Medical Image Computing and Computer-Assisted Intervention – MICCAI 2007 pot

VIII PrefaceMICCAI 2007 Papers by Topic General Biological Image Computing 3% Computational Physiology 6% Computer Assisted Interventional Robotics 14% Computational Anatomy 8% General M

Lecture Notes in Computer Science 4792

Commenced Publication in 1973

Founding and Former Series Editors:

Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen

Anthony Maeder (Eds.)

Medical

Image Computing

and Computer-Assisted Intervention –

Volume Editors

Nicholas Ayache

INRIA, Asclepios Project-Team

2004 Route des Lucioles, 06902 Sophia-Antipolis, France

E-mail: nicholas.ayache@inria.fr

Sébastien Ourselin

Anthony Maeder

CSIRO ICT Centre, e-Health Research Centre

20/300 Adelaide St., Brisbane, Queensland 4000, Australia

E-mail: {sebastien.ourselin, anthony.maeder}@csiro.au

Library of Congress Control Number: 2007937392

CR Subject Classification (1998): I.5, I.4, I.3.5-8, I.2.9-10, J.3, J.6

LNCS Sublibrary: SL 6 – Image Processing, Computer Vision, Pattern Recognition,and Graphics

ISBN-10 3-540-75758-9 Springer Berlin Heidelberg New York

ISBN-13 978-3-540-75758-0 Springer Berlin Heidelberg New York

This work is subject to copyright All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in any other way, and storage in data banks Duplication of this publication

or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965,

in its current version, and permission for use must always be obtained from Springer Violations are liable

to prosecution under the German Copyright Law.

Springer is a part of Springer Science+Business Media

The 10th International Conference on Medical Imaging and Computer AssistedIntervention, MICCAI 2007, was held at the Brisbane Convention and ExhibitionCentre, South Bank, Brisbane, Australia from 29th October to 2nd November2007.

MICCAI has become a premier international conference in this domain, within-depth papers on the multidisciplinary fields of biomedical image computing,computer assisted intervention and medical robotics The conference brings to-gether biological scientists, clinicians, computer scientists, engineers, mathemati-cians, physicists and other interested researchers and offers them a forum toexchange ideas in these exciting and rapidly growing fields

The conference is both very selective and very attractive: this year we received

a record number of 637 submissions from 35 countries and 6 continents, fromwhich 237 papers were selected for publication Some interesting facts about thedistribution of submitted and accepted papers are shown graphically at the end

Pro-2 New key words regrouped within 7 new categories were introduced to scribe the content of the submissions and the expertise of the reviewers

de-3 Each submitted paper was assigned to 3 Program Committee members whoseresponsibility it was to assign each paper to 3 external experts (outside of theProgram Committee membership) who provided scores and detailed reports

in a double blind procedure

4 Program Committee members provided a set of normalized scores for thewhole set of papers for which they were responsible (typically 27 papers).They did this using the external reviews and their own reading of the pa-pers and had to complete missing reviews themselves Program Committeemembers eventually had to provide a recommendation for acceptance of thetop 35% of their assigned papers

5 During a 2 day meeting of about 20 members of the Program Committee

in Sophia-Antipolis, France, borderline papers were examined carefully andthe final set of papers was accepted to appear in the LNCS proceedings Atop list of about 100 papers was scrutinized to provide the Program Chairand Co-chair with a list of 54 potential podium presentations

6 From this list, the Program Chair and Co-chair selected 38 podium tations to create a program with a reasonable number of oral sessions andspread of content

presen-VI Preface

7 Because 199 excellent contributions would be presented as posters, it wasdecided in consultation with the MICCAI Society Board to augment thetime allocated to the poster sessions, and replace the oral poster teasers bycontinuous video teasers run on large screens during the conference.The selection procedure was very selective, and many good papers remainedamong the 400 rejected We received 9 factual complaints from the authors

of rejected papers A subcommittee of the Program Committee treated all ofthem equally, checking carefully that no mistake had been made during theselection procedure In a few cases, an additional review was requested from anindependent Program Committee member In the end, all the original decisionswere maintained, but some additional information was provided to the authors

to better explain the final decision

Seven MICCAI Young Scientist Awards were presented by the MICCAI ciety on the last day of the conference The selection was made before the con-ference by nominating automatically the 21 eligible papers with the highestnormalized scores (provided by the Program Committee during the reviewingprocedure), and regrouping them into the 7 main categories of the conference

So-A subgroup of the Program Committee had to vote to elect one paper out of 3

in each category

The 2007 MedIA-MICCAI Prize was offered by Elsevier to the first author of

an outstanding article in the special issue of the Medical Image Analysis Journaldedicated to the previous conference MICCAI 2006 The selection was organized

by the guest-editors of this special issue

We want to thank wholeheartedly all Program Committee members for theirexceptional work, as well as the numerous external expert reviewers (who arelisted on the next pages) We should also acknowledge the substantial contribu-tion made towards the successful execution of MICCAI 2007 by the BioMedicalImage Analysis Laboratory team at the CSIRO ICT Centre / e-Health ResearchCentre

It was our pleasure to welcome MICCAI 2007 attendees in Brisbane This wasthe first time the conference had been held in Australia, indeed only the secondtime outside of Europe/North America, the other being MICCAI 2002 in Japan.This trend will continue with MICCAI 2010, which is planned for Beijing Thevibrant sub-tropical river city of Brisbane with its modern style and world-classconference venue was a popular choice and a convenient point of departure fordelegates who took the opportunity while there to see more of the Australianoutback

We thank our two invited keynote speakers, Prof Peter Hunter from theBioengineering Institute at the University of Auckland, New Zealand, and Prof.Stuart Crozier from Biomedical Engineering at the University of Queensland,Brisbane, whose excellent presentations were a highlight of the conference Wealso acknowledge with much gratitude the contributions of Terry Peters, MIC-CAI 2007 General Co-Chair, whose strong connection with the MICCAI Societyand past MICCAI conferences proved invaluable to us We also note our thanks

to our sponsors, without whose financial assistance the event would have been afar lesser one.

We look forward to welcoming you to MICCAI 2008, to be held 4-8 September

in New York City, USA, and MICCAI 2009, scheduled to be held in London,UK

S´ebastien OurselinAnthony Maeder

VIII Preface

MICCAI 2007 Papers by Topic

General Biological Image Computing 3%

Computational Physiology 6%

Computer Assisted Interventional Robotics 14%

Computational Anatomy 8%

General Medical Image Computing 46%

None Specified 1%

Innovative Clinical and Biological Applications 11%

Neuroscience Image Computing 8%

Visualization and Interaction 3%

General Medical Image Computing

19 1 3 9 21 3 4

39 9

Extraction of visual features : texture, shape, connectivity, motion, etc Grid-enabled image processing algorithms Methodological tools for validation Morphometric and functional segmentation Non linear registration and fusion Other (General Medical Image Computing) PDEs and Level Sets methods

Processing X-ray, CT, MR (anatomical, functional, diffusion, spectroscopic), SPECT, Statistical image analysis

Computer Assisted Interventional Systems and

Robotics

4 4

17 9

Advanced Medical Robotics

Image-guided robotized intervention

Instrument & Patient Localization and

Tracking

Other (Computer Assisted Interventional

Systems and Robotics)

Fig 1 View at a glance of MICCAI 2007 accepted submissions based on the declared

primary keyword A total of 237 full papers were presented

Full paper submissions: 637

Asia

22 19 7 1 5 24

6 7 2

China

Hongg In a Iran IsraelJa n

Kore

a, South

Singore

Taiwan

Ger

many

Greece

Hungary Italy

Neth

erla

nds

Norway Portugal Sl

enia Spain

SwedenSwi

tzer

landTurke

y UK

North America

58 206

Canada USA

Others

1 31

1 2 2

Egypt

Austra

New Z

ealand Br il

Colo

mbia

Fig 2 Distribution of MICCAI 2007 submissions (637 in total) by continent

The MICCAI Young Scientist Award is a prize of US$500 awarded to the firstauthor (in person) for the best paper in a particular topic area, as judged byreviewing and presentation (oral or poster) At MICCAI 2007, up to 7 prizeswere available, in the topic areas publicised in the conference CFP:

