Data-driven modelling of neurodegenerative disease progression: thinking outside the black box.

Data-driven disease progression models are an emerging set of computational tools that reconstruct disease timelines for long-term chronic diseases, providing unique insights into disease processes and their underlying mechanisms. Such methods combine a priori human knowledge and assumptions with large-scale data processing and parameter estimation to infer long-term disease trajectories from short-term data. In contrast to 'black box' machine learning tools, data-driven disease progression models typically require fewer data and are inherently interpretable, thereby aiding disease understanding in addition to enabling classification, prediction and stratification. In this Review, we place the current landscape of data-driven disease progression models in a general framework and discuss their enhanced utility for constructing a disease timeline compared with wider machine learning tools that construct static disease profiles. We review the insights they have enabled across multiple neurodegenerative diseases, notably Alzheimer disease, for applications such as determining temporal trajectories of disease biomarkers, testing hypotheses about disease mechanisms and uncovering disease subtypes. We outline key areas for technological development and translation to a broader range of neuroscience and non-neuroscience applications. Finally, we discuss potential pathways and barriers to integrating disease progression models into clinical practice and trial settings.

Online clinical tools to support the use of new plasma biomarker diagnostic technology in the assessment of Alzheimer's disease: a narrative review.

Recent advances in new diagnostic technologies for Alzheimer's disease have improved the speed and precision of diagnosis. However, accessing the potential benefits of this technology poses challenges for clinicians, such as deciding whether it is clinically appropriate to order a diagnostic test, which specific test or tests to order and how to interpret test results and communicate these to the patient and their caregiver. Tools to support decision-making could provide additional structure and information to the clinical assessment process. These tools could be accessed online, and such 'e-tools' can provide an interactive interface to support patients and clinicians in the use of new diagnostic technologies for Alzheimer's disease. We performed a narrative review of the literature to synthesize information available on this research topic. Relevant studies that provide an understanding of how these online tools could be used to optimize the clinical utility of diagnostic technology were identified. Based on these, we discuss the ways in which e-tools have been used to assist in the diagnosis of Alzheimer's disease and propose recommendations for future research to aid further development.

Registration-based methods applied to serial high-resolution T1-weighted magnetic resonance imaging for the assessment of brain volume changes in anorexia nervosa of the restricting type.

We aimed to determine whether variation in the body mass index (BMI)­a marker of anorexia nervosa (AN) severity­is associated with brain volume changes longitudinally estimated using registration-based methods on serial high-resolution T1-weighted magnetic resonance images (MRI). Fifteen female patients (mean age = 21 years; standard deviation [SD] = 5.7; range: 15­33 years) with the diagnosis of AN of the restricting type (AN-r)­according to the Diagnostic and Statistic Manual of Mental Disorders, 5th edition criteria­underwent T1-weighted MRI at baseline and after a mean follow-up period of 11 months (SD = 6.4). We used the brain boundary shift integral (BSI) and the ventricular BSI (VBSI) to estimate volume changes after registering voxels of follow-up onto baseline MRI. Very significant and strong correlations were found between BMI variation and the brain BSI, as well as between BMI variation and the VBSI. After adjustment for age at onset, duration of illness, and the BMI rate of change before baseline MRI, the statistical significance of both associations persisted. Registration-based methods on serial MRI represent an additional tool to estimate AN severity, because they provide measures of brain volume change strongly associated with BMI variation.

BRain health and healthy AgeINg in retired rugby union players, the BRAIN Study: study protocol for an observational study in the UK.

