The European Board of Physical and Rehabilitation Medicine : the past, the present and the future [Europejska Rada Medycyny Fizykalnej i Rehabilitacji: przeszłość, teraźniejszość i perspektywy]

Right from the beginning of the development of the medical specialty of Physical and Rehabilitation Medicine (PRM) the harmonization of the fields of competence and the specialist training across Europe was always an important issue. The initially informal European collaboration was formalized in 1963 under the umbrella of the European Federation of PRM. The European Academy of PRM and the UEMS section of PRM started to contribute in 1969 and 1974 respectively. In 1991 the European Board of Physical and Rehabilitation Medicine (EBPRM) was founded with the specific task of harmonizing education and training in PRM in Europe. The EBPRM has progressively defined curricula for the teaching of medical students and for the postgraduate education and training of PRM specialists. It also created a harmonized European certification system for medical PRM specialists, PRM trainers and PRM training sites. European teaching initiatives for PRM trainees (European PRM Schools) were promoted and learning material for PRM trainees and PRM specialists (e-learning, books and e-books, etc.) was created. For the future the Board will have to ensure that a minimal specific undergraduate curriculum on PRM based on a detailed European catalogue of learning objectives will be taught in all medical schools in Europe as a basis for the general medical practice. To stimulate the harmonization of national curricula, the existing postgraduate curriculum will be expanded by a syllabus of competencies related to PRM and a catalogue of learning objectives to be reached by all European PRM trainees. The integration of the certifying examination of the PRM Board into the national assessment procedures for PRM specialists will also have to be promoted.

Learning spatial filters for multispectral image segmentation

We present a novel filtering method for multispectral satellite image classification. The proposed method learns a set of spatial filters that maximize class separability of binary support vector machine (SVM) through a gradient descent approach. Regularization issues are discussed in detail and a Frobenius-norm regularization is proposed to efficiently exclude uninformative filters coefficients. Experiments carried out on multiclass one-against-all classification and target detection show the capabilities of the learned spatial filters.