1 General Medical Image Computing

2 Computer Assisted Intervention Systems and Robotics

3 Visualization and Interaction

4 General Biological and Neuroscience Image Computing

5 Computational Anatomy

6 Computational Physiology

7 Innovative Clinical and Biological Applications

All current first author students and early career scientists attending MICCAI

2007 were eligible The awards were announced and presented at the closingsession of the conference on Thursday, 1st November 2007

MICCAI 2005 Student Awards

Image Segmentation and Analysis: Pingkun Yan, “MRA Image Segmentation

with Capillary Active Contour”

Image Registration: Ashraf Mohamed, “Deformable Registration of Brain

Tu-mor Images via a Statistical Model of TuTu-mor Induced Deformation”

Computer-Assisted Interventions and Robotics: Henry C Lin, “Automatic

De-tection and Segmentation of Robot Assisted Surgical Motions”

Simulation and Visualization: Peter Savadjiev, “3D Curve Inference for

Dif-fusion MRI Regularization”

Clinical Application: Srinivasan Rajagopalan, “Schwarz Meets Schwann: Design

and Fabrication of Biomorphic Tissue Engineering Scaffolds”

MICCAI 2006 Student Awards

Image Segmentation and Registration: Delphine Nain, “Shape-Driven 3D

Seg-mentation Using Spherical Wavelets”

Image Analysis: Karl Sj¨ostrand, “The Entire Regularization Path for the port Vector Domain Description”

Sup-X MICCAI Young Scientist Awards

Simulation and Visualization: Andrew W Dowsey, “Motion-Compensated MR

Valve Imaging with COMB Tag Tracking and Super-Resolution Enhancement”

Computer-Assisted Interventions and Robotics: Paul M Novotny, “GPU Based

Real-Time Instrument Tracking with Three Dimensional Ultrasound”

Clincial Applications: Jian Zhang, “A Pilot Study of Robot-Assisted Cochlear

Implant Surgery Using Steerable Electrode Arrays”

The 2007 MedIA-MICCAI Prize

This prize is awarded each year by Elsevier to the first author of an outstandingarticle of the previous MICCAI conference, which is published in the MICCAIspecial issue of the Medical Image Analysis Journal

In 2006, the prize was awarded to T Vercauteren, first author of the article:Vercauteren, T., Perchant, A., Pennec, X., Malandain, G., Ayache, N.: Robustmosaicing with correction of motion distortions and tissue deformations for invivo fibered microscopy Med Image Anal 10(5), 673–692 (2006)

In 2005, the prize was awarded to D Burschka and M Jackowski who are thefirst authors of the articles:

Burschka, D., Li, M., Ishii, M., Taylor, R.H., Hager, G.D.: Scale invariant istration of monucular endoscopic images to CT-Scans for sinus surgery Med.Image Anal 9(5), 413–426 (2005)

reg-Jackowski, M., Kao, C.Y., Qiu, M., Constable, R.T., Staib, L.H.: White mattertractography by anisotropic wave front evolution and diffusion tensor imaging.Med Image Anal 9(5), 427–440 (2005)

Executive Committee

General Co-chair Terry Peters (Robarts Research Institute,

Canada)

Program Co-chair S´ebastien Ourselin (CSIRO, Australia)

Program Committee

Elsa Angelini (ENST, Paris, France)

Simon R Arridge (University College London, UK)

Leon Axel (University Medical Centre, USA)

Christian Barillot (IRISA, Rennes, France)

Margrit Betke (Boston University, USA)

Elizabeth Bullitt (University of North Carolina, Chapel Hill , USA)Albert Chung (Hong Kong University of Science and Technology, China)Ela Claridge (The University of Birmingham, UK)

Stuart Crozier (University of Queensland, Australia)

Christos Davatzikos (University of Pennsylvania, USA)

Marleen de Bruijne (University of Copenhagen, Denmark)

Rachid Deriche (INRIA, Sophia Antipolis, France)

Etienne Dombre (CNRS, Montpellier, France)

James S Duncan (Yale University, USA)

Gary Egan (Howard Florey Institute, Australia)

Randy Ellis (Queens University, Canada)

Gabor Fichtinger (Johns Hopkins University, USA)

Alejandro Frangi (Pompeu Fabra University, Barcelona, Spain)

Guido Gerig (University of North Carolina, Chapel Hill, USA)

Polina Golland (Massachusetts Institute of Technology, USA)

Miguel Angel Gonzalez Ballester (University of Bern, Switzerland)Richard Hartley (Australian National University, Australia)

David Hawkes (University College London, UK)

Pheng Ann Heng (The Chinese University of Hong Kong, China)

Robert Howe (Harvard University, USA)

Peter Hunter (The University of Auckland, New Zealand)

Tianzi Jiang (The Chinese Academy of Sciences, China)

Sarang Joshi (University of Utah, USA)

XII Organization

Leo Joskowicz (The Hebrew University of Jerusalem, Israel)

Hans Knustsson (Linkoping University, Sweden)

Rasmus Larsen (Technical University of Denmark, Denmark)

Boudewijn Lelieveldt (Leiden University Medical Centre, Netherlands)

Cristian Lorenz (Philips, Hamburg, Germany)

Frederik Maes (Katholieke Universiteit Leuven, Belgium)

Gregoire Malandain (INRIA, Sophia Antipolis, France)

Jean-Francois Mangin (CEA, SHFJ, Orsay, France)

Dimitris Metaxas (Rutgers University, New Jersey, USA)

Kensaku Mori (Mori Nagoya University, Japan)

Nassir Navab (TUM, Munich, Germany)

Poul Nielsen (The University of Auckland, New Zealand)

Wiro Niessen (Erasmus Medical School, Rotterdam, Netherlands)

Alison Noble (Oxford University, UK)

Jean-Christophe Olivo-Marin (Institut Pasteur, Paris, France)

Nikos Paragios (Ecole Centrale de Paris, France)

Xavier Pennec (INRIA, Sophia Antipolis, France)

Franjo Pernus (University of Ljubljana, Slovenia)

Josien Pluim (University Medical Center, Utrecht, Netherlands)

Jean-Baptiste Poline (CEA, SHFJ, Orsay, France)

Jerry L Prince (Johns Hopkins University, USA)

Richard A Robb (Mayo Clinic, College of Medicine, Rochester, Minnesota, USA)Daniel Rueckert (Imperial College, London, UK)

Tim Salcudean (The University of British Columbia, Canada)

Yoshinobu Sato (Osaka University, Japan)

Achim Schweikard (Institute for Robotics and Cognitive Systems, Germany)Pengcheng Shi (Hong Kong University of Science and Technology, China)Stephen Smith (Oxford University, UK)

Lawrence Staib (Yale University, USA)

Colin Studholme (University of California, San Francisco, USA)

Gabor Sz´ekely (ETH, Zurich, Switzerland)

Russell Taylor (Johns Hopkins University, USA)

Jean-Philippe Thiran (EPFL, Lausanne, Switzerland)

Jocelyne Troccaz (CNRS, Grenoble, France)

Bram van Ginneken (University Medical Center, Utrecht, Netherlands)

Koen Van Leemput (HUS, Helsinki, Finland)

Baba Vemuri (University of Florida, USA)

Simon Warfield (Harvard University, USA)

Sandy Wells (Massachusetts Institute of Technology, USA)

