Rudi Deklerck

A novel computer‐aided lung nodule detection system for CT images

PURPOSE The paper presents a complete computer-aided detection (CAD) system for the detection of lung nodules in computed tomography images. A new mixed feature selection and classification methodology is applied for the first time on a difficult medical image analysis problem. METHODS The CAD system was trained and tested on images from the publicly available Lung Image Database Consortium (LIDC) on the National Cancer Institute website. The detection stage of the system consists of a nodule segmentation method based on nodule and vessel enhancement filters and a computed divergence feature to locate the centers of the nodule clusters. In the subsequent classification stage, invariant features, defined on a gauge coordinates system, are used to differentiate between real nodules and some forms of blood vessels that are easily generating false positive detections. The performance of the novel feature-selective classifier based on genetic algorithms and artificial neural networks (ANNs) is compared with that of two other established classifiers, namely, support vector machines (SVMs) and fixed-topology neural networks. A set of 235 randomly selected cases from the LIDC database was used to train the CAD system. The system has been tested on 125 independent cases from the LIDC database. RESULTS The overall performance of the fixed-topology ANN classifier slightly exceeds that of the other classifiers, provided the number of internal ANN nodes is chosen well. Making educated guesses about the number of internal ANN nodes is not needed in the new feature-selective classifier, and therefore this classifier remains interesting due to its flexibility and adaptability to the complexity of the classification problem to be solved. Our fixed-topology ANN classifier with 11 hidden nodes reaches a detection sensitivity of 87.5% with an average of four false positives per scan, for nodules with diameter greater than or equal to 3 mm. Analysis of the false positive items reveals that a considerable proportion (18%) of them are smaller nodules, less than 3 mm in diameter. CONCLUSIONS A complete CAD system incorporating novel features is presented, and its performance with three separate classifiers is compared and analyzed. The overall performance of our CAD system equipped with any of the three classifiers is well with respect to other methods described in literature.

Context aware inactivity recognition for visual fall detection

This paper introduces a method for automatic fall detection, targeted towards the monitoring of elderly people in a nursing home or in their natural home environment. The method uses information extracted from images obtained using novel 3D camera technology, combined with a context model. Visual information consists of body orientation calculated from posture extraction and of periods of inactivity. The context model allows for a different interpretation of the visual fall detection results, depending on the exact location, time and duration of the detected event. The context model is learnt during the ongoing monitoring task without human intervention and automatically adapts to the changing activity patterns of the monitored subject

Vascular Active Contour for Vessel Tree Segmentation

In this paper, a novel active contour model is proposed for vessel tree segmentation. First, we introduce a region competition-based active contour model exploiting the Gaussian mixture model, which mainly segments thick vessels. Second, we define a vascular vector field to evolve the active contour along its center line into the thin and weak vessels. The vector field is derived from the eigenanalysis of the Hessian matrix of the image intensity in a multiscale framework. Finally, a dual curvature strategy, which uses a vesselness measure-dependent function selecting between a minimal principal curvature and a mean curvature criterion, is added to smoothen the surface of the vessel without changing its shape. The developed model is used to extract the liver and lung vessel tree as well as the coronary artery from high-resolution volumetric computed tomography images. Comparisons are made with several classical active contour models and manual extraction. The experiments show that our model is more accurate and robust than these classical models and is, therefore, more suited for automatic vessel tree extraction.

Segmentation of medical images

Abstract Segmentation and labelling remains the weakest step in many medical vision applications. This paper illustrates an approach based on generic modules which are designed to solve typical problems encountered in various applications, and which are controllable through adaptation of their parameters. Two of these modules are presented: the cavity detector , a method for the segmentation of regions which are not completely surrounded by walls and edgmentation , a modified split-and-merge algorithm for edge preserving image enhancement, segmentation and data reduction. We describe the principles of the algorithms and illustrate their generic properties by discussing the results of various applications in 2D and 3D cardiac MRI, in 3D and 4D cardiac SPECT. and in 2D brain X-ray CT.

Evaluation of the radiation dose in micro-CT with optimization of the scan protocol

INTRODUCTION Micro-CT provides non-invasive anatomic evaluation of small animals. Serial micro-CT measurements are, however, hampered by the severity of ionizing radiation doses cumulating over the total period of follow-up. The dose levels may be sufficient to influence experimental outcomes such as animal survival or tumor growth. AIM This study was designed to evaluate the radiation dose of micro-CT and to optimize the scanning protocol for longitudinal micro-CT scans. METHODS AND MATERIALS Normal C57Bl/6 mice were euthanized. Radiation exposure was measured using individually calibrated lithium fluoride thermoluminescent dosimeters (TLDs). Thirteen TLDs were placed in the mice at the thyroid, lungs, liver, stomach, colon, bladder and near the spleen. Micro-CT (SkyScan 1178) was performed using two digital X-ray cameras which scanned over 180 degrees at a resolution of 83 microm, a rotation step of 1.08 degrees , 50 kV, 615 microA and 121 s image acquisition time. The TLDs were removed after each scan. CTDI(100) was measured with a 100 mm ionization chamber, centrally positioned in a 2.7 cm diameter water phantom, and rotation steps were increased to reduce both scan time and radiation dose. RESULTS Internal TLD analysis demonstrated median organ dose of 5.5 +/- 0.6 mGy per mA s, confirmed by CTDI(100) with result of 6.6 mGy per mA s. A rotation step of 2.16 resulted in qualitatively accurate images. At a resolution of 83 microm the scan time is reduced to 63 s with an estimated dose of 2.9 mGy per mA s. At 166 microm resolution, the scan time is limited to 27 s, with a concordant dose of 1.2 mGy per mA s. CONCLUSIONS The radiation dose of a standard micro-CT scan is relatively high and could influence the experimental outcome. We believe that the presented adaptation of the scan protocol allows for accurate imaging without the risk of interfering with the experimental outcome of the study.

