May 2015 - Issue #4
CominLabs at a glance
CominLabs is a Labex involving ten partner labs distributed over Brittany-Nantes, gathering about 500 researchers. Its scientific scope ranges from optics and electronics to over-the-Web applications, with a focus on networks, security and privacy, connected things, the social Web, ICT for health, multimedia, ICT & social sciences, ICT for education. Due to this distributed nature and wide scientific scope, CominLabs decided to support specific actions aligned with its strategic program for research, education, and innovation. Thus, CominLabs runs a program of highly ambitious multi-lab and multi-disciplinary projects. Thirty four research and education projects, including the CominWeb platform, are now in operation as the result of eight successive project calls in the period 2012-2016. Several projects involve Labex MER and Labex Centre Henri Lebesgue (maths). Through its leadership and International Advisory Committee, CominLabs seeds, selects, monitors and manages its projects, way beyond what a traditional funding program does, thus offering a novel and unparalleled service to its research and education community. The second edition of CominLabs Days was held in March 2015 with more than 200 attendees including our international experts. CominLabs activities benefit from a high-quality cooperative platform (CominWeb) and maintains a well-documented web site. CominLabs can thus be seen as an exploratory structure about how to manage networked research. Relations have been established with SATT Ouest-Valorisation and IRT b<>com.
Two first research projects ended fall 2015. POSEIDON project dealt with protection (security policies) of outsourced or mutualized data and content, with an application to cloud and peer-to-peer networks. PREDICTIVE project was concerned with predictive models for patient-personalized treatment management, and adaptive radiotherapy for prostate cancer. These are samples of exciting research that would not exist without CominLabs.
CominLabs Activities
Events
Pasts events
June 2, 2015 Hearing of the Labex by an international jury
On June 2, 2015, Cominlabs was heared by an international jury in Paris for the mid-term evaluation of the 171 labex managed by ANR. CominLabs organization, activities and achievements were presented by Albert Benveniste, Patrick Bouthemy and Ramesh Pyndiah. Based on this hearing and the mid-term report sent on March 2015, the international jury for the ICT domain issued its evaluation report in July 2015. The evaluation was clearly positive: “a number of interdisciplinary projects of high international quality”, “no specific weaknesses have been identified”, “the research community landscape in Bretagne has been deeply changed”, “increasing the chances of sustainability though the formation of tracks is promising”. The jury recommendations are in line with our roadmap for 2016-19 regarding innovation and industrial collaboration, education and international activities. As for the jury comment on of the uselessness of developing yet another MOOC platform, there was probably a lack of presentation or explanation. CominLabs does not develop its own MOOC platforms, but, on the contrary, the partners of the Cominlabs education projects have established contacts with national platforms (France Digital University –FUN in French-, ClassCode, Eduthèque Portal of the Ministry of Education) with a view to transfer technological modules realized in these projects.
June 24, seminar « SHS et numérique » organized in Rennes
On June 24, 2015, a one-day workshop on “Social Sciences and Digital Sciences” was organized by Eric Jamet, Maryline Boizard and Patrick Bouthemy in Inria Rennes premises. The overall goal was to strengthen the involvement of social-sciences research teams in CominLabs and their interactions with ICT research teams, to encourage proposals for the 2016 call for projects. The workshop gathered 50 attendees, included 7 presentations and time for exchange. Addressed topics were related to ergonomics, psychology, law, economics, and uses of digital technology.
January 5, Workshop oranized by Yann Busnel and Patrick Bouthemy in collaboration with DGA-MI and ENSAI
On January 5, 2016, a one-day workshop was organized by Yann Busnel and Patrick Bouthemy in collaboration with DGA-MI and ENSAI in Inria Rennes premises. The overall goal was to cross-fertilize maths, computer science, data processing aspects and application needs related to big data analysis and management, with a view to build one or more proposals to forthcoming calls. The workshop gathered 30 attendees, 12 CominLabs teams were represented. It ended up with the launch of a proposal to ASTRID call (ANR-DGA), and possibly of two working groups.
Upcoming events for 2016
FADEx (French-American Doctoral Exchange)
CominLabs will participate in the FADEx (French-American Doctoral Exchange) event, devoted this year to the scientific domain of cyber-physical systems (CPS) and jointly organized in Grenoble, Paris and Rennes on July 4-8, 2016. Patrick Bouthemy will present CominLabs on July 6 in Paris. CominLabs has provided financial support for this event. See here for more information.
