The main intention of content based medical image retrieval (CBMIR) is to efficiently retrieve medical images that are visually similar to a query image. Medical images are usually retrieved on the basis of low level and high level features. This work deals with the concept of texture based spine MRI image retrieval in the wavelet compressed domain. We use two statistical methods such as Haralick features and texture??spectrum features for spine MRI image feature extraction and project the features to a set of signatures. The obtained statistical features are classifying, according to various types of spine MRI images using k-means clustering algorithm. Then the image retrieval is carried out by calculating the distance between the signatures in the database images and the query image. This method is applied around 500 spine MRI images and improvements of retrieval efficiency are found with standard precision and recall analysis.

Dieser Artikel gibt eine Einfuhrung in die inhaltsbasierte Suche in medizinischen Bilddatenbanken. Es wird klar, dass neue bildgebende Verfahren in der Medizin nicht nur fur eine erhohte Variabilitat der Bildoder Multimedia-Daten sorgen, sondern auch die Anzahl der erzeugten Dokumente und deren Grose standig steigt (Radiologie Genf: ~40'000 Bilder pro Tag im Jahr 2006, ). Wie in der Radiologie, wo mittlerweile fast alle Bilder in digitaler Form verfugbar sein konnen, werden auch in anderen Bereichen (Pathologie, Dermatologie) verstarkt Bilder digital erzeugt, die uber kurz oder lang ebenfalls im PACS (Picture Archival and Communication System) gespeichert werden mussen. Bisher wurden medizinische multimediale Daten nur patientenbezogen gespeichert und genutzt. Da in den Daten vergangener Falle auch viel Information auch fur die Zukunft gespeichert ist, drangt die Forschung verstarkt dazu, Daten gezielt fur die Diagnoseunterstutzung zu extrahieren und zu benutzen. Das birgt auf der einen Seite Chancen fur eine verbesserte Diagnoseunterstutzung, auf der anderen Seite Risiken fur die Privatsphare der Patienten. Bei der Erstellung von Datenbanken muss also vorsichtig vorgegangen werden, und vor allem mussen an Kliniken klare Richtlinien fur die Weiterbenutzung von klinischen Daten existieren um die geplante Benutzung zu erleichtern. Inhaltsbasierte Bildsuche ist eine Technik, die es erlaubt, in Datenbanken nach visuell ahnlichen Bildern zu suchen, bzw. im medizinischen Bereich Falle mit ahnlichen Bildern zu finden . Dies entspricht einem Mediziner, der alte Falle ahnlich zu seinem aktuell behandelten Fall benotigt, um sie mit der jetzigen Situation zu vergleichen. Alte Falle mussen dabei anonymisiert sein, allerdings so viel Information wie moglich enthalten, z.B. Bilder, Alter, Medikamente, Diagnose, Behandlung aber auch, falls moglich, das Resultat der Behandlung. Dieser Artikel fuhrt multimediale Datenquellen im Krankenhaus ein und beschreibt die inhaltsbasierte Suche mit all Ihren Vorteilen und Problemen. Am Ende werden Forschungsrichtungen beschrieben, die den aktuellen Techniken weiterhelfen durften.

In dental implantology more than one hundred enossal implant systems are in use. Once embedded, the dental X{ray examination is the most important tool for determing implants' producer, name, and type. In this paper, we present a system for automatical detection and identiication of dental xtures in intraoral X rays (IDEFIX) combining common direct digital image acquisition techniques with specially designed image analysis. IDEFIX can process any digital radiograph (e.g. RVG, Sens-A-Ray, Schick, Sidexis, Digora) as well as digitized dental lms. A reference database has been generated by precise measurement on the implant systems used so far (eight implants) including parameters like length, diameter, and cross section area. After binarization of the current digital X{ray image, a parameter set is extracted from each detected object applying mathematical morphology. All objects are classiied using a simpliied nearest neighbor method and the Euclidean distance metric. If the distance of the objects' parameter set to one of the reference sets is below a given threshhold, name and type of the identiied dental xture are displayed on the screen. Otherwise, the actual object will be rejected as a non{implant. IDEFIX has been evaluated by processing various in{vitro acquired radiographs. Diierent implants were classiied captured with identical conditions as well as acquired varying the angulation of the X{ray tube. It is shown that misangulations up to twenty degree are tolerable preserving correct identiication. Other image structures like teeth or llings result in large distances to all reference parameter sets and therefore, they are reliably recognized as non{implants.

