Visualization in medicine and life sciences

329 Pages·2008·8.62 MB·English

by  Lars Linsen, Hans Hagen, Bernd Hamann

Mathematics and Visualization
SeriesEditors
GeraldFarin
Hans-ChristianHege
DavidHoffman
ChristopherR.Johnson
KonradPolthier
MartinRumpf
Lars Linsen
Hans Hagen
Bernd Hamann
Editors
Visualization in Medicine
and Life Sciences
With 76 Figures, 47 in Color
ABC
Lars Linsen Hans Hagen
School of Engineering and Science Technische Universität Kaiserslautern
Jacobs University Bremen 67653 Kaiserslautern, Germany
P.O. Box 750561 E-mail: [email protected]
28725 Bremen, Germany
E-mail: [email protected]
Bernd Hamann
Department of Computer Science
University of California
One Shields Avenue
Davis, CA 95616-8562, U.S.A
E-mail: [email protected]
LibraryofCongressControlNumber:2007935102
MathematicsSubjectClassification:68-06, 68U05
ISBN-13 978-3-540-72629-6SpringerBerlinHeidelbergNewYork
Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis
concerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,broadcasting,
reproductiononmicrofilmorinanyotherway,andstorageindatabanks.Duplicationofthispublication
orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,
1965,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violationsare
liableforprosecutionundertheGermanCopyrightLaw.
SpringerisapartofSpringerScience+BusinessMedia
springer.com
(cid:1)c Springer-VerlagBerlinHeidelberg2008
Theuseofgeneraldescriptivenames,registerednames,trademarks,etc.inthispublicationdoesnotimply,
evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotectivelaws
andregulationsandthereforefreeforgeneraluse.
TypesettingbytheauthorsandSPiusingaSpringerLATEXmacropackage
Coverdesign:design&productionGmbH,Heidelberg
Printedonacid-freepaper SPIN:12066520 46/SPi/3100 543210
Preface
Visualizationtechnologyhasbecomeacrucialcomponentofmedicalandbio-
medical data processing and analysis. This technology complements tradi-
tionalimageprocessingmethodsasitallowsscientistsandpracticingmedical
doctors to visually interact with large, high-resolution three-dimensional im-
age data. Further, an ever increasing number of new data acquisition meth-
ods is being used in medicine and the life sciences, in particular in genomics
and proteomics. The book contains papers discussing some of the latest data
processing and visualization techniques and systems for effective analysis of
diverse, large, complex, and multi-source data.
Internationally leading experts in the area of data visualization came to-
getherforaworkshopdedicatedtovisualization inmedicineandlifesciences,
held on the island of Ru¨gen, Germany, in July 2006. About 40 participants
presented state-of-the-art research on this topic. Research and survey papers
were solicited and carefully refereed, resulting in this collection.
The research topics covered by the papers in this book deal with these
themes:
• Segmentation and Feature Detection
• Surface Extraction
• Volume Visualization
• Graph and Network Visualization
• Visual Data Exploration
• Multivariate and Multidimensional Data Visualization
• Large Data Visualization
The workshop was supported, in part, by the Deutsche Forschungsgemein-
schaft (DFG).
