Glioblastoma multiforme (GBM) is a fast-growing and highly invasive brain tumor, it tends to occur in adults between the ages of 45 and 70 and it accounts for 52 percent of all primary brain tumors. Usually, GBMs are detected by magnetic resonance images (MRI). Among MRI images, Fluid-attenuated inversion recovery (FLAIR) sequence produces high quality digital tumor representation. This sequence is very sensitive to pathology and makes the differentiation between cerebrospinal fluid (CSF) and an abnormality much easier. Fast detection and segmentation techniques are needed for overcoming subjective medical doctors (MDs) judgment. In this work, a new methodology for fast detection and segmentation of GBM on FLAIR images is presented. The methodology leverages topological data analysis, textural features and interpretable machine learning algorithm, it was evaluated on a public available dataset. The machine learning classifier uses only eight input numerical features and it reaches up to the 97% of accuracy on the detection task and up to 95% of accuracy on the segmentation task. Tools from information theory were used for interpreting, in a human readable format, what are the main numerical characteristics of an image to be classified ill or healthy.
View on arXiv