BIOIMAGING 2015 Abstracts

Full Papers Short Papers
Full Papers
Paper Nr: 5
Title:

Crutchfield Information Metric for Quantifying the Inter-sequence Relationship of Multiparametric MRI Data
Authors:

Jens Kleesiek, Armin Biller and Kai Ueltzhöffer
Abstract: A plethora of different MRI sequences exists. To automatically structure this ’zoo’ of available sequences we propose the usage of a framework rooted in information theory. In this paper we show that the Crutchfield information metric is a suitable distance measure for this purpose. It is demonstrated that the physical relationship can be inferred with this metric solely based on the voxel intensities. As future applications we envisage MRI sequence quality control and standardization.
Download
Paper Nr: 6
Title:

Content Based Retrieval of MRI Based on Brain Structure Changes in Alzheimer’s Disease
Authors:

Katarina Trojacanec, Ivan Kitanovski, Ivica Dimitrovski and Suzana Loshkovska
Abstract: The aim of the paper is to present Content Based Retrieval of MRI based on the brain structure changes characteristic for Alzheimer’s Disease (AD). The approach used in this paper aims to improve the retrieval performance while using smaller number of features in comparison to the descriptor dimensionality generated by the traditional feature extraction techniques. The feature vector consists of the measurements of cortical and subcortical brain structures, including volumes of the brain structures and cortical thickness. Two main stages are required to obtain these features: segmentation and calculation of the quantitative measurements. The feature subset selection is additionally applied using Correlation-based Feature Selection (CFS) method. Euclidean distance is used as a similarity measurement. The retrieval performance is evaluated using MRIs provided by the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Experimental results show that the strategy used in this research outperforms the traditional one despite its simplicity and small number of features used for representation.
Download
Paper Nr: 8
Title:

Meshing Meristems - An Iterative Mesh Optimization Method for Modeling Plant Tissue at Cell Resolution
Authors:

Guillaume Cerutti and Christophe Godin
Abstract: We address in this paper the problem of reconstructing a mesh representation of plant cells in a complex, multi-layered tissue structure, based on segmented images obtained from confocal microscopy of shoot apical meristem of model plant Arabidopsis thaliana. The construction of such mesh structures for plant tissues is currently a missing step in the existing image analysis pipelines. We propose a method for optimizing the surface triangular meshes representing the tissue simultaneously along several criteria, based on an initial low-quality mesh. The mesh geometry is deformed by iteratively minimizing an energy functional defined over this discrete surface representation. This optimization results in a light discrete representation of the cell surfaces that enables fast visualization, and quantitative analysis, and gives way to in silico physical and mechanical simulations on real-world data. We provide a framework for evaluating the quality of the cell tissue reconstruction, that underlines the ability of our method to fit multiple optimization criteria.
Download
Paper Nr: 13
Title:

Optical Imaging for Diagnosis of Rheumatoid Arthritis - Automatic Versus Human Evaluation
Authors:

Pouyan Mohajerani, Reinhard Meier, Ernst Rummeny and Vasilis Ntziachristos
Abstract: Successful detection of rheumatoid arthritis (RA) at the early stages of development can significantly enhance the chances of effective therapy. The early onset of RA is often marked with inflammation of the synovial lining of the joint, a condition known as synovitis. Effective imaging of synovitis is therefore of critical importance. While dynamic, contrast-enhanced magnetic resonance imaging (MRI) is capable of effective imaging of synovitis, it is a costly modality. As an alternative, inexpensive approach, optical imaging post injection of the near-infrared fluorescent dye indocynine green (ICG) has been recently proposed for imaging RA. Evaluation of the obtained optical images is performed via examination by trained human readers. However, optical imaging has yet to achieve the diagnostic accuracy of MRI. In this paper we present a method for automatic evaluation of the fluorescence images and compare its performance with the human-based evaluation. Our method relies on our previous work on spatiotemporal analysis of image sequence with principal component analysis (PCA) to seek synovitis signal components with the help of a segmentation method. The results for a group of 600 joints, obtained from 20 patients, suggest improved diagnostic performance using the automatic approach in comparison to human-based evaluation.
Download
Short Papers
Paper Nr: 1
Title:

Segmentation of LG Enhanced Cardiac MRI

Authors:

