Image Management System Research Articles

Robust CoHOG Feature Extraction in Human-Centered Image/Video Management System

Many human-centered image and video management systems depend on robust human detection. To extract robust features for human detection, this paper investigates the following shortcomings of co-occurrence histograms of oriented gradients (CoHOGs) which significantly limit its advantages: 1) The magnitudes of the gradients are discarded, and only the orientations are used; 2) the gradients are not smoothed, and thus, aliasing effect exists; and 3) the dimensionality of the CoHOG feature vector is very large (e.g., 200,000). To deal with these problems, in this paper, we propose a framework that performs the following: 1) utilizes a novel gradient decomposition and combination strategy to make full use of the information of gradients; (2) adopts a two-stage gradient smoothing scheme to perform efficient gradient interpolation; and (3) employs incremental principal component analysis to reduce the large dimensionality of the CoHOG features. Experimental results on the two different human databases demonstrate the effectiveness of the proposed method.

A modular fuzzy inference system approach in integrating qualitative and quantitative analysis of store image

The image situation in a store includes various stimuli, such as color, sound, scent, taste, layout and space, which are important clues for buyers. This paper describes store image response and a fuzzy logic model developed by comprehensive literature studies on image measurements (including atmospheric factors) and perceptual measures. Here, a fuzzy inference system is proposed as an alternative approach to handle effectively the impreciseness and uncertainty that are normally found in store image selection processes. This paper also shows that the proposed decision-making model is application to modified stimulus–organism–response (S–O–R) framework for integrating qualitative and quantitative analysis. The result of the simulation indicates a numerical and linguistic change in the store image perception after analyzing three input parameters. This finding is able to provide a solid foundation on which retailers and decision makers can base suitable strategies for ensuring the efficiency and stability of store image management system.

Towards Building High Performance Medical Image Management System for Clinical Trials.

Medical image based biomarkers are being established for therapeutic cancer clinical trials, where image assessment is among the essential tasks. Large scale image assessment is often performed by a large group of experts by retrieving images from a centralized image repository to workstations to markup and annotate images. In such environment, it is critical to provide a high performance image management system that supports efficient concurrent image retrievals in a distributed environment. There are several major challenges: high throughput of large scale image data over the Internet from the server for multiple concurrent client users, efficient communication protocols for transporting data, and effective management of versioning of data for audit trails. We study the major bottlenecks for such a system, propose and evaluate a solution by using a hybrid image storage with solid state drives and hard disk drives, RESTful Web Services based protocols for exchanging image data, and a database based versioning scheme for efficient archive of image revision history. Our experiments show promising results of our methods, and our work provides a guideline for building enterprise level high performance medical image management systems.

Persistent Chest Pain in Absence of Angiographic Significant Coronary Artery Disease is Associated with Permanent Myocardial Perfusion Defects in Magnetic Resonance Imaging in Post-menopausal Women

We studied a population of post-menopausal women with persistent chest pain (PChP) in order to investigate the relationship between myocardial perfusion at rest and during a stress test using magnetic resonance imaging (MRI). Our goals were to document whether transient myocardial perfusion is induced by dipyridamole infusion and if perfusion defects are also present at rest. The study population consisted of 45 consecutive women (mean age 57.6±8.7 years), who reported chest pain symptoms. PChP was defined as self-reported continuing chest pain after one year. We compared the results of the perfusion MRI studies in subgroups with and without obstructive coronary artery disease (CAD). The latest tools and technologies of Synapse™ Cardiovascular – Fujifilm’s cardiovascular (CV) image and information management system – helped us to achieve clear and comprehensive outcomes. In the group of women with PChP and non-obstructive CAD, 16 of 34 (48%) showed a well-evident left ventricular perfusion defect at baseline (four in one segment; eight in two segments and four in three or more segments). The localisation of the perfusion defects – seen using Synapse Cardiovascular – were anteroapical (n=6); septal (n=10); and inferoor inferolateral (n=4). These defects were ‘permanent’ or ‘fixed’, i.e. were present at rest and were neither induced nor modified by the administration of dipyridamole. In any of the women with CAD we found these anomalies. ‘Fixed’ perfusion defects at MRI – probably due to permanent damage of the coronary microcirculation – suggest a disease state typical for post-menopausal women with PChP.