In this paper, the automatic annotation task of the 2005 CLEF cross-language image retrieval campaign (ImageCLEF) is described. This paper focuses on the database used, the task setup, and the plans for further medical image annotation tasks in the context of ImageCLEF. Furthermore, a short summary of the results of 2005 is given. The automatic annotation task was added to ImageCLEF in 2005 and provides the first international evaluation of state-of-the-art methods for completely automatic annotation of medical images based on visual properties.The aim of this task is to explore and promote the use of automatic annotation techniques to allow for extracting semantic information from little-annotated medical images. A database of 10.000 images was established and annotated by experienced physicians resulting in 57 classes, each with at least 10 images. Detailed analysis is done regarding the (i) image representation, (ii) classification method, and (iii) learning method. Based on the strong participation of the 2005 campain, future benchmarks are planned.

The diagnostics of oncological diseases is based on histological specimens in hematoxilin-eosin staining. Since manual evaluation of microscopy images is time consuming and depends on the human expert, several approaches to automatic image analysis and classification have been published. In such systems, feature extraction usually relies on a fixed resolution and a small number of numerical features. Contrarily, this framework is based on a morphometric study using two levels of optical magnification (50 and 200 times, correspondingly). In this paper, we propose a principle scheme for automation of the oncological diagnostics, and an algorithm of morphometric feature extraction of tissue fragment at low magnification. In particular, patterns of cells, vessels, and fragments of tissue are considered individually and combined for correct identification of objects extracted from the specimen. The fact of invasion is established automatically after this procedure as well as polymorphism, polychromism and anaplasia. Using this method, diagnostics of 86 out of 100 patients was confirmed.

A combination of several classifiers using global features for the content description of medical images is proposed. Beside two texture features, downscaled representations of the original images are used, which preserve spatial information and utilize distance measures which are robust regarding common variations in radiation dose, translation, and local deformation. No query refinement mechanisms are used. The single classifiers are used within a parallel combination scheme, with the optimization set being used to obtain the best weighing parameters. For the medical automatic annotation task, a categorization rate of 78.6% is obtained, which ranks 12th among 28 submissions. When applied in the medical retrieval task, this combination of classifiers yields a mean average precision (MAP) of 0.0172, which is rank 11 of 11 submitted runs for automatic, visual only systems.

Multiscale analysis provides a complete hierarchical partitioning of images into visually plausible regions. Each of them is formally characterized by a feature vector describing shape, texture and scale properties. Consequently, object extraction becomes a classification of the feature vectors. Classifiers are trained by relevant and irrelevant regions labeled as object and remaining partitions, respectively. A trained classifier is applicable to yet uncategorized partitionings to identify the corresponding region's classes. Such an approach enables retrieval of a-priori unknown objects within a point-and-click interface. In this work, the classification pipeline consists of a framework for data selection, feature selection, classifier training, classification of testing data, and evaluation. According to the no-free-lunch-theorem of supervised learning, the appropriate classification pipeline is determined experimentally. Therefore, each of the steps is varied by state-of-the-art methods and the respective classification quality is measured. Selection of training data from the ground truth is supported by bootstrapping, variance pooling. virtual training data, and cross validation. Feature selection for dimension reduction is performed by linear discriminant analysis, principal component analysis, and greedy selection. Competing classifiers are k-nearest-neighbor: Bayesian classifier, and the support vector machine. Quality is measured by precision and recall to reflect the retrieval task. A set of 105 hand radiographs from clinical routine serves as ground truth, where the metacarpal bones have been labeled manually. In total, 368 out of 39.017 regions are identified as relevant. In initial experiments for feature selection with the support vector machine have been obtained recall, precision and F-measure of 0.58, 0.67, and 0,62, respectively.