Carl-Fredrik Westin (Westin Harvard University, USA)

Ross Whitaker (University of Utah, USA)

Chenyang Xu (Siemens Corporate Research, USA)

Guang Zhong Yang (Imperial College, London, UK)

MICCAI Board

Nicholas Ayache, INRIA, Sophia Antipolis, France

Alan Colchester, University of Kent, Canterbury, UK

James Duncan, Yale University, New Haven, Connecticut, USA

Gabor Fichtinger, Johns Hopkins University, Baltimore, Maryland, USAGuido Gerig, University of North Carolina, Chapel Hill, North Carolina, USAAnthony Maeder, University of Queensland, Brisbane, Australia

Dimitris Metaxas, Rutgers University, Piscataway Campus, New Jersey, USANassir Navab, Technische Universit¨at, Munich, Germany

Mads Nielsen, IT University of Copenhagen, Copenhagen, Denmark

Alison Noble, University of Oxford, Oxford, UK

Terry Peters, Robarts Research Institute, London, Ontario, Canada

Richard Robb, Mayo Clinic College of Medicine, Rochester, Minnesota, USA

MICCAI Society

Society Officers

President and Board Chair Alan Colchester

Executive Secretary Nicholas Ayache

Boston, USA

Local Planning Committee

Sponsors and Exhibitors Oscar Acosta-Tamayo

Registration and VIP Liaison Tony Adriaansen

Tutorials and Workshops Pierrick Bourgeat

Technical Proceedings Support Jason Dowling

Professional Society Liaison Brian Lovell

Student & Travel Awards Olivier Salvado

XIV Organization

Sponsors

CSIRO ICT Centre

e-Health Research Centre

Northern Digital, Inc

Medtronic, Inc

The Australian Pattern Recognition Society

CSIRO Preventative Health Flagship

Siemens Corporate Research

Chui, Chee Kong

Chui, Yim Pan

Corso, JasonCotin, StephaneCoulon, OlivierCoupe, PierrickCrouch, JessicaCrum, WilliamD’Agostino, EmilianoDam, Erik

Dan, IppeitaDarkner, SuneDauguet, JulienDavis, BradDawant, Benoit

De Craene, MathieuDeguchi, DaisukeDehghan, EhsanDelingette, Herv´eDeLorenzo, ChristineDeng, Xiang

Desai, JaydevDescoteaux, MaximeDey, Joyoni

Diamond, Solomon GilbertDieterich, Sonja

Dijkstra, JoukeDillenseger, Jean-LouisDiMaio, Simon

Dirk, LoeckxDodel, SilkeDornheim, JanaDorval, ThierryDouiri, AbdelDuan, QiDuay, Val´erieDubois, Marie-DominiqueDuchesne, Simon

Dupont, PierreDurrleman, StanleyEcabert, OlivierEdwards, PhilipEggers, HolgerEhrhardt, JanEl-Baz, Ayman

Gu, LixuGuerrero, JulianGuimond, AlexandreHager, Gregory DHahn, HorstHall, MattHamarneh, GhassanHan, Xiao

Hansen, KlausHanson, DennisHarders, MatthiasHata, Nobuhiko

He, Huiguang

He, YongHeckemann, RolfHeintzmann, RainerHellier, Pierre

Ho, HonPongHodgson, AntonyHoffmann, KennethHolden, MarkHoldsworth, DavidHolmes, DavidHornegger, JoachimHorton, Ashley

Hu, Mingxing

Hu, QingmaoHua, JingHuang, JunzhouHuang, XiaoleiHuang, HengHutton, BrianIglesias, Juan EugenioJ¨ager, Florian

Jain, Ameet

Lazar, MarianaLee, Su-LinLee, BryanLeemans, AlexanderLekadir, KarimLenglet, ChristopheLepore, NatashaLeung, K Y EstherLevman, Jacob

Li, Kang

Li, Shuo

Li, MingLiao, ShuLiao, RuiLieby, PauletteLikar, BostjanLin, FuchunLinguraru, Marius GeorgeLinte, Cristian

Liu, YanxiLiu, HuafengLiu, JiminLohmann, GabrieleLoog, MarcoLorenzen, PeterLueders, EileenLum, Mitchell

Ma, BurtonMacq, BenoitMadabhushi, AnantManduca, ArmandoManniesing, RashindraMarchal, MaudMarchesini, RenatoMarsland, Stephen

´Olafsd´ottir, HildurOliver, ArnauOlsen, Ole FoghOost, ElcoOtake, YoshitoOzarslan, EvrenPadfield, DirkPadoy, NicolasPalaniappan, KannappanPang, Wai-Man

Papademetris, XeniosPapadopoulo, Th´eoPatriciu, AlexandruPatronik, NicholasPavlidis, IoannisPechaud, MickaelPeine, WilliamPeitgen, Heinz-OttoPekar, VladimirPenney, GraemePerperidis, DimitriosPeters, Terry

Petit, YvanPham, DzungPhillips, RogerPichon, EricPitiot, AlainPizer, StephenPlaskos, ChristopherPock, ThomasPohl, Kilian MariaPoignet, PhilippePoupon, CyrilPrager, RichardPrastawa, MarcelPrause, GuidoPreim, BernhardPrima, SylvainQian, ZhenQian, XiaoningRaaymakers, BasRadaelli, AlessandroRajagopal, Vijayaraghavan

San Jose Estepar, Raul

Sanchez Castro, Francisco Javier

Sundar, HariSzczerba, DominikSzilagyi, LaszloTagare, HemantTalbot, HuguesTalib, HaydarTalos, Ion-FlorinTanner, ChristineTao, XiaodongTarte, SegoleneTasdizen, TolgaTaylor, ZeikeTaylor, JonathanTek, HuseyinTendick, FrankTerzopoulos, DemetriTh´evenaz, PhilippeThirion, BertrandTieu, KinhTodd-Pokropek, AndrewTodman, Alison

van Assen, Hans

van de Ville, Dimitri

van der Bom, Martijn

van der Geest, Rob

van Rikxoort, Eva

van Walsum, Theo

von Berg, Jens

von Lavante, Etienne

von Siebenthal, Martin

Westenberg, MichelWestermann, RuedigerWhitcher, BrandonWiemker, RafaelWiest-Daessle, NicolasWigstrom, LarsWiles, AndrewWink, OnnoWong, KenWong, KennethWong, StephenWong, Tien-TsinWong, WilburWood, BradfordWood, FionaWorsley, KeithW¨orz, StefanWˇsrn, Heinz

Wu, JueXia, YanXie, Jun

Xu, Sheng

Xu, YeXue, HuiXue, ZhongYan, PingkunYang, KingYang, LinYang, YihongYaniv, ZivYeo, Boon ThyeYeung, Sai-KitYogesan, KanagasingamYoshida, Hiro

