This work validates the generalizability of MRI-based classification of Alzheimer's disease (AD) patients and controls (CN) to an external data set and to the task of prediction of conversion to AD... Show moreThis work validates the generalizability of MRI-based classification of Alzheimer's disease (AD) patients and controls (CN) to an external data set and to the task of prediction of conversion to AD in individuals with mild cognitive impairment (MCI). We used a conventional support vector machine (SVM) and a deep convolutional neural network (CNN) approach based on structural MRI scans that underwent either minimal pre-processing or more extensive preprocessing into modulated gray matter (GM) maps. Classifiers were optimized and evaluated using cross validation in the Alzheimer's Disease Neuroimaging Initiative (ADNI; 334 AD, 520 CN). Trained classifiers were subsequently applied to predict conversion to AD in ADNI MCI patients (231 converters, 628 non converters) and in the independent Health-RI Parelsnoer Neurodegenerative Diseases Biobank data set. From this multi-center study representing a tertiary memory clinic population, we included 199 AD patients, 139 participants with subjective cognitive decline, 48 MCI patients converting to dementia, and 91 MCI patients who did not convert to dementia. AD-CN classification based on modulated GM maps resulted in a similar area-under-the-curve (AUC) for SVM (0.940; 95%CI: 0.924-0.955) and CNN (0.933; 95%CI: 0.918-0.948). Application to conversion prediction in MCI yielded significantly higher performance for SVM (AUC = 0.756; 95%CI: 0.720-0.788) than for CNN (AUC = 0.742; 95%CI: 0.709-0.776) (p < 0.01 for McNemar's test). In external validation, performance was slightly decreased. For AD-CN, it again gave similar AUCs for SVM (0.896; 95%CI: 0.855-0.932) and CNN (0.876; 95%CI: 0.836-0.913). For prediction in MCI, performances decreased for both SVM (AUC = 0.665; 95%CI: 0.576-0.760) and CNN (AUC = 0.702; 95%CI: 0.624-0.786). Both with SVM and CNN, classification based on modulated GM maps significantly outperformed classification based on minimally processed images (p = 0.01). Deep and conventional classifiers performed equally well for AD classification and their performance decreased only slightly when applied to the external cohort. We expect that this work on external validation contributes towards translation of machine learning to clinical practice. Show less
Beek, M. van de; Steenoven, I. van; Ramakers, I.H.G.B.; Aalten, P.; Koek, H.L.; Rikkert, M.G.M.O.; ... ; Flier, W.M. van der 2019
Background: Quality of Life (QoL) is an important outcome measure in dementia, particularly in the context of interventions. Research investigating longitudinal QoL in dementia with Lewy bodies ... Show moreBackground: Quality of Life (QoL) is an important outcome measure in dementia, particularly in the context of interventions. Research investigating longitudinal QoL in dementia with Lewy bodies (DLB) is currently lacking.Objective: To investigate determinants and trajectories of QoL in DLB compared to Alzheimer's disease (AD) and controls.Methods: QoL was assessed annually in 138 individuals, using the EQ5D-utility-score (0-100) and the health-related Visual Analogue Scale (VAS, 0-100). Twenty-nine DLB patients (age 69 +/- 6), 68 AD patients (age 70 +/- 6), and 41 controls (age 70 +/- 5) were selected from the Dutch Parelsnoer Institute-Neurodegenerative diseases and Amsterdam Dementia Cohort. We examined clinical work-up over time as determinants of QoL, including cognitive tests, neuropsychiatric inventory, Geriatric Depression Scale (GDS), and disability assessment of dementia (DAD).Results: Mixed models showed lower baseline VAS-scores in DLB compared to AD and controls (AD: beta +/- SE = - 7.6 +/- 2.8, controls: beta +/- SE = -7.9 +/- 3.0, p < 0.05). An interaction between diagnosis and time since diagnosis indicated steeper decline on VAS-scores for AD patients compared to DLB patients (beta +/- SE = 2.9 +/- 1.5, p < 0.1). EQ5D-utility-scores over time did not differ between groups. Higher GDS and lower DAD-scores were independently associated with lower QoL in dementia patients (GDS: VAS beta +/- SE = -1.8 +/- 0.3, EQ5D-utility beta +/- SE = -3.7 +/- 0.4; DAD: VAS = 0.1 +/- 0.0, EQ5D-utility beta +/- SE = 0.1 +/- 0.1, p < 0.05). No associations between cognitive tests and QoL remained in the multivariate model.Conclusion: QoL is lower in DLB, while in AD QoL shows steeper decline as the disease advances. Our results indicate that non-cognitive symptoms, more than cognitive symptoms, are highly relevant as they impact QoL. Show less