INTRODUCTION: Relatively little is known about the long-term health of former elite rugby players, or former sportspeople more generally. As well as the potential benefits of being former elite sportspersons, there may be potential health risks from exposures occurring during an individual's playing career, as well as following retirement. Each contact sport has vastly different playing dynamics, therefore exposing its players to different types of potential traumas. Current evidence suggests that these are not necessarily comparable in terms of pathophysiology, and their potential long-term adverse effects might also differ. There is currently limited but increasing evidence that poorer age-related and neurological health exists among former professional sportsmen exposed to repetitive concussions; however the evidence is limited on rugby union players, specifically. METHODS AND ANALYSIS: We present the protocol for a cross-sectional study to assess the association between self-reported history of concussion during a playing career, and subsequent measures of healthy ageing and neurological and cognitive impairment. We are recruiting a sample of approximately 200 retired rugby players (former Oxford and Cambridge University rugby players and members of the England Rugby International Club) aged 50 years or more, and collecting a number of general and neurological health-related outcome measures though validated assessments. Biomarkers of neurodegeneration (neurofilaments and tau) will be also be measured. Although the study is focusing on rugby union players specifically, the general study design and the methods for assessing neurological health are likely to be relevant to other studies of former elite sportspersons. ETHICS AND DISSEMINATION: The study has been approved by the Ethical Committee of London School of Hygiene and Tropical Medicine (reference: 11634-2). It is intended that results of this study will be published in peer-reviewed medical journals, communicated to participants, the general public and all relevant stakeholders.

The Edinburgh Consensus: preparing for the advent of disease-modifying therapies for Alzheimer's disease.

CONTEXT: This commentary discusses the implications of disease-modifying treatments for Alzheimer's disease which seem likely to appear in the next few years and results from a meeting of British experts in neurodegenerative diseases in Edinburgh. The availability of such treatments would help change public and professional attitudes and accelerate engagement with the prodromal and preclinical populations who might benefit from them. However, this would require an updated understanding of Alzheimer's disease, namely the important distinction between Alzheimer's disease and Alzheimer's dementia. CONSENSUS: Since treatments are likely to be most effective in the early stages, identification of clinically relevant brain changes (for example, amyloid burden using imaging or cerebrospinal fluid biomarkers) will be crucial. While current biomarkers could be useful in identifying eligibility for new therapies, trial data are not available to aid decisions about stopping or continuing treatment in clinical practice. Therefore, effective monitoring of safety and effectiveness when these treatments are introduced into clinical practice will be necessary to inform wide-scale use. Equity of access is key but there is a tension between universal access for everyone with a diagnosis of Alzheimer's disease and specifying an eligible population most likely to respond. We propose the resources necessary for an optimal care pathway as well as the necessary education and training for primary and secondary care. CONCLUSION: The majority of current services in the UK and elsewhere would not be able to accommodate the specialist investigations required to select patients and prescribe these therapies. Therefore, a stepped approach would be necessary: from innovating sentinel clinical-academic centres that already have capacity to deliver the necessary phase IV trials, through early adoption in a hub and spoke model, to nationwide adoption for true equity of access. The optimism generated by recent and anticipated developments in the understanding and treatment of Alzheimer's disease presents a great opportunity to innovate and adapt our services to incorporate the next exciting development in the field of dementia.

Operationalizing hippocampal volume as an enrichment biomarker for amnestic mild cognitive impairment trials: effect of algorithm, test-retest variability, and cut point on trial cost, duration, and sample size.

The objective of this study was to evaluate the effect of computational algorithm, measurement variability, and cut point on hippocampal volume (HCV)-based patient selection for clinical trials in mild cognitive impairment (MCI). We used normal control and amnestic MCI subjects from the Alzheimer's Disease Neuroimaging Initiative 1 (ADNI-1) as normative reference and screening cohorts. We evaluated the enrichment performance of 4 widely used hippocampal segmentation algorithms (FreeSurfer, Hippocampus Multi-Atlas Propagation and Segmentation (HMAPS), Learning Embeddings Atlas Propagation (LEAP), and NeuroQuant) in terms of 2-year changes in Mini-Mental State Examination (MMSE), Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-Cog), and Clinical Dementia Rating Sum of Boxes (CDR-SB). We modeled the implications for sample size, screen fail rates, and trial cost and duration. HCV based patient selection yielded reduced sample sizes (by ∼40%-60%) and lower trial costs (by ∼30%-40%) across a wide range of cut points. These results provide a guide to the choice of HCV cut point for amnestic MCI clinical trials, allowing an informed tradeoff between statistical and practical considerations.