Objectives: The number of articles published annually in the fields of biomedical signal and image acquisition and processing is increasing. Based on selected examples, this survey aims at comprehensively demonstrating he recent trends and development.Methods: Four articles are selected for biomedical data acquisition covering topics such as dose saving in CT, C-arm X-ray imaging systems far volume imaging, and the replacement of dose-intensive CT-based diagnostic with harmonic ultrasound imaging. Regarding biomedical signal analysis (BSA), he four selected articles discuss the equivalence of different time-frequency approaches for signal analysis, an application to Cochlea implants, where time-frequency analysis is applied far controlling the replacement system, recent trends far fusion of different modalities, and the role of BSA as part of a brain machine interfaces. To cover the broad spectrum of publications in the field of biomedical image processing, six papers are focused. Important topics are content-based image retrieval in medical applications, automatic classification of tongue photographs from traditional Chinese medicine, brain perfusion analysis in single photon emission computed tomography (SPECT), model-based visualization of vascular trees, and virtual surgery, where enhanced visualization and haptic feedback techniques are combined with a sphere-filled model of he organ.Results: The selected papers emphasize the five fields forming the chain of biomedical data processing: (1) data acquisition, (2) data reconstruction and pre-processing, (3) data handling, (4) data analysis, and (5) data Visualization. Fields 1 and 2 form he sensor informartics, while fields 2 to 5 form signal or image informatics with respect to the nature of the data considered.Conclusions: Biomedical data acquisition and pro-processing, as well as data handling, analysis and Visualization aims at providing reliable tools for decision support that improve the quality of health care. Comprehensive evaluation of he processing methods and their reliable integration in routine applications are future challenges in the hold of sensor, signal and image informatics.

With the increasing amount of medical images that are routinely acquired in today’s healthcare institutions, common methods of image management become inefficient. Even in modern picture archiving and communication systems that are based on the DICOM standard, image data is addressed by alphanumerical indexes such as patient name and examination date. In this software demonstration, the capabilities of content-based access to medical images is exemplified. Radiographs that are arbitrarily selected from the internet can be compared to the reference database by means of numerical features extracted from the pixel values. This allows for automatic identification of image content such as body region or the biomedical system images, as well as the determination of the modality used for image acquisition. Since the features of the system capture global, local, and structural image properties, content-based access impacts research, teaching and patient care.

Die automatisierte explizite Extraktion von Objekten aus Bildserien, erfordert eine reproduzierbare Beschreibung der entsprechenden Bildregionen. Ein hierarchisches Partitionierungsverfahren zerlegt dazu ein Bild in seine visuell plausiblen Regionen, für die dann ein Vektor mit beschreibenden ordinalen Merkmalen berechnet wird. Objektextraktion entspricht damit der Klassifikation entsprechender Merkmalsvektoren, die vom Anwender durch markieren, trainiert werden. Da für die Klassifikation das „No-Free-Lunch-Theorem“ gilt, müssen Klassifikator und Merkmalsauswahl für jede Domäne experimentell ermittelt werden. Beim Vergleich des Nearest-Neighbor Klassifikators mit dem Bayes Klassifikator mit Gaußschen Mischverteilungen und der Supportvektormaschine liefert letztere für die Handknochenextraktion aus Röntgenaufnahmen das beste Ergebnis.

The increasing number of digital imaging modalities results in data volumes of several Tera Bytes per year that must be transferred and archived in a common-sized hospital. Hence, data compression is an important issue for picture archiving and communication systems (PACS). The effect of lossy image compression is frequently analyzed with respect to images from a certain modality supporting a certain diagnosis. However, novel compression schemes have been developed recently allowing efficient but lossless compression. In this study, we compare the lossless compression schemes embedded in the tagged image file format (TIFF), graphics interchange format (GIF), and Joint Photographic Experts Group (JPEG 2000 II) with the Borrows-Wheeler compression algorithm (BWCA) with respect to image content and origin. Repeated measures ANOVA was based on 1.200 images in total. Statistically significant effects (p < 0,0001) of compression scheme, image content, and image origin were found. Best mean compression factor of 3.5 (2.272 bpp) is obtained applying BTW to secondarily digitized radiographs of the head, while the lowest factor of 1,05 (7.587 bpp) resulted from the TIFF packbits algorithm applied to pelvis images captured digitally. Over all, the BWCA is slightly but significantly more effective than JPEG 2000. Both compression schemes reduce the required bits per pixel (bpp) below 3. Also, secondarily digitized images are more compressible than the directly digital ones. Interestingly, JPEG outperforms BWCA for directly digital images regardless of image content, while BWCA performs better than JPEG on secondarily digitized radiographs. In conclusion, efficient lossless image compression schemes are available for PACS.

