Pulmonary embolism (PE) affects up to 600,000 patients and contributes to at least 100,000 deaths every year in the United States alone. Diagnosis of PE can be difficult as most symptoms are unspecific and early diagnosis is essential for successful treatment. Computed Tomography (CT) images can show morphological anomalies that suggest the existence of PE. Various image-based procedures have been proposed for improving computer-aided diagnosis of PE. We propose a novel method for detecting PE based on localized vessel-based features computed in Dual Energy CT (DECT) images. DECT provides 4D data indexed by the three spatial coordinates and the energy level. The proposed features encode the variation of the Hounsfield Units across the different levels and the CT attenuation related to the amount of iodine contrast in each vessel. A local classification of the vessels is obtained through the classification of these features. Moreover, the localization of the vessel in the lung provides better comparison between patients. Results show that the simple features designed are able to classify pulmonary embolism patients with an AUC (area under the receiver operating curve) of 0.71 on a lobe basis. Prior segmentation of the lung lobes is not necessary because an automatic atlas-based segmentation obtains similar AUC levels (0.65) for the same dataset. The automatic atlas reaches 0.80 AUC in a larger dataset with more control cases.
Medical images contain a large amount of visual information about structures and anomalies in the human body. To make sense of this information, human interpretation is often essential. On the other hand, computer-based approaches can exploit information contained in the images by numerically measuring and quantifying specific visual features. Annotation of organs and other anatomical regions is an important step before computing numerical features on medical images. In this paper, a texture-based organ classification algorithm is presented, which can be used to reduce the time required for annotating medical images. The texture of organs is analyzed using a combination of state-of-the-art techniques: the Riesz transform and a bag of meaningful visual words. The effect of a meaningfulness transformation in the visual word space yields two important advantages that can be seen in the results. The number of descriptors is enormously reduced down to 10% of the original size, whereas classification accuracy is improved by up to 25% with respect to the baseline approach.
Advances in medical knowledge give clinicians more objective information for a diagnosis. Therefore, there is an increasing need for bibliographic search engines that can provide services helping to facilitate faster information search.
The ImageCLEFmed benchmark proposes a medical case-based retrieval task. This task aims at retrieving articles from the biomedical literature that are relevant for differential diagnosis of query cases including a textual description and several images. In the context of this campaign many approaches have been investigated showing that the fusion of visual and text information can improve the precision of the retrieval. However, fusion does not always lead to better results.
In this paper, a new query-adaptive fusion criterion to decide when to use multi-modal (text and visual) or only text approaches is presented. The proposed method integrates text information contained in MeSH (Medical Subject Headings) terms extracted and visual features of the images to find synonym relations between them. Given a text query, the query-adaptive fusion criterion decides when it is suitable to also use visual information for the retrieval.
Results show that this approach can decide if a text or multi{modal approach should be used with 77.15% of accuracy.
Distinct texture classes are often sharing several visual concepts. Texture instances from different classes are sharing regions in the feature hyperspace, which results in ill-defined classification configurations. In this work, we detect rotation-covariant visual concepts using steerable Riesz wavelets and bags of visual words. In a first step, K-means clustering is used to detect visual concepts in the hyperspace of the energies of steerable Riesz wavelets. The coordinates of the clusters are used to construct templates from linear combinations of the Riesz components that are corresponding to visual concepts. The visualization of these templates allows verifying the relevance of the concepts modeled. Then, the local orientations of each template are optimized to maximize their response, which is carried out analytically and can still be expressed as a linear combination of the initial steerable Riesz templates. The texture classes are learned in the feature space composed of the concatenation of the maximum responses of each visual concept using support vector machines. An experimental evaluation using the Outex TC 00010 test suite allowed a classification accuracy of 97.5%, which demonstrates the feasibility of the proposed approach. An optimal number K = 20 of clusters is required to model the visual concepts, which was found to be fewer than the number of classes. This shows that higher-level classes are sharing low-level visual concepts. The importance of rotation-covariant visual concept modeling is highlighted by allowing an absolute gain of more than 30% in accuracy. The visual concepts are modeling the local organization of directions at various scales, which is in accordance with the bottom{up visual information processing sequence of the primal sketch in Marr's theory on vision.
Volumetric medical images contain an enormous amount of visual information that can discourage the exhaustive use of local descriptors for image analysis, comparison and retrieval. Distinctive features and patterns that need to be analyzed for finding diseases are most often local or regional, often in only very small parts of the image. Separating the large amount of image data that might contain little important information is an important task as it could reduce the current information overload of physicians and make clinical work more efficient. In this paper a novel method for detecting key-regions is introduced as a way of extending the concept of keypoints often used in 2D image analysis. In this way also computation is reduced as important visual features are only extracted from the detected key regions. The region detection method is integrated into a platform-independent, web-based graphical interface for medical image visualization and retrieval in three dimensions. This web-based interface makes it easy to deploy on existing infrastructures in both small and large-scale clinical environments. By including the region detection method into the interface, manual annotation is reduced and time is saved, making it possible to integrate the presented interface and methods into clinical routine and workflows, analyzing image data at a large scale.
