Abstract. This work presents the design and analysis of the first searchable symmetric encryption (SSE) protocol that supports conjunctive search and general Boolean queries on outsourced symmetricallyencrypted data and that scales to very large databases and arbitrarilystructured data including free text search. To date, work in this area has focused mainly on single-keyword search. For the case of conjunctive search, prior SSE constructions required work linear in the total number of documents in the database and provided good privacy only for structured attribute-value data, rendering these solutions too slow and inflexible for large practical databases.In contrast, our solution provides a realistic and practical trade-off between performance and privacy by efficiently supporting very large databases at the cost of moderate and well-defined leakage to the outsourced server (leakage is in the form of data access patterns, never as direct exposure of plaintext data or searched values). We present a detailed formal cryptographic analysis of the privacy and security of our protocols and establish precise upper bounds on the allowed leakage. To demonstrate the real-world practicality of our approach, we provide performance results of a prototype applied to several large representative data sets, including encrypted search over the whole English Wikipedia (and beyond).

Evidence before this study Using PubMed and Google Scholar the authors reviewed prior work on longitudinal neuroimaging markers of Alzheimer pathology with a focus on autosomal dominant Alzheimer disease (ADAD). We searched for all articles prior to October 31 st , 2017 with no language restrictions for the keywords Alzheimer's, Alzheimer, longitudinal, positron emission tomography, PET, MRI, atrophy, FDG, hypometabolism, familial, and autosomal. Theories proposed initially in 2010 by Jack and colleagues and revised in 2013 posited temporal trajectories of Alzheimer biomarkers relative to each other and clinical decline. Work by Bateman and colleagues in 2012, Benzinger and colleagues in 2013, and Fleisher and colleagues in 2015 depict such temporal ordering of biomarkers in ADAD populations derived from cross-sectional analyses. There was also a small subset of longitudinal ADAD studies, but these had one or more limitation such as small populations (n<50), examination of only one biomarker, not accounting for regional differences or correlations in the brain, or had a short duration of longitudinal followup. Added value of this studyOur study presents the first known work examining both the longitudinal temporal trajectories and spatial patterns of Alzheimer pathology in ADAD cohorts using neuroimaging. This work also presents the largest known cohort to date of ADAD individuals studied longitudinally with multiple neuroimaging biomarkers. Longitudinal analyses can provide a more accurate and powerful way to model the temporal emergence of pathology in ADAD. We find that mutation carriers first display Aβ accumulation, followed by hypometabolism, and finally structural atrophy; this is consistent with theoretical models and cross-sectional estimates from ADAD. Most importantly we consider such temporal relationships not in one singular summary measure, but characterize these trajectories throughout the brain. We found that the accrual of pathology varied throughout the brain and by modality in terms of the time of initial emergence and the rates of longitudinal change. These findings suggest region specific vulnerabilities to β-amyloidosis, metabolic decline, and atrophy that change over the course of the disease. Implications of all the available evidenceOur results build upon existing evidence characterizing biomarkers in clinical and preclinical Alzheimer disease. Our findings suggest that imaging biomarkers follow a sequential pattern, with β-amyloidosis, hypometabolism, and structural atrophy emerging more than twenty, fifteen, and ten years respectively before the expected onset of dementia. Although there is a general hierarchical pattern, there was considerable regional heterogeneity. Most commonly, regions demonstrated an increase in β-amyloidosis and structural atrophy, but there was not evidence of metabolic declines. Further, rather than being homogenous, the same biomarker often demonstrates different longitudinal trajectories across brain regions. Characterizing the temporal and regional dynamics...

We design and implement dynamic symmetric searchable encryption schemes that efficiently and privately search server-held encrypted databases with tens of billions of record-keyword pairs. Our basic theoretical construction supports single-keyword searches and offers asymptotically optimal server index size, fully parallel searching, and minimal leakage. Our implementation effort brought to the fore several factors ignored by earlier coarse-grained theoretical performance analyses, including lowlevel space utilization, I/O parallelism and goodput. We accordingly introduce several optimizations to our theoretically optimal construction that model the prototype's characteristics designed to overcome these factors. All of our schemes and optimizations are proven secure and the information leaked to the untrusted server is precisely quantified. We evaluate the performance of our prototype using two very large datasets: a synthesized census database with 100 million records and hundreds of keywords per record and a multi-million webpage collection that includes Wikipedia as a subset. Moreover, we report on an implementation that uses the dynamic SSE schemes developed here as the basis for supporting recent SSE advances, including complex search queries (e.g., Boolean queries) and richer operational settings (e.g., query delegation), in the above terabyte-scale databases.

No abstract