Young, AlistairYoung, Stewart

Yu, YangYue, NingYuen, Shelten

Table of Contents – Part II

Computer Assisted Intervention and Robotics - II

Real-Time Tissue Tracking with B-Mode Ultrasound Using Speckle

and Visual Servoing . 1

Alexandre Krupa, Gabor Fichtinger, and Gregory D Hager

Intra-operative 3D Guidance in Prostate Brachytherapy Using a

Non-isocentric C-arm . 9

Ameet K Jain, A Deguet, Iulian I Iordachita,

Gouthami Chintalapani, J Blevins, Y Le, E Armour,

C Burdette, Danny Y Song, and Gabor Fichtinger

A Multi-view Opto-Xray Imaging System: Development and First

Application in Trauma Surgery . 18

Joerg Traub, Tim Hauke Heibel, Philipp Dressel,

Sandro Michael Heining, Rainer Graumann, and Nassir Navab

Towards 3D Ultrasound Image Based Soft Tissue Tracking: A

Transrectal Ultrasound Prostate Image Alignment System . 26

Michael Baumann, Pierre Mozer, Vincent Daanen, and

Jocelyne Troccaz

A Probabilistic Framework for Tracking Deformable Soft Tissue in

Minimally Invasive Surgery . 34

Peter Mountney, Benny Lo, Surapa Thiemjarus,

Danail Stoyanov, and Guang Zhong-Yang

Precision Targeting of Liver Lesions with a Needle-Based Soft Tissue

Navigation System . 42

L Maier-Hein, F Pianka, A Seitel, S.A M¨ uller, A Tekbas,

M Seitel, I Wolf, B.M Schmied, and H.-P Meinzer

Dynamic MRI Scan Plane Control for Passive Tracking of Instruments

and Devices . 50

Simon P DiMaio, E Samset, Gregory S Fischer,

Iulian I Iordachita, Gabor Fichtinger, Ferenc A Jolesz, and

Clare MC Tempany

Design and Preliminary Accuracy Studies of an MRI-Guided

Transrectal Prostate Intervention System . 59

Axel Krieger, Csaba Csoma, Iulian I Iordachita, Peter Guion,

Anurag K Singh, Gabor Fichtinger, and Louis L Whitcomb

Thoracoscopic Surgical Navigation System for Cancer Localization in

Collapsed Lung Based on Estimation of Lung Deformation . 68

Masahiko Nakamoto, Naoki Aburaya, Yoshinobu Sato, Kozo Konishi,

Ichiro Yoshino, Makoto Hashizume, and Shinichi Tamura

Visualization and Interaction

Clinical Evaluation of a Respiratory Gated Guidance System for Liver

Punctures . 77

S.A Nicolau, Xavier Pennec, Luc Soler, and Nicholas Ayache

Rapid Voxel Classification Methodology for Interactive 3D Medical

Image Visualization . 86

Qi Zhang, Roy Eagleson, and Terry M Peters

Towards Subject-Specific Models of the Dynamic Heart for

Image-Guided Mitral Valve Surgery . 94

Cristian A Linte, Marcin Wierzbicki, John Moore, Stephen H Little,

G´ erard M Guiraudon, and Terry M Peters

pq-space Based Non-Photorealistic Rendering for Augmented Reality 102

Mirna Lerotic, Adrian J Chung, George P Mylonas, and

Guang-Zhong Yang

Eye-Gaze Driven Surgical Workflow Segmentation . 110

A James, D Vieira, Benny Lo, Ara Darzi, and Guang-Zhong Yang

Neuroscience Image Computing - I

Prior Knowledge Driven Multiscale Segmentation of Brain MRI . 118

Ayelet Akselrod-Ballin, Meirav Galun, John Moshe Gomori,

Achi Brandt, and Ronen Basri

Longitudinal Cortical Registration for Developing Neonates . 127

Hui Xue, Latha Srinivasan, Shuzhou Jiang, Mary A Rutherford,

A David Edwards, Daniel Rueckert, and Joseph V Hajnal

Regional Homogeneity and Anatomical Parcellation for fMRI Image

Classification: Application to Schizophrenia and Normal Controls . 136

Feng Shi, Yong Liu, Tianzi Jiang, Yuan Zhou, Wanlin Zhu,

Jiefeng Jiang, Haihong Liu, and Zhening Liu

Probabilistic Fiber Tracking Using Particle Filtering . 144

Fan Zhang, Casey Goodlett, Edwin Hancock, and Guido Gerig

SMT: Split and Merge Tractography for DT-MRI . 153

U˘ gur Bozkaya and Burak Acar

Table of Contents – Part II XXV

Tract-Based Morphometry . 161

Lauren J O’Donnell, Carl-Fredrik Westin, and Alexandra J Golby

Towards Whole Brain Segmentation by a Hybrid Model . 169

Zhuowen Tu and Arthur W Toga

Computational Anatomy - II

A Family of Principal Component Analyses for Dealing with Outliers . 178

J Eugenio Iglesias, Marleen de Bruijne, Marco Loog,

Fran¸ cois Lauze, and Mads Nielsen

Automatic Segmentation of Articular Cartilage in Magnetic Resonance

Images of the Knee . 186

Jurgen Fripp, Stuart Crozier, Simon K Warfield, and

S´ ebastien Ourselin

Automated Model-Based Rib Cage Segmentation and Labeling in CT

Images . 195

Tobias Klinder, Cristian Lorenz, Jens von Berg,

Sebastian P.M Dries, Thomas B¨ ulow, and J¨ orn Ostermann

Efficient Selection of the Most Similar Image in a Database for Critical

Structures Segmentation . 203

Olivier Commowick and Gr´ egoire Malandain

Unbiased White Matter Atlas Construction Using Diffusion Tensor

Images . 211

Hui Zhang, Paul A Yushkevich, Daniel Rueckert, and James C Gee

Innovative Clinical and Biological Applications - II

Real-Time SPECT and 2D Ultrasound Image Registration . 219

Marek Bucki, Fabrice Chassat, Francisco Galdames, Takeshi Asahi,

Daniel Pizarro, and Gabriel Lobo

A Multiphysics Simulation of a Healthy and a Diseased Abdominal

Aorta . 227

Robert H.P McGregor, Dominik Szczerba, and G´ abor Sz´ ekely

New Motion Correction Models for Automatic Identification of Renal

Transplant Rejection . 235

Ayman S El-Baz, Georgy Gimel’farb, and Mohamed A El-Ghar

Detecting Mechanical Abnormalities in Prostate Tissue Using FE-Based

Image Registration . 244

Patrick Courtis and Abbas Samani

Real-Time Fusion of Ultrasound and Gamma Probe for Navigated

Localization of Liver Metastases . 252

Thomas Wendler, Marco Feuerstein, Joerg Traub, Tobias Lasser,

Jakob Vogel, Farhad Daghighian, Sibylle I Ziegler, and Nassir Navab

Fast and Robust Analysis of Dynamic Contrast Enhanced MRI

Datasets . 261

Olga Kubassova, Mikael Boesen, Roger D Boyle, Marco A Cimmino,

Karl E Jensen, Henning Bliddal, and Alexandra Radjenovic

Spectroscopic and Cellular Imaging

Functional Near Infrared Spectroscopy in Novice and Expert

Surgeons – A Manifold Embedding Approach . 270

Daniel Richard Leff, Felipe Orihuela-Espina, Louis Atallah,

Ara Darzi, and Guang-Zhong Yang

A Hierarchical Unsupervised Spectral Clustering Scheme for Detection

of Prostate Cancer from Magnetic Resonance Spectroscopy (MRS) . 278

Pallavi Tiwari, Anant Madabhushi, and Mark Rosen

A Clinically Motivated 2-Fold Framework for Quantifying and

Classifying Immunohistochemically Stained Specimens . 287

Bonnie Hall, Wenjin Chen, Michael Reiss, and David J Foran

Cell Population Tracking and Lineage Construction with Spatiotemporal

Context . 295

Kang Li, Mei Chen, and Takeo Kanade

Spatio-Temporal Registration

Spatiotemporal Normalization for Longitudinal Analysis of Gray

Matter Atrophy in Frontotemporal Dementia . 303

Brian Avants, Chivon Anderson, Murray Grossman, and

James C Gee

Population Based Analysis of Directional Information in Serial

Deformation Tensor Morphometry . 311

Colin Studholme and Valerie Cardenas

Non-parametric Diffeomorphic Image Registration with the Demons

Algorithm . 319

Tom Vercauteren, Xavier Pennec, Aymeric Perchant, and

Nicholas Ayache

Three-Dimensional Ultrasound Mosaicing . 327

Christian Wachinger, Wolfgang Wein, and Nassir Navab

Table of Contents – Part II XXVII

General Medical Image Computing - III

Automated Extraction of Lymph Nodes from 3-D Abdominal CT

Images Using 3-D Minimum Directional Difference Filter . 336

Takayuki Kitasaka, Yukihiro Tsujimura, Yoshihiko Nakamura,

Kensaku Mori, Yasuhito Suenaga, Masaaki Ito, and Shigeru Nawano

Non-Local Means Variants for Denoising of Diffusion-Weighted and

Diffusion Tensor MRI . 344

Nicolas Wiest-Daessl´ e, Sylvain Prima, Pierrick Coup´ e,

Sean Patrick Morrissey, and Christian Barillot

Quantifying Calcification in the Lumbar Aorta on X-Ray Images . 352

Lars A Conrad-Hansen, Marleen de Bruijne, Fran¸ cois Lauze,

L´ aszl´ o B Tank´ o, Paola C Pettersen, Qing He, Jianghong Chen,

Claus Christiansen, and Mads Nielsen

Physically Motivated Enhancement of Color Images for Fiber

Endoscopy . 360

Christian Winter, Thorsten Zerfaß, Matthias Elter,

Stephan Rupp, and Thomas Wittenberg

Signal LMMSE Estimation from Multiple Samples in MRI and

C.J Rose, S Mills, J.P.B O’Connor, G.A Buonaccorsi,

C Roberts, Y Watson, B Whitcher, G Jayson, A Jackson, and

G.J.M Parker

Improving Temporal Fidelity in k-t BLAST MRI Reconstruction 385

Andreas Sigfridsson, Mats Andersson, Lars Wigstr¨ om,

John-Peder Escobar Kvitting, and Hans Knutsson

Segmentation and Classification of Breast Tumor Using Dynamic

Contrast-Enhanced MR Images . 393

Yuanjie Zheng, Sajjad Baloch, Sarah Englander,

Mitchell D Schnall, and Dinggang Shen

Automatic Whole Heart Segmentation in Static Magnetic Resonance

Image Volumes . 402

Jochen Peters, Olivier Ecabert, Carsten Meyer, Hauke Schramm,

Reinhard Kneser, Alexandra Groth, and J¨ urgen Weese

PCA-Based Magnetic Field Modeling: Application for On-Line MR

Temperature Monitoring . 411

G Maclair, B Denis de Senneville, M Ries, B Quesson,

P Desbarats, J Benois-Pineau, and C.T.W Moonen

A Probabilistic Model for Haustral Curvatures with Applications to

Colon CAD . 420

John Melonakos, Paulo Mendon¸ ca, Rahul Bhotka, and Saad Sirohey

LV Motion Tracking from 3D Echocardiography Using Textural and

Structural Information . 428

Andriy Myronenko, Xubo Song, and David J Sahn

A Novel 3D Multi-scale Lineness Filter for Vessel Detection . 436

H.E Bennink, H.C van Assen, G.J Streekstra, R ter Wee,

J.A.E Spaan, and Bart M ter Haar Romeny

Live-Vessel: Extending Livewire for Simultaneous Extraction of

Optimal Medial and Boundary Paths in Vascular Images . 444

Kelvin Poon, Ghassan Hamarneh, and Rafeef Abugharbieh

A Point-Wise Quantification of Asymmetry Using Deformation Fields:

