Research Axes
In the continuity of the previous works related to natural and medical image/information access, IPAL researches are focusing now on cognitive approaches applied both to
medical image understanding — by proposing a cognitive/pervasive medical image exploration for medical prognosis and treatment, and to the ambient assistive living — more oriented
toward innovative services related to aging people.
The global framework is related to pervasive access to information, involving continuous learning algorithms, reasoning (visual reasoning, meta-rules and second-order rules…) and
human-machine confluence. In the pervasive exploration/understanding processes, aspects related to explicit and implicit knowledge modeling will be developed by extension of the
actual semantic indexing and retrieval algorithms developed in IPAL. The exploration approach needs to be confluent for the user, enabling the generation of a new knowledge
(cognitive medical image) or an active ubiquitous assistance approach (AAL, ambient assistive living).
The application domains are medical prognosis and treatment support and the ambient assistive living. IPAL research domain will cover the whole cycle, starting from the healthcare
to the wellness, with a strong interaction between them, using the imaging techniques to support (validate/consolidate/quantify) the AAL policies and using the AAL to induce an impact
at the healthcare (diagnosis, treatment assistance) level.
Strong collaborations with the Singaporean/ASEAN and French/European hospitals and companies are already in place and will be strengthened in the future, in order to increase the impact
of the IPAL research outcome, opening up more IP opportunities and supporting our collaborators in their effort to the market and to the patients/people.
According to these global objectives, the research activity of IPAL is naturally structured along two axes; one related to Medical Image Understanding (MIU) and the second one,
Pervasive Access and Wellbeing Management (PAWM).
Medical Image Understanding (MIU)
Keywords: Cognitive Medical Image, Pervasive Exploration of Medical Image, Prognosis, Cognitive Vision, Medical Information Access, Multi-Scale,
Trans-Modalities, Medical Information Fusion, Uncertainties Management, Traceability, Medical ontologies for image interpretation.
PI France: Daniel Racoceanu (CNRS) - daniel.racoceanu at ens2m.fr
PI Singapore: Wee Kheng Leow (NUS) - leowwk at comp.nus.edu.sg
Abstract: The scope of the MIU research axis is to set the bases of a cognitive/pervasive exploration of the medical image for prognosis and treatment assistance,
by including cognitive vision, pervasive access and human-machine confluence algorithms.
Challenges: Cognitive and Pervasive Content-Based Medical Image Indexing and Retrieval for Validation and Prognosis Support:
A crucial step toward the translational implementation of medical image indexing and retrieval systems is related to narrowing the content-base image retrieval gaps
(content, features, performance and usability). Cognitive vision paradigms are studied and new methods are developed for medical image exploration, by integ rating pervasive
approaches and more particular continuous active learning techniques. Medical Trans-Modalities and Multi-Scale exploration for Knowledge Discovery: Multi-Scale and more generally,
Trans-modality exploration, are used to establish the basis of datamining processes, providing a real knowledge discovery to the physicists. Use of the Medical Ontologies and Visual
Reasoning for Prognosis Traceability: Use of structured a priori medical knowledge support is crucial for establishing a confluent exploration approach, enabling a traceable reliable
second opinion for medical image analysis. Visual reasoning approaches are developed in order to support the medical image exploration and allow the integration of the medical feedback
and validation in the analysis algorithms.
Related projects
MMedWeb (Multimedia Medical Conceptual Web for Intelligent Information Access): A*STAR / SERC project (2007-2010)Onco-Media (ONtology and COntext related MEdical image Distributed Intelligent Access): ICT-ASIA Programme (2006-2010)
“An Intelligent Vision System for Quantitative Microscopy in Neural Stem Cells Progenitor Growth and Differentiation”: A*STAR Cross Council Office (CCO) Project (2009-2011)
Pervasive Access and Wellbeing Management (PAWM)
Keywords: Ambient Assistive Living, disabled people, cognitive impairment, dynamic user interface, sensors networks,
mobile information access, context modeling, multi-modal interaction, life memory context management and access, cognitive memory assistance
PI France: Mounir Mokhtari (CNRS / TELECOM SudParis) - mounir.mokhtari at int-edu.eu
PI Singapore: Joo Hwee Lim (I2R/A*STAR) - joohwee at i2r.a-star.edu.sg
The challenges to quality of life for the elderly are multi-faceted. First, old people's physical and mental functions deteriorate with time and it is a great challenge to maximize the diminishing reserve in an old person to keep him active and independent. Second, once he becomes dependent on caregivers to any appreciable extent, that feeling of dependency presents a de-motivating influence towards independent living. Third, there is a genuine concern of safety for elderly living alone because of numerous factors that threaten his safety and well-being. In addition, there are social factors to be taken into account, which for an increasingly urbanized population alienate the elderly from the mainstream of life and make communal or group activity difficult. There is a great need for making the transition from hospital/nursing home back to the community at large, as well as preserving some degree of autonomy as a means to upholding quality of life for those residing in nursing homes. Assistive technology holds much promise to help increase the autonomy and daily functioning of older persons so that they can either return to independent living in their own homes and the surrounding environment or continue to lead good quality lives in the nursing home. The assistance given must however span across a multi-dimensional set of factors and the approach must be holistic to encompass physical, cognitive, emotional and social needs of the elderly. This research axis focuses specifically on the Human-Environment interaction modeling in an ambient intelligent living space, including semantic user, event, service, and system modeling with multi-modal interaction (e.g. GPS location, orientation and touch sensors environmental imagery etc) to enrich user’s daily activities. We focus particularly on people having motor disabilities and ageing people having cognitive impairment (e.g. Mild Dementia). This imposes a multidisciplinary research vision dealing with:
- User profile modeling
- Service semantic and service continuity
- System Reasoning based on ontology
- User interface and multi-modality modeling (abstracted user interface)
- Environment modeling and autonomous environment recognition
- Life memory management and access via rich media