefficient (DSC) was used to calculate the agreement between ground truth and automatically detected ROIs.ResultsThe automated algorithm was tested on 3900 scans. Detection accuracy was 99.4% for AA and 99.8% for PDA. Aorta localization showed excellent agreement with the ground truth, with DSC ≥ 0.9 in 94.8% of AA (DSC = 0.97 ± 0.04) and 99.5% of PDA cases (DSC = 0.98 ± 0.03). AA×PDA detection probabilities could discriminate scans with IQ ≥ 1 from those severely corrupted by artefacts (AUC = 90.6%). If scans with detection probability < 0.75 were excluded (350 scans), the algorithm was able to correctly detect and localize AA and PDA in all the remaining 3550 scans (100% accuracy).ConclusionThe proposed method for automated AA and PDA localization was extremely accurate and the automatically derived detection probabilities provided a robust mechanism to detect low quality scans for further human review. Applying the proposed localization and quality control techniques promises at least a ten-fold reduction in human involvement without sacrificing any accuracy.

Patient-Reported Outcome Measures of Long-Term Swallowing Function after Oropharyngeal Radiation Therapy: A Cross-Sectional Methods Comparison

Definitive radiation therapy (RT) plays an important role in the curative treatment of many patients with oropharyngeal cancer (OPC), but carries a risk of long-term toxicity including the development of severe dysphagia. Here we characterize the relationship of patient-reported outcomes (PRO) of swallowing to quality of life (QOL) and radiation dose in long-term OPC survivors following definitive RT. Patients with non-metastatic OPC treated with definitive RT at a single institution between January 2000 and April 2014 were examined. All patients completed the MD Anderson Symptom Inventory—Head and Neck (MDASI-HN), the MD Anderson Dysphagia Inventory (MDADI), and the EuroQol five dimension questionnaire (EQ-5D VAS) 12 months or more following end of RT in a cross-sectional survivorship survey. A single dysphagia question from MDASI-HN (designated MDASI-HN-S) was used. An MDASI-HN-S score ≥6 and composite MDADI score <60 were considered representative of moderate-severe swallowing dysfunction. The Bayesian information criteria (BIC) were used to test the predictive power of MDASI-HN-S and composite MDADI with QOL as measured by EQ-5D VAS. The modified Breiman recursive partitioning analysis (RPA) was used to identify an MDASI-HN-S cut off score for poor QOL. To assess validity of PRO cutpoints, the mean dose to candidate regions of interest (ROI) were compared in patients with and without moderate/severe dysphagia symptoms. A total of 714 patients met criteria for inclusion, with 91.9% treated with intensity modulated radiotherapy (IMRT) and a mean time from treatment of 6.7 years. Moderate/severe dysphagia was reported by 17% and 16% of respondents by MDASI-HN-S and composite MDADI, respectively. QOL as measured by EQ-5D VAS was lower in those with moderate/severe dysphagia per both MDASI-HN-S (63.4 vs 83.5, p<0.001) and composite MDADI scores (64.5 vs 83.2, p< 0.001). Although both inventories were predictive of QOL, the model including MDASI-HN-S was slightly more parsimonious for discrimination of EQ-5D VAS scores compared to composite MDADI (BIC of 6062 vs 6076). Recursive partitioning analysis showed that an MDASI-HN-S cutoff score of ≥6 correlated best with EQ-5D VAS decline (LogWorth 2.7, p<0.001). Among 320 patients with available DICOM-RT plans, a higher mean dose to the contralateral anterior digastric muscle, intrinsic tongue muscles, and larynx was associated with moderate-severe dysphagia by both MDASI-HN-S and composite MDADI (all p <0.05). In long-term survivors of OPC following definitive RT, the single item dysphagia MDASI-HN-S question performed as well as the 20-item MDADI for discrimination of patient-reported dysphagia and QOL. Dose-response relationships to candidate ROIs were similar for the two questionnaires.

Identifying outliers using multiple kernel canonical correlation analysis with application to imaging genetics

Identifying significant outliers or atypical objects from multimodal datasets is an essential and challenging issue for biomedical research. This problem is addressed, using the influence function of multiple kernel canonical correlation analysis. First, the influence function (IF) of the kernel mean element, the kernel covariance operator, the kernel cross-covariance operator and kernel canonical correlation analysis (kernel CCA) are studied. Second, an IF of multiple kernel CCA is proposed, which can be applied to multimodal datasets. Third, a visualization method is proposed to detect influential observations of multiple sources of data based on the IF of kernel CCA and multiple kernel CCA. Finally, to validate the method, experiments on both synthesized and imaging genetics data (e.g., SNP, fMRI, and DNA methylation) are performed. To examine the outliers, both the stem-and-leaf display and distribution based technique are used. The performance of the proposed approach is illustrated on 116 candidate regions of interest (ROIs) from the fMRI data of schizophrenia study to identify significant ROIs. The proposed method and two state-of-the-art statistical methods have identified 8, 34, and 10 ROIs, respectively. Based on an online database, the brain mappings of the selected common 7 ROIs indicate the irregular brain regions susceptible to schizophrenia. The results demonstrate that the proposed method is capable of analyzing outliers and the influence of observations, and can be applicable to many other biomedical data which are often high-dimensional and multi-modal.

