@article {873, title = {Comparison and evaluation of methods for liver segmentation from CT datasets}, journal = {IEEE TRANSACTIONS ON MEDICAL IMAGING}, volume = {28}, year = {2009}, note = {ScopusID: 68249121543doi: 10.1109/TMI.2009.2013851}, month = {Aug 2009}, pages = {1251 - 1265}, type = {Journal article}, address = {Price, K., Anything you can do, I can do better (no you can{\textquoteright}t) (1986) Comput. Vis. Graph. Image Process, 36 (2-3), pp. 387-391;S. G. Armato, G. McLennan, M. F. McNitt-Gray, C. R. Meyer, D. Yankelevitz, D. R. Aberle, C. I. Henschke, E. A. Hoffman, E. A. Ka}, abstract = {

This paper presents a comparison study between 10 automatic and six interactive methods for liver segmentation from contrast-enhanced CT images. It is based on results from the "MICCAI 2007 Grand Challenge" workshop, where 16 teams evaluated their algorithms on a common database. A collection of 20 clinical images with reference segmentations was provided to train and tune algorithms in advance. Participants were also allowed to use additional proprietary training data for that purpose. All teams then had to apply their methods to 10 test datasets and submit the obtained results. Employed algorithms include statistical shape models, atlas registration, level-sets, graph-cuts and rule-based systems. All results were compared to reference segmentations five error measures that highlight different aspects of segmentation accuracy. All measures were combined according to a specific scoring system relating the obtained values to human expert variability. In general, interactive methods reached higher average scores than automatic approaches and featured a better consistency of segmentation quality. However, the best automatic methods (mainly based on statistical shape models with some additional free deformation) could compete well on the majority of test images. The study provides an insight in performance of different segmentation approaches under real-world conditions and highlights achievements and limitations of current image analysis techniques. {\textcopyright} 2009 IEEE.

}, isbn = {0278-0062}, doi = {10.1109/TMI.2009.2013851}, author = {Tobias Heimann and Brahm Van Ginneken and Martin A Styner and Yulia Arzhaeva and Volker Aurich and Christian Bauer and Andreas Beck and Christoph Becker and Reinhardt Beichel and Gy{\"o}rgy Bekes and Fernando Bello and Gerd Binnig and Horst Bischof and Alexander Bornik and Peter MM Cashman and Ying Chi and Andres C{\'o}rdova and Benoit M Dawant and M{\'a}rta Fidrich and Jacob D Furst and Daisuke Furukawa and Lars Grenacher and Joachim Hornegger and Dagmar Kainm{\"u}ller and Richard I Kitney and Hidefumi Kobatake and Hans Lamecker and Thomas Lange and Jeongjin Lee and Brian Lennon and Rui Li and Senhu Li and Hans-Peter Meinzer and G{\'a}bor N{\'e}meth and Daniela S Raicu and Anne-Mareike Rau and Eva M Van Rikxoort and Mikael Rousson and L{\'a}szl{\'o} Rusk{\'o} and Kinda A Saddi and G{\"u}nter Schmidt and Dieter Seghers and Akinobi Shimizu and Pieter Slagmolen and Erich Sorantin and Grzegorz Soza and Ruchaneewan Susomboon and Jonathan M Waite and Andreas Wimmer and Ivo Wolf} } @inbook {1043, title = {Whole Body MRI Intensity Standardization}, booktitle = {Bildverarbeitung f{\"u}r die Medizin 2007}, series = {Informatik aktuell}, year = {2007}, note = {doi: 10.1007/978-3-540-71091-2_92}, month = {March 2007}, pages = {459 - 463}, publisher = {Springer-Verlag}, organization = {Springer-Verlag}, type = {Conference paper}, address = {M{\"u}nchen, Germany}, abstract = {

A major problem of segmentation of magnetic resonance images isthat intensities are not standardized like in computed tomography. This article deals with the correction of inter volume intensity differences that lead to a missing anatomical meaning of the observed gray values. We present a method for MRI intensity standardization of whole body MRI scans. The approach is based on the alignment of a learned reference and the current histogram. Each of these histograms is at least 2-d and represents two or more MRI sequences (e.g., T1- and T2-weighted images). From the matching a non-linear correction function is gained which describes a mapping between the intensity spaces and consequently adapts the image statistics to a known standard. As the proposed intensity standardization is based on the statistics of the data sets only, it is independent from spatial coherences or prior segmentations of the reference and newly acquired images. Furthermore, it is not designed for a particular application, body region or acquisition protocol. The method was evaluated on whole body MRI scans containing data sets acquired by T1/FL2D and T2/TIRM sequences. In order to demonstrate the applicability, examples from noisy and pathological image series acquired on a whole body MRI scanner are given.

}, isbn = {978-3-540-71090-5}, issn = {1431-472X}, doi = {10.1007/978-3-540-71091-2_92}, author = {Florian J{\"a}ger and L{\'a}szl{\'o} G{\'a}bor Ny{\'u}l and Bernd Frericks and Frank Wacker and Joachim Hornegger}, editor = {Alexander Horsch and Thomas Martin Deserno and Heinz Handels and Hans-Peter Meinzer and Thomas Tolxdorff} }