Armazenamento distribuído de imagens médicas DICOM no formato de dados HDF5

  • Authors:
  • Douglas D. J. de Macedo;Eros Comunello;Aldo von Wangenheim;M. A. R. Dantas

  • Affiliations:
  • Universidade Federal de Santa Catarina -- UFSC, Florianópolis -- SC -- Brasil;Universidade Federal de Santa Catarina -- UFSC, Florianópolis -- SC -- Brasil;Universidade Federal de Santa Catarina -- UFSC, Florianópolis -- SC -- Brasil;Universidade Federal de Santa Catarina -- UFSC, Florianópolis -- SC -- Brasil

  • Venue:
  • Proceedings of the 14th Brazilian Symposium on Multimedia and the Web
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

DICOM medical images in telemedicine systems are characterized by the distribution, storage and recovery employing relational database systems. In addition, medical image databases can reach the order of Terabytes. Therefore, performance and access issues can be poorly during the manipulation of these databases. In this paper we adopt the approach to convert DICOM images in the scientific format of data HDF5. In this distributed paradigm, images and it attributes are converted to plain text, stored and distributed using a file system called as PVFS. Empirical experiments using DICOM images were evaluated. A model of operation and access, coined as H5WL, was developed making possible the conversion, storage and construction of images in a hierarchy form. Our tests indicate the advantage of using the new approach in comparison to a conventional database system, when the storage and recovery operations are used.