Implementing a Departmental Outpatient Image Management System

Implementing a Departmental Outpatient Image Management System

Guidelines for fracture healing assessments in clinical trials. Part II: Electronic data capture and image management systems—Global Adjudicator™ system

Adjudication of outcomes helps to reduce variation in outcome assessment when clinical judgement is used. Adjudication is extremely important when the outcome being measured is subjective, as in the assessment of fracture healing. Adjudication Committees have the potential to provide a systematic, independent, and unbiased method of determining fracture healing. In order to make the adjudication process less time consuming and more efficient, Global Research Solutions™, a Contract Research Organization which specialises in orthopaedic and surgical clinical trials, developed the Global Adjudicator™, a web-based system, to serve as both an image management system and electronic data capture system. This manuscript highlights the features of the Global Adjudicator™ and demonstrates the benefits of using a web-based adjudication system.

Exchanging information: current United Kingdom developments in the infrastructure to support post-mortem imaging

Through joint working across the Home Office, Ministry of Justice and Department of Health the United Kingdom (UK) has developed a single programme of research and development in the field of less invasive autopsy, forensic and mass fatality imaging. By supporting research and development within these fields the UK has developed an outline for a national core training programme in this field along with the development of the National Forensic Image Management System (FiMag). This symposium paper is a brief overview of this development.

Management System for Tagged Image Databases Using Lattice Structures

Tagged Image Management System based on Lattice Structures (TaILaS), the next-generation management we propose for a large number of images, constructs a concept lattice from image tags using formal concept analysis, then visualizes the Hasse diagram of this lattice for any numbers of images. TaILaS enables users to avoid having to classify images which are difficult to divide into folders. TaILaS runs on a conventional computer (CPU = 2.13 GHz,MM= 2 GB) and is implemented in Processing as a Java-based programming language. Experiments involving 11 subjects using 100 images selected from the Corel Image Gallery confirmed that TaILaS manages huge image databases effectively.

Integration of Arthroscopic and C-Arm Imaging Into the Arthroscopic Image Management System

Integration of Arthroscopic and C-Arm Imaging Into the Arthroscopic Image Management System

Bisque: a platform for bioimage analysis and management

Advances in the field of microscopy have brought about the need for better image management and analysis solutions. Novel imaging techniques have created vast stores of images and metadata that are difficult to organize, search, process and analyze. These tasks are further complicated by conflicting and proprietary image and metadata formats, that impede analyzing and sharing of images and any associated data. These obstacles have resulted in research resources being locked away in digital media and file cabinets. Current image management systems do not address the pressing needs of researchers who must quantify image data on a regular basis. We present Bisque, a web-based platform specifically designed to provide researchers with organizational and quantitative analysis tools for 5D image data. Users can extend Bisque with both data model and analysis extensions in order to adapt the system to local needs. Bisque's extensibility stems from two core concepts: flexible metadata facility and an open web-based architecture. Together these empower researchers to create, develop and share novel bioimage analyses. Several case studies using Bisque with specific applications are presented as an indication of how users can expect to extend Bisque for their own purposes.

