Glioblastoma Differential Diagnosis
Tumors are classically distinguished based on biopsy of the tumor itself, as well as a radiological interpretation using diverse MRI modalities.
As its historical name glioblastoma multiforme implies, glioblastoma is a histologically diverse, World Health Organization grade IV astrocytic neoplasm. In spite of its simple definition of presence of vascular proliferation and/or necrosis in a diffuse astrocytoma, the wide variety of cytohistomorphologic appearances overlap with many other neoplastic or non-neoplastic lesions 1).
General imaging differential considerations include:
may look identical
both may appear multifocal
metastases usually are centered on grey-white matter junction and spare the overlying cortex rCBV in the 'edema' will be reduced
Cerebral abscess central restricted diffusion is helpful, however, if Glioblastoma is hemorrhagic then the assessment may be difficult presence of smooth and complete SWI low-intensity rim presence of dual rim sign
Anaplastic astrocytoma should not have central necrosis consider histology sampling bias
Tumefactive demyelination lesion can appear similar often has an open ring pattern of enhancement usually younger patients
Subacute cerebral infarction history is essential in suggesting the diagnosis should not have elevated choline should not have elevated rCBV
Cerebral toxoplasmosis especially in patients with AIDS
In a study, Samani et al. of the overarching goal are to demonstrate that primary glioblastomas and secondary (brain metastases) malignancies can be differentiated based on the microstructure of the peritumoral region. This is achieved by exploiting the extracellular water differences between vasogenic edema and infiltrative tissue and training a convolutional neural network (CNN) on the Diffusion Tensor Imaging (DTI)-derived free water volume fraction. They obtained 85% accuracy in discriminating extracellular water differences between local patches in the peritumoral area of 66 glioblastomas and 40 metastatic patients in a cross-validation setting. On an independent test cohort consisting of 20 glioblastomas and 10 metastases, we got 93% accuracy in discriminating metastases from glioblastomas using majority voting on patches. This level of accuracy surpasses CNNs trained on other conventional DTI-based measures such as fractional anisotropy (FA) and mean diffusivity (MD), which have been used in other studies. Additionally, the CNN captures the peritumoral heterogeneity better than conventional texture features, including Gabor filter and radiomic features. The results demonstrate that the extracellular water content of the peritumoral tissue, as captured by the free water volume fraction, is best able to characterize the differences between infiltrative and vasogenic peritumoral regions, paving the way for its use in classifying and benchmarking peritumoral tissue with varying degrees of infiltration 2).