CominLabs Days
The next CominLabs Days will be held on November 28-29-30 in Inria Rennes premises.
CominWeb plateform
The CominWeb CominWeb platform is built on top of Liferay, a framework for constructing Web portals, collaborative platforms, and social networks on a single platform. CominWeb hosts the CominLabs web site. The current version dated Jan 2015 has the following features. CominWeb offers advanced search services (LookinLabs) relying on big data technologies, see below. CominWeb offers also an intranet with the following services, globally or for each CominLabs group, all services stemming directly from Liferay:
- A document warehouse
- A wiki
- A Readme, a FAQ, a forum
- Public pages for reporting activities
For this version, access is managed by using the classical techniques offered by Liferay, namely roles and assignment of rights to roles. People are manually registered together with their role. Registering CominLabs people was cumbersome. Updates and modifications would require a full time web master, something that we want to avoid.
LookingLabs
The main recent developments focused on the LookinLabs service. LookinLabs offers competence mapping for CominLabs through the following features:
- LookinLabs responds to scientific semantic queries (who’s doing what on what? What’s the profile of Albert Benveniste, who are researchers most similar to him, etc.)
- The responses rely on semantic analyses, it’s not just word matching.
- The service requires zero-maintenance and is always to date.
To achieve this, a data base of all publications of CominLabs researchers is exploited using data mining techniques (title, authors, abstract, and meta-data). No input is expected from CominLabs researchers besides their personal link to standard bibliographical data bases, including IEEE Explore, DBLP, HAL, and more. These data bases are then regularly and automatically crawled, which keeps the document data base to date.
The publications of each author related to that topic are made accessible. The following open source packages were used for text mining
Tuning these algorithms was by no means trivial. Additional pre- and post-processing was necessary. The service was deployed January 2016 and is still under improvements.
Plans for 2016 and beyond
We will continue improving the quality of the service based on user’s feedback.
Short term: LookinLabs4HAL
We are currently deploying LookinLabs for the entire HAL French data base covering all disciplines (about 1 million documents). When this will be done, the following services will be provided to HAL users:
- Extend to all HAL authors what we currently have for CominLabs researchers. Since HAL contains more meta-data than most bibliographical data-bases, the corresponding version of LookinLabs is being upgraded to exploit these meta-data.
- Handle laboratories and teams the same way we are currently handling individuals (returning a list of labs working on the queried topic, providing profiles for labs).
For this extension we expect the following difficulties: there are homonym authors and a same lab may be referenced in several different ways.
Mid term plans
Two directions can be considered – we did not prioritize them yet.
A first direction consists in searching for other uses of the LookinLabs concept, e.g., for large companies or networks of companies. We would then exploit in-house data bases of documents. The following difficulties are expected. First, in-house information may not be available as well written documents such as publications; quite often in industry, information is stored in slideware with emphasis on graphics, not reports with emphasis on text. Second, in-house documents may not be directly linked to individuals; nor may they reflect the competencies of their authors. Finally, issues of data protection or security will certainly arise (these could possibly be handled, we think). Depending on these issues, this track will be either developed or abandoned.
A second direction consists in deploying the activity monitor for which we have a mockup. This service crawls the flow of email of a community (in our case, each CominLabs project) and mines it to discover mentions of activities of interest (meetings...) and attachments of interest (draft papers under writing). So far we have delayed the deployment and experimentation of this service. The reason for this is that project members do not seem to systematically use their alias when exchanging information related to their project. As a consequence the flow of collected email was too thin in our first experiments.
Research
Seven new research projects were selected in February 2016: BBC, BIGCLIN, INTERCOM, M5HSTIA, PRIVGEN, PROFILE, RAMPART.
