Large Image Databases Research Articles

A fresh look at spinal alignment and deformities: Automated analysis of a large database of 9832 biplanar radiographs.

We developed and used a deep learning tool to process biplanar radiographs of 9,832 non-surgical patients suffering from spinal deformities, with the aim of reporting the statistical distribution of radiological parameters describing the spinal shape and the correlations and interdependencies between them. An existing tool able to automatically perform a three-dimensional reconstruction of the thoracolumbar spine has been improved and used to analyze a large set of biplanar radiographs of the trunk. For all patients, the following parameters were calculated: spinopelvic parameters; lumbar lordosis; mismatch between pelvic incidence and lumbar lordosis; thoracic kyphosis; maximal coronal Cobb angle; sagittal vertical axis; T1-pelvic angle; maximal vertebral rotation in the transverse plane. The radiological parameters describing the sagittal alignment were found to be highly interrelated with each other, as well as dependent on age, while sex had relatively minor but statistically significant importance. Lumbar lordosis was associated with thoracic kyphosis, pelvic incidence and sagittal vertical axis. The pelvic incidence-lumbar lordosis mismatch was found to be dependent on the pelvic incidence and on age. Scoliosis had a distinct association with the sagittal alignment in adolescent and adult subjects. The deep learning-based tool allowed for the analysis of a large imaging database which would not be reasonably feasible if performed by human operators. The large set of results will be valuable to trigger new research questions in the field of spinal deformities, as well as to challenge the current knowledge.

Analysis of Machine Learning and Deep Learning Models to Classify ASD Using fMRI

Diagnosis of autism spectrum disorders (ASD) is a complex task, the solution of which usually depends on the experience of the physicians due to the lack of specific quantitative biomarkers. Machine learning and Deep Learning approaches are increasingly being used as a diagnostic tool for ASD classification, with the potential to improve discrimination validity among ASD and typically developing (TD) individuals. This paper describes the use of feature selection and two classification techniques to successfully distinguish between individuals with ASD and individuals without ASD, using data from a large resting-state functional magnetic resonance imaging (rs-fMRI) database.

Unsupervised Appearance Map Abstraction for Indoor Visual Place Recognition With Mobile Robots

Visual Place Recognition (VPR), the task of identifying the place where an image has been taken from, is at the core of important robotic problems as relocalization, loop-closure detection or topological navigation. Even for indoors, the focus of this work, VPR is challenging for a number of reasons, including real-time performance when dealing with large image databases ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim10^{4}$</tex-math></inline-formula> ) (probably captured by different robots), or the avoidance of Perceptual Aliasing in environments with repetitive structures and scenes. In this letter, we tackle these issues by proposing an off-line mapping technique that abstracts a dense database of georeferenced images without particular order into a Multivariate Gaussian Mixture Model, by creating soft clusters in terms of their similarity in both pose and appearance. This abstract representation is obtained through an Expectation-Maximization algorithm and plays the role of a simplified map. Since querying this map yields a probability of being in a cluster, we exploit this “belief” within a Bayesian filter that regards previous query images and a topological map between clusters to perform more robust VPR. We evaluate our proposal in two different indoor datasets, demonstrating comparable VPR precision to querying the full database while incurring in shorter query times and handling Perceptual Aliasing for sequential navigation.

Optimization of Content Based Image Retrieval Using Hybrid Approach

With the dawn of multimedia technology and the social web, retrieval from large image databases become possible. As a consequence of rising usage and incredible enthusiasm for inquiry about on Content-Based Image Retrieval (CBIR) systems, it needs improvement in the accuracy of CBIR systems in addition to the decline in monotonous results. Currently, the majority of research has focused on the representation and differentiation of images by an arrangement of low-level visual features. However, most of the retrieval systems produce repetitive or unnecessary retrieval results, termed a redundancy factor. In addition, content-based retrieval with reduced redundancy has a direct correlation with high-level semantics which is overlooked. Accurate content retrieval with reduced redundancy enables the user to focus on the actual problem in preference to nurture the retrieval results, augments the efficiency, and improves overall system performance. To enhance the retrieval accuracy and diminish the redundancy factor in image retrieval, we proposed an optimization-based technique that is blended with classification. In the proposed novel hybrid approach for CBIR, we used a two-tier architecture model. The first tier belongs to the feature extraction process via Particle Swarm Optimization with a Support Vector Machine as a classifier. On the way to reduce the redundancy factor, the K-Nearest Neighbor as a classifier is used with the Genetic Algorithm in the second tier. We noted a noteworthy increase in retrieval accuracy for images that is up to 25% (approx.). The proposed hybrid model is effective to enhance accuracy and reduce redundancy factors in CBIR systems. We used the WANG dataset for experimentation. Henceforward, improving the retrieval accuracy and reducing redundancy factor using unsupervised learning techniques is part of our future work.

