Dominik Blum1, Inga Liepelt-Scarfone2,3, Daniela Berg4, Thomas Gasser2,3, Christian la Fougère5, Matthias Reimold5. 1. Institute for Nuclear Medicine and Clinical Molecular Imaging, Eberhard Karls University, Tuebingen, Germany. dominik.blum@med.uni-tuebingen.de. 2. German Center of Neurodegenerative Diseases, Eberhard Karls University, Tuebingen, Germany. 3. Hertie-Institute for Clinical Brain Research, Eberhard Karls University, Tuebingen, Germany. 4. Department of Neurology, Christian-Albrechts-University, Kiel, Germany. 5. Institute for Nuclear Medicine and Clinical Molecular Imaging, Eberhard Karls University, Tuebingen, Germany.
Abstract
OBJECTIVE: The pattern expression score (PES), i.e., the degree to which a pathology-related pattern is present, is frequently used in FDG-brain-PET analysis and has been shown to be a powerful predictor of conversion to Alzheimer's disease (AD) in mild cognitive impairment (MCI). Since, inevitably, the PES is affected by non-pathological variability, our aim was to improve classification with the simple, yet novel approach to identify patterns of non-pathological variance in a separate control sample using principal component analysis and removing them from patient data (controls-based denoising, CODE) before calculating the PES. METHODS: Multi-center FDG-PET from 220 MCI patients (64 non-converter, follow-up ≥ 4 years; 156 AD converter, time-to-conversion ≤ 4 years) were obtained from the ADNI database. Patterns of non-pathological variance were determined from 262 healthy controls. An AD pattern was calculated from AD patients and controls. We predicted AD conversion based on PES only and on PES combined with neuropsychological features and ApoE4 genotype. We compared classification performance achieved with and without CODE and with a standard machine learning approach (support vector machine). RESULTS: Our model predicts that CODE improves the signal-to-noise ratio of AD-PES by a factor of 1.5. PES-based prediction of AD conversion improved from AUC 0.80 to 0.88 (p= 0.001, DeLong's method), sensitivity 69 to 83%, specificity 81% to 88% and Matthews correlation coefficient (MCC) 0.45 to 0.66. Best classification (0.93 AUC) was obtained when combining the denoised PES with clinical features. CONCLUSIONS: CODE, applied in its basic form, significantly improved prediction of conversion based on PES. The achieved classification performance was higher than with a standard machine learning algorithm, which was trained on patients, explainable by the fact that CODE used additional information (large sample of healthy controls). We conclude that the proposed, novel method is a powerful tool for improving medical image analysis that offers a wide spectrum of biomedical applications, even beyond image analysis.
OBJECTIVE: The pattern expression score (PES), i.e., the degree to which a pathology-related pattern is present, is frequently used in FDG-brain-PET analysis and has been shown to be a powerful predictor of conversion to Alzheimer's disease (AD) in mild cognitive impairment (MCI). Since, inevitably, the PES is affected by non-pathological variability, our aim was to improve classification with the simple, yet novel approach to identify patterns of non-pathological variance in a separate control sample using principal component analysis and removing them from patient data (controls-based denoising, CODE) before calculating the PES. METHODS: Multi-center FDG-PET from 220 MCI patients (64 non-converter, follow-up ≥ 4 years; 156 AD converter, time-to-conversion ≤ 4 years) were obtained from the ADNI database. Patterns of non-pathological variance were determined from 262 healthy controls. An AD pattern was calculated from AD patients and controls. We predicted AD conversion based on PES only and on PES combined with neuropsychological features and ApoE4 genotype. We compared classification performance achieved with and without CODE and with a standard machine learning approach (support vector machine). RESULTS: Our model predicts that CODE improves the signal-to-noise ratio of AD-PES by a factor of 1.5. PES-based prediction of AD conversion improved from AUC 0.80 to 0.88 (p= 0.001, DeLong's method), sensitivity 69 to 83%, specificity 81% to 88% and Matthews correlation coefficient (MCC) 0.45 to 0.66. Best classification (0.93 AUC) was obtained when combining the denoised PES with clinical features. CONCLUSIONS: CODE, applied in its basic form, significantly improved prediction of conversion based on PES. The achieved classification performance was higher than with a standard machine learning algorithm, which was trained on patients, explainable by the fact that CODE used additional information (large sample of healthy controls). We conclude that the proposed, novel method is a powerful tool for improving medical image analysis that offers a wide spectrum of biomedical applications, even beyond image analysis.
Entities:
Keywords:
Denoising; Pattern expression score; Physiological variance; Principal component analysis
Authors: Paula T Trzepacz; Peng Yu; Jia Sun; Kory Schuh; Michael Case; Michael M Witte; Helen Hochstetler; Ann Hake Journal: Neurobiol Aging Date: 2013-08-15 Impact factor: 4.673
Authors: María Elena Martino; Juan Guzmán de Villoria; María Lacalle-Aurioles; Javier Olazarán; Isabel Cruz; Eloisa Navarro; Verónica García-Vázquez; José Luis Carreras; Manuel Desco Journal: Ann Nucl Med Date: 2013-04-13 Impact factor: 2.668
Authors: R C Petersen; P S Aisen; L A Beckett; M C Donohue; A C Gamst; D J Harvey; C R Jack; W J Jagust; L M Shaw; A W Toga; J Q Trojanowski; M W Weiner Journal: Neurology Date: 2009-12-30 Impact factor: 9.910
Authors: Timo Grimmer; Carolin Wutz; Panagiotis Alexopoulos; Alexander Drzezga; Stefan Förster; Hans Förstl; Oliver Goldhardt; Marion Ortner; Christian Sorg; Alexander Kurz Journal: J Nucl Med Date: 2015-11-19 Impact factor: 10.057