Contrastive learning make great success in multiple large language vision self-supervised tasks and gain the same attention in Multivariate Time Series (MTS) data analysis field. However, many existing MTS works focus on improving the representation of learning capabilities by a single supervised learning paradigm, inevitably ignoring the potential for jointly optimizing models across diverse supervised signals. To remedy the limitations, we propose a novel Class-Aware Spatio-Temporal Graph Contrasting Learning (CA-STGCL) framework, with a Multivariate Time Series Feature Extraction (MFE) Encoder-Decoder structure, and Contrastive Learning (CL) Decoder. Specifically, we first design an MFE Encoder-Decoder module to process the raw data and use the obtained representations to acquire different supervised signals. Then, we introduce an effective CL Decoder to obtain two distinct additional supervised signals by two data augmentation strategies and contrastive approaches. Further, we jointly optimize the model using three different supervised signals, i.e., the original representations signals by the MFE Encoder-Decoder module, and temporal and contextual contrasting signals from the CL Decoder. Finally, extensive experiments and comparisons are conducted on multiple dominant MTS datasets, e.g., UCI-HAR, C-MAPSS, and ISRUC-S3. The results show that the proposed framework can surpass current promising methods and achieve the state-of-the-art performance(1).

3D spatial data management is increasingly vital across various application scenarios, such as GIS, digital twins, human atlases, and tissue imaging. However, the inherent complexity of 3D spatial data, primarily represented by 3D geometries in real-world applications, hinders the efficient evaluation of spatial relationships through resource-intensive geometric computations. Geometric simplification algorithms have been developed to reduce the complexity of 3D representations, albeit at the cost of querying accuracy. Previous work has aimed to address precision loss by leveraging the spatial relationship between the simplified and original 3D object representations. However, this approach relied on specialized geometric simplification algorithms tailored to regions with specific criteria. In this paper, we introduce a novel approach to achieve highly efficient and accurate 3D spatial queries, incorporating geometric computation and simplification. We present a generalized progressive refinement methodology applicable to general geometric simplification algorithms, involving accurate querying of 3D geometry data using low-resolution representations and simplification extents quantified using Hausdorff distances at the facet level. Additionally, we propose techniques for calculating and storing Hausdorff distances efficiently. Extensive experimental evaluations validate the effectiveness of the proposed method which outperforms state-of-the-art systems by a factor of 4 while minimizing computational and storage overhead.

We released open-source software Hadoop-GIS in 2011, and presented and published the work in VLDB 2013. This work initiated the development of a new spatial data analytical ecosystem characterized by its large-scale capacity in both computing and data storage, high scalability, compatibility with low-cost commodity processors in clusters and open-source software. After more than a decade of research and development, this ecosystem has matured and is now serving many applications across various fields. In this paper, we provide the background on why we started this project and give an overview of the original Hadoop-GIS software architecture, along with its unique technical contributions and legacy. We present the evolution of the ecosystem and its current state-ofthe-art, which has been influenced by the Hadoop-GIS project. We also describe the ongoing efforts to further enhance this ecosystem with hardware accelerations to meet the increasing demands for low latency and high throughput in various spatial data analysis tasks. Finally, we will summarize the insights gained and lessons learned over more than a decade in pursuing high-performance spatial data analytics.

Spatial join has become a frequently used yet resource-intensive operation in geospatial applications, driven by the increasing volume and complexity of geospatial data. With Hadoop and Spark becoming the de facto standard platforms for distributed computing, scalable spatial data processing is primarily achieved by partitioning the input space to form parallel units on these platforms. Effective spatial data partitioning is critical for task parallelization and load balancing, but it faces significant challenges due to data skew and the geometric and topological complexity of spatial objects, particularly in supporting spatial joins. This paper examines the interplay among query performance, spatial data partitioning, query types, data, and system characteristics. We qualitatively and quantitatively analyze the features of representative partitioning algorithms that impact overall query performance. Along with these analyses, we propose a data sampling-based approach for selecting optimized partitioning strategies. Extensive experiments on large and complex datasets using MapReduce frameworks are conducted to validate the correctness of our analysis and the effectiveness of our optimization approach.