Many image retrieval systems, and the evaluation methodologies of these systems, make use of either visual or textual information only. Only few combine textual and visual features for retrieval and evaluation. If text is used, it is often relies upon having a standardised and complete annotation schema for the entire collection. This, in combination with high-level semantic queries, makes visual/textual combinations almost useless as the information need can often be solved using just textual features. In reality, many collections do have some form of annotation but this is often heterogeneous and incomplete. Web-based image repositories such as FlickR even allow collective, as well as multilingual annotation of multimedia objects. This article describes an image retrieval evaluation campaign called ImageCLEF. Unlike previous evaluations, we offer a range of realistic tasks and image collections in which combining text and visual features is likely to obtain the best results. In particular, we offer a medical retrieval task which models exactly the situation of heterogenous annotation by combining four collections with annotations of varying quality, structure, extent and language. Two collections have an annotation per case and not per image, which is normal in the medical domain, making it difficult to relate parts of the accompanying text to corresponding images. This is also typical of image retrieval from the web in which adjacent text does not always describe an image. The ImageCLEF benchmark shows the need for realistic and standardised datasets, search tasks and ground truths for visual information retrieval evaluation.

Medizinische Bildinhalte sind häufig aus mehreren Bildregionen zusammengesetzt, die zueinander in bestimmten Verhältnissen stehen. Diese lassen sich über Wahrscheinlichkeitsverteilungen regionaler und relationaler Attribute zu einem Strukturprototypen verallgemeinern und mit Szenen unbekannten Inhalts vergleichen. Die semiautomatische Synthese eines Strukturprototyps sowie seine Anwendung zur Szenenanalyse wird am Beispiel von Handradiographien vorgestellt. Durch Verwendung der Strukturinformation wird hier eine Steigerung der Erkennungsrate um 17% erzielt.

In 1977, Reichertz has defined medical informatics from the viewpoint of information logistics: to provide the right information at the right time at the right place [1; 2]. The analysis of medical information and (bio-)signal processing contributes substantially to an improved information provision. Therefore, medical informatics is an applied science in the best sense of the word. The special requirements of the medical domain continuously promote innovative processes in computer science, e. g., we refer to the invention of expert systems in the context of medical diagnosis. Almost 30 years later, medical informatics is still an emerging field of research and applications that aims at information logistics, but the process also continues beyond this goal [3]. Just providing information is not sufficient anymore. Since the amount of information available in electronic form is exploding, information selection must also be supported by medical informatics. Furthermore, information needs interpretation in order to gain knowledge. Reflecting an intensive discussion in scientific literature ten years ago [4–8], Haux has defined the following grand challenges for research in health informatics [9]: 1. diagnostics (‘the visible body’); 2. therapy (‘medical intervention with as little strain on the patient as possible’); 3. therapy simulation; 4. early-recognition and prevention; 5. compensating physical handicaps; 6. health consulting (‘the informed patient’); 7. health reporting; 8. healthcare information systems; 9. medical documentation; 10. comprehensive documentation of medical knowledge and knowledge-based decision support.

The aim of this study was to investigate the impact of wait days (WDs) on missed outpatient MRI appointments across different demographic and socioeconomic factors.An institutional review board–approved retrospective study was conducted among adult patients scheduled for outpatient MRI during a 12-month period. Scheduling data and demographic information were obtained. Imaging missed appointments were defined as missed scheduled imaging encounters. WDs were defined as the number of days from study order to appointment. Multivariate logistic regression was applied to assess the contribution of race and socioeconomic factors to missed appointments. Linear regression was performed to assess the relationship between missed appointment rates and WDs stratified by race, income, and patient insurance groups with analysis of covariance statistics.A total of 42,727 patients met the inclusion criteria. Mean WDs were 7.95 days. Multivariate regression showed increased odds ratio for missed appointments for patients with increased WDs (7-21 days: odds ratio [OR], 1.39; >21 days: OR, 1.77), African American patients (OR, 1.71), Hispanic patients (OR, 1.30), patients with noncommercial insurance (OR, 2.00-2.55), and those with imaging performed at the main hospital campus (OR, 1.51). Missed appointment rate linearly increased with WDs, with analysis of covariance revealing underrepresented minorities and Medicaid insurance as significant effect modifiers.Increased WDs for advanced imaging significantly increases the likelihood of missed appointments. This effect is most pronounced among underrepresented minorities and patients with lower socioeconomic status. Efforts to reduce WDs may improve equity in access to and utilization of advanced diagnostic imaging for all patients.