Application to the Study of the Crouzon Mouse Model . 452

Hildur ´ Olafsd´ ottir, Stephanie Lanche, Tron A Darvann,

Nuno V Hermann, Rasmus Larsen, Bjarne K Ersbøll,

Estanislao Oubel, Alejandro F Frangi, Per Larsen, Chad A Perlyn,

Gillian M Morriss-Kay, and Sven Kreiborg

Object Localization Based on Markov Random Fields and Symmetry

Interest Points . 460

Ren´ e Donner, Branislav Micusik, Georg Langs, Lech Szumilas,

Philipp Peloschek, Klaus Friedrich, and Horst Bischof

2D Motion Analysis of Long Axis Cardiac Tagged MRI . 469

Ting Chen, Sohae Chung, and Leon Axel

MCMC Curve Sampling for Image Segmentation . 477

Ayres C Fan, John W Fisher III, William M Wells III,

James J Levitt, and Alan S Willsky

Automatic Centerline Extraction of Irregular Tubular Structures Using

Probability Volumes from Multiphoton Imaging . 486

A Santamar´ıa-Pang, C.M Colbert, P Saggau, and

Table of Contents – Part II XXIX

Is a Single Energy Functional Sufficient? Adaptive Energy Functionals

and Automatic Initialization . 503

Chris McIntosh and Ghassan Hamarneh

A Duality Based Algorithm for TV-L1-Optical-Flow Image

Registration . 511

Thomas Pock, Martin Urschler, Christopher Zach,

Reinhard Beichel, and Horst Bischof

Deformable 2D-3D Registration of the Pelvis with a Limited Field of

View, Using Shape Statistics . 519

Ofri Sadowsky, Gouthami Chintalapani, and Russell H Taylor

Segmentation-driven 2D-3D Registration for Abdominal Catheter

Ben Glocker, Nikos Komodakis, Nikos Paragios, Christian Glaser,

Georgios Tziritas, and Nassir Navab

Similarity Metrics for Groupwise Non-rigid Registration . 544

Kanwal K Bhatia, Joseph V Hajnal, Alexander Hammers, and

Daniel Rueckert

A Comprehensive System for Intraoperative 3D Brain Deformation

Recovery . 553

Christine DeLorenzo, Xenophon Papademetris, Kenneth P Vives,

Dennis D Spencer, and James S Duncan

Bayesian Tracking of Tubular Structures and Its Application to Carotid

Arteries in CTA . 562

Michiel Schaap, Rashindra Manniesing, Ihor Smal,

Theo van Walsum, Aad van der Lugt, and Wiro Niessen

Automatic Fetal Measurements in Ultrasound Using Constrained

Probabilistic Boosting Tree . 571

Gustavo Carneiro, Bogdan Georgescu, Sara Good, and

Dorin Comaniciu

Quantifying Effect-Specific Mammographic Density . 580

Jakob Raundahl, Marco Loog, Paola C Pettersen, and Mads Nielsen

Revisiting the Evaluation of Segmentation Results: Introducing

Confidence Maps . 588

Christophe Restif

Error Analysis of Calibration Materials on Dual-Energy

Mammography . 596

Xuanqin Mou and Xi Chen

Computer Assisted Intervention and Robotics - III

A MR Compatible Mechatronic System to Facilitate Magic Angle

Experiments in Vivo 604

Haytham Elhawary, Aleksandar Zivanovic, Marc Rea,

Zion Tsz Ho Tse, Donald McRobbie, Ian Young, Martyn Paley,

Brian Davies, and Michael Lamp´ erth

Variational Guidewire Tracking Using Phase Congruency . 612

Greg Slabaugh, Koon Kong, Gozde Unal, and Tong Fang

Endoscopic Navigation for Minimally Invasive Suturing . 620

Christian Wengert, Lukas Bossard, Armin H¨ aberling, Charles Baur,

G´ abor Sz´ ekely, and Philippe C Cattin

On Fiducial Target Registration Error in the Presence of Anisotropic

Noise . 628

Burton Ma, Mehdi Hedjazi Moghari, Randy E Ellis, and

Purang Abolmaesumi

Rotational Roadmapping: A New Image-Based Navigation Technique

for the Interventional Room . 636

Markus Kukuk and Sandy Napel

Bronchoscope Tracking Without Fiducial Markers Using Ultra-tiny

Electromagnetic Tracking System and Its Evaluation in Different

Environments . 644

Kensaku Mori, Daisuke Deguchi, Kazuyoshi Ishitani,

Takayuki Kitasaka, Yasuhito Suenaga, Yosihnori Hasegawa,

Kazuyoshi Imaizumi, and Hirotsugu Takabatake

Online Estimation of the Target Registration Error for n-Ocular

Optical Tracking Systems . 652

Tobias Sielhorst, Martin Bauer, Oliver Wenisch,

Gudrun Klinker, and Nassir Navab

Assessment of Perceptual Quality for Gaze-Contingent Motion

Stabilization in Robotic Assisted Minimally Invasive Surgery . 660

George P Mylonas, Danail Stoyanov, Ara Darzi, and

Table of Contents – Part II XXXI

Multi-criteria Trajectory Planning for Hepatic Radiofrequency

Ablation . 676

Claire Baegert, Caroline Villard, Pascal Schreck, and Luc Soler

General Biological Imaging Computing

A Bayesian 3D Volume Reconstruction for Confocal Micro-rotation

Cell Imaging . 685

Yong Yu, Alain Trouv´ e, and Bernard Chalemond

Bias Image Correction Via Stationarity Maximization . 693

T Dorval, A Ogier, and A Genovesio

Toward Optimal Matching for 3D Reconstruction of Brachytherapy

Seeds . 701

Christian Labat, Ameet K Jain, Gabor Fichtinger, and

Jerry L Prince

Alignment of Large Image Series Using Cubic B-Splines Tessellation:

Application to Transmission Electron Microscopy Data . 710

Julien Dauguet, Davi Bock, R Clay Reid, and Simon K Warfield

Quality-Based Registration and Reconstruction of Optical Tomography

Volumes . 718

Wolfgang Wein, Moritz Blume, Ulrich Leischner,

Hans-Ulrich Dodt, and Nassir Navab

Simultaneous Segmentation, Kinetic Parameter Estimation, and

Uncertainty Visualization of Dynamic PET Images . 726

Ahmed Saad, Ben Smith, Ghassan Hamarneh, and Torsten M¨ oller

