Abstract #2586

Classification of Parkinson’s disease based on multi-parametric data derived from MR Fingerprinting measurements

Thomas Amthor¹, Peter Koken¹, Mariya Doneva¹, Vera Catharina Keil², Stilyana Peteva Bakoeva², Alina Jurcoane^2,3, Wolfgang Block², Ullrich Wüllner^4,5, Burkhard Mädler⁶, and Elke Hattingen^2,3

¹Philips Research Europe, Hamburg, Germany, ²Department of Radiology, University Hospital Bonn, Bonn, Germany, ³Institute for Neuroradiology, University Hospital Frankfurt/Main, Frankfurt, Germany, ⁴Department of Neurology, University Hospital Bonn, Bonn, Germany, ⁵German Centre for neurodegenerative disease research (DZNE), Bonn, Germany, ⁶Philips Healthcare, Bonn, Germany

We investigated the potential of multi-parametric MR Fingerprinting measurements for the classification of Parkinson’s disease. For each measured quantity (T1, T2, proton density) and each segmented brain region, several statistical parameters were determined and used to train a Random Forest classification algorithm. An AUC of 0.92 was achieved for distinguishing Parkinson patients from healthy control subjects.

This abstract and the presentation materials are available to members only; a login is required.

Join Here