Learning Token-Based Representation for Image Retrieval

In image retrieval, deep local features learned in a data-driven manner have been demonstrated effective to improve retrieval performance. To realize efficient retrieval on large image database, some approaches quantize deep local features with a large codebook and match images with aggregated match kernel. However, the complexity of these approaches is non-trivial with large memory footprint, which limits their capability to jointly perform feature learning and aggregation. To generate compact global representations while maintaining regional matching capability, we propose a unified framework to jointly learn local feature representation and aggregation. In our framework, we first extract local features using CNNs. Then, we design a tokenizer module to aggregate them into a few visual tokens, each corresponding to a specific visual pattern. This helps to remove background noise, and capture more discriminative regions in the image. Next, a refinement block is introduced to enhance the visual tokens with self-attention and cross-attention. Finally, different visual tokens are concatenated to generate a compact global representation. The whole framework is trained end-to-end with image-level labels. Extensive experiments are conducted to evaluate our approach, which outperforms the state-of-the-art methods on the Revisited Oxford and Paris datasets.

Geometrically transformed image retrieval with transfer learning

Image retrieval as per their features is an important topic since last decades. For image processing, classification and retrieval deep learning (DL) techniques utilized as state-of-the art techniques. For the images geometric transformation especially, rotation is common thing. However, Convolutional Neural Networks (CNNs) not supporting rotation invariance features due to use of anisotropic filters for convolution. So, it an important emerging field in image processing that, rotations in images should not affect on its characteristics or features. In this paper, we propose to use features pre-trained CNN model combinations, which trained for large image database containing rotated as well as scaled images for classification &amp; similar image retrieval. This approach produces superior result of classification and retrieval for Corel dataset for original query images and comparably good results by considering rotated &amp; scaled query images.

French Imaging Database Against Coronavirus (FIDAC): A large COVID-19 multi-center chest CT database

PurposeDuring the first wave of the COVID-19 pandemic, the French Society of Radiology and the French College of Radiology, in partnership with NEHS Digital, have set up a system to collect chest computed tomography (CT) examinations with clinical, virological and radiological metadata, from patients clinically suspected of COVID-19 pneumonia. This allowed the constitution of an anonymized multicenter database, named FIDAC (French Imaging Database Against Coronavirus). The aim of this report was to describe the content of this public database. Materials and methodsTwenty-two French radiology centers participated to the data collection. The data collected were chest CT examinations in DICOM format associated with the following metadata: patient age and sex, originating facility identifier, originating facility region, time from symptom onset to CT examination, indication for CT examination, reverse transcription-polymerase chain reaction (RT-PCR) results and normalized CT report performed by a senior radiologist. All the data were anonymized and sent through a NEHS Digital system to a centralized data center. ResultsA total of 5944 patients were included from the 22 centers aggregated into 8 regions with a mean number of patients of 743 ± 603.3 [SD] per region (range: 102–1577 patients). Reasons for CT examination and normalized CT reports were provided for all patients. RT-PCR results were provided in 5574 patients (93.77%) with a positive result of RT-PCR in 44.6% of patients. ConclusionThe FIDAC project allowed the creation of a large database of chest CT images and metadata available, under conditions, in open access through the CERF-SFR website.

Reverse Image Search Using Deep Unsupervised Generative Learning and Deep Convolutional Neural Network

Reverse image search has been a vital and emerging research area of information retrieval. One of the primary research foci of information retrieval is to increase the space and computational efficiency by converting a large image database into an efficiently computed feature database. This paper proposes a novel deep learning-based methodology, which captures channel-wise, low-level details of each image. In the first phase, sparse auto-encoder (SAE), a deep generative model, is applied to RGB channels of each image for unsupervised representational learning. In the second phase, transfer learning is utilized by using VGG-16, a variant of deep convolutional neural network (CNN). The output of SAE combined with the original RGB channel is forwarded to VGG-16, thereby producing a more effective feature database by the ensemble/collaboration of two effective models. The proposed method provides an information rich feature space that is a reduced dimensionality representation of the image database. Experiments are performed on a hybrid dataset that is developed by combining three standard publicly available datasets. The proposed approach has a retrieval accuracy (precision) of 98.46%, without using the metadata of images, by using a cosine similarity measure between the query image and the image database. Additionally, to further validate the proposed methodology’s effectiveness, image quality has been degraded by adding 5% noise (Speckle, Gaussian, and Salt pepper noise types) in the hybrid dataset. Retrieval accuracy has generally been found to be 97% for different variants of noise