Special corner: Visual Categorization and Image Management Systems

This special corner consists of three selected papers that were presented at the Visual Categorisation and Image Management Systems (VCIMS) workshop organised by V. Bonnardel, M. Oakes and J. Tait and sponsored by the Multimedia Knowledge Management Network chaired by S. Rueger that took place at the University of Sunderland on the 18 June 2006. The VCIMS workshop offered the opportunity to bring together Cognitive Psychologists, Visual Neuroscientists, Information Retrieval scientists to promote integration from interdisciplinary approaches on visual categorisation in natural and artificial systems. Categorization is the ability to determine that an entity belongs to a particular group of objects and by this mean to recognize, differentiate and understand objects. Objects are classified in function of attributes defining their category memberships. There are different theoretical views that specify the nature of category-relevant attributes and the strategy in distinguishing between exemplars and nonexemplars of a given category. In the classical view, categories are mutually exclusive and collectively exhaustive discrete entities. They are characterized by a list of necessary and sufficient attributes, so that any object belongs univocally to one and only one category. More recent theories, such as prototype theory, acknowledge that natural categories are graded and that some members of the category, based on their properties, are more central than others. For these theories it is essential to evaluate the relative relevance of the different properties. Hammer, Hertz, Hochstein and Weinshall propose a novel approach to determine the ‘dimension weighting’ (property relevance) in category learning. In their approach, learning strategies are based on a restriction indicating whether two exemplars belong (positive equivalence constraint, PEC) or not (negative equivalence constraint, NEC) to the same category. PEC differs from NEC on several aspects. PEC specifies within-category variations whereas NEC specifies between-category variations, and if NEC are more common than PEC (i.e., number of pair comparisons from between category objects is larger than that from within category objects) they are less informative. Indeed, if two objects are from the same category, it can be deduced that at least some of the shared dimensions are relevant to that category and all non-shared dimensions are irrelevant, but if two objects from different categories differ by more than one dimension, it is impossible to know which of these dimensions is relevant for discriminating between categories. Probably because of this difference in the amount of information, performances in a categorization task are better in PEC condition. When the amount of information is made equal in the PEC and NEC conditions, participants are divided in two groups: those who used the NEC strategy successfully and those who do not. These individual differences highlight fundamental distinctions between the two strategies. PEC is used more intuitively but not perfectly while NEC is potentially more accurate but not implemented by all participants. Interestingly, a similar asymmetry is observed in machine-learning where NEC is V. Bonnardel (&) Department of Psychology, University of Winchester, Winchester, UK e-mail: valerie.bonnardel@winchester.ac.uk

TU‐A‐210A‐01: Quality Assurance for Digital Radiography

Diagnostic medical physicists have traditionally played a key role in the radiology department's Quality Assurance (QA) program. Procedures for QA of digital radiography (DR) systems are less well‐established than those designed for conventional screen‐film radiography. Some methods that were appropriate for conventional systems do not yield accurate results for digital systems. Methods that rely on images on film cannot be performed in a facility devoid of chemical processing. Methods that do not consider effects of all stages of the digital imaging chain from acquisition to display are likely to provide erroneous results. Digital systems often do not provide data in a convenient form for assessing performance. Nevertheless, medical physicists are responsible for designing and directing efforts to assure the safety and efficacy of DR.Development of standards for performance verification of digital radiographic systems has been outpaced by commercialization of the technology. Practitioners, operators, vendors, and medical physicists are uncertain as to how to configure, calibrate, and verify systems for the best diagnostic quality images at the lowest practical radiation exposure to the patient. Several manufacturers have fielded commercial systems for performance verification, but long term data indicating trends and action limits is rare. Sources of definitive information on practical clinical methods are also scarce in the scientific literature. This has led to the current situation where clinical imaging operations are routinely conducted using systems that are not optimized.The good news is that methods in common practice for conventional radiography can be applied with modification to digital radiography. Some of these methods are described in recent literature. The AAPM has recognized the dilemma and has initiated an effort to provide practical guidance to clinical physicists, i.e. Task Group 150.This lecture will discuss some components of a QA program for DR, including adaptation of conventional QA processes. The lecture will discuss methods of evaluating system performance and provide examples of potential interferences. The lecture will review commercial devices for performance evaluation and their limitations. Finally, some ongoing standardization efforts will be described.Learning Objectives:1. Review components of a QA program and show how they apply to DR.2. Understand how some conventional tests should be modified for a digital radiographic system integrated into an electronic image management system.3. Identify key references and standards that can be useful in QA of DR.

TU-C-213A-01: Adapting Traditional Clinical Medical Physics to Digital Radiography