LoAd: a locally adaptive cortical segmentation algorithm.

Thickness measurements of the cerebral cortex can aid diagnosis and provide valuable information about the temporal evolution of diseases such as Alzheimer's, Huntington's, and schizophrenia. Methods that measure the thickness of the cerebral cortex from in-vivo magnetic resonance (MR) images rely on an accurate segmentation of the MR data. However, segmenting the cortex in a robust and accurate way still poses a challenge due to the presence of noise, intensity non-uniformity, partial volume effects, the limited resolution of MRI and the highly convoluted shape of the cortical folds. Beginning with a well-established probabilistic segmentation model with anatomical tissue priors, we propose three post-processing refinements: a novel modification of the prior information to reduce segmentation bias; introduction of explicit partial volume classes; and a locally varying MRF-based model for enhancement of sulci and gyri. Experiments performed on a new digital phantom, on BrainWeb data and on data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) show statistically significant improvements in Dice scores and PV estimation (p<10(-3)) and also increased thickness estimation accuracy when compared to three well established techniques.

11C-PiB PET assessment of change in fibrillar amyloid-beta load in patients with Alzheimer's disease treated with bapineuzumab: a phase 2, double-blind, placebo-controlled, ascending-dose study.

BACKGROUND: Carbon-11-labelled Pittsburgh compound B ((11)C-PiB) PET is a marker of cortical fibrillar amyloid-beta load in vivo. We used (11)C-PiB PET to investigate whether bapineuzumab, a humanised anti-amyloid-beta monoclonal antibody, would reduce cortical fibrillar amyloid-beta load in patients with Alzheimer's disease. METHODS: Patients with mild-to-moderate Alzheimer's disease were randomly assigned to receive intravenous bapineuzumab or placebo in a ratio of seven to three in three ascending dose groups (0.5, 1.0, or 2.0 mg/kg). Each dose group was enrolled after safety review of the previous group. Randomisation was by interactive voice response system; masking was achieved with numbered kit allocation. Patients, investigators, study site personnel, sponsor staff, and carers were masked to treatment. Patients received up to six infusions, 13 weeks apart, and had (11)C-PiB PET scans at baseline and at weeks 20, 45, and 78. The primary outcome was the difference between the pooled bapineuzumab group and the pooled placebo group in mean change from screening to week 78 in (11)C-PiB cortical to cerebellar retention ratio averaged across six cortical regions of interest. Analysis was by modified intention to treat. This study is registered with EudraCT, number 2004-004120-12; ISRCTN17517446. FINDINGS: 28 patients were assigned to bapineuzumab (n=20) or placebo (n=8). 19 patients in the bapineuzumab group and seven in the placebo group were included in the modified intention-to-treat analysis. Estimated mean (11)C-PiB retention ratio change from baseline to week 78 was -0.09 (95% CI -0.16 to -0.02; p=0.014) in the bapineuzumab group and 0.15 (95% CI 0.02 to 0.28; p=0.022) in the placebo group. Estimated mean difference in (11)C-PiB retention ratio change from baseline to week 78 between the bapineuzumab group and the placebo group was -0.24 (95% CI -0.39 to -0.09; p=0.003). Differences between the bapineuzumab group and the placebo group in the individual regions of interest were similar to the overall mean difference. Adverse events were typically mild to moderate in severity and transient. Two patients in the 2.0 mg/kg bapineuzumab group had transient cerebral vasogenic oedema. INTERPRETATION: Treatment with bapineuzumab for 78 weeks reduced cortical (11)C-PiB retention compared with both baseline and placebo. (11)C-PiB PET seems to be useful in assessing the effects of potential Alzheimer's disease treatments on cortical fibrillar amyloid-beta load in vivo. FUNDING: Elan Pharmaceuticals and Wyeth Research.

