Automated Volumes-of-Interest Identification for Classical and Atypical Parkinsonian Syndrome Differentiation Using T2’ MR Imaging

Summary

Objectives: In clinical routine, patients with classical Parkinsonian syndromes (CPS) need to be differentiated from those with atypical Parkinsonian syndromes (APS), particularly with respect to prognosis and treatment decision. To date, this diagnosis is mainly based on clinical criteria, leading to failure rates up to 25%, motivating the development of image-based decision support systems. Magnetic resonance imaging (MRI) and in particular T2´ image sequences have been suggested as a potential marker for differential diagnosis. The aim of this study was to investigate whether automatically identified T2´ volumes-of-interest (VOIs) can be used for an automatic differentiation of CPS and APS patients.

Material and Methods: 74 MRI datasets were available for this hypothesis generating trial, including image sequences from 24 healthy subjects, 33 CPS and 17 APS patients. First, a problem-specific reference atlas was generated using the healthy control datasets. Next, patients’ datasets were registered to the atlas. Voxel-wise t-tests, reflecting significance levels of T2´ value differences between CPS and APS patients, were then applied for calculation of a p-map. Finally, the calculated p-map was thresholded and a connected component analysis was performed for final VOI detection. In parallel, manually defined VOIs were determined in grey and white matter for comparison.

Results: Three VOIs in parts of the basal ganglia and the left occipital lobe were automatically identified by the presented method. There was a trend for higher area under the curve on multivariable receiver operating characteristic curves for automatically determined VOIs over manually defined VOIs (0.939 vs. 0.818, p = 0.0572).

Conclusion: The diagnostic role of automatically defined VOIs in differentiation of CPS and APS patients based on T2´ image sequences should be further investigated.

Keywords

Magnetic resonance imaging - T2’ image sequences - computer-assisted image analysis - Parkinsonian syndrome

• References

• 1 Van Den Eeden SK, Tanner CM, Bernstein AL, Fross RD, Leimpeter A, Bloch DA. et al Incidence of Parkinson’s disease: variation by age, gender, and race/ethnicity. Am J Epidemiol 2003; 157 (11) 1015-1022.
• 2 Davie CA. A review of Parkinson’s disease. Br Med Bull 2008; 86: 109-127.
• 3 Tolosa E, Wenning G, Poewe W. The diagnosis of Parkinson’s disease. Lancet Neurol 2006; 5 (01) 75-86.
• 4 Olanow CW, Hauser RA, Jankovic J, Langston W, Lang A, Poewe W. et al A randomized, double-blind, placebo-controlled, delayed start study to assess rasagiline as a disease modifying therapy in Parkinson’s disease (the ADAGIO study): rationale, design, and baseline characteristics. Mov Disord 2008; 23 (15) 2194-2201.
• 5 Tang CC, Poston KL, Eckert T, Feigin A, Frucht S, Gudesblatt M. et al Differential diagnosis of parkinsonism: a metabolic imaging study using pattern analysis. Lancet Neurol 2010; 9 (02) 149-158.
• 6 Duchesne S, Rolland Y, Verin M. Automated computer differential classification in Parkinsonian Syndromes via pattern analysis on MRI. Acad Radiol 2009; 16 (01) 61-70.
• 7 Graham JM, Paley MN, Grunewald RA, Hoggard N, Griffiths PD. Brain iron deposition in Parkinson’s disease imaged using the PRIME magnetic resonance sequence. Brain 2000; 123 Pt 12 2423-2431.
• 8 Forkert ND, Schmidt-Richberg A, Holst B, Munchau A, Handels H, Boelmans K. Image-based Classification of Parkinsonian Syndromes Using T2´-Atlases. Stud Health Technol Inform 2011; 169: 465-469.
• 9 Boelmans K, Holst B, Hackius M, Finsterbusch J, Gerloff C, Fiehler J. et al Brain iron deposition fingerprints in Parkinson’s disease and progressive supranuclear palsy. Mov Disord 2012; 27 (03) 421-427.
• 10 Schenck JF, Zimmerman EA. High-field magnetic resonance imaging of brain iron: birth of a biomarker?. NMR Biomed 2004; 17 (07) 433-445.
• 11 Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992; 55 (03) 181-184.
• 12 Gilman S, Wenning GK, Low PA, Brooks DJ, Mathias CJ, Trojanowski JQ. et al Second consensus statement on the diagnosis of multiple system atrophy. Neurology 2008; 71 (09) 670-676.
• 13 Litvan I, Bhatia KP, Burn DJ, Goetz CG, Lang AE, McKeith I. et al Movement Disorders Society Scientific Issues Committee report: SIC Task Force appraisal of clinical diagnostic criteria for Parkinsonian disorders. Mov Disord 2003; 18 (05) 467-486.
• 14 Guimond A, Meunier J, Thirion J-P. Average Brain Models: A Convergence Study. Comput Vis Image Underst 2000; 77 (02) 192-210.
• 15 Ehrhardt J, Werner R, Schmidt-Richberg A, Handels H. Statistical modeling of 4D respiratory lung motion using diffeomorphic image registration. IEEE Trans Med Imaging 2011; 30 (02) 251-265.
• 16 Schmidt-Richberg A, Ehrhardt J, Werner R, Handels H. Diffeomorphic Diffusion Registration of Lung CT Images. In van Ginneken B. MK, Heimann T, Pekar V, Deng X. editor Image Analysis for the Clinic - A Grand Challenge, EMPIRE10 - Evaluation of Methods for Pulmonary Image Registration; 2010; Bejing. 2010; 55-62.
• 17 Forkert ND, Saring D, Fiehler J, Illies T, Moller D, Handels H. Automatic brain segmentation in Time-of-Flight MRA images. Methods Inf Med 2009; 48 (05) 399-407.
• 18 DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988; 44 (03) 837-845.
• 19 Obuchowski NA, McClish DK. Sample size determination for diagnostic accuracy studies involving binormal ROC curve indices. Stat Med 1997; 16 (13) 1529-1542.