Abstract #2262
Computer Aided Radiological Diagnostics: Random Forest Classification of Glioma Tumor Progression using Image Texture Parameters derived from ADC-Maps.
Johannes Slotboom 1 , Nuno Pedrosa de Barros 1 , Stefan Bauer 2 , Urspeter Knecht 1 , Nicole Porz 3 , Philippe Schucht 3 , Pica Pica 4 , Andreas Raabe 3 , Roland Wiest 5 , and Beate Sick 6
1
DRNN, Institute of Diagnostic and
Interventional Neuroradiology, University Hospital Bern,
Bern, Switzerland,
2
Institute
of Surgical Technology and Biomechanics, University
Bern, Bern, Switzerland,
3
DKNS-Neurosurgery,
University Hospital Bern, Bern, Switzerland,
4
DOLS-Radiooncology,
University Hospital Bern, Bern, Switzerland,
5
1DRNN,
Institute of Diagnostic and Interventional
Neuroradiology, University Hospital Bern, Bern,
Switzerland,
6
Division
of Biostatistics, ISPM, University Zrich, Zrich,
Switzerland
Despite the huge amount of information provided by an
MR-examination, the initial diagnosis and grading of
frequently extremely heterogeneous brain tumors by
visual inspection remains a difficult task. A diagnostic
text often lists a number of most likely diagnoses, e.g.
anaplastic astrocytoma or glioblastoma multiforme. Here
we discuss a method for computer aided radiologic
diagnostics on how texture parameter analysis in
combination with the advanced statistical classification
random forest algorithm can be used to solve important
differential diagnosis problem for individual
diagnostics.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