Kjersti Engan, Valery Naranjo, Trygve Eftestøl, Stein Ørn and Leik Woie
Abstract: In this paper a method for segmentation of the endocardium in Late Gadolinium Enhanced Cardiac Magnetic Resonance (LGE-CMR) images is presented and combined with a previously proposed method for segmentation of the epicardium. The method is fully automatic and based on utilizing a priori knowledge about the type of images. No other image modalities, like CINE images, are used. Using a combination of a rough a priori model and preprocessed images, an a posteriori model is built. The final segmentation step is performed in the polar domain. We compare our results on a set of 395 images from 54 patients with segmentation using marker controlled watershed with different gradient images and different markers. The proposed method gives a mean Dice and Jaccard indices over all images as 0.85 and 0.74 respectively.
Download
Paper Nr: 3
Title:

A Block-based Approach for Malignancy Detection within the Prostate Peripheral Zone in T2-weighted MRI
Authors:

Andrik Rampun, Paul Malcolm and Reyer Zwiggelaar
Abstract: In this paper, a computer-aided diagnosis method is proposed for the detection of prostate cancer within the peripheral zone. Firstly, the peripheral zone is modelled according to the generic 2D mathematical model from the literature. In the training phase, we captured 334 samples of malignant blocks from cancerous regions which were already defined by an expert radiologist. Subsequently, for every unknown block within the peripheral zone in the testing phase we compare its global, local and attribute similarities with training samples captured previously. Next we compare the similarity between subregions and find which of the subregion has the highest possibility of being malignant. An unknown block is considered to be malignant if it is similar in comparison to one of the malignant blocks, its location is within the subregion which has the highest possibility of being malignant and there is a significant difference in lower grey level distributions within the subregions. The initial evaluation of the proposed method is based on 260 MR images from 40 patients and we achieved 90% accuracy and sensitivity and 89% specificity with 5% and 6% false positives and false negatives, respectively.
Download
Paper Nr: 12
Title:

Detection of Root Knot Nematodes in Microscopy Images
Authors:

Faroq AL-Tam, António dos Anjos, Stephane Bellafiore and Hamid Reza Shahbazkia
Abstract: Object detection in microscopy image is essential for further analysis in many applications. However, images are not always easy to analyze due to uneven illumination and noise. In addition, objects may appear merged together with debris. This work presents a method for detecting rice root knot nematodes in microscopy images. The problem involves four subproblems which are dealt with separately. The uneven illumination is corrected via polynomial fitting. The nematodes are then highlighted using mathematical morphology. A binary image is obtained and the microscope lines are removed. Finally, the detected nematodes are counted after thresholding the non-nematode particles. The results obtained from the performed tests show that this is a reliable and effective method when compared to manual counting.
Download
Paper Nr: 4
Title:

Influence of Muscle Cross-sectional Area in Skin Temperature
Authors:

Eduardo Borba Neves, Fabio Bandeira, Leandra Ulbricht, José Vilaça-Alves and Victor Machado Reis
Abstract: The present study aimed to determine the correlations among the arm subcutaneous fat percentage (SFP), arm muscle cross-sectional area (MCSA), arm total cross-sectional area (TCSA), and the difference between core temperature and skin temperature in biceps and triceps areas, measured using thermography. This research focused on a cross-sectional study using a quantitative approach with participants consisting of young, untrained volunteers from the city of Curitiba, Brazil. The total sample size was 20 volunteers including 13 males and 7 females. A statistical correlation between MSCA and core and skin Δ temperature for the right and left biceps (r = -0.487, p = 0.030 / r = -0.518, p = 0.019), and also between TCSA core and skin Δ temperature for the right and left biceps (r = -0.513, p = 0.021 / r = -0.554, p = 0.011) was identified. These results confirmed that arm muscle cross-sectional area influenced skin temperature at the biceps region. This result can also be generalized to other areas of the skin, which show similar characteristics to the studied area.
Download
Paper Nr: 9
Title:

A Fully Automatic Tool for Counting Virchow-Robin Spaces in Magnetic Resonance Imaging for Lacunar Stroke Study
Authors:

