Fraunhofer Center for Assistive Information and Communication Solutions – AICOSAICOS’ researcher Liliana Ferreira was awarded with the “Best PhD Dissertation Award” in the PROPOR 2012 – the 10th International Conference on Computational Processing of the Portuguese Language, which took place in Coimbra between 17-20 April.
PROPOR is the main event in the area of Natural Language Processing that is focused on Portuguese and in the theoretical and technological issues related to this specific language. The main objective of the MSc/PhD dissertations contest is to recognize the best academic work on processing of the Portuguese language and related issues, concluded in the three years preceding the event.
Liliana Ferreira’s thesis, entitled “Medical Information Extraction in European Portuguese”, was considered “Best PhD Dissertation” by an international evaluation committee. The main objective of this PhD work, developed in the Institute of Electronics and Telematics Engineering of Aveiro (IEETA), Universidade de Aveiro, was to develop MedInX, a Medical Information eXtraction System designed to process textual clinical discharge records written in Portuguese in order to perform automatic and accurate mapping of free text reports onto a structured representation.
MedInX is a pioneer system that brings many advantages for both patients and health professionals, as it allows to improve access to clinical reports, and, consequently, to enable faster and more accurate creation and analysis of statistical data. The developed system also allows to reduce medical errors and to increase safety and efficiency, while the creation of a semantically informed interpretation of texts contributes to enable patient mobility and borderless access to health care.
Moreover, all these objectives are achieved while allowing physicians to continue to practice using their current descriptive language in free-text reports without a requirement to enter structured data in a complex, time consuming computer-based system. The system was evaluated on a set of authentic patient discharge letters and performed with 95% precision and recall.
The researcher’s expertise is an added value to Fraunhofer AICOS’ research activities, as it already launched some projects in the areas of Medical Language Processing and Information Extraction.
Namely, the projects “Information Extraction from Medication Leaflets”, an MSc dissertation developed in AICOS, and also FACT, the Fraunhofer Portugal Assisted Coding Technologies. This last project has as main objective the development of a knowledge-intensive assisted coding system capable of assigning ICD (International Classification of Diseases) codes to diseases, symptoms, injuries and other external causes of illness described in medical records, being a project with a huge market potential, as it will reduce significantly the costs of manual annotation.
Source: 30.04.2012, Fraunhofer Portugal