Neuroscience Image Computing - II

Nonlinear Analysis of BOLD Signal: Biophysical Modeling,

Physiological States, and Functional Activation . 734

Zhenghui Hu and Pengcheng Shi

Effectiveness of the Finite Impulse Response Model in Content-Based

fMRI Image Retrieval . 742

Bing Bai, Paul Kantor, and Ali Shokoufandeh

Sources of Variability in MEG . 751

Wanmei Ou, Polina Golland, and Matti H¨ am¨ al¨ ainen

Customised Cytoarchitectonic Probability Maps Using Deformable

Registration: Primary Auditory Cortex . 760

Lara Bailey, Purang Abolmaesumi, Julian Tam, Patricia Morosan,

Rhodri Cusack, Katrin Amunts, and Ingrid Johnsrude

Segmentation of Q-Ball Images Using Statistical Surface Evolution . 769

Maxime Descoteaux and Rachid Deriche

Evaluation of Shape-Based Normalization in the Corpus Callosum for

White Matter Connectivity Analysis . 777

Hui Sun, Paul A Yushkevich, Hui Zhang, Philip A Cook,

Jeffrey T Duda, Tony J Simon, and James C Gee

Accuracy Assessment of Global and Local Atrophy Measurement

Techniques with Realistic Simulated Longitudinal Data . 785

Oscar Camara, Rachael I Scahill, Julia A Schnabel,

William R Crum, Gerard R Ridgway, Derek L.G Hill, and

Nick C Fox

Combinatorial Optimization for Electrode Labeling of EEG Caps . 793

Micka¨ el P´ echaud, Renaud Keriven, Th´ eo Papadopoulo, and

Jean-Michel Badier

Computational Anatomy - III

Analysis of Deformation of the Human Ear and Canal Caused by

Mandibular Movement . 801

Sune Darkner, Rasmus Larsen, and Rasmus R Paulsen

Shape Registration by Simultaneously Optimizing Representation and

Transformation . 809

Yifeng Jiang, Jun Xie, Deqing Sun, and Hungtat Tsui

Landmark Correspondence Optimization for Coupled Surfaces . 818

Lin Shi, Defeng Wang, Pheng Ann Heng, Tien-Tsin Wong,

Winnie C.W Chu, Benson H.Y Yeung, and Jack C.Y Cheng

Mean Template for Tensor-Based Morphometry Using Deformation

Tensors . 826

Natasha Lepor´ e, Caroline Brun, Xavier Pennec, Yi-Yu Chou,

Oscar L Lopez, Howard J Aizenstein, James T Becker,

Arthur W Toga, and Paul M Thompson

Shape-Based Myocardial Contractility Analysis Using Multivariate

Wai-Man Pang, Jing Qin, Yim-Pan Chui, Tien-Tsin Wong,

Kwok-Sui Leung, and Pheng Ann Heng

Table of Contents – Part II XXXIII

Interactive Contacts Resolution Using Smooth Surface

Representation . 850

J´ er´ emie Dequidt, Julien Lenoir, and St´ ephane Cotin

Using Statistical Shape Analysis for the Determination of Uterine

Deformation States During Hydrometra . 858

M Harders and G´ abor Sz´ ekely

Predictive K-PLSR Myocardial Contractility Modeling with Phase

Contrast MR Velocity Mapping . 866

Su-Lin Lee, Qian Wu, Andrew Huntbatch, and Guang-Zhong Yang

A Coupled Finite Element Model of Tumor Growth and

Vascularization . 874

Bryn A Lloyd, Dominik Szczerba, and G´ abor Sz´ ekely

Innovative Clinical and Biological Applications - III

Autism Diagnostics by 3D Texture Analysis of Cerebral White Matter

Gyrifications . 882

Ayman S El-Baz, Manuel F Casanova, Georgy Gimel’farb,

Meghan Mott, and Andrew E Switala

3-D Analysis of Cortical Morphometry in Differential Diagnosis of

Parkinson’s Plus Syndromes: Mapping Frontal Lobe Cortical Atrophy

in Progressive Supranuclear Palsy Patients . 891

Duygu Tosun, Simon Duchesne, Yan Rolland, Arthur W Toga,

Marc V´ erin, and Christian Barillot

Tissue Characterization Using Fractal Dimension of High Frequency

Ultrasound RF Time Series . 900

Mehdi Moradi, Parvin Mousavi, and Purang Abolmaesumi

Towards Intra-operative 3D Nuclear Imaging: Reconstruction of 3D

Radioactive Distributions Using Tracked Gamma Probes . 909

Thomas Wendler, Alexander Hartl, Tobias Lasser, Joerg Traub,

Farhad Daghighian, Sibylle I Ziegler, and Nassir Navab

Instrumentation for Epidural Anesthesia . 918

King-wei Hor, Denis Tran, Allaudin Kamani, Vickie Lessoway, and

Robert Rohling

Small Animal Radiation Research Platform: Imaging, Mechanics,

Control and Calibration . 926

Mohammad Matinfar, Owen Gray, Iulian I Iordachita,

Chris Kennedy, Eric Ford, John Wong, Russell H Taylor, and

Peter Kazanzides

Proof of Concept of a Simple Computer–Assisted Technique for

Correcting Bone Deformities . 935

Burton Ma, Amber L Simpson, and Randy E Ellis

Global Registration of Multiple Point Sets: Feasibility and Applications

in Multi-fragment Fracture Fixation . 943

Mehdi Hedjazi Moghari and Purang Abolmaesumi

Precise Estimation of Postoperative Cup Alignment from Single

Standard X-Ray Radiograph with Gonadal Shielding . 951

Guoyan Zheng, Simon Steppacher, Xuan Zhang, and Moritz Tannast

Fully Automated and Adaptive Detection of Amyloid Plaques in

Stained Brain Sections of Alzheimer Transgenic Mice . 960

Abdelmonem Feki, Olivier Teboul, Albertine Dubois,

Bruno Bozon, Alexis Faure, Philippe Hantraye, Marc Dhenain,

Benoit Delatour, and Thierry Delzescaux

Non-rigid Registration of Pre-procedural MR Images with

Intra-procedural Unenhanced CT Images for Improved Targeting of

Tumors During Liver Radiofrequency Ablations . 969

N Archip, S Tatli, P Morrison, Ferenc A Jolesz,

Simon K Warfield, and S Silverman

Author Index 979

Real-Time Tissue Tracking with B-Mode Ultrasound Using Speckle and Visual Servoing