Accuracy assessment of global and local atrophy measurement techniques with realistic simulated longitudinal Alzheimer's disease images.

The evaluation of atrophy quantification methods based on magnetic resonance imaging have been usually hindered by the lack of realistic gold standard data against which to judge these methods or to help refine them. Recently [Camara, O., Schweiger, M., Scahill, R., Crum, W., Sneller, B., Schnabel, J., Ridgway, G., Cash, D., Hill, D., Fox, N., 2006. Phenomenological model of diffuse global and regional atrophy using finite-element methods. IEEE Trans. Med.l Imaging 25, 1417-1430], we presented a technique in which atrophy is realistically simulated in different tissue compartments or neuroanatomical structures with a phenomenological model. In this study, we have generated a cohort of realistic simulated Alzheimer's disease (AD) images with known amounts of atrophy, mimicking a set of 19 real controls and 27 probable AD subjects, with an improved version of our atrophy simulation methodology. This database was then used to assess the accuracy of several well-known computational anatomy methods which provide global (BSI and SIENA) or local (Jacobian integration) estimates of longitudinal atrophy in brain structures using MR images. SIENA and BSI results correlated very well with gold standard data (Pearson coefficient of 0.962 and 0.969 respectively), achieving small mean absolute differences with respect to the gold standard (percentage change from baseline volume): BSI of 0.23%+/-0.26%; SIENA of 0.22%+/-0.28%. Jacobian integration was guided by both fluid and FFD-based registration techniques and resulting deformation fields and associated Jacobians were compared, region by region, with gold standard ones. The FFD-based technique outperformed the fluid one in all evaluated structures (mean absolute differences from the gold standard in percentage change from baseline volume): whole brain, FFD=0.31%, fluid=0.58%; lateral ventricles, FFD=0.79%; fluid=1.45%; left hippocampus, FFD=0.82%; fluid=1.42%; right hippocampus, FFD=0.95%; fluid=1.62%. The largest errors for both local techniques occurred in the sulcal CSF (FFD=2.27%; fluid=3.55%) regions. For large structures such as the whole brain, these mean absolute differences, relative to the applied atrophy, represented similar percentages for the BSI, SIENA and FFD techniques (controls/patients): BSI, 51.99%/16.36%; SIENA, 62.34%/21.59%; FFD, 41.02%/24.95%. For small structures such as the hippocampi, these percentages were larger, especially for controls where errors were approximately equal to the small applied changes (controls/patients): FFD, 92.82%/43.61%. However, these apparently large relative errors have not prevented the global or hippocampal measures from finding significant group separation in our study. The evaluation framework presented here will help in quantifying whether the accuracy of future methodological developments is sufficient for analysing change in smaller or less atrophied local brain regions. Results obtained in our experiments with realistic simulated data confirm previously published estimates of accuracy for both evaluated global techniques. Regarding Jacobian Integration methods, the FFD-based one demonstrated promising results and potential for being used in clinical studies alongside (or in place of) the more common global methods. The generated gold standard data has also allowed us to identify some stages and sets of parameters in the evaluated techniques--the brain extraction step in the global techniques and the number of multi-resolution levels and the stopping criteria in the registration-based methods--that are critical for their accuracy.

Accuracy assessment of global and local atrophy measurement techniques with realistic simulated longitudinal data.

The main goal of this work was to assess the accuracy of several well-known methods which provide global (BSI and SIENA) or local (Jacobian integration) estimates of longitudinal atrophy in brain structures using Magnetic Resonance images. For that purpose, we have generated realistic simulated images which mimic the patterns of change obtained from a cohort of 19 real controls and 27 probable Alzheimer's disease patients. SIENA and BSI results correlate very well with gold standard data (BSI mean absolute error < 0.29%; SIENA < 0.44%). Jacobian integration was guided by both fluid and FFD-based registration techniques and resulting deformation fields and associated Jacobians were compared, region by region, with gold standard ones. The FFD registration technique provided more satisfactory results than the fluid one. Mean absolute error differences between volume changes given by the FFD-based technique and the gold standard were: sulcal CSF < 2.49%; lateral ventricles < 2.25%; brain < 0.36%; hippocampi < 1.42%.

