[image 02771] 講演会(11/30)のご案内

[Hideo Saito]

皆様

北アイルランド Ulster UniversityのBryan W. 
Scotney教授（慶大特別招聘教授（国際））の講演会を東大生研で開催させて頂くことになりました．直前のご案内で恐縮ですがお知らせ致します．事前登録不要、参加費不要ですのでお気軽にご参加ください．よろしくお願いします．

斎藤英雄／慶大

------------------------------------------------------------
日時：  2017年11月30日(木) 17:30 - 18:30
場所：　東京大学生産技術研究所An棟4階An401&402
講演者：Bryan W. Scotney
         Professor of Informatics, Ulster University, UK
         Guest Professor (Global), Keio University

タイトル: Decision Support for Breast Cancer Units through Image-based 
Breast and Tumour Characterisation

Abstract:

This talk will describe recent work on Image-based Breast and Tumour 
Characterisation being carried out in the project DESIREE, funded by the 
European Union’s HORIZON 2020 
programme (under grant agreement No 
690238). Multidisciplinary Breast Units (BUs) were first introduced into 
the healthcare system in order to deal efficiently with breast cancer 
cases, setting guideline-based quality procedures, clinical decisions on 
cases based on consensus and a high standard of care. However, daily 
clinical practice and case presentation in BUs is hampered by the 
complexity of the disease and the ever-growing amount of patient and 
disease data available.
The DESIREE project is developing a decision support system (DSS) that 
incorporates experience from previous cases and outcomes, thus going 
beyond the limitations of existing guideline-based decision support 
systems. The DSS is based on a knowledge model in which patient cases 
are represented using a Digital Breast Cancer Patient (DBCP) model that 
incorporates information about the patient’s clinical history and 
diagnostic and therapeutic procedures in cycles that may last for many 
years. A vast amount of digital information is generated from diagnostic 
and therapeutic procedures, including medical imaging data from 
different modalities, biological and genetic data, novel diagnostic 
tests and biomarkers, risk factors and clinical trials. The DESIREE 
project aims to incorporate information from multiple sources in the 
DBCP model to be exploited prospectively for decision support. Within 
the project, automated analysis of medical images is being developed 
with a view to identifying specific biomarkers for image-based breast 
tissue characterization and image-based tumour characterization. Breast 
tissue characterisation is important in breast cancer risk assessment 
and has demonstrated predictive value for expected outcomes of some 
treatments.
Breast and pectoral muscle segmentation is an essential pre-processing 
step for the subsequent processes in computer-aided diagnosis systems. 
Estimating the breast and pectoral boundaries can be a difficult task, 
especially in mammograms, due to artifacts, homogeneity between the 
pectoral and breast regions, and low contrast. We describe an approach 
developed using a range of image processing techniques that can achieve 
high accuracy in this essential preprocessing step. We also show a 
method for automatic breast pectoral muscle segmentation in mediolateral 
oblique mammograms using Convolutional Neural Networks. For breast 
tissue characterization we show that methods based on Local Ternary 
Patterns (LTP) and Local Quinary Patterns (LQP) can be effective for 
breast density classification in mammograms. We develop a 
multi-resolution and multi-orientation approach and investigate the 
effects of neighbourhood topologies, dominant pattern selection, and the 
use of only the fibro-glandular disk region rather than the whole beast 
area.

Professor Bryan W. Scotney (BSc Mathematics, University of Durham, UK; 
PhD Mathematics, University of Reading, UK) is Professor of Informatics 
at the University of Ulster. He has over 300 publications, spanning a 
range of research interests in mathematical computation, especially in 
digital image processing and computer vision, pattern recognition and 
classification, statistical databases, reasoning under uncertainty, and 
applications to healthcare informatics, official statistics, biomedical 
and vision sciences, and telecommunications network management. He has 
collaborated widely with academic, government and commercial partners, 
and much of his work has been supported by funding from the European 
Union Framework Programmes and the UK Research Councils. Most recently 
he has been an investigator on the EPSRC NETWORK in Next Generation 
Networks Systems and Services (EP/F030118/1), the EPSRC–DST funded 
India-UK Advanced Technology Centre (IU-ATC) of Excellence in Next 
Generation Networks Systems and Services (EP/G051674/1 and 
EP/J016748/1), an ESRC-funded project on Design for Ageing Well 
(RES-353-25-0004), the SAVASA project on a Standards-based Approach to 
Video Archive Search and Analysis, funded by the EU FP7 Security 
programme, and a Principal Investigator on the recently completed EU FP7 
project: Security System for Language and Image Analysis (SLANDAIL). 
Currently he is an investigator on two EU H2020 projects: DESIREE: 
Decision Support and Information Management System for Breast Cancer 
(Feb 2016 – Jan 2019) and ASGARD - Analysis System for Gathered Raw Data 
(Sep 2016 – Mar 2020), working on breast tissue characterisation from 
medical images, and biometrics from video data (face recognition), 
respectively.
He was Director of Ulster University’s Computer Science Research 
Institute since its formation in 2005 until May 2015, and recently 
resumed this role in May 2017, with management responsibility for 50 
academic staff, approximately 30 contract research staff and 80 PhD 
students. In this capacity he led the University’s periodic UK Research 
Assessment submissions for Computer Science and Informatics, RAE2008 and 
REF2014. He was President of the Irish Pattern Recognition and 
Classification Society 2007-2014 (re-elected in 2010), and a member of 
the Governing Board of the International Association for Pattern 
Recognition (IAPR), 2007-2014.

以上です．