Classification of mastoid air cells by CT scan images using deep learning method

PurposeMastoid abnormalities show different types of ear illnesses, however inadequacy of experts and low accuracy of diagnostic demand a new approach to detect these abnormalities and reduce human mistakes. The manual analysis of mastoid CT scans is time-consuming and labor-intensive. In this paper the first and robust deep learning-based approaches is introduced to diagnose mastoid abnormalities using a large database of CT images obtained in the clinical center with remarkable accuracy.MethodsIn this paper, mastoid abnormalities are classified using the Xception based Convolutional Neural Network (CNN) model, with optimizer Adamax into five categories (Complete pneumatized, Opacification in pneumatization, Partial pneumatization, Opacification in partial pneumatization, None pneumatized). For this reason, a total of 24,800 slides of 152 patients were selected that include the mastoid from most upper to the lowest part of the middle ear cavity to complete the construction of the proposed deep neural network model.ResultsThe proposed model had the best accuracy of 87.80% (based on grader 1) and 88.44% (based on grader 2) on the 20th epoch and 87.70% (based on grader 1) and 87.56% (based on grader 2) on average and also significantly faster than other types of implemented architectures in terms of the computer running time (in seconds). The 99% confidence interval of the average accuracy was 0.012 which means that the true accuracy is 87.80% and 87.56% ± 1.2% that indicates the power of the model.ConclusionsThe manual analysis of ear cavity CT scans is often time-consuming and prone to errors due to various inter- or intra operator variability studies. The proposed method can be used to automatically analyze the middle ear cavity to classify mastoid abnormalities, which is markedly faster than most types of models with the highest accuracy.

Computerized Analysis of Mammogram Images for Early Detection of Breast Cancer

Breast cancer is widespread worldwide and can be cured if diagnosed early. Using digital mammogram images and image processing with artificial intelligence can play an essential role in breast cancer diagnosis. As many computerized algorithms for breast cancer diagnosis have significant limitations, such as noise handling and varying or low contrast in the images, it can be difficult to segment the abnormal region. These challenges could be overcome by proposing a new pre-processing model, exploring its impact on the post-processing module, and testing it on an extensive database. In this research work, the three-step method is proposed and validated on large databases of mammography images. The first step corresponded to the database classification, followed by the second step, which removed the pectoral muscle from the mammogram image. The third stage utilized new image-enhancement techniques and a new segmentation module to detect abnormal regions in a well-enhanced image to diagnose breast cancer. The pre-and post-processing modules are based on novel image processing techniques. The proposed method was tested using data collected from different hospitals in the Qassim Health Cluster, Qassim Province, Saudi Arabia. This database contained the five categories in the Breast Imaging and Reporting and Data System and consisted of 2892 images; the proposed method is analyzed using the publicly available Mammographic Image Analysis Society database, which contained 322 images. The proposed method gives good contrast enhancement with peak-signal to noise ratio improvement of 3 dB. The proposed method provides an accuracy of approximately 92% on 2892 images of Qassim Health Cluster, Qassim Province, Saudi Arabia. The proposed method gives approximately 97% on the Mammographic Image Analysis Society database. The novelty of the proposed work is that it could work on all Breast Imaging and Reporting and Data System categories. The performance of the proposed method demonstrated its ability to improve the diagnostic performance of the computerized breast cancer detection method.

Protocol of a Multi-Center Initiative to Create a Standardized Ovarian Ultrasound Image Database in China

Background: At present there is no large, multi-center and standardized database of ovarian ultrasound images for teaching and research in China. Methods: A standardized ovarian ultrasound image database is being created in a collaborative initiative of about 20 centers proficient in gynecological ultrasound imaging. The database will include both adults and children in China. Results: Each center will provide cases that meet the submission requirements, including standard normal cases (SNC), standard abnormal cases (SAC) and historical classic cases (HCC). This database will cover the majority of the spectrum of ovarian lesions in China. Conclusions: This comprehensive database of ovarian lesions will be a valuable resource for diagnosis and education.