Development of standards for performance verification of digital radiographic systems has been outpaced by commercialization of the technology. Practitioners, operators, vendors, and medical physicists are uncertain as to how to configure, calibrate, and verify systems for the best diagnostic quality images at the lowest practical radiation exposure to the patient. Several manufacturers have fielded commercial systems for performance verification, but long term data indicating trends and action limits is rare. Sources of definitive information on practical clinical methods are also scarce in the scientific literature. This has led to the current situation where clinical imaging operations are routinely conducted using systems that are not optimized. This workshop will discuss four important aspects of traditional medical physics practice that are modified when dealing with DR systems. These processes are of primary importance to medical physicists, because they affect both the quality of DR images and the ionizing radiation dose to patients. First, we consider how to perform an inspection on a radiographic unit using CR or DR. Second, we discuss how to set up the AEC for CR or DR exams. Third, we address how to verify the exposure indicator for CR and DR. Finally, we discuss available phantoms and their characteristics and limitations when used with DR. Learning Objectives: 1. Understand how some conventional tests should be modified for a digital radiographic system that is integrated into an electronic image management system. 2. Identify key references and standards that can be useful in testing of DR. 3. Appreciate opportunities for improvement of testing of DR systems.

Digital image management project for dermatological health care environments: a new dedicated software and review of the literature

Because the skin is the only organ completely accessible to visual examination, digital technology has therefore attracted the attention of dermatologists for documenting, monitoring, measuring and classifying morphological manifestations. To describe a digital image management system dedicated to dermatological health care environments and to compare it with other existing softwares for digital image storage. We designed a reliable hardware structure that could ensure future scaling, because storage needs tend to grow exponentially. For the software, we chose a client-web server application based on a relational database and with a 'minimalist' user interface. We developed a software with a ready-made, adaptable index of skin pathologies. It facilitates classification by pathology, patient and visit, with an advanced search option allowing access to all images according to personalized criteria. The software also offers the possibility of comparing two or more digital images (follow-up). The fact that the archives of years of digital photos acquired and saved on PCs can easily be entered in the program distinguishes it from the others in the market. This option is fundamental for accessing all the photos taken in years of practice in the program without entering them one by one. The program is available to any user connected to the local Intranet and the system may directly be available in the future from the Internet. All clinics and surgeries, especially those that rely on digital images, are obliged to keep up with technological advances. It is therefore hoped that our project will become a model for medical structures intending to rationalise digital and other data according to statutory requirements.

The Lifecycle of Embedded Image Metadata within Digital Photographs: Challenges and Best Practices. (or The Secret Life of Photo Metadata)

The value of metadata embedded in digital images is recognized by communities at different stages in the lifecycle of an image. Photographers working for news media and publishers embed text information (metadata) into photographs to communicate what is happening in the image and to find those images later. Software developers build applications which expose the metadata in images to assist businesses with image management, search, discovery, and distribution. As more and more images arrive at museums, libraries and other archival institutions in digital form, it may be prudent to reevaluate how digital images and their supporting information are acquired in order to manage the potential glut of new arrivals. This may require, at minimum, exploring the idea of importing metadata that is already embedded when importing digital files into image and database management systems. How this data is extracted and moved into these other systems is important, and should be done in accordance with documented procedures using current industry standards and best practices.

1A2-D03 分散染色法によるアスベスト定性分析作業支援のための偏光顕微鏡画像管理システムの開発

1A2-D03 分散染色法によるアスベスト定性分析作業支援のための偏光顕微鏡画像管理システムの開発

Budgeting for PACS

There are a number of models for the acquisition of digital image management systems. The specific details for development of a budget for a PACS/RIS acquisition will depend upon the acquisition model – although there are similarities in the overarching principles and general information, particularly concerning the radiology service requirements that will drive budget considerations.While budgeting for PACS/RIS should follow the same principles as budgeting for any new technology, it is important to understand how far the implementation of digital image management systems can reach in a healthcare setting. Accurate identification of those elements of the healthcare service that will be affected by a PACS/RIS implementation is a critical component of successful budget formation and of the success of any business case and subsequent project that relies on those budget estimates.A budget for a PACS/RIS capital acquisition project should contain capital and recurrent elements. The capital is associated with the acquisition of the system in a purchase model and capital budget may also be required for upgrade – depending upon a facility’s financial management processes.The recurrent (or operational) cost component for the PACS/RIS is associated with maintaining the system(s) in a sustainable operational state.It is also important to consider the service efficiencies, cost savings and service quality improvements that PACS/RIS can generate and include these factors into the economic analysis of any proposal for a PACS/RIS project.