BBC : Wireless Interconnect Network on chip BRoadcast-Based parallel Computing
Partners:
UMR 6285 LAB-STICC- Laboratoire des sciences et Techniques de l’Information, de la Communication et de la Connaissance (UBO BREST/ UBS LORIENT),
INRIA Rennes-Bretagne Atlantique-IRISA UMR CNRS 6074
External collaborator : CEA-LETI Grenoble
The evolution of microelectronics and requirement in data rate for High Performance Computing (HPC), involve large number of computation resources and faster components (e.g. processor, memories) to support the application needs. Adding more execution resources in a chip leads to increase the need for efficient communication media between them and in this context, the introduction of new kinds of interconnects becomes one of the majors challenges for the next MPSoC (Multiprocessor System-on-Chip). Today, the manycore architectures are the standard and they imply impressive gain in the domains of HPC and servers but also in the area of embedded systems. Applications in all these categories are greedy for parallelism and integration progress will allow the increase of numbers of cores on chips. So, this trend will continue but only if solutions can be found to overcome the interconnect bottleneck that now increase the cost of exchange information in term of crosstalk, energy consumption, data rate limitations, useless area and limited flexibility. The challenge is now to provide new communication medium between cores, chips and PCB boards.
In this context, the aim of the “BBC” (on-chip wireless Broadcast-Based parallel Computing) project is to evaluate the feasibility of using wireless link inside Chips and to define new paradigms which could be associated to these radio communications. Using wireless communications easily enables broadcast capabilities for Wireless Networks on Chip (WiNoC) and new management techniques for memories and parallelism. So, these new paradigms will be defined and evaluated during this project. The key elements which will be taken into account, will concern the improvement of power consumption, the estimation of achievable data rates, the flexibility and the reconfigurability, the size reduction and the easiness of parallelism and memory hierarchy management.
Thanks to the use of wireless links associated with CDMA (Code Division Multiple Access) access techniques and new broadcast capabilities, limitations due to interconnect can be largely reduced. So, two complementary originalities of this proposal are:
- the evaluation of the contribution of RF-radio link for the intra-chips interconnects,
- the definition of new opportunities for parallelism management and concurrent memory accesses.
Furthermore, the objective of this project will be to study how this specific communication medium can be combined with some other possible solutions, such as classical electrical or optical NoCs which could provide interesting characteristics. In particular, optical and wireless communications can provide very attractive solutions but maybe not in the same context. So, this project will provide us the possibility to answer to the question:
"In which cases the RF wireless links are attractive and in which cases the optical links are preferable?”.
To address this global question, collaborations with other CominLabs projects (“3D optical manycore” and “EPOC”) which address optical interconnects are planned to achieve comparisons between the two approaches. But before doing comparisons between different communication technologies, this project will study the different parts of a complete wireless infrastructure. To do that, the project is divided into three main work-packages:
i) Physical Work-Package (WP) will focus on the RF-transceiver specification and capabilities as well as the characterization of channels with the aim to evaluate the energy needed to transport one bit. (This WP will be done in close collaboration between Lab-STICC/DIM team and the CEA/LETI team);
ii) The second Work-Package will address the development of new low-power MAC (media access control) technique based on CDMA access. (This WP will be done in close collaboration between Lab-STICC/IAS team and the IRISA/CAIRN team);
iii) The third Work-Package will concern the definition of new broadcast-based fast cooperation protocol designed for resource sharing (bandwidth, distributed memory, cache coherency) and parallel programming. (This WP will be done in close collaboration between Lab-STICC/MOCS team and the IRISA/CAIRN team).
Figure 1 illustrates this intra-chip aspect and summarizes the different protocols that will be analyzed and developed during the “BBC” project.

BIGCLIN Big data analytics for unstructured Clinical data
Partners:
Inserm/LTSI (HBD team), CNRS-IRISA (LinkMedia, Cidre & Dionysos teams).
External collaborator: CNRS-STL
Health Big Data (HBD) is more than just a very large amount of data or a large number of data sources. The data collected or produced during the clinical care process can be exploited at different levels and across different domains, especially concerning questions related to clinical and translational research. To leverage these big, heterogeneous, sensitive and multi-domain clinical data, new infrastructures are arising in most of the academic hospitals, which are intended to integrate, reuse and share data for research.
Yet, a well-known challenge for secondary use of HBD is that much of detailed patient information is embedded in narrative text, mostly stored as unstructured data. The lack of efficient Natural Language Processing (NLP) resources dedicated to clinical narratives, especially for French, leads to the development of ad-hoc NLP tools with limited targeted purposes. Moreover, the scalability and real-time issues are rarely taken into account for these possibly costly NLP tools, which make them inappropriate in real-world scenarios.
Some other today’s challenges when reusing Health data are still not resolved: data quality assessment for research purposes, scalability issues when integrating heterogeneous HBD or patient data privacy and data protection. These barriers are completely interwoven with unstructured data reuse and thus constitute an overall issue which must be addressed globally.
