Search

Abstract High-resolution global flood risk maps are increasingly used to inform disaster risk planning and response, particularly in lower income countries with limited data or capacity. However, current approaches do not adequately account for spatial variation in social vulnerability, which is a key determinant of variation in outcomes for exposed populations. Here we integrate annual average exceedance probability estimates from a high-resolution fluvial flood model with gridded population and poverty data to create a global vulnerability-adjusted risk index for flooding (VARI Flood) at 90-meter resolution. The index provides estimates of relative risk within or between countries and changes how we understand the geography of risk by identifying ‘hotspots’ characterised by high population density and high levels of social vulnerability. This approach, which emphasises risks to human well-being, could be used as a complement to traditional population or asset-centred approaches.

High resolution poverty mapping supports evidence-based policy and research, yet about half of all countries lack the survey data needed to generate useful poverty maps. To overcome this challenge, new non-traditional data sources and deep learning techniques are increasingly used to create small-area estimates of poverty in low- and middle-income countries (LMICs). Convolutional Neural Networks (CNN) trained on satellite imagery are emerging as one of the most popular and effective approaches. However, the spatial resolution of poverty estimates has remained relatively coarse, particularly in rural areas. To address this problem, we use a transfer learning approach to train three CNN models and use them in an ensemble to predict chronic poverty at 1 km2 scale in rural Sindh, Pakistan. The models are trained with spatially noisy georeferenced household survey containing poverty scores for 1.67 million anonymized households in Sindh Province and publicly available inputs, including daytime and nighttime satellite imagery and accessibility data. Results from both hold-out and k-fold validation exercises show that the ensemble provides the most reliable spatial predictions in both arid and non-arid regions, outperforming previous studies in key accuracy metrics. A third validation exercise, which involved ground-truthing of predictions from the ensemble model with original survey data of 7000 households further confirms the relative accuracy of the ensemble model predictions. This inexpensive and scalable approach could be used to improve poverty targeting in Pakistan and other low- and middle-income countries.

Background: Accurate registration between microscopy and MRI data is necessary for validating imaging biomarkers against neuropathology, and to disentangle complex signal dependencies in microstructural MRI. Existing registration methods often rely on serial histological sampling or significant manual input, providing limited scope to work with a large number of stand-alone histology sections. Here we present a customisable pipeline to assist the registration of stand-alone histology sections to whole-brain MRI data. Methods: Our pipeline registers stained histology sections to whole-brain post-mortem MRI in 4 stages, with the help of two photographic intermediaries: a block face image (to undistort histology sections) and coronal brain slab photographs (to insert them into MRI space). Each registration stage is implemented as a configurable stand-alone Python script using our novel platform, Tensor Image Registration Library (TIRL), which provides flexibility for wider adaptation. We report our experience of registering 87 PLP-stained histology sections from 14 subjects and perform various experiments to assess the accuracy and robustness of each stage of the pipeline. Results: All 87 histology sections were successfully registered to MRI. Histology-to-block registration (Stage 1) achieved 0.2–0.4 mm accuracy, better than commonly used existing methods. Block-to-slice matching (Stage 2) showed great robustness in automatically identifying and inserting small tissue blocks into whole brain slices with 0.2 mm accuracy. Simulations demonstrated sub-voxel level accuracy (0.13 mm) of the slice-to-volume registration (Stage 3) algorithm, which was observed in over 200 actual brain slice registrations, compensating 3D slice deformations up to 6.5 mm. Stage 4 combined the previous stages and generated refined pixelwise aligned multi-modal histology-MRI stacks. Conclusions: Our open-source pipeline provides robust automation tools for registering stand-alone histology sections to MRI data with sub-voxel level precision, and the underlying framework makes it readily adaptable to a diverse range of microscopy-MRI studies.

The acquisition of MRI and histology in the same post-mortem tissue sample enables direct correlation between MRI and histologically-derived parameters. However, there still lacks a standardised automated pipeline to process histology data, with most studies relying on manual intervention. Here, we introduce an automated pipeline to extract a quantitative histological measure for staining density (stain area fraction, SAF) from multiple immunohistochemical (IHC) stains. The pipeline is designed to directly address key IHC artefacts related to tissue staining and slide digitisation. Here, the pipeline was applied to post-mortem human brain data from multiple subjects, relating MRI parameters (FA, MD, RD, AD, R2*, R1) to IHC slides stained for myelin, neurofilaments, microglia and activated microglia. Utilising high-quality MRI-histology co-registrations, we then performed whole-slide voxelwise comparisons (simple correlations, partial correlations and multiple regression analyses) between multimodal MRI- and IHC-derived parameters. The pipeline was found to be reproducible, robust to artefacts and generalisable across multiple IHC stains. Our partial correlation results suggest that some simple MRI-SAF correlations should be interpreted with caution, due to the co-localisation of other tissue features (e.g., myelin and neurofilaments). Further, we find activated microglia—a generic biomarker of inflammation—to consistently be the strongest predictor of high DTI FA and low RD, which may suggest sensitivity of diffusion MRI to aspects of neuroinflammation related to microglial activation, even after accounting for other microstructural changes (demyelination, axonal loss and general microglia infiltration). Together, these results show the utility of this approach in carefully curating IHC data and performing multimodal analyses to better understand microstructural relationships with MRI.

Post-mortem magnetic resonance imaging (MRI) provides the opportunity to acquire high-resolution datasets to investigate neuroanatomy and validate the origins of image contrast through microscopy comparisons. We introduce the Digital Brain Bank (open.win.ox.ac.uk/DigitalBrainBank), a data release platform providing open access to curated, multimodal post-mortem neuroimaging datasets. Datasets span three themes—Digital Neuroanatomist: datasets for detailed neuroanatomical investigations; Digital Brain Zoo: datasets for comparative neuroanatomy; and Digital Pathologist: datasets for neuropathology investigations. The first Digital Brain Bank data release includes 21 distinctive whole-brain diffusion MRI datasets for structural connectivity investigations, alongside microscopy and complementary MRI modalities. This includes one of the highest-resolution whole-brain human diffusion MRI datasets ever acquired, whole-brain diffusion MRI in fourteen nonhuman primate species, and one of the largest post-mortem whole-brain cohort imaging studies in neurodegeneration. The Digital Brain Bank is the culmination of our lab’s investment into post-mortem MRI methodology and MRI-microscopy analysis techniques. This manuscript provides a detailed overview of our work with post-mortem imaging to date, including the development of diffusion MRI methods to image large post-mortem samples, including whole, human brains. Taken together, the Digital Brain Bank provides cross-scale, cross-species datasets facilitating the incorporation of post-mortem data into neuroimaging studies.