The purpose of this study was to evaluate the role of diffusion weighted-magnetic resonance imaging (DW-MRI) in the examination and classification of brain tumors, namely, glioma and meningioma. Our hypothesis was that as signal intensity variations on diffusion weighted (DW) images depend on histology and cellularity of the tumor, analysing the signal intensity characteristics on DW images may allow differentiating between the tumor types. Towards this end the signal intensity variations on DW images of the entire tumor volume data of 20 subjects with glioma and 12 subjects with meningioma were investigated and quantified using signal intensity gradient (SIG) parameter. The relative increase in the SIG values (RSIG) for the subjects with glioma and meningioma was in the range of 10.08-28.36 times and 5.60-9.86 times, respectively, compared to their corresponding SIG values on the contralateral hemisphere. The RSIG values were significantly different between the subjects with glioma and meningioma (P < 0.01), with no overlap between RSIG values across the two tumors. The results indicate that the quantitative changes in the RSIG values could be applied in the differential diagnosis of glioma and meningioma, and their adoption in clinical diagnosis and treatment could be helpful and informative.
Keyphrases
- diffusion weighted
- contrast enhanced
- magnetic resonance imaging
- deep learning
- high intensity
- computed tomography
- magnetic resonance
- convolutional neural network
- diffusion weighted imaging
- optic nerve
- optical coherence tomography
- electronic health record
- machine learning
- artificial intelligence
- big data
- high resolution