The BigClin project thus proposes to address the essential need to leverage the above barriers when reusing unstructured clinical data at a large scale:1) We propose to develop new clinical records representation relying on fine-grained semantic annotation thanks to new NLP tools dedicated to French clinical narratives.
2) Since, the aim is to efficiently map this added semantic information to existing structured data to be further analyzed in a Big data infrastructure, the project also addresses distributed systems issues: scalability, management of uncertain data and privacy, stream processing at runtime...
In this project, we will demonstrate how clinical research might leverage NLP, information retrieval (IR) and automatic reasoning methods in order to address different use cases. The specific objectives are:
-To develop methods of information extraction and indexing dedicated to clinical texts;
-To exploit data mining techniques to handle conjointly the generated representation of the unstructured information of the clinical records and the existing structured clinical information;
- To develop distributed methods to ensure both the scalability and the online processing of these NLP/IR and data mining techniques;
- To evaluate the added value of these methods in several real clinical data and on real use-cases, including epidemilology and pharmaco-vigilance, clinical practice assessment and health care quality research, clinical trials.

INTERCOM: Interactive Communication Massive random access to subsets of compressed correlated data
Partners:
Inria, Sirocco team: Aline Roumy, Thomas Maugey
LabSTICC, Télécom Bretagne, Signal & Communications Department: Elsa Dupraz, Karine Amis
Inria, i4S team: Jean Dumoulin
External partner: Michel Kieffer L2S, CentraleSupelec, Univ. Paris Sud.
This project aims to develop novel compression techniques allowing massive random access to large databases. Indeed, we consider a database that is so large that, to be stored on a single server, the data have to be compressed efficiently, meaning that the redundancy/correlation between the data have to be exploited. The dataset is then stored on a server and made available to users that may want to access only a subset of the data. Such a request for a subset of the data is indeed random, since the choice of the subset is user-dependent. Finally, massive requests are made, meaning that, upon request, the server can only perform low complexity operations (such as bit extraction but no decompression/compression).
Algorithms for two emerging applications of this problem will be developed: Free-viewpoint Television (FTV) and massive requests to a database collecting data from a large-scale sensor network (such as Smart Cities).

Random access to a database: the user can choose any subset of the compressed correlated data.
M5HSTIA
Partners:
INSA of Rennes, IETR : Maryline Hélard, Matthieu Crussière
Télécom-Bretagne, Lab-STICC: François Gallée, Patrice Pajusco, Camilla Karnfelt, Daniel Bourreau
University of Rennes 1, IETR : Ronan Sauleau, Bernard Uguen
B-Com: Rodolphe Legouable, Jean Dion, Stéphane Paquelet
Orange Labs: Nadine Malhouroux, Christian Gallard, Philippe Ratajczak, Jean-Pierre Rossi
Exploiting millimetre-wave (mmW) radio spectrum will unlock ultra-large bandwidths for future 5th generation (5G) – and beyond – wireless systems. Developments of innovative cost-effective antenna / beamforming architectures and new processing techniques play a key role in those frequency bands, in order to allow massive deployment of consumer products in the coming years.
To meet growing demand for higher throughputs, advanced digital communication techniques based on multicarrier modulations, multiple antenna systems (MIMO) and their extension to massive MIMO (M-MIMO), powerful coding schemes or interference coordination are under study and could be combined with solutions based on network densification and deployment of heterogeneous infrastructures. An alternative but complementary way to increase throughputs is to deploy cellular systems operating in mmW bands (typically in V-band from 57 to 66 GHz, and in E-bands from 71-76 GHz and 81-86 GHz). These high frequency bands offer very large bandwidths (some of them are unlicensed) that are one of the simplest ways to increase system capacity, and also lead to enhanced miniaturisation of radio-frequency architectures. In such a context, M-MIMO systems, with up to hundreds of radiating elements at the access point (AP), are extremely attractive solutions to achieve very high data rates (multi-gigabit / sec) for multiple users sharing the same spectrum at the same time, with low power consumption thanks to the use of specific analogue/digital precoding techniques. Moreover, any effective hardware implementation of such systems must rely on a realistic knowledge of channel impairments and mmW propagation / antenna characteristics, especially for outdoor and mobile communications for which the results available in the most recent literature are very limited.