Alexandre Krupa1, Gabor Fichtinger2, and Gregory D Hager2

1 IRISA - INRIA Rennes, Francealexandre.krupa@irisa.fr

2

Engineering Research Center, Johns Hopkins University, USA

{gabor,hager}@cs.jhu.edu 

Abstract We present a method for real-time tracking of moving soft

tissue with B-mode ultrasound (US) The method makes use of thespeckle information contained in the US images to estimate the in-planeand out-of-plane motion of a fixed target relative to the ultrasound scanplane The motion information is then used as closed-loop feedback to arobot which corrects for the target motion The concept is demonstratedfor translation motions in an experimental setup consisting of an ultra-sound speckle phantom, a robot for simulating tissue motion, and a robotthat performs motion stabilization from US images This concept showspromise for US-guided procedures that require real-time motion trackingand compensation

of moving soft tissues during US scanning or to synchronize the insertion of aneedle into a moving target during biopsy or local therapy

In this paper, we present a system that is capable of fully automatic, time tracking and motion compensation of a moving soft tissue target using

real-a sequence of B-mode ultrreal-asound imreal-ages Contrreal-ary to prior work in this real-arereal-a,which has relied on segmenting structures of interest [1,2], we make direct use ofthe speckle information contained in the US images While US speckle is usuallyconsidered to be noise from an imaging point of view, it in fact results from the

Fig 1 (left) Experimental decorrelation curves obtained by measuring the correlation

value between 25 patches of B-scan I 1 and their corresponding patches in B-scan I 2

along the elevation distance d (right)

coherent reflection of microscopic structures contained in soft tissue As such,

it is spatially coherent Furthermore, an US beam is several mm wide As aresult, there is substantial overlap between US scan planes with small lateraldisplacements and, therefore, substantial correlation of the speckle informationbetween successive images Speckle correlation occurs for both in-plane and out-of-plane motion, thereby making it possible to track both out-of plane and in-plane motion, and raising the possibility of calculating full 6-DOF relative pose

of speckle patches

Initially, speckle information has been used to estimate multi-dimensional flow

in 2D ultrasound image ([3]) Recently several authors ([4,5]) have publishedspeckle decorrelation techniques to allow freehand 3D US scanning without aposition sensor on the US probe Their techniques depend on experimentallycalibrating speckle decorrelation curves from real soft tissues and/or specklesimulating phantoms These curves (Fig 1) are obtained by capturing B-mode

images at known distances d along the elevation direction (i.e orthogonal to the

image plane) and measuring the normalized correlation coefficients for a finitenumber of rectangular patches fixed in the images The imaging procedure thenentails capturing an US stream by moving the probe in a given direction Therelative in-plane and out-of-plane position between each image is then estimated,off-line, from the estimated elevation distances from at least 3 non-collinearpatches in the image plane These distances are computed from the calibrateddecorrelation curves using the measured inter-patch correlation value for eachimage patch

In our experimental scenario, we also perform an offline calibration procedure

to relate speckle decorrelation to elevation motion However, we subsequentlyservo the US probe to track a user-selected B-scan target in a fully automatic,online manner The 6-DOF motion of the target B-scan is extracted by an es-timation method using the speckle information and an image region trackingalgorithm based on grey level intensity A visual servoing scheme is then used

Real-Time Tissue Tracking with B-Mode Ultrasound 3

to control the probe displacement Section 2 presents the methods used to tract 6-DOF rigid motion of the target B-scan image The visual servoing controllaws are developed in section 3 and section 4 presents first results obtained fromex-vivo experiments where only translation motions are considered

The overall tracking problem is to minimize the relative position between thecurrent B-scan (denoted by a Cartesian frame{c}) and a target B-scan (denoted

by a Cartesian frame{i}) The full 6 DOF target plane position can be

decom-posed by two successive homogeneous transformations:cHi = cHp pHi where

cHp andpHi describing the in-plane and out-of-plane displacement of the get, respectively Note that{p} corresponds to an intermediate “virtual” plane The in-plane displacement corresponds to the translations x and y along the X and Y axes of the current image plane and the angular rotation γ around the Z

tar-axis (orthogonal to the image), such that:

We use a classical template tracking technique [6] to extract the in-plane motion

parameters x, y, γ This information is then used to relate the image coordinates

of patches in the two images for the purposes of estimating out-of-plane motionusing speckle decorrelation

To extract the out-of-plane motion, we use the Gaussian model introduced

in [4] From experimental observations (Fig 1), we found that the elevationdistance between a patch in the target plane and the corresponding patch in thecurrent image can be estimated by ˆd =

−2ˆσ2ln(ρ), where ˆ σ = 0.72 mm is the

mean resolution cell width (identified from experimental decorrelation curves)

To compute the full out-of-plane motion, we compute the elevation distance

of a grid of patches (25 in our current system), and fit a plane to this data

How-ever, the Gaussian model does not detect the sign of the elevation distance for

a given patch Thus, we employ the following algorithm to estimate the plane position of the target plane with respect to the virtual plane{p} We first

out-of-set a random sign on each inter-patch distance and estimate (with a least-squarealgorithm) an initial position of the target plane using these signs We thenuse the iterative algorithm we presented in [7] to determine the correct signeddistances and the associated plane This algorithm, which minimizes the least-square error of the estimated target plane, converges to two stable solutions thatare symmetrical around plane{p} The two solutions correspond to the positive and negative elevation distances z, respectively Note that from one solution

we can easily determine the second By formulating the out-of-plane relative

position as a combination of a translation z along the Z axis of plane {p} and

two successive rotations α, β around the Y and X axes of {p}, we obtain the

following homogeneous transformation matrix for out-of-plane motion:

ˆ

θ = −asin(ˆb) and (-) indicates the solution corresponding to ˆz < 0 with

ˆ

α = atan( −ˆa/ˆc), ˆθ = −asin(−ˆb) Here (ˆa, ˆb, ˆc) is the normal vector of the

estimated target plane that is obtained for the solution ˆz > 0 The subscript ˆ

denotes values provided by the template tracking and plane estimation methods

It will be purposely dropped in the next of the paper for clarity of presentation.This method works only locally about the target region due to the rapid rate

of speckle decorrelation with out-of-plane motion Therefore, in order to increasethe range of convergence, we augment the basic algorithm with a FIFO buffer

of intermediate planes{i} between the target {t} and current plane {c} These

planes, which are acquired online as the probe moves, are chosen to be closeenough to be well “speckle correlated” and thus provide a “path” of ultrasoundimages that can be traced back to the target