3D human pose recognition for home monitoring of elderly

A toolbox for the automatic monitoring of elderly in a nursing home or in the natural home environment is proposed. Rather than monitoring vital signs or other biomedical parameters, the toolbox is focussed on the monitoring of activity patterns and changes therein. Activity information is derived from visual information using image processing algorithms. The visual information is acquired using 3D camera technology. Besides a traditional visual image, 3D cameras also provide highly accurate depth information. The 3D position of the subject is derived and serves as the primary information source for the different components in the toolbox.

Region competition based active contour for medical object extraction

In this paper, a probabilistic and level set model for three-dimensional medical object extraction is proposed, which is called region competition based active contour. The algorithms are derived by minimizing a region based probabilistic energy function and implemented in a level set framework. An additional speed-controlling term makes the active contour quickly convergent to the actual contour on strong edges, whereas a probabilistic model makes the active contour performing well for weak edges. Prior knowledge about the initial contour and the probabilistic distribution contributes to more efficient extraction. The developed model has been applied to a variety of medical images, from CTA and MRA of the coronary to rotationally scanned and real-time three-dimensional echocardiography images of the mitral valve. As the results show, the algorithm is fast, convergent, adapted to a broad range of medical objects and produces satisfactory results.

Markov random field-based clustering applied to the segmentation of masses in digital mammograms

In this paper we propose a new pixel clustering model applied to the analysis of digital mammograms. The clustering represents here the first step in a more general method and aims at the creation of a concise data-set (clusters) for automatic detection and classification of masses, which are typically among the first symptoms analysed in early diagnosis of breast cancer. For the purpose of this work, a set of mammographic images has been employed, that are 12-bit gray level digital scans and as such, are inherently inhomogeneous and affected by the noise resulting from the film scanning. The image pixels are described only by their intensity (gray level), therefore, the available information is limited to one dimension. We propose a Markov random field (MRF)-based technique that is suitable for performing clustering in an environment which is described by poor or limited data. The proposed method is a statistical classification model, that labels the image pixels based on the description of their statistical and contextual information. Apart from evaluating the pixel statistics, that originate from the definition of the K-means clustering scheme, the model expands the analysis by the description of the spatial dependence between pixels and their labels (context), hence leading to the reduction of the inhomogeneity of the output. Moreover, we define a probabilistic description of the model, that is characterised by a remarkable simplicity, such that its realisation can be easily and efficiently implemented in any high- or low-level programming language, thus allowing it to be run on virtually any kind of platform. Finally, we evaluate the algorithm against the classical K-means clustering routine. We point out similarities between the two methods and, moreover, show the advantages and superiority of the MRF scheme.

CT analysis of defects of the cement mantle and alignment of the stem IN VITRO COMPARISON OF CHARNLEY-KERBOUL FEMORAL HIP IMPLANTS INSERTED LINE-TO-LINE AND UNDERSIZED IN PAIRED FEMORA

Using a modern cementing technique, we implanted 22 stereolithographic polymeric replicas of the Charnley-Kerboul stem in 11 pairs of human cadaver femora. On one side, the replicas were cemented line-to-line with the largest broach. On the other, one-size undersized replicas were used (radial difference, 0.89 mm sd 0.13).CT analysis showed that the line-to-line stems without distal centralisers were at least as well aligned and centered as undersized stems with a centraliser, but were surrounded by less cement and presented more areas of thin (< 2 mm) or deficient (< 1 mm) cement. These areas were located predominantly at the corners and in the middle and distal thirds of the stem. Nevertheless, in line-to-line stems, penetration of cement into cancellous bone resulted in a mean thickness of cement of 3.1 mm (sd 0.6) and only 6.2% of deficient and 26.4% of thin cement. In over 90% of these areas, the cement was directly supported by cortical bone or cortical bone with less than 1 mm of cancellous bone interposed. When Charnley-Kerboul stems are cemented line-to-line, good clinical results are observed because cement-deficient areas are limited and are frequently supported by cortical bone.

Signal Representation and Processing of Nucleotide Sequences

Sets of related signals can be represented by separating their joint variation and showing the individual signal offsets with respect to this reference. An example is the genomic signal analysis of pathogen variability. The conversion of symbolic nucleotide sequences to genomic signals allows to use signal processing methods to analyze genomic data. This approach reveals striking regularities in the distribution of nucleotides and pair of nucleotides along the sequences, in both prokaryotes and eukaryotes. Genomic signals can also be used for sequence prediction, similarly to time series prediction. The methodology is also adequate for studying the development of pathogen multiple resistance to drugs.