Visual assessment of atrophy on magnetic resonance imaging in the diagnosis of pathologically confirmed young-onset dementias.

OBJECTIVES: To investigate the diagnostic accuracy of visual inspection of magnetic resonance imaging (MRI) in a range of pathologically confirmed diseases causing young-onset dementia and to assess the sensitivity and specificity of atrophy patterns for Alzheimer disease (AD) and frontotemporal lobar degeneration (FTLD). DESIGN: Sixty-two patients with pathologically confirmed diseases that may present as young-onset dementia were selected from a biopsy and postmortem series. The first diagnostic T1-weighted volumetric MRI was obtained for each patient, together with images from 22 healthy control subjects. All MRIs were assessed for regional atrophy independently by 3 neuroradiologists, blinded to all clinical details except age. Observers were also asked to use their clinical judgment to form a diagnosis. RESULTS: Eighty-seven percent of dementia cases were distinguished from controls after visual inspection of MRI, and a correct pathologically confirmed diagnosis was given in 58% of cases. Hippocampal atrophy was noted in 92% of AD cases but was commonly seen in other dementias and controls. A bilateral symmetrical pattern of hippocampal atrophy discriminated AD from FTLD with 47% specificity, while posterior greater than anterior gradient of atrophy was 92% specific for AD. Atrophy of the anterior, inferior, and lateral temporal lobes was suggestive of FTLD pathology (> or =90% sensitivity), while anterior greater than posterior gradient of atrophy and hemispheric asymmetry of atrophy were each at least 85% specific for FTLD. CONCLUSION: Despite variation and overlap of atrophy patterns, visual inspection of regional atrophy on MRI may aid in discriminating AD and FTLD.

An automated algorithm for the computation of brain volume change from sequential MRIs using an iterative principal component analysis and its evaluation for the assessment of whole-brain atrophy rates in patients with probable Alzheimer's disease.

This article introduces an automated method for the computation of changes in brain volume from sequential magnetic resonance images (MRIs) using an iterative principal component analysis (IPCA) and demonstrates its ability to characterize whole-brain atrophy rates in patients with Alzheimer's disease (AD). The IPCA considers the voxel intensity pairs from coregistered MRIs and identifies those pairs a sufficiently large distance away from the iteratively determined PCA major axis. Analyses of simulated and real MRI data support the underlying assumption of a linear relationship in paired voxel intensities, identify an outlier distance threshold that optimizes the trade-off between sensitivity and specificity in the detection of small volume changes while accounting for global intensity changes, and demonstrate an ability to detect changes as small as 0.04% of brain volume without confounding effects of between-scan shifts in voxel intensity. In eight patients with probable AD and eight age-matched normal control subjects, the IPCA was comparable to the established but partly manual digital subtraction (DS) method in characterizing annual rates of whole-brain atrophy: resulting rates were correlated (Spearman rank correlation = 0.94, P < 0.0005) and comparable in distinguishing probable AD from normal aging (IPCA-detected atrophy rates: 2.17 +/- 0.52% per year in the patients vs. 0.41 +/- 0.22% per year in the controls [Wilcoxon-Mann-Whitney test P = 7.8 x 10(-4)]; DS-detected atrophy rates: 3.51 +/- 1.31% per year in the patients vs. 0.48 +/- 0.29% per year in the controls [P = 7.8 x 10(-4)]). The IPCA could be used in tracking the progression of AD, evaluating the disease-modifying effects of putative treatments, and investigating the course of other normal and pathological changes in brain morphology.