Stephanie Mangesius1, Anna Hussl2, Florian Krismer3, Philipp Mahlknecht4, Eva Reiter5, Susanne Tagwercher6, Atbin Djamshidian7, Michael Schocke8, Regina Esterhammer9, Gregor Wenning10, Christoph Müller11, Christoph Scherfler12, Elke R Gizewski13, Werner Poewe14, Klaus Seppi15. 1. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria; Department of Neuroradiology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: stephanie.mangesius@tirol-kliniken.at. 2. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: anna.hussl@tirol-kliniken.at. 3. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: florian.krismer@i-med.ac.at. 4. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: philipp.mahlknecht@i-med.ac.at. 5. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: eva-magdalena.reiter@student.i-med.ac.at. 6. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: susanne.tagwercher@gmx.at. 7. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: atbin.djamshidian-tehrani@i-med.ac.at. 8. Department of Neuroradiology, Medical University Innsbruck, Innsbruck, Austria; Neuroimaging Core Facility, Medical University Innsbruck, Innsbruck, Austria. Electronic address: michael.schocke@rku.de. 9. Department of Neuroradiology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: regina.esterhammer@i-med.ac.at. 10. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: gregor.wenning@i-med.ac.at. 11. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria. Electronic address: christoph.mueller@tirol-kliniken.at. 12. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria; Neuroimaging Core Facility, Medical University Innsbruck, Innsbruck, Austria. Electronic address: christoph.scherfler@i-med.ac.at. 13. Department of Neuroradiology, Medical University Innsbruck, Innsbruck, Austria; Neuroimaging Core Facility, Medical University Innsbruck, Innsbruck, Austria. Electronic address: elke.gizewski@tirol-kliniken.at. 14. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria; Neuroimaging Core Facility, Medical University Innsbruck, Innsbruck, Austria. Electronic address: werner.poewe@i-med.ac.at. 15. Department of Neurology, Medical University Innsbruck, Innsbruck, Austria; Neuroimaging Core Facility, Medical University Innsbruck, Innsbruck, Austria. Electronic address: klaus.seppi@tirol-kliniken.at.
Abstract
INTRODUCTION: Several previous studies examined different brainstem-derived MR planimetric measures with regards to their diagnostic accuracy in separating patients with neurodegenerative parkinsonian disorders and reported conflicting results. The current study aimed to compare their performance in a well-characterized sample of patients with neurodegenerative parkinsonian disorders. METHODS: MR planimetric measurements were assessed in a large retrospective cohort of 55 progressive supranuclear palsy (PSP), 194 Parkinson's disease (PD) and 63 multiple system atrophy (MSA) patients. This cohort served as a training set used to build C4.5 decision tree models to discriminate PSP, PD and MSA. The models were validated in two independent test sets. The first test set comprised 84 patients with early, clinically unclassifiable parkinsonism (CUP). A prospective cohort of patients with PSP (n = 23), PD (n = 40) and MSA (n = 22) was exploited as a second test-set. RESULTS: The pons-to-midbrain diameter ratio, the midbrain diameter, the middle cerebellar peduncle width and the pons area were identified as the most predictive parameters to separate PSP, MSA and PD in C4.5 decision tree models derived from the training set. Using these decision models, AUCs in discriminating PSP, MSA and PD were 0.90, 0.57 and 0.73 in the CUP-cohort and 0.95, 0.61 and 0.87 in the prospective cohort, respectively. CONCLUSION: We were able to demonstrate that brainstem-derived MR planimetric measures yield high diagnostic accuracy for the discrimination of PSP from related disorders when decision tree algorithms are applied, even at early, clinically uncertain stages. However, their diagnostic accuracy in discriminating PD and MSA was suboptimal.
INTRODUCTION: Several previous studies examined different brainstem-derived MR planimetric measures with regards to their diagnostic accuracy in separating patients with neurodegenerative parkinsonian disorders and reported conflicting results. The current study aimed to compare their performance in a well-characterized sample of patients with neurodegenerative parkinsonian disorders. METHODS: MR planimetric measurements were assessed in a large retrospective cohort of 55 progressive supranuclear palsy (PSP), 194 Parkinson's disease (PD) and 63 multiple system atrophy (MSA) patients. This cohort served as a training set used to build C4.5 decision tree models to discriminate PSP, PD and MSA. The models were validated in two independent test sets. The first test set comprised 84 patients with early, clinically unclassifiable parkinsonism (CUP). A prospective cohort of patients with PSP (n = 23), PD (n = 40) and MSA (n = 22) was exploited as a second test-set. RESULTS: The pons-to-midbrain diameter ratio, the midbrain diameter, the middle cerebellar peduncle width and the pons area were identified as the most predictive parameters to separate PSP, MSA and PD in C4.5 decision tree models derived from the training set. Using these decision models, AUCs in discriminating PSP, MSA and PD were 0.90, 0.57 and 0.73 in the CUP-cohort and 0.95, 0.61 and 0.87 in the prospective cohort, respectively. CONCLUSION: We were able to demonstrate that brainstem-derived MR planimetric measures yield high diagnostic accuracy for the discrimination of PSP from related disorders when decision tree algorithms are applied, even at early, clinically uncertain stages. However, their diagnostic accuracy in discriminating PD and MSA was suboptimal.
Authors: Florian Krismer; Klaus Seppi; Georg Göbel; Ruth Steiger; Isabel Zucal; Sylvia Boesch; Elke R Gizewski; Gregor K Wenning; Werner Poewe; Christoph Scherfler Journal: Mov Disord Date: 2019-03-28 Impact factor: 10.338
Authors: Marina Picillo; Filomena Abate; Sara Ponticorvo; Maria Francesca Tepedino; Roberto Erro; Daniela Frosini; Eleonora Del Prete; Paolo Cecchi; Mirco Cosottini; Roberto Ceravolo; Gianfranco Di Salle; Francesco Di Salle; Fabrizio Esposito; Maria Teresa Pellecchia; Renzo Manara; Paolo Barone Journal: Front Neurol Date: 2020-11-12 Impact factor: 4.003
Authors: Andrea Quattrone; Maria G Bianco; Angelo Antonini; David E Vaillancourt; Klaus Seppi; Roberto Ceravolo; Antonio P Strafella; Gioacchino Tedeschi; Alessandro Tessitore; Roberto Cilia; Maurizio Morelli; Salvatore Nigro; Basilio Vescio; Pier Paolo Arcuri; Rosa De Micco; Mario Cirillo; Luca Weis; Eleonora Fiorenzato; Roberta Biundo; Roxana G Burciu; Florian Krismer; Nikolaus R McFarland; Christoph Mueller; Elke R Gizewski; Mirco Cosottini; Eleonora Del Prete; Sonia Mazzucchi; Aldo Quattrone Journal: Mov Disord Date: 2022-04-11 Impact factor: 9.698
Authors: Stephanie Mangesius; Anna Hussl; Susanne Tagwercher; Eva Reiter; Christoph Müller; Lukas Lenhart; Florian Krismer; Philipp Mahlknecht; Michael Schocke; Elke R Gizewski; Werner Poewe; Klaus Seppi Journal: Eur Radiol Date: 2020-01-17 Impact factor: 5.315
Authors: T Janjic; S Pereverzyev; M Hammerl; V Neubauer; H Lerchner; V Wallner; R Steiger; U Kiechl-Kohlendorfer; M Zimmermann; A Buchheim; A E Grams; E R Gizewski Journal: Eur Radiol Date: 2020-07-18 Impact factor: 7.034