M5HESTIA project aims at designing advanced M-MIMO antennas, characterizing / modelling the outdoor mmW channel and proposing innovative algorithms in order to demonstrate, a full M-MIMO hardware (HW) platform operating in the 60-GHz band.
To reach the very ambitious goals of the M5HESTIA project, a closed collaboration has been set up with a complementary project funded by IRT b<>com (internal project at IRT) and also dedicated to mmW transmissions; this project is entitled 5M (Mm-Waves Multi-User Massive MIMO).

PRIVGEN: Privacy-preserving sharing and processing of genetic data
Partners :
Gouenou Coatrieux, LaTIM Inserm UMR1101 / Télécom Bretagne, Brest
Emmanuelle Genin, Inserm UMR 1078, Brest
Mario Südholt, École des Mines de Nantes – Inria - Lina
Labex GENMED (http://www.genmed.fr/index.php/fr/)
Cloud computing has emerged as a successful paradigm allowing individuals and companies to flexibly store and process large amounts of data. However, cloud applications are subject to new security risks and risks to the privacy of data concerning the disclosure, ownership, and integrity of data. These problems are particularly important in the context of the sharing and processing of genetic data that require a multitude of security and privacy properties to be satisfied.
The PrivGen project aims at providing new techniques for making secure and protect the privacy of shared genetic data that is processed using distributed applications. To do so, PrivGen proposes to develop: (i) new means for the combination of watermarking, encryption and fragmentation techniques to ensure the security and protection of privacy of shared genetic data, (ii) a composition theory for security mechanisms that allows the enforcement of security and privacy properties in a constructive manner on the programming level, and (iii) new service-based techniques for the distributed processing of shared genetic data.
PROFILE: Analyzing and mitigating the risks of online profiling: building a global perspective at the intersection of law, computer science and sociology
Partners:
ASAP team (INRIA), DiverSE team (INRIA), DRUID team (IRISA), IODE (Université de Rennes 1), PREFics (Université de Rennes 2).
External collaborator: Sébastien Gambs (Université du Québec à Montréal, Canada).
The practice of online profiling, which can be defined as the tracking and collection of user information on computer networks, has grown massively during the last decade, and is now affecting the vast majority of citizens. Despite its importance and impact, profiling remains largely unregulated, with no legal provisions determining its lawful use and limits under either the French and European law. This has encouraged market players to exploit a wide range of tracking technologies to collect user information, including personal data. Consequently, most online companies are now routinely violating the fundamental rights of their users, especially with respect to their privacy, with little or no oversight.
The PROFILE project brings together experts from law, computer science and sociology to address the challenges raised by online profiling, following a multidisciplinary approach. More precisely, the project will pursue two complementary and mutually informed lines of research:
- Investigate, design, and introduce a new right of opposition into the legal framework of data protection to better regulate profiling and to modify the behavior of commercial companies towards being more respectful of the privacy of their users.
- Provide users with the technical means they need to detect stealthy profiling techniques as well as to control the extent of the digital traces they routinely produce. As a case study, we focus on browser fingerprinting, a new profiling technique for targeted advertisement. The project will develop a generic framework to reason on the data collected by profiling algorithms, to uncover their inner working, and make them more accountable to users. PROFILE will also propose an innovative protection to mitigate browser fingerprinting, based on the collaborative reconfiguration of browsers.
The legal model developed in PROFILE will be informed by our technological efforts (e.g., what is technologically possible or not), while our technological research will incorporate the legal and sociological insights produced by the project (e.g., what is socially and legally desirable / acceptable). The resulting research lies at the crossing of three fields of expertise (namely Law, Computer Science and Sociology), and—we believe—forms a proposal that is timely, ambitious, and immediately relevant to our modern societies.

RAMPART
Partners
LTSI Inserm 1099, Rennes
LATIM Inserm UMR 1101 Brest
Inria Rennes
External partner Lebesgue Labex
The project RAMPART (RAdiomics and Modeling for ProstAte RadioTherapy) builds on the results obtained in PREDICTIVE and takes a major step forward by further investigating the issue of both therapy response and toxicity prediction in prostate cancer by exploiting and combining big data (biological, biomarkers, dosimetry, imaging of 2500 patients), modeling and Radiomics on CT, CBCT, MRI and PET images.
EDUCATION
One education project was selected on December 2015: the SUNSET project.