Emergence of the obesity epidemic: 6-decade visualization with humanoid avatars

Visualizations of the emerging obesity epidemic, such as with serial US color prevalence maps, provide graphic images that extend informative public health messages beyond those in written communications. Advances in low-cost 3D optical technology now allow for development of large image databases that include participants varying in race/ethnicity, body mass, height, age, and circumferences. When combined with contemporary statistical methods, these data sets can be used to create humanoid avatar images with prespecified anthropometric features. The current study aimed to develop a humanoid avatar series with characteristics of representative US adults extending over the past 6 decades. 3D optical scans were conducted on a demographically diverse sample of 570 healthy adults. Image data were converted to principal components and manifold regression equations were then developed with body mass, height, age, and waist circumference as covariates. Humanoid avatars were generated for representative adults with these 4 characteristics as reported in CDC surveys beginning in 1960-1962 up to 2015-2018. There was a curvilinear increase in adult US population body mass, waist circumference, and BMI in males and females across the 9 surveys spanning 6 decades. A small increase in average adult population age was present between 1960 and 2018; height changes were inconsistent. A series of 4 avatars developed at ∼20-y intervals for representative males and females reveal the changes in body size and shape consistent with the emergence of the obesity epidemic. An additional series of developed avatars portray the shapes and sizes of males and females at key BMI cutoffs. New mathematical approaches and accessible 3D optical technology combined with increasingly available large and diverse data sets across the life span now make unique visualization of body size and shape possible on a previously unattainable scale. This study is registered at https://clinicaltrials.gov/ct2/show/NCT03637855 as NCT03637855.

Deciphering the Colonization Strategies in Roots of Long-Term Fertilized Festuca rubra

Festuca rubra is one of the dominant species in mountain high natural values grasslands. Most herbaceous plants are involved in a symbiotic partnership with arbuscular mycorrhizas for maintaining their abundance and cover. This research was conducted to explore the changes in mycorrhizal colonization patterns and structure development due to the long-term application of treatments. A large database of microscopic images was deeply analyzed with the MycoPatt tool, for the extraction of clear mycorrhizal maps that present particular colonization strategies. The overall colonization frequency and intensity varied largely between treatments, with a maximum in organic conditions. The presence of both arbuscules and vesicles in the same root area indicate a continuous alternance of fungal strategy, from storage to enhanced transfer of nutrients. A low-mineral organic treatment permits a clear separation of colonization strategy in different parts of roots. The nutrient availability due to mineral treatment induced a restriction in mycorrhizal development, which was maintained colonization by a resistance conditions strategy. The use of mycorrhizal maps permits a deep scanning of colonized roots, identifying the real positioning of fungal structures, along with their developmental potential and the assessment of the colonization strategy.

No-Reference Quality Assessment of Pan-Sharpening Images with Multi-Level Deep Image Representations

The Pan-Sharpening (PS) techniques provide a better visualization of a multi-band image using the high-resolution single-band image. To support their development and evaluation, in this paper, a novel, accurate, and automatic No-Reference (NR) PS Image Quality Assessment (IQA) method is proposed. In the method, responses of two complementary network architectures in a form of extracted multi-level representations of PS images are employed as quality-aware information. Specifically, high-dimensional data are separately extracted from the layers of the networks and further processed with the Kernel Principal Component Analysis (KPCA) to obtain features used to create a PS quality model. Extensive experimental comparison of the method on the large database of PS images against the state-of-the-art techniques, including popular NR methods adapted in this study to the PS IQA, indicates its superiority in terms of typical criteria.

FSeCBIR: A Faster Secure Content-Based Image Retrieval for Cloud

<h2>Abstract</h2> In recent years, content-based image retrieval in encrypted domain has emerged as an important challenge in cloud environment. We present FSeCBIR, a prototype that implements Faster and Secure Content-Based Image Retrieval ensuring secure search and retrieval in large encrypted image databases. This prototype is implemented using MATLAB on a server/high-end PC and it incorporates a copy deterrence module to identify data leaks through untrustworthy query users. This simulates a system that can be used in real-time cloud applications involving sensitive information which demand high level of security, retrieval precision, and less time/space complexity.

Development of training image database using web crawling for vision-based site monitoring

As most of the state-of-the-art technologies for vision-based monitoring were originated from machine learning or deep learning algorithms, it is crucial to build a large and rich training image database (DB). For this purpose, this paper proposes an automated framework that builds a large, high-quality training DB for construction site monitoring. The framework consists of three main processes: (1) automated construction image collection using web crawling, (2) automated image labeling using an image segmentation model, and (3) fully randomized foreground-background cross-oversampling. Using the developed framework, it was possible to automatically construct a training DB, composed of 5864 images, for the detection of construction objects in 53.5 min. The deep learning model trained by the DB successfully detected construction resources with an average precision of 92.71% and a recall rate of 88.14%. The findings of this study can reduce the time and effort required to develop vision-based site monitoring technologies.