Geometric computation can be heavy duty for spatial queries, in particular for complex geometries such as polygons with many edges based on a vector-based representation. While many techniques have been provided for spatial partitioning and indexing, they are mainly built on minimal bounding boxes or other approximation methods, which will not mitigate the high cost of geometric computation. In this paper, we propose a novel vector-raster hybrid approach through rasterization, where pixel-centric rich information is preserved to help not only filtering out more candidates but also reducing geometry computation load. Based on the hybrid model, we develop an efficient rasterization based ray casting method for point-in-polygon queries and a circle buffering method for point-to-polygon distance calculation, which is a common operation for distance based queries. Our experiments demonstrate that the hybrid model can boost the performance of spatial queries on complex polygons by up to one order of magnitude.

BACKGROUND:Opioid addiction and overdose have a large burden of disease and mortality in New York State (NYS). The medication naloxone can reverse an overdose, and buprenorphine can treat opioid use disorder. Efforts to increase the accessibility of both medications include a naloxone standing order and a waiver program for prescribing buprenorphine outside a licensed drug treatment program. However, only a slim majority of NYS pharmacies are listed as participating in the naloxone standing order, and less than 7% of prescribers in NYS have a buprenorphine waiver. Therefore, there is a significant opportunity to increase access. OBJECTIVE:Identifying the geographic regions of NYS that are farthest from resources can help target interventions to improve access to naloxone and buprenorphine. To maximize the efficiency of such efforts, we also sought to determine where these underserved regions overlap with the largest numbers of actual patients who have experienced opioid overdose. METHODS:We used address data to assess the spatial distribution of naloxone pharmacies and buprenorphine prescribers. Using the home addresses of patients who had an opioid overdose, we identified geographic locations of resource deficits. We report findings at the high spatial granularity of census tracts, with some neighboring census tracts merged to preserve privacy. RESULTS:We identified several hot spots, where many patients live far from the nearest resource of each type. The highest density of patients in areas far from naloxone pharmacies was found in eastern Broome county. For areas far from buprenorphine prescribers, we identified subregions of Oswego county and Wayne county as having a high number of potentially underserved patients. CONCLUSIONS:Although NYS is home to thousands of naloxone pharmacies and potential buprenorphine prescribers, access is not uniform. Spatial analysis revealed census tract areas that are far from resources, yet contain the residences of many patients who have experienced opioid overdose. Our findings have implications for public health decision support in NYS. Our methods for privacy can also be applied to other spatial supply-demand problems involving sensitive data.

The human body is made up of about 37 trillion cells (adults). Each cell has its own unique role and is affected by its neighboring cells and environment. The NIH Human BioMolecular Atlas Program (HuBMAP) aims at developing a 3D atlas of human body consisting of organs, vessels, tissues to singe cells with all 3D spatially registered in a single 3D human atlas using tissues obtained from normal individuals across a wide range of ages. A critical step of building the atlas is to register 3D tissue blocks in real-time to the right location of a human organ, which itself consists of complex 3D sub-structures. The complexity of the 3D organ model, e.g., 35 meshes for a typical kidney, poses a significant computational challenge for the registration. In this paper, we propose a comprehensive framework TICKET (TIssue bloCK rEgisTration) to support tissue block registration for 3D human atlas, including (1) 3D mesh pre-processing, (2) spatial queries on intersection relationship and (3) intersection volume computation between organs and tissue blocks. To minimize search space and computation cost, we develop multi-level indexing at both the anatomical structure level and mesh level, and utilize OpenMP for parallel computing. Considering cuboid based shape of the tissue block, we propose an efficient voxelization-based method to estimate the intersection volume. Our experiments demonstrate that the proposed framework is efficient and practical. TICKET is being integrated into the HuBMAP CCF registration portal [1].

3D spatial data has been generated at an extreme scale from many emerging applications, such as high definition maps for autonomous driving and 3D Human BioMolecular Atlas. In particular, 3D digital pathology provides a revolutionary approach to map human tissues in 3D, which is highly promising for advancing computer-aided diagnosis and understanding diseases through spatial queries and analysis. However, the exponential increase of data at 3D leads to significant I/O, communication, and computational challenges for 3D spatial queries. The complex structures of 3D objects such as bifurcated vessels make it difficult to effectively support 3D spatial queries with traditional methods. In this article, we present our work on building an efficient and scalable spatial query system, iSPEED, for large-scale 3D data with complex structures. iSPEED adopts effective progressive compression for each 3D object with successive levels of detail. Further, iSPEED exploits structural indexing for complex structured objects in distance-based queries. By querying with data represented in successive levels of details and structural indexes, iSPEED provides an option for users to balance between query efficiency and query accuracy. iSPEED builds in-memory indexes and decompresses data on-demand, which has a minimal memory footprint. iSPEED provides a 3D spatial query engine that can be invoked on-demand to run many instances in parallel implemented with, but not limited to, MapReduce. We evaluate iSPEED with three representative queries: 3D spatial joins, 3D nearest neighbor query, and 3D spatial proximity estimation. The extensive experiments demonstrate that iSPEED significantly improves the performance of existing spatial query systems.

