• Summary
• Related Documents
  • Active Bibliography
  • Co-citation
• Version History

Popular Tags

Add a tag:

No tags have been applied to this document.

BibTeX | Add To MetaCart

@MISC{Paris-France02computationalintelligence,
    author = {Symposium Neuroradiologicum Paris-France and Georgia Panagi Md and Lia A. Moulopoulos Md and George Dounias Phd and Thomas Maris Phd and Panourgias Md and Athanasios Tsakonas Msc and Meletios A. Dimopoulos Md},
    title = {COMPUTATIONAL INTELLIGENCE BASED ANALYSIS OF dMRI, FOR DETECTION OF SPINAL BONE MARROW MALIGNANCIES},
    year = {2002}
}

Bookmarks

         

OpenURL

 

Abstract:

This study deals with the problem of detecting spinal bone marrow malignancies with the aid of dynamic contrast enhanced MRI (dMRI). Detection of spinal bone marrow infiltration has improved with the aid of MRI, even though conventional MRI may not be helpful in the presence of red marrow or benign disorders of the vertebral bodies, which often complicate the course of disease in cancer patients. In most of these cases, dMRI may identify underlying malignant infiltration. Modern computational intelligence based methods are applied in order to uncover possible hidden relations among the ROI measurements of dMRI used to describe the problem of spinal bone marrow malignancies, i.e. signal intensity of contrast medium in discrete time intervals and specific measurements (wash-in and wash-out rates, TTPK and TMSP values). The methods used for discovering knowledge, hidden inside the imaging data, are inductive machine learning and genetic programming. A group of 92 patients divided in three sub-groups (normal, abnormal and normal appearing bone marrow) underwent dMRI of the lumbosacral spine. Meaningful sets of diagnostic rules and decision trees are produced by analyzing the parameters corresponding to the sequences of dMRI, which not only classify correctly the already proven normal and abnormal group of patients, but also suggest a classification for the group of patients with proven malignant dissemination and apparently normal appearance of the bone marrow on conventional MR images. Furthermore comparisons are given between the results acquired by the computational intelligence based methods and the standard statistical analysis performed on the same data, in order to validate generalized conclusions arising from the proposed analysis.

Citations