21/04/2017
Marie Skłodowska-Curie Actions

MSCA Individual Fellowships (IF): Improving Automatic Medical Diagnosis by Deep Learning

This hosting has expired


  • OFFER DEADLINE
    27/04/2018 00:00 - Europe/Brussels
  • EU RESEARCH FRAMEWORK PROGRAMME
    H2020 / Marie Skłodowska-Curie Actions
  • LOCATION
    Spain, Córdoba
  • ORGANISATION/COMPANY
    Fundación para la Investigación Biomédica de Córdoba (FIBICO)
  • DEPARTMENT
    IMIBIC (Maimonides Biomedical Research Institute)
  • LABORATORY
    Knowledge Discovery and Intelligence Systems

Aim of the call

The Maimonides Biomedical Research Institute of Cordoba (IMIBIC) is seeking to develop proposals with experienced researchers for submission under the Horizon 2020 Marie Skłodowska-Curie Actions: 

http://ec.europa.eu/research/participants/portal/desktop/en/opportunitie...

IMPORTANT: Applicants should check their CV against the eligibility and mobility conditions of Marie Skłodowska-Curie Actions.

 

Brief description of the Research Group

Knowledge Discovery and Intelligence Systems (GG-25)

PI: Sebastían Ventura

The Knowledge Discovery and Intelligent Systems (KDIS, http://www.uco.es/kdis ) research group, established in 2009, focuses its research on Data Science (Data Analysis, Machine Learning, Data Mining, Big Data, etc.) and its applications to diverse problems in different areas, with special interest in biomedicine.

Project description:

Improving Automatic Medical Diagnosis by Deep Learning

The knowledge discovery in databases is a crucial process with a high relevance in many areas of the society; the huge volume of data generated by the digital and information systems have made impossible that humans can analyze, in a traditional manner, the current plethora of data. One of the most important task into the data mining process is the construction of predictive models able to estimate the value of a variable(s) of interest, which commonly has a remarkable impact on the object of study.

In the last few years, the deep learning methods have gained an increasing attention in the research community, mainly due to their capacity to model high-level relationships existing in data by means of applying a sequence of non-linear transformations, thus enhancing the discovering and learning of more complex patterns. Another field of study that has recently gained in popularity is the development of methods for text mining, which can derive high-quality information from text, and the deep learning methods have been also proved very useful for this task. As for the possible real-world applications, biomedicine is one of the areas of study that has taken advantage of the benefits of these two learning paradigms; we can find applications that range from the detection of pathologies by analyzing biomedical images to the prediction of diseases through the automatic analysis of personal electronic health records.

The main objective of this project is to develop deep learning models for text mining and their application to the automatic diagnostic of diseases from existing texts in the personal health records. In the development of this project, the candidate will design and implement methods able to correctly scale on big data problems, that is, where data are overabundant, and they can be available from multiple sources. On the other hand, it will be necessary that the developed methods can learn the predictive models from data having more flexible representations, such as in data that objects are described by more than one sample (multi-instance) and/or simultaneously associated to several classes (multi-label). These methods will be developed in the TensorFlow framework, enabling their correctly parallelization over a cluster of machines and/or GPUs.

Profile

Skills/Qualifications:

  • PhD in Computer Science or Bioinformatics
  • Other related fields (Mathematics, Biology, etc.) with accredited experience and interest in the topics of the project may also be accepted, please contact the PI.

Specific Requirements:

  • Experience in Deep Learning Models or Artificial Neural Networks

Required Research Experience:

  • Experience in Natural Language Processing
  • Any preliminary experience in the application of data analysis (statistics, machine learning, data mining) to bioinformatics and/or biomedicine will be welcome.

Required Languages:

  • Excellent level of spoken and written English.
  • Basic level of written Spanish will be welcome (for dealing with Spanish documents in the research project).

Eligibility criteria:

The candidate must fulfil the eligibility and mobility conditions of Marie Skłodowska-Curie Actions.

Selection Process:

The process consists of an analysis, evaluation and ranking of all CVs received. Following the evaluation, the highest ranked applicants may be called for a personal interview in order to evaluate more precisely the skills of the candidate.

How to Apply:

Applicants should send their CV to the following address: personal@imibic.org stating clearly in the subject of the email the reference “PostdocMSCA2018GC25”. Extended Deadline for sending your CV: 27th April, 2018.

Warning: Application emails that do not include reference will not be considered.

 

About the Host Institution:

The biomedical research institute, IMIBIC, located in Cordoba, southern Spain, is a partnership between the University of Cordoba (UCO) and the Reina Sofia University Hospital. IMIBIC offers a multidisciplinary environment focused on results-oriented research and based on precision medicine and excellence in science. IMIBIC is accredited with the Excellence distinction from the Spanish National Institute of Health Carlos III, and has received the HR Excellence in Research award of the European Commission (HRS4R).

The Institute is structured in research groups that cooperate in the implementation of various scientific programmes with the major goal to promote biomedical innovation as a powerful engine for economic and social development. IMIBIC offers an active environment in which to conduct high-level scientific research. Regular seminars and research events offer the opportunity to meet with national and international experts covering a diverse range of topics in biomedicine. IMIBIC provides an ideal environment to strengthen the skills and knowledge of young scientists and fosters translational research by bringing together experimental basic and medical sciences.

IMIBIC is young and fast-growing research center. It operates in a brand new building located within the UCO Health Sciences Campus and next to the University Hospital. It hosts a wide variety of core facilities for over 600 researchers, including the Biomedical Research Support Units with the latest technology equipment and laboratories to support the technical needs of the research community, as well as a Clinical Research Unit to support clinical trial research.

Cordoba is a medium sized city located in the hearth of Andalusia, with excellent transport connections to the main cities of Spain and both to the Mediterranean and Atlantic coasts. Rich in history and culture, Cordoba is a friendly and welcoming city with high standard of living, making it an ideal location for investigators wishing to carry out their research in Spain.