SUNSET
The SunSet project aims to develop an innovative training software suite based on immersive and collaborative virtual reality technology for training and evaluating non-technical skills. This approach will be implemented and evaluated in the context of training neurosurgical scrub nurse. It will integrate methods and systems developed in the S3PM project on procedural knowledge. By relying on Human Factors approaches, the project will address training and evaluation of interpersonal skills. Whereas the developed technologies and approaches will be generic and adaptable to any surgical specialty, the project will evaluate the developed system within training sessions performed with scrub nurses. The ambition of the project is double: to propose novel approaches for surgical non-technical skill learning and assessment, and to implement and install the developed training factory at the University Hospital of Rennes, and evaluate it with real scale user experiments. This will be ensured by 1) involving medical professionals from the early design stage of the project and for specifying the pedagogical content of the system and 2) by a strong engineering effort for developing a prototype of a surgical simulation system. Limitations of existing systems include costly and subjective human based training in sensitive clinical environments or non-realistic simulated environments, assessment of technical skills only, lack of quantitative metrics, lack of realistic scenarios, and lack of flexibility. To overcome these limitations, the SunSet project aims at proposing a new paradigm in surgical teaching and assessment of non-technical skills. It aims at reducing costs of teaching as well as improving quality of teaching.
GOVERNANCE
COORDINATOR OF COMINLABS AND STEERING BOARD
Following the end of UEB in December 2015 and the creation of UBL (Université Bretagne – Loire) in January 2016, UBL is now the coordinating institution with respect to ANR for CominLabs.
In January 2016, UBL head appointed Professor Jean-François Carpentier (Full Professor at University of Rennes 1 and recently Vice-president of Rennes 1 in charge of Research) as Chairman of CominLabs steering board.
INTERNATIONAL ADVISORY COMMITTEE
Peter Hall decided to quit IAC since he retired. We warmly thank him for his contribution to CominLabs. Professor Yang Hao from Queen Mary College, University of London, and Professor Alexander Pretschner from TU München have kindly accepted to join the International Advisory Committee. .
CREATION OF TRACKS
We decided in 2015 to launch specific programs over the most important clusters of CominLabs projects, and we called them tracks. Two were launched in September 2015: the “ICT for Health” track comprising now 9 projects, headed by Jean-Louis Coatrieux, and the “Security and Privacy” track comprising 7 projects and headed by Thomas Jensen. A third one, “Communication” track, comprising 8 projects will be soon launched. With tracks over the 2015-19 period, we want to capitalize on the skills developed, to allow for cross-fertilization and cross-project interactions, and to pursue the following goals :
- Strong and distinctive competencies of CominLabs must be pushed forward;
- The transforming and structuring effect of CominLabs must be made explicit and further exploited;
- We want to prepare for the future after 2019.
FOCUS ON COMINLABS Ph-D STUDENTS

I received the Master1 degree in computer science from faculty of science, Lebanese university and Master 2 degree in Bioinformatics and modeling from AZM center, Lebanese University, Lebanon in 2013. Currently, I am a third year PhD student between the Lebanese University (AZM center) and the University of Rennes1(LTSI). My PhD thesis takes place in the context of the “Neural Coding” project. The topic of my thesis is: “Methods for graphs classification. Application to neural networks involved in memory processes”.
The main objective of the thesis is the development and evaluation of methods for the classification of graphs. So far, a number of approaches have been proposed in the literature to compare graphs. These methods need to be evaluated in the specific context of our Neural Coding project in which identified graphs have a given topology that is the key element for comparison. Therefore, we may also expect that new algorithms are to be elaborated to account for these topological properties.
My research interests include EEG signal processing, brain connectivity, graph analysis and graphs similarity.
I want to commend the role of Labex Cominlabs in the scientific research and the development of the capacity of PhD students through scientific activities and financial support.
My publications:
Journal
[1]A. Mheich, M. Hassan, M. Khalil, C. Berrou, and F. Wendling, "A new algorithm for spatiotemporal analysis of brain functional connectivity," Journal of neuroscience methods, 2015.
Conferences
1] Mheich A., Hassan M., Gripon V., Dufor O., Khalil M., Berrou C., Wendling F: “A novel algorithm for measuring graph similarity: application to brain networks” 7th International IEEE EMBS Neural engineering Conference, Montpellier, France: 2015
[2] A. Mheich, M. Hassan, O. Dufor, M. Khalil, C. Berrou, and F. Wendling, "Spatiotemporal Analysis of Brain Functional Connectivity," in 6th European Conference of the International Federation for Medical and Biological Engineering, 2015, pp. 934-937.