Bremen, Germany Lars Linsen
Kaiserslautern, Germany Hans Hagen
Davis, California, U.S.A. Bernd Hamann
June 2007
Contents
Part I Surface Extraction Methods from Medical Imaging Data
Towards Automatic Generation of 3D Models of Biological
Objects Based on Serial Sections
Vincent Jasper Dercksen, Cornelia Bru¨ß, Detlev Stalling, Sabine
Gubatz, Udo Seiffert, and Hans-Christian Hege ...................... 3
A Topological Approach to Quantitation of Rheumatoid
Arthritis
Hamish Carr, John Ryan, Maria Joyce, Oliver Fitzgerald, Douglas
Veale, Robin Gibney, and Patrick Brennan .......................... 27
3D Visualization of Vasculature: An Overview
Bernhard Preim and Steffen Oeltze................................. 39
3D Surface Reconstruction from Endoscopic Videos
Arie Kaufman and Jianning Wang ................................. 61
Part II Geometry Processing in Medical Applications
A Framework for the Visualization of Cross Sectional Data
in Biomedical Research
Enrico Kienel, Marek Vanˇco, Guido Brunnett, Thomas Kowalski,
Roland Clauß, and Wolfgang Knabe ................................ 77
Towards a Virtual Echocardiographic Tutoring System
Gerd Reis, Bernd Lapp´e, Sascha K¨ohn, Christopher Weber,
Martin Bertram, and Hans Hagen.................................. 99
Supporting Depth and Motion Perception
in Medical Volume Data
Jennis Meyer-Spradow, Timo Ropinski, and Klaus Hinrichs ...........121
VIII Contents
Part III Visualization of Multi-channel Medical Imaging Data
Multimodal Image Registration for Efficient Multi-resolution
Visualization
Joerg Meyer.....................................................137
A User-friendly Tool for Semi-automated Segmentation
and Surface Extraction from Color Volume Data
Using Geometric Feature-space Operations
Tetyana Ivanovska and Lars Linsen ................................153
Part IV Vector and Tensor Visualization in Medical Applications
Global Illumination of White Matter Fibers
from DT-MRI Data
David C. Banks and Carl-Fredrik Westin............................173
Direct Glyph-based Visualization of Diffusion MR Data
Using Deformed Spheres
Martin Domin, So¨nke Langner, Norbert Hosten, and Lars Linsen ......185
Visual Analysis of Bioelectric Fields
Xavier Tricoche, Rob MacLeod, and Chris R. Johnson ................205
MRI-based Visualisation of Orbital Fat Deformation
During Eye Motion
Charl P. Botha, Thijs de Graaf, Sander Schutte, Ronald Root, Piotr
Wielopolski, Frans C.T. van der Helm, Huibert J. Simonsz, and
Frits H. Post ....................................................221
Part V Visualizing Molecular Structures
Visual Analysis of Biomolecular Surfaces
Vijay Natarajan, Patrice Koehl, Yusu Wang, and Bernd Hamann ......237
BioBrowser – Visualization of and Access to Macro-Molecular
Structures
Lars Offen and Dieter Fellner .....................................257
Visualization of Barrier Tree Sequences Revisited
Christian Heine, Gerik Scheuermann, Christoph Flamm,
Ivo L. Hofacker, and Peter F. Stadler ..............................275
Contents IX
Part VI Visualizing Gene Expression Data
Interactive Visualization of Gene Regulatory Networks
with Associated Gene Expression Time Series Data
Michel A. Westenberg, Sacha A. F. T. van Hijum, Andrzej T. Lulko,
Oscar P. Kuipers, and Jos B. T. M. Roerdink .......................293
Segmenting Gene Expression Patterns of Early-stage
Drosophila Embryos
Min-Yu Huang, Oliver Ru¨bel, Gunther H. Weber, Cris L. Luengo
Hendriks, Mark D. Biggin, Hans Hagen, and Bernd Hamann ..........313
Color Plates ...................................................329
Towards Automatic Generation of 3D Models
of Biological Objects Based on Serial Sections
Vincent Jasper Dercksen1, Cornelia Bru¨ß2, Detlev Stalling3, Sabine
Gubatz2, Udo Seiffert2, and Hans-Christian Hege1
1 Zuse Institute Berlin, Germany {dercksen,hege}@zib.de
2 Leibniz Institute of Plant Genetics and Crop Plant Research, Gatersleben,
Germany {bruess,gubatz,seiffert}@ipk-gatersleben.de
3 Mercury Computer Systems Inc., Berlin, Germany [email protected]
Summary. Wepresentasetofcoherentmethodsforthenearlyautomaticcreation
of 3D geometric models from large stacks of images of histological sections. Three-
dimensional surface models facilitate the visual analysis of 3D anatomy. They also
formabasisforstandardizedanatomicalatlasesthatallowresearcherstointegrate,
accumulate and associate heterogeneous experimental information, like functional
or gene-expression data, with spatial or even spatio-temporal reference. Models are
created by performing the following steps: image stitching, slice alignment, elastic
registration,imagesegmentationandsurfacereconstruction.Theproposedmethods
are to a large extent automatic and robust against inevitably occurring imaging
artifacts. The option of interactive control at most stages of the modeling process
complements automatic methods.
