Sunday, 24 September 2017

Yves-Rémi VAN EYCKE

From Image

Jump to: navigation, search
YR.jpg

Contact

mail:yveycke _AT_ ulb.ac.be
tel.:+32 2 650 93 56
tel.:+32 2 650 26 91
fax:+32 2 650 22 98

ULB - LISA CP165/57
50, Av. F.Roosevelt
1050 Bruxelles
Belgium

Research

Developing image analysis and machine learning tools to quantify data extracted from whole slide (project page:Quantitative immunohistochemical analysis) and fluorescence microscopy images. Developing image registration tools to perform multi-biomarker analysis on whole slide images (colocalization and coexpression).

Publications

Papers

Y. Van Eycke, J. Allard, I. Salmon, O. Debeir, C. Decaestecker,
Image processing in digital pathology: an opportunity to solve inter-batch variability of immunohistochemical staining
Scientific Reports, Vol. 7, 2017
Bibtex
Bibtex : info:hdl:2013/246463
Note : SCOPUS: ar.j
Abstract : Immunohistochemistry (IHC) is a widely used technique in pathology to evidence protein expression in tissue samples. However, this staining technique is known for presenting inter-batch variations. Whole slide imaging in digital pathology offers a possibility to overcome this problem by means of image normalisation techniques. In the present paper we propose a methodology to objectively evaluate the need of image normalisation and to identify the best way to perform it. This methodology uses tissue microarray (TMA) materials and statistical analyses to evidence the possible variations occurring at colour and intensity levels as well as to evaluate the efficiency of image normalisation methods in correcting them. We applied our methodology to test different methods of image normalisation based on blind colour deconvolution that we adapted for IHC staining. These tests were carried out for different IHC experiments on different tissue types and targeting different proteins with different subcellular localisations. Our methodology enabled us to establish and to validate inter-batch normalization transforms which correct the non-relevant IHC staining variations. The normalised image series were then processed to extract coherent quantitative features characterising the IHC staining patterns.


X. Moles Lopez, P. Barbot, Y. Van Eycke, L. Verset, A. Trepant, L. Larbanoix, I. Salmon, C. Decaestecker,
Registration of whole immunohistochemical slide images: an efficient way to characterize biomarker colocalization.
Journal of the American Medical Informatics Association, Vol. 22, 1, pp. 86-99, 2015
Bibtex
Bibtex : info:hdl:2013/177042
Note : JOURNAL ARTICLE
Abstract : Extracting accurate information from complex biological processes involved in diseases, such as cancers, requires the simultaneous targeting of multiple proteins and locating their respective expression in tissue samples. This information can be collected by imaging and registering adjacent sections from the same tissue sample and stained by immunohistochemistry (IHC). Registration accuracy should be on the scale of a few cells to enable protein colocalization to be assessed.


Conferences

Y. Van Eycke, J. Allard, M. Derock, I. Salmon, O. Debeir, C. Decaestecker,
Image normalization for quantitative immunohistochemistry in digital pathology
2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI), pp. 795 - 798, Prague, Czech Republic, 2016
Bibtex
Bibtex : info:hdl:2013/233019
Note : SCOPUS: cp.p
Abstract : We propose to adapt to immunohistochemistry (IHC) some methods proposed to normalize images from histological slices stained with hematoxylin-eosin (H&E). Our final aim is to provide a coherent quantitative characterization of IHC biomarkers across different IHC batches with possible staining variations. In contrast to H&E, IHC staining strongly varies with the tissue analyzed and the protein targeted, making image normalization challenging. To solve this problem, we added in each IHC batch a slice from a reference tissue microarray (TMA) and then digitalized it to establish an inter-batch normalization transform. A comparison of two methods adapted to the specificity of IHC-stained slides evidences some normalization requirements to make valid IHC biomarker quantification across different staining batches.