When physicians are searching for articles in the medical literature, images of the articles can help determining relevance of the article content for a specific information need. The visual image representation can be an advantage in effectiveness (quality of found articles) and also in efficiency (speed of determining relevance or irrelevance) as many articles can likely be excluded much quicker by looking at a few representative images. In domains such as medical information retrieval, allowing to determine relevance quickly and accurately is an important criterion. This becomes even more important when small interfaces are used as it is frequently the case on mobile phones and tablets to access scientific data whenever information needs arise. In scientific articles many figures are used and particularly in the biomedical literature only a subset may be relevant for determining the relevance of a specific article to an information need. In many cases clinical images can be seen as more important for visual appearance than graphs or histograms that require looking at the context for interpretation. To get a clearer idea of image relevance in articles, a user test with a physician was performed who classified images of biomedical research articles into categories of importance that can subsequently be used to evaluate algorithms that automatically select images as representative examples. The manual sorting of images of 50 journal articles of BioMedCentral with each containing more than 8 figures by importance also allows to derive several rules that determine how to choose images and how to develop algorithms for choosing the most representative images of specific texts. This article describes the user tests and can be a first important step to evaluate automatic tools to select representative images for representing articles and potentially also images in other contexts, for example when representing patient records or other medical concepts when selecting images to represent RadLex terms in tutorials or interactive interfaces for example. This can help to make the image retrieval process more efficient and effective for physicians.
Journal images represent an important part of the knowledge stored in the medical literature. Figure classification has received much attention as the information of the image types can be used in a variety of contexts to focus image search and filter out unwanted information or ”noise”, for example non–clinical images. A major problem in figure classification is the fact that many figures in the biomedical literature are compound figures and do often contain more than a single figure type. Some journals do separate compound figures into several parts but many do not, thus requiring currently manual separation. In this work, a technique of compound figure separation is proposed and implemented based on systematic detection and analysis of uniform space gaps. The method discussed in this article is evaluated on a dataset of journal figures of the open access literature that was created for the ImageCLEF 2012 benchmark and contains about 3000 compound figures. Automatic tools can easily reach a relatively high accuracy in separating compound figures. To further increase accuracy efforts are needed to improve the detection process as well as to avoid over–separation with powerful analysis strategies. The tools of this article have also been tested on a database of approximately 150’000 compound figures from the biomedical literature, making these images available as separate figures for further image analysis and allowing to filter important information from them.
Comparing several series of images is not always easy as the corresponding slices often need
to be selected manually. In times where series contain an ever-increasing number of slices this
can mean manual work when moving several series to the corresponding slice. Particularly two
situations were identified in this context: (1) patients with a large number of image series over
time (such as patients with cancers that are monitored) frequently need to compare the series,
for example to compare tumor growth over time. Manually adapting two series is possible but
with four or more series this can mean loosing time. Having automatically the closest slice
by comparing visual similarity also in older series with differing slice thickness and inter slice
distance can save time and synchronize the viewing instantly. (2) analyzing visually similar
image series of several patients can profit from being viewed in a synchronized way to compare
the cases, so when sliding through the slices in one volume, the corresponding slices in the other
volumes are shown. This application could be employed after content-based 3D image retrieval
has found similar series, for example. Synchronized viewing can help finding or confirming the
most relevant cases quickly.
To allow for synchronized viewing of several image volumes, the test image series are first
registered applying affine transformation for the global registration of images followed by diffeomorphic
image registration. Then corresponding slices in the two volumes are estimated based
on a visual similarity. Once the registration is finished, the user can subsequently move inside
the slices of one volume (reference volume) and can view the corresponding slices in the other
volumes. These corresponding slices are obtained after a correspondence match in the registration
procedure. These volumes are synchronized in that the slice closest to the original reference
volume is shown even when the slice thicknesses or inter slice distances differ, and this is automatically
done by comparing the visual image content of the slices. The tool has the potential to
help in a variety of situations and it is currently being made available as a plugin for the popular
Osirix image viewer.
Segmentation of the various parts of the brain is a challenging area in medical imaging and it is a prerequisite
for many image analysis tasks useful for clinical research. Advances have been made in generating brain image
templates that can be registered to automatically segment regions of interest in the human brain. However, these
methods may fail with some subjects if there is a significant shape distortion or difference from the proposed
models. This is also the case of newborns, where the developing brain strongly differs from adult magnetic
resonance imaging (MRI) templates.
In this article, a texture-based cerebellum segmentation method is described. The algorithm presented does
not use any prior spatial knowledge to segment the MRI images. Instead, the system learns the texture features
by means of a multi-scale filtering and visual words feature aggregation. Visual words are a commonly used
technique in image retrieval. Instead of using visual features directly, the features of specific regions are modeled
(clustered) into groups of discriminative features. This means that the final feature space can be reduced in size
and also that the visual words in local regions are really discriminative for the given data set. The system is
currently trained and tested with a dataset of 18 adult brain MRIs. An extension to the use with newborn brain
images is being foreseen as this could highlight the advantages of the proposed technique.
Results show that the use of texture features can be valuable for the task described and can lead to good
results. The use of visual words can potentially improve robustness of existing shape-based techniques for cases
with significant shape distortion or other differences from the models. As the visual words based techniques are
not assuming any prior knowledge such techniques could be used for other types of segmentations as well using
a large variety of basic visual features.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.