[3] M. Hassan, A. Mheich, F. Wendling, O. Dufor, and C. Berrou, "Graph-based analysis of brain connectivity during spelling task," in Advances in Biomedical Engineering (ICABME), 2013 2nd International Conference on, 2013, pp. 191-194.
Riswan Masood (BOWI project)

My name is Rizwan Masood and I am currently working as a PhD student at Telecom Bretagne, France. My principal research encompasses antennas and propagation for Body Centric Wireless Communications for BoWI project.
BoWI stands for Body-World Interaction (http://www.bowi.cominlabs.ueb.eu/) and is a novel and innovative inspiration of Wireless Body Area Sensor Networks (WBASNs) and is focused on the society challenge called Digital Environment for the Citizen as well as the social challenge of ICT for personalized medicine. The principle goal of BoWI project is the design of pioneer interfaces for an accurate gesture and body movement estimation with extremely severe constraints in terms of footprint and power consumption. This interactive concept of 3D reconstruction opens new vistas to develop number of new features and usages for a plethora of exciting and innovative applications while interacting with the body language with smart environments such as home, media, information systems, entertainment and health systems. This will also influence all existing motion capture technologies (e.g., as those used in films such as The Lord of the Rings, The Hobbit, Avatar etc.) which are not only composed of several cumbersome devices but are also limited by their high-cost constraint and operating environments. BoWI was undertaken by a multidisciplinary research consortium with a tight research alliance between three different French CNRS labs namely Lab-STICC, IETR and IRISA located in various cities in France. The first prototype of the project along with the simulator interface (on Android platform) has already been demonstrated to IAC by the BoWI consortium in March 2015. My PhD dissertation finds its place in the antennas and radio sub-project of BoWI which is supervised by Professor Christian Person (Lab-STICC, Brest) and co-supervised by Professor Ronan Sauleau (IETR, Rennes). More information about my thesis orientation and results can be found in (http://www.bowi.cominlabs.ueb.eu/fr/antenna-design-and-wireless-channel-models).
My PhD thesis involved cutting-edge research for antenna design and also interaction of antennas with human body which is considered to be a hostile environment from propagation perspective due to its abnormally high tissues parameters at microwave frequencies. The thesis involved rigorous simulation and measurement studies involving antenna design and propagation studies including advanced techniques such as Multiple-Input Multiple-Output (MIMO). The thesis resulted in new antenna prototypes with improved performance metrics and tunable features for optimizing performance in real-time conditions which involve WBASN user mobility. The thesis also led to the development of a powerful wireless channel fading algorithm developed in MATLAB ® along with a numerical channel simulator developed using commercial tool CST Microwave Studio (https://www.cst.com/). This simulator incorporates cutting-edge and intelligent simulation enhancements to simulate any arbitrary WBASN scenario with limited simulation resources. Working in partnership with Labex CominLabs was also a wonderful experience for me. Interactive cooperation with a multidisciplinary research consortium for many fruitful discussions and brainstorming helped me to refine the technical and global aspect of my research.
Anis Bkakria ( POSEIDON project)

Anis Bkakria received his PhD in computer sciences from Telecom Bretagne of Institut Mines-TELECOM and his master’s degree in computer security and cryptology from University of Limoges. He is currently a post-doctoral researcher at Telecom Bretagne of Institut Mines-TELECOM. His research interest is on outsourced data security.
Recent advance in cloud computing has transformed the way information is managed and consumed. Cloud service providers are increasingly required to take responsibility for the storage as well as the efficient and reliable sharing of information, thus carrying out a "data outsourcing" architecture. Despite that outsourcing information on Cloud service providers may cut down data owners’ responsibility of managing data while increasing data availability, data owners hesitate to fully trust Cloud service providers to protect their outsourced data. Recent data breaches on Cloud storage providers have exacerbated these security concerns. In response, security designers defined approaches that provide high level security assurance, such as encrypting data before outsourcing them to Cloud servers. Such traditional approaches bring however the disadvantage of prohibiting useful information release. This raises then the need to come up with new models and approaches for defining and enforcing security and utility policies on outsourced data. This thesis aims to address this trade-off, while considering two kind of security policies. In the first hand, we focus on confidentiality policies specification and enforcement, which requires enforcing the secrecy of outsourced data stored by an untrusted Cloud service provider, while providing an efficient use (e.g., searching and computing) of the outsourced data by different authorized users. On the other hand, we address the problem of heterogeneous security policies (e.g., confidentiality requirements, privacy requirements, ownership requirements, etc.) specification and deployment.