Sérgio Pereira, José Mariz, Nuno Sousa, J. H. Correia and Carlos Silva
Abstract: Virchow-Robin Spaces surround the perforating arteries of the brain and sometimes they become dilated. Studies suggest that those structures are correlated with some conditions such as lacunar strokes, small vessel diseases, multiple sclerosis or even normal aging. However, the majority of those studies are based on the detection of those structures by a human expert, in some regions of interest, which is prone to the subjectivity of the person doing the task. Moreover, dilated Virchow-Robin Spaces may look similar to lacunar strokes, making them difficult to identify. Few works have been proposed on the computational detection of dilated Virchow-Robin Spaces. In this paper, we propose a fully automatic tool, capable of preprocessing the magnetic resonance images, extract the most relevant regions of interest and detect dilated Virchow-Robin Spaces. Such a tool may be useful to eliminate human subjectivity, but also to improve the reproducibility of the studies, leading to more reliable correlations. An application to visualize and count the detected structures was also built, with the aim of helping in a study of the correlation of lacunar strokes, Virchow-Robin Spaces and vascular dementia.
Download
Paper Nr: 16
Title:

Low-rank and Sparse Matrix Decomposition with a-priori Knowledge for Dynamic 3D MRI Reconstruction
Authors:

Dornoosh Zonoobi, Shahrooz Faghigh roohi and Ashraf. A. Kassim
Abstract: It has been recently shown that incorporating priori knowledge significantly improves the performance of basic compressive sensing based approaches. We have managed to successfully exploit this idea for recovering a matrix as a summation of a Low-rank and a Sparse component from compressive measurements. When applied to the problem of construction of 4D Cardiac MR image sequences in real-time from highly under-sampled k-space data, our proposed method achieves superior reconstruction quality compared to the other state-of-the-art methods.
Download
Paper Nr: 22
Title:

Robust Image Analysis of BeadChip Microarrays
Authors:

Jan Kalina and Anna Schlenker
Abstract: Microarray images in molecular genetics are heavily contaminated by noise and outlying measurements. This paper is devoted to analysis of Illumina BeadChip microarray images, primarily to their low-level preprocessing. We point out that standard image analysis procedures, which are implemented in the beadarray package of BioConductor software, are highly sensitive to contamination by severe noise and outliers. Therefore, the habitually used methodology does not discover many of the outliers. We illustrate this on real data and show that the standard background correction method may actually amplify the noise in the image. A robust image analysis tailor-made for this type of microarray images is highly desirable. We explain principles and show preliminary results of our robust alternative to the standard approach, which aims to be robust to noise and outliers in each its step.
Download
Paper Nr: 25
Title:

Encoded Total Focusing Method for Improving Data Acquisition Rate
Authors:

César Gutiérrez Fernández, Ana Jiménez and Carlos Julián Martín-Arguedas
Abstract: Synthetic aperture imaging techniques are capable to obtain high quality images, fully focused in both transmission and reception. However, these techniques require to perform so many emissions as elements in the array to acquire RF data. This requirement decreases acquisition rate and can result in tissue motion artifacts because of the phase misalignments between signals acquired in different emissions. Such inconvenience claims for alternatives that reduce the total number of emissions needed to obtain the data. This work proposes the use of Code Division Multiple Access (CDMA) techniques to attain this goal. By encoding the ultrasound excitation signal emitted, through a pesudo-random Kasami code, several elements can emit simultaneously and the amount of data acquired in every emission increases. The encoded proposal attains a Kx speed up in acquisition rate compared to conventional total focusing method, being K the number of Kasami sequences available in the set.
Download
Paper Nr: 27
Title:

Quantitative Scoring of Muscle Involvement in MRI of Neuromuscular Diseases
Authors:

Maria Evelina Fantacci, Guja Astrea, Roberta Battini, Alessandra Retico, Chiara Sottocornola and Michela Tosetti
Abstract: An automated method to evaluate the fat infiltration in tissues has been developed and applied to images of the human leg. The final aim is to obtain a quantitative evaluation of fat infiltration percentage and to relate it to the grade of muscle impairment in subjects affected by Neuro-Muscular Diseases (NMD). Through a muscle segmentation algorithm on structural T1-weighted magnetic resonance images (MRIs), the estimated non-muscle percentage (eNMP) in the segmented muscle area has been evaluated in healthy subjects as a reference value. A semi-automated procedure allows extending the algorithm to MRIs of NMD patients. A strong correlation has been demonstrated between the eNMP index and the disease severity.
Download