Grant: Travel Awards - Institute Community Support

Magnetic resonance (MR) imaging has been largely used as a diagnostic imaging modality. Especially in the brain, MR imaging allows the detection and analysis of subtle abnormalities, aiding in patient diagnosis and follow-up. Our research group have been developing methods and approaches for medical image processing of large, multi-center data sets, with focus on dementia research using machine learning techniques. In this work, we developed a method to better understanding the possible origin of cerebral white matter hyperintensities (WMH). These brain abnormalities are observed in a number of neurological and psychiatric disorders, on both symptomatic and asymptomatic patients. A better understanding of the WMHs origin may aid in patient diagnosis and treatment strategies. We propose a texture-based machine learning (ML) method to distinguish WMH of presumed vascular origin (small vessel disease (SVD) patients) from those of inflammatory origin (such as in multiple sclerosis (MS) patients). Our ML method was evaluated in an adult population, reporting the observed prevalence of WMH vascular, inflammatory, or uncertain origin. Our work combines expertises from different areas, such as computer engineering, radiology, neuroscience, among others. In order to develop valuable research in such interdisciplinary area, it is required to attend and present work in conferences with both technical and clinical backgrounds, growing networking and possible collaborations with other research from multiple areas.