The complete algorithm summarizing our method for extracting target planeposition is described in Fig 2 (for positive elevation distances) and Fig 3(for negative elevation distances.) At initialization, the target plane is cap-

tured in the initial B-scan image and stored in a FIFO buffer (plane) ing with index i = 0 The current image is also stored as the target image (imageref erence = currentplane) A small negative elevation displacement is

start-then applied to the probe in order to obtain an initial positive elevation distance

z[0] ≥ s > 0 of plane[0] with respect to the current B-scan plane Here s is a small

threshold distance fixed to guarantee speckle correlation between US images.The algorithm goes to the case of positive elevation distance The array index is

then incremented and an intermediate plane is stored (plane[i] = currentplane)

with the homogeneous matrix iHi −1 = cHi −1(+) describing the position of

plane[i − 1] with respect to plane[i] and given by (3) Each time an

intermedi-ate plane is added, the target image used by the in-plane motion tracker is also

updated (imageref erence = currentplane) After initialization, the

configura-tion of planes corresponds to case 1 in Fig 2, where the target plane posiconfigura-tion

iscHt= cHi(+) 1i kHk −1 Now, we suppose that the target plane moves forsome reason By computing (3) for cHi and cHi −1, we can: 1) determine the

consistent pair of solutions that express the current plane relative to plane[i] and plane[i − 1], 2) determine which of cases 1, 2 or 3 is valid and 3) compute

the target elevation position cH accordingly As shown, the three cases are:

Real-Time Tissue Tracking with B-Mode Ultrasound 5

Fig 2 (top) possible planes configurations and (bottom) process used to manage the

intermediates planes when the target elevation distance is positive

1) if the current plane moves a distance s beyond the top of the FIFO array,

then a new intermediate plane is added or 2) if the current plane is between thetop two planes of the FIFO array, then no change occurs, or 3) if the elevationdistance decreases, then the last intermediate plane is removed from the FIFOarray In the latter case, a special situation arises when there are only two planes

(i = 1) in the array In this case, if the absolute value of the target elevation distance reaches the threshold s, then the algorithm switches to the second mode

described in Fig 3 which is the symmetric logic for negative elevations For thismode, the possible configurations of planes are illustrated by cases 4 to 6 in Fig

3 The algorithm switches back to the first mode when the target plane elevationposition becomes positive again

Now, as the position of the B-scan target with respect to the current planehas been estimated, we move the robot (holding the probe) in order to followthe target plane In our approach, a 3D visual servoing control scheme is used

to minimize the relative position between the current and target planes The

error vector is the 6 dimensional pose vector x = (tP T, θuT)T describing the

Fig 3 (top) possible planes configurations and (bottom) process used to manage the

intermediates planes when the target elevation distance is negative

position of the current plane frame{c} with respect to the target plane frame {t} Here tP cis the translation vector obtained directly from the 4th column of

t , and θu is the angle-axis representation of the rotation tRc [8]

The variation of x is related to the velocity screw v = (v x , v y , v z , ω x , ω y , ω z)T

of the ultrasound probe by ˙x = Lsv In visual servoing, Lsis called the tion matrix and is given in this case by (cf [9]):

where I3 is the 3× 3 identity matrix and [u] × is the skew matrix of the vector

preproduct linked with u The visual servoing task (cf [9]) can then be expressed

as a regulation to zero of the pose x and is performed by applying the following control law: v =−λL −1

s x where λ is the proportional coefficient involved for a

exponential convergence

We have tested the motion stabilization method on 2-DOF motions combining

a translation along the image X axis (in-plane translation) and elevation Z axis

Real-Time Tissue Tracking with B-Mode Ultrasound 7

time (s)

Tracking error (mm)

x (in−plane)

z (out−plane)

Fig 4 (top) experimental setup (bottomleft) evolution of the robots positions

-(bottom-right) tracking error

(out-of-plane translation) The experimental, setup, shown in Fig 4, consists

of two X-Z Cartesian robots fixed and aligned on an optical table The firstrobot provides a ground truth displacement for an US speckle phantom Thesecond robot holds a transrectal 6.5 Mhz US transducter and is controlled asdescribed above to track a target plane The US image is 440× 320 pixels with

resolution of 0.125 mm/pixel A laptop computer (Pentium IV 2 Ghz) capturesthe US stream at 10 fps, extracts the target plane position by using a grid

of 25 patches and computes the velocity control vector applied to the probeholding robot The plots in Fig 4 show the evolution of the robots positionsand the tracking error when sinusoidal motions (magnitude of 30 mm on eachaxis) were applied to the phantom The dynamic tracking error was below 3 mmfor in-plane translation and 3.5 mm for the elevation translation This error isattributed the dynamics of the target motion, time delays in the control scheme,and the dynamics of the probe holding robot These errors could be reduced if aprediction of its variation was introduced into the control law by some methodsuch as Kalman filter or generalized predictive controller [10] Adopting recent

methods [11] for more accurate and efficient identification of fully developedspeckle patches should also improve on tracking performance and may allowestimation of relative motion between different soft tissue elements In order todetermine the static accuracy of the tracking robotic task, we applied a set of 140random positions to the phantom by using ramp trajectories while tracking thetarget plane by the robotized probe When the probe stabilized at a position, thephantom was held motionless for 2 seconds and the locations of the two robotswere recorded We recorded a static error of 0.0219±0.05 mm (mean ± standard

deviation) for the in-plane tracking and 0.0233±0.05 mm for the out-of-plane

tracking, which is close to the positioning accuracy of the robots (± 0.05 mm).

In conclusion, results obtained from 2-DOF in-plane and out-of-plane motionsdemonstrated the potential of our approach We are presently adding rotationalstages to the robots to experimentally validate full 6-DOF motion tracking andvisual servoing capabilities of the current algorithm described in this paper

References

1 Abolmaesumi, P., Salcudean, S.E., Zhu, W.H., Sirouspour, M., DiMaio, S.: guided control of a robot for medical ultrasound IEEE Trans Robotics and Au-tomation 18, 11–23 (2002)

Image-2 Hong, J., Dohi, T., Hashizume, M., Konishi, K., Hata, N.: An ultrasound-drivenneedle insertion robot for percutaneous cholecystostomy Physics in Medicine andBiology 49(3), 441–455 (2004)

3 Bohs, L.N., Geiman, B.J., Anderson, M.E., Gebhart, S.C., Trahey, G.E.: Speckletracking for multi-dimensional flow estimation Ultrasonics 28(1-8), 369–375 (2000)

4 Gee, A.H., Housden, R.J., Hassenpflug, P., Treece, G.M., Prager, R.W.: less freehand 3D ultrasound in real tissues: Speckle decorrelation without fullydeveloped speckle Medical Image Analysis 10(2), 137–149 (2006)

Sensor-5 Chang, R.-F., Wu, W.-J., Chen, D.-R., Chen, W.-M., Shu, W., Lee, J.-H., Jeng,L.-B.: 3-D US frame positioning using speckle decorrelation and image registration.Ultrasound in Med & Bio 29(6), 801–812 (2003)

6 Hager, G.D., Belhumeur, P.N.: Efficient region tracking with parametric models ofgeometry and illumination IEEE Transactions on Pattern Analysis and MachineIntelligence 20(10), 1025–1039 (1998)

7 Krupa, A., Fichtinger, G., Hager, G.D.: Full Motion Tracking in Ultrasound UsingImage Speckle Information and Visual Servoing In: ICRA 2007 IEEE Int Conf

on Robotics and Automation, Roma, Italy, IEEE Computer Society Press, LosAlamitos (2007)

8 Craig, J.J.: Introduction to Robotics: Mechanics and Control, 2nd edn Wesley, London, UK (1989)

Addison-9 Chaumette, F., Hutchinson, S.: Visual Servo Control, Part I: Basic Approaches.IEEE Robotics and Automation Magazine 13(4), 82–90 (2006)

10 Ginhoux, R., Gangloff, J., de Mathelin, M., Soler, L., Sanchez, M.M.A., Marescaux,J.: Active Filtering of Physiological Motion in Robotized Surgery Using PredictiveControl IEEE Transactions on Robotics 21(1), 67–79 (2005)

11 Rivaz, H., Boctor, E., Fichtinger, G.: Ultrasound Speckle Detection Using Low der Moments In: IEEE International Ultrasonics Symposium, Vancouver, Canada,IEEE Computer Society Press, Los Alamitos (2006)