L. Lebrun, D. Milowich, M. Le Mercier, J. Allard, Y. Van Eycke, T. Roumeguere, C. Decaestecker, I. Salmon, S. Rorive,
UCA1 Overexpression is a New Independent Prognostic Marker in Bladder Cancer Associated with Increased Patient Survival
Conference: Joint Meeting of the British Division of the International Academy of Pathology and the Pathological Society of Great Britain & Ireland(9th: 28 June – 1 July 2016: Nottingham), 2016
Bibtex
Bibtex : info:hdl:2013/242095
Note : Language of publication: en


A. Trepant, N. D'Haene, J. Allard, Y. Van Eycke, C. Decaestecker, I. Salmon, P. Demetter,
Vascular Insulin-Like Growth Factor Receptor Type 2 (IGF2R) Expression is Upregulated in Malignant Tumours
Conference: Joint Meeting of the British Division of the International Academy of Pathology and the Pathological Society of Great Britain & Ireland(9th: 28 June – 1 July 2016: Nottingham), 2016
Bibtex
Bibtex : info:hdl:2013/242086
Note : Language of publication: en


Y. Van Eycke, O. Debeir, L. Verset, P. Demetter, I. Salmon, C. Decaestecker,
Automated Tissue Microarray Image Processing in Digital Pathology.
Proceedings of the Fifth joint WIC/IEEE SP Symposium on Information Theory and Signal Processing in the Benelux, Brussels, Belgium, 2015
Bibtex
Bibtex : info:hdl:2013/199489
Note : Language of publication: en


Y. Van Eycke, O. Debeir, L. Verset, P. Demetter, I. Salmon, C. Decaestecker,
High-Throughput Analysis of Tissue-Based Biomarkers in Digital Pathology
EMBC'15 Proceedings (IEEE Engineering in Medicine and Biology Society), pp. 7732 - 7735, Milan, 2015
Bibtex
Bibtex : info:hdl:2013/205314
Note : Language of publication: en


Y. Van Eycke, X. Moles Lopez, J. Allard, M. Derock, S. Rorive, I. Salmon, C. Decaestecker,
Multiresolution registration of whole slide images to evidence and quantify virtual colocalization of tissue-based biomarkers
Conference: Digital Pathology Congress(4-5 December 2014: London), 2014
Bibtex
Bibtex : info:hdl:2013/193430
Note : Language of publication: en
Abstract : Extracting relevant information from actors involved in complex biological processes (such as cancers or treatment responses) requires to target different antigens simultaneously. Multichromogenic (brightfield) immunohistochemistry (IHC) suffers from limitations that notably prevent the analysis of proteins expressed in the same cellular compartment. We thus developed an alternative based on the analysis of adjacent, or serial, tissue sections on which different proteins were targeted by means of standard IHC. For this analysis, we developed a multiresolution method for registering serial slide images. This method uses the pyramidal and multimodal registration framework of the elastix software to optimize two parametric registrations. The first, low-resolution registration, is applied on the 1X equivalent magnification images (4,000 by 3,000 pixels). The result of this first step was then used to initialize high-resolution registrations independently performed on 20X equivalent fields of view. Our method shows accuracy levels compatible with biomarker colocalization characterization. Indeed, registration error on serial slides was evaluated to be at most between 20 µm and 80 µm, depending respectively of the presence or absence of histological structures in the tissue (e.g. in colonic tumor as compared to brain gliomas). In the latter situation, a sequential IHC technique applied on the same slide can be usefully employed. This “Sequential Immunoperoxidase Labeling and Erasing” (SIMPLE) method is based on cycles of staining/digitization/erasing, where after IHC staining and slide digitization, staining is erased through an antibody elution technique. We improved the original SIMPLE method and successfully utilized it to identify antigens expressed in the same cellular compartment of high-grade glioma samples. We tested our registration on the virtual slides so obtained and achieved very good results, i.e. about 5 µm of registration error. We then implemented a method to extract biomarker colocalization measurements taking the level of registration error into account and validated our complete procedure by comparison to colocalization information obtained by means of double staining (with different cell locations).