Auréline Fargeas (PREDICTIVE project)

Auréline FARGEAS was born in Saint-Dié-Des-Vosges, France, in 1989. After a DESS in Mathematics, she received in 2012 a master degree in computer science engineering, image and signal processing, from ESIR (Ecole Supérieure d’Ingénieurs de Rennes) school at the university of Rennes 1. From 2012 to 2016, she is committed for a study at the University of Rennes 1 and is affiliated with the research group LTSI (Laboratoire Traitement du Signal et de l'Image). She will defend her thesis in June 2016. Moreover, Auréline FARGEAS has shared statistical teaching activities in French Universities.
“Labex Cominlabs is a key component of communicating your science and an important element in a successful scientific career. This community exchange encourages the students for a high-quality science by organizing different meetings and posters sessions. It should be considered as a way to engage colleagues in a dialog about your work.”
Prostate cancer is among the most common types of cancer worldwide. One of the standard treatments is external radiotherapy, which involves delivering ionizing radiation to a clinical target, in this instance the prostate and seminal vesicles. The goal of radiotherapy is to achieve a maximal local control while sparing neighboring organs (mainly the rectum and the bladder) to avoid normal tissue complications.
Classification, feature extraction and prediction of side effects in prostate cancer radiotherapy
Understanding the dose/toxicity relationships is a central question for improving treatment reliability at the inverse planning step. Normal tissue complication probability (NTCP) toxicity prediction models have been developed in order to predict toxicity events using dosimetric data. The main considered information are dose-volume histograms (DVH), which provide an overall representation of dose distribution based on the dose delivered per percentage of organ volume. Nevertheless, current dose-based models display limitations as they are not fully optimized; most of them do not include additional non-dosimetric information (patient, tumor and treatment characteristics). Furthermore, they do not provide any understanding of local relationships between dose and effect (dose-space/effect relationship) as they do not exploit the rich information from the 3D planning dose distributions. In the context of rectal bleeding prediction after prostate cancer external beam radiotherapy, the objectives of my work are: i) to extract relevant information from DVH and non-dosimetric variables, in order to improve existing NTCP models and ii) to analyze the spatial correlations between local dose and side effects allowing a characterization of 3D dose distribution at a sub-organ level. Thus, strategies aimed at exploiting the information from the radiotherapy planning (DVH and 3D planned dose distributions) were proposed. Firstly, based on independent component analysis, a new model for rectal bleeding prediction, by combining dosimetric and non-dosimetric information in an original manner, was proposed. Secondly, we have developed new approaches aimed at jointly taking advantage of the 3D planning dose distributions that may unravel the subtle correlation between local dose and side effects to classify and/or predict patients at risk of suffering from rectal bleeding, and identify regions which may be at the origin of this adverse event. More precisely, we proposed three stochastic methods based on principal component analysis, independent component analysis and discriminant nonnegative matrix factorization, and one deterministic method based on canonical polyadic decomposition of fourth order array containing planned dose. The obtained results show that our new approaches exhibit in general better performances than state-of-the-art predictive methods.
Semi-automatic prostate cancer tumor segmentation based on multiparametric MRI
Moreover, currently, there is not a consensus to the best way to integrate several MRI image modalities at identifying prostate cancer (PCa) in transitional (TZ) and peripheral zone (PZ), neither the relevance to include all the available sequences. With the aim of improving the prediction of PCa tumor tissue taking into account all the information from MRI images (MRI-T2, ADC and DCE-MRI), a voxel-based classification method on relevant extracted features was investigated. Thirty-four patients with localized prostate adenocarcinoma underwent a pre-treatment 3.0T MRI before external beam radiotherapy. Several textures and functional features were extracted from each modality and ranked according to their minimum redundancy and maximal relevance in discriminating tumor tissue. Finally, a voxel-based classification was performed on those features. The combination of MRI image modalities features appear as strong tool for discriminating PCa tumor and healthy tissue, especially when combining each modality together.