Flame Detection Using Generic Color Model and Improved Block-Based PCA in Active Infrared Camera

In this paper, we proposed an all-weather flame detection algorithm which could make full use of active infrared cameras presently installed in many public places for surveillance purposes. Firstly, according to the different spectral imaging results in day and night, we propose a video type classification algorithm (VTCA) via imaging clues. VTCA could help us select different flame visual features in color image and infrared image. Secondly, we use a generic YCbCr-color-space-based chrominance model to extract regions of interest (ROI) of flame. Thirdly, two flame dynamic features are used to verify the candidate ROIs, which are common flame flicker feature and an improved block-based PCA in consecutive frames. The experimental results show that the proposed flame detection model has been successfully applied to various situations, including day and night, indoor and outdoor on our test video datasets, and it gives a better performance compared with other state-of-the-art methods.

Runway Detection in SAR Images Based on Fusion Sparse Representation and Semantic Spatial Matching

In this paper, a novel algorithm is presented for runway detection in synthetic aperture radar images. It involves two steps: runway assessment and confirmation. In the first step, the primary runway (PR) and the auxiliary runway (AR) of the airport are assessed by using a sparse representation fusion frame. A set of residuals, for each feature from PR or AR, is first generated by performing sparse reconstructions over two training dictionaries constructed by a set of discriminative features. Based on all residuals for all types of features, two residual sequences for PR and AR are, respectively, built. To improve the assessment performance, these two residual sequences are normalized and further linearly fused. An assessment criterion is applied to the fusion result to infer an optimal target estimate. In the second step, the histogram of oriented gradient feature descriptors of PR, AR, and the entire runway region constructed by PR, AR, and taxiways are first generated. Afterward, two semantic spatial rules are developed to verify each candidate region of interest. If a perfect match is achieved, the candidate target can be confirmed. Since the PR and AR are selected based on the residual fusion related with the concatenation of multiple features, the presented algorithm has a good representation ability to the runway. By introducing semantic spatial relationships into the confirmation scheme, this algorithm can well discriminate other runway-like targets. The test results using real scene data demonstrate that the presented method has superiority to some state-of-the-art alternatives.

Cascaded Segmentation-Detection Networks for Text-Based Traffic Sign Detection

In this paper, we propose a novel text-based traffic sign detection framework with two deep learning components. More precisely, we apply a fully convolutional network to segment candidate traffic sign areas providing candidate regions of interest (RoI), followed by a fast neural network to detect texts on the extracted RoI. The proposed method makes full use of the characteristics of traffic signs to improve the efficiency and accuracy of text detection. On one hand, the proposed two-stage detection method reduces the search area of text detection and removes texts outside traffic signs. On the other hand, it solves the problem of multi-scales for the text detection part to a large extent. Extensive experimental results show that the proposed method achieves the state-of-the-art results on the publicly available traffic sign data set: Traffic Guide Panel data set. In addition, we collect a data set of text-based traffic signs including Chinese and English traffic signs. Our method also performs well on this data set, which demonstrates that the proposed method is general in detecting traffic signs of different languages.

Multi-Instance Multi-Label Learning for Multi-Class Classification of Whole Slide Breast Histopathology Images.

Digital pathology has entered a new era with the availability of whole slide scanners that create the high-resolution images of full biopsy slides. Consequently, the uncertainty regarding the correspondence between the image areas and the diagnostic labels assigned by pathologists at the slide level, and the need for identifying regions that belong to multiple classes with different clinical significances have emerged as two new challenges. However, generalizability of the state-of-the-art algorithms, whose accuracies were reported on carefully selected regions of interest (ROIs) for the binary benign versus cancer classification, to these multi-class learning and localization problems is currently unknown. This paper presents our potential solutions to these challenges by exploiting the viewing records of pathologists and their slide-level annotations in weakly supervised learning scenarios. First, we extract candidate ROIs from the logs of pathologists' image screenings based on different behaviors, such as zooming, panning, and fixation. Then, we model each slide with a bag of instances represented by the candidate ROIs and a set of class labels extracted from the pathology forms. Finally, we use four different multi-instance multi-label learning algorithms for both slide-level and ROI-level predictions of diagnostic categories in whole slide breast histopathology images. Slide-level evaluation using 5-class and 14-class settings showed average precision values up to 81% and 69%, respectively, under different weakly labeled learning scenarios. ROI-level predictions showed that the classifier could successfully perform multi-class localization and classification within whole slide images that were selected to include the full range of challenging diagnostic categories.