Enhancing Scalability of Image Retrieval Using Visual Fusion of Feature Descriptors

Content-Based Image Retrieval (CBIR) is an approach of retrieving similar images from a large image database. Recently CBIR poses new challenges in semantic categorization of the images. Different feature extraction technique have been proposed to overcome the semantic breach problems, however these methods suffer from several shortcomings. This paper contributes an image retrieval system to extract the local features based on the fusion of scale-invariant feature transform (SIFT) and KAZE. The strength of local feature descriptor SIFT complements global feature descriptor KAZE. SIFT concentrates on the complete region of an image using high fine points of features and KAZE ponders on details of a boundary. The fusion of local feature descriptor and global feature descriptor boost the retrieval of images having diverse semantic classification and also helps in achieving the better results in large scale retrieval. To enhance the scalability of image retrieval bag of visual words (BoVW) is mainly used. The fusion of local and global feature representations are selected for image retrieval for the reason that SIFT effectively captures shape and texture and robust towards the change in scale and rotation, while KAZE have strong response towards boundary and changes in illumination. Experiments conducted on two image collections, namely, Caltech-256 and Corel 10k demonstrate the proposed scheme appreciably enhanced the performance of the CBIR compared to state-of-the-art image retrieval techniques.

Region-based image retrieval using region of interest (ROI) according to incremental frame and clustering color image

Content-based image retrieval involves the extraction of global feature images for their retrieval performance in large image databases. Extraction of global features image cause problem of the semantic gap between the high-level meaning and low-level visual features images. In this study RBIR, Region of Interest Based (ROI) Image Retrieval Using Incremental Frame of Color Image was proposed. It combines several methods, including filtering process, image partitioning using clustering and incremental frame formation, complementation law of theory set to generate ROI, NROI, or ER of the region. The concept of weighting as well as a significant query is also incorporated as a query strategy. Extensive experiments were also conducted on the Wang database and the color model selected was the CIE lab. Experimental results show the proposed method is efficient in image retrieval. The performance of the proposed method shows a better average IPR value of 3.51% compared to RGB and 22.92% with the HSV color model. Meanwhile, it also performs better by 36%, 5%, and 24% compared to methods CH (8,2,2), CH (8,3,3), and CH (16,4,4).

Hybrid Evolutionary Algorithm Based Relevance Feedback Approach for Image Retrieval

Searching images from the large image databases is one of the potential research areas of multimedia research. The most challenging task for nay CBIR system is to capture the high level semantic of user. The researchers of multimedia domain are trying to fix this issue with the help of Relevance Feedback (RF). However existing RF based approaches needs a number of iteration to fulfill user's requirements. This paper proposed a novel methodology to achieve better results in early iteration to reduce the user interaction with the system. In previous research work it is reported that SVM based RF approach generating better results for CBIR. Therefore, this paper focused on SVM based RF approach. To enhance the performance of SVM based RF approach this research work applied Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) before applying SVM on user feedback. The main objective of using these meta-heuristic was to increase the positive image sample size from SVM. Firstly steps PSO is applied by incorporating the user feedback and secondly GA is applied on the result generated through PSO, finally SVM is applied using the positive sample generated through GA. The proposed technique is named as Particle Swarm Optimization Genetic Algorithm- Support Vector Machine Relevance Feedback (PSO-G A-SVM-RF). Precisions, recall and F-score are used as performance metrics for the assessment and validation of PSO-GA-SVM-RF approach and experiments are conducted on coral image dataset having 10908 images. From experimental results it is proved that PSO-GA-SVM-RF approach outperformed then various well known CBIR approaches.

Systematic biases when using deep neural networks for annotating large catalogs of astronomical images

Deep convolutional neural networks (DCNNs) have become the most common solution for automatic image annotation due to their non-parametric nature, good performance, and their accessibility through libraries such as TensorFlow. Among other fields, DCNNs are also a common approach to the annotation of large astronomical image databases acquired by digital sky surveys. One of the main downsides of DCNNs is the complex non-intuitive rules that make DCNNs act as a “black box”, providing annotations in a manner that is unclear to the user. Therefore, the user is often not able to know what information is used by the DCNNs for the classification. Here we demonstrate that the training of a DCNN is sensitive to the context of the training data such as the location of the objects in the sky. We show that for basic classification of elliptical and spiral galaxies, the sky location of the galaxies used for training affects the behavior of the algorithm, and leads to a small but consistent and statistically significant bias. That bias exhibits itself in the form of cosmological-scale anisotropy in the distribution of basic galaxy morphology. Therefore, while DCNNs are powerful tools for annotating images of extended sources, the construction of training sets for galaxy morphology should take into consideration more aspects than the visual appearance of the object. In any case, catalogs created with deep neural networks that exhibit signs of cosmological anisotropy should be interpreted with the possibility of consistent bias.