OSPACS: Ultrasound image management system

BackgroundUltrasound scanning uses the medical imaging format, DICOM, for electronically storing the images and data associated with a particular scan. Large health care facilities typically use a picture archiving and communication system (PACS) for storing and retrieving such images. However, these systems are usually not suitable for managing large collections of anonymized ultrasound images gathered during a clinical screening trial.ResultsWe have developed a system enabling the accurate archiving and management of ultrasound images gathered during a clinical screening trial. It is based upon a Windows application utilizing an open-source DICOM image viewer and a relational database. The system automates the bulk import of DICOM files from removable media by cross-validating the patient information against an external database, anonymizing the data as well as the image, and then storing the contents of the file as a field in a database record. These image records may then be retrieved from the database and presented in a tree-view control so that the user can select particular images for display in a DICOM viewer or export them to external media.ConclusionThis system provides error-free automation of ultrasound image archiving and management, suitable for use in a clinical trial. An open-source project has been established to promote continued development of the system.

Quality Control Management and Communication Between Radiologists and Technologists

The greatest barrier to quality control (QC) in the digital imaging environment is the lack of communication and documentation between those who interpret images and those who acquire them. Paper-based QC methods are insufficient in a digital image management system. Problem work flow must be incorporated into reengineering efforts when migrating to a digital practice. The authors implemented a Web-based QC feedback tool to document and facilitate the communication of issues identified by radiologists. The goal was to promote a responsive and constructive tool that contributes to a culture of quality. The hypothesis was that by making it easier for radiologists to submit quality issues, the number of QC issues submitted would increase. The authors integrated their Web-based quality tracking system with a clinical picture archiving and communication system so that radiologists could report quality issues without disrupting clinical work flow. Graphical dashboarding techniques aid supervisors in using this database to identify the root causes of different types of issues. Over the initial 12-month rollout period, starting in the general section, the authors recorded 20 times more QC issues submitted by radiologists, accompanied by a rise in technologists' responsiveness to QC issues. For technologists with high numbers of QC issues, the incorporation of data from this tracking system proved useful in performance appraisals and in driving individual improvement. This tool is an example of the types of information technology innovations that can be leveraged to support QC in the digital imaging environment. Initial data suggest that the result is not only an improvement in quality but higher levels of satisfaction for both radiologists and technologists.

WE-A-352-01: Quality Assurance Procedures for Digital Radiography

Diagnostic medical physicists have traditionally played a key role in the radiology department's Quality Assurance (QA) program. Procedures for QA of digital radiography (DR) systems are less well-established than those designed for conventional screen-film radiography. Some methods that were appropriate for conventional systems do not yield accurate results for digital systems. Methods that rely on images on film cannot be performed in a facility devoid of chemical processing. Methods that do not consider effects of all stages of the digital imaging chain from acquisition to display are likely to provide erroneous results. Digital systems often do not provide data in a convenient form for assessing performance. Nevertheless, medical physicists are responsible for designing and directing efforts to assure the safety and efficacy of DR. Development of standards for performance verification of digital radiographic systems has been outpaced by commercialization of the technology. Practitioners, operators, vendors, and medical physicists are uncertain as to how to configure, calibrate, and verify systems for the best diagnostic quality images at the lowest practical radiation exposure to the patient. Several manufacturers have fielded commercial systems for performance verification, but long term data indicating trends and action limits is rare. Sources of definitive information on practical clinical methods are also scarce in the scientific literature. This has led to the current situation where clinical imaging operations are routinely conducted using systems that are not optimized. The good news is that methods in common practice for conventional radiography can be applied with modification to digital radiography. Some of these methods are described in recent literature. The AAPM has recognized the dilemma and has initiated an effort to provide practical guidance to clinical physicists, i.e. Task Group 150. This lecture will discuss some components of a QA program for DR, including adaptation of conventional QA processes. The lecture will discuss methods of evaluating system performance and provide examples of potential interferences. The lecture will review commercial devices for performance evaluation and their limitations. Finally, some ongoing standardization efforts will be described. Educational Objectives: 1. Review components of a QA program and show how they apply to DR. 2. Understand how some conventional tests should be modified for a digital radiographic system integrated into an electronic image management system. 3. Identify key references and standards that can be useful in QA of DR.