Key words: Geometry reconstruction, surface representations, registration,
segmentation, neural nets
1 Introduction
Three-dimensional models help scientists to gain a better understanding of
complex biomedical objects. Models provide fundamental assistance for the
analysisofanatomy,structure,functionanddevelopment.Theycanforexam-
plesupportphenotypingstudies[J.T06]toanswerquestionsabouttherelation
between genotypeandphenotype.Anothermajoraimistoestablish anatom-
ical atlases. Atlases enable researchers to integrate (spatial) data obtained by
different experiments on different individuals into one common framework. A
multitude of structural and functional properties can then jointly be visual-
izedandanalyzed,revealingnewrelationships.4Datlasescanprovideinsight
into temporal development and spatio-temporal relationships.
Weintendtoconstructhigh-resolution4Datlasesofdevelopingorganisms,
thatallowtheintegrationofexperimentaldata,e.g.geneexpressionpatterns.
4 V.J. Dercksen et al.
Such atlases can support the investigation of morphogenesis and gene expres-
sion during development. Their creation requires population-based averages
of 3D anatomical models representing different developmental stages. By in-
terpolating these averages the development over time can be visualized.
In this work we focus on the mostly automatic creation of individual 3D
modelsasanessentialsteptowardsastandardatlas.Individualmodelscanbe
createdfromstacksof2Dimagesofhistologicalsectionsorfromtrue3Dimage
data.Withhistologicalsectionshigherresolutionsarepossiblethanforexam-
plewith3DNuclearMagneticResonance(NMR)imaging.Furthermore,fluo-
rescentdyepenetrationproblems,whichcanoccurwhenimaging(thick)plant
tissue with for example (3D) Confocal Laser Scanning Microscopy (CLSM),
canbeavoided.Duringdataacquisitionoftenmultipleimagespersectionhave
to be created to achieve the desired resolution. These sub-images have to be
stitched in an initial mosaicing step. The 3D model construction then contin-
ues with the alignment of the image stack to restore the 3D coherence. In the
following segmentation step, the structures of interest need to be identified,
delimited and labeled. The model is completed by creating a polygonal sur-
face, marking the object boundary and separating the structures it consists
of. This sequence of nontrivial processing steps is here called the geometry
reconstruction pipeline (see Fig. 1).
Withourflexiblegeneral-purpose3Dvisualizationsystem[SWH05],three
highlyresolved3Dmodelsofgrainshaverecentlybeencreatedbasicallyman-
ually [S. 07] (see Fig. 2(a)). Experience gained during the highly interactive
modelinghoweverclearlyshowedtheneedfortheautomationandfacilitation
of the many repetitive, time-consuming, and work-intensive steps.
When creating such detailed models, the size of the data sets is a com-
plicating factor. Due to the high resolution of the images and the size of the
subject, the data sets frequently do not fit into the main memory of common
workstations and must therefore be processed out-of-core. Another problem
is that due to the cutting and handling, histological sections are susceptible
to imaging artifacts, like cracks, contrast differences and pollution. Very ro-
bust processing methods that are able to deal with such imperfect data are
therefore required.
Fig. 1. Fromphysicalgrainsto3Dmodels:processingstepsofthegeometryrecon-
struction pipeline.