Large-scale three-dimensional spatial data has gained increasing attention with the development of self-driving, mineral exploration, CAD, and human atlases. Such 3D objects are often represented with a polygonal model at high resolution to preserve accuracy. This poses major challenges for 3D data management and spatial queries due to the massive amounts of 3D objects, e.g., trillions of 3D cells, and the high complexity of 3D geometric computation. Traditional spatial querying methods in the Filter-Refine paradigm have a major focus on indexing-based filtering using approximations like minimal bounding boxes and largely neglect the heavy computation in the refinement step at the intra-geometry level, which often dominates the cost of query processing. In this paper, we introduce 3DPro, a system that supports efficient spatial queries for complex 3D objects. 3DPro uses progressive compression of 3D objects preserving multiple levels of details, which significantly reduces the size of the objects and has the data fit into memory. Through a novel Filter-Progressive-Refine paradigm, 3DPro can have query results returned early whenever possible to minimize decompression and geometric computations of 3D objects in higher resolution representations. Our experiments demonstrate that 3DPro out-performs the state-of-the-art 3D data processing techniques by up to an order of magnitude for typical spatial queries.

With the advent of IoT and emerging 5G technology, real-time streaming data are being generated at unprecedented speed and volume having both temporal and spatial dimensions. Effective analysis at such scale and speed require support for dynamically adjusting querying capabilities in real-time. In the spatio-temporal domain, this warrants data as well as query optimization strategies especially for objects with changing motion states. Contemporary spatio-temporal data stream management systems in the distributed domain are mostly dominated by specify-once-apply-continuously query model. Any modification in query state requires query restart limiting system responsiveness and producing outdated or in worst case erroneous results. In this paper, we propose adaptations of principles from streaming databases, spatial data management, and distributed computing to support dynamic spatio-temporal query processing over high-velocity big data streams. We first formulate a set of spatio-temporal data types and functions to seamlessly handle changes in distributed query states. We develop a comprehensive set of streaming spatio-temporal querying methods and propose geohash based dynamic spatial partitioning for effective parallel processing. We implement a prototype on top of Apache Flink, where the in-memory stream processing fits nicely with our spatio-temporal models. Comparative evaluation of our prototype demonstrates the effectiveness of our strategy by maintaining high consistent processing rates for both stationary as well as moving queries over high velocity spatio-temporal big data streams.

Contact tracing is gaining its importance in controlling the spread of COVID-19. However, the enormous volume of the frequently sampled tracing data brings major challenges for real-time processing. In this paper, we propose a GPU-based real-time contact tracing system based on spatial proximity queries with temporal constraints using location data. We provide dynamic indexing of moving objects using an adaptive partitioning schema on GPU with extremely low overhead. Our system optimizes the retrieval of contacted pairs to match both the requirements of contact tracing scenarios and GPU centered parallelism. We propose an efficient contacts evaluation mechanism to keep only the spatially and temporally valid contacts. Our experiments demonstrate that the system can achieve sub-second level response for large-scale contact tracing of tens of millions of people, with two magnitudes of performance boost over CPU based approach.

Spatial cross-matching operation over geospatial polygonal datasets is a highly compute-intensive yet an essential task to a wide array of real-world applications. At the same time, modern computing systems are typically equipped with multiple processing units capable of task parallelization and optimization at various levels. This mandates for the exploration of novel strategies in the geospatial domain focusing on efficient utilization of computing resources, such as CPUs and GPUs. In this paper, we present a CPU-GPU hybrid platform to accelerate the cross-matching operation of geospatial datasets. We propose a pipeline of geospatial subtasks that are dynamically scheduled to be executed on either CPU or GPU. To accommodate geospatial datasets processing on GPU using pixelization approach, we convert the floating point-valued vertices into integer-valued vertices with an adaptive scaling factor as a function of the area of minimum bounding box. We present a comparative analysis of GPU enabled cross-matching algorithm implementation in CUDA and OpenACC accelerated C++. We test our implementations over Natural Earth Data and our results indicate that although CUDA based implementations provide better performance, OpenACC accelerated implementations are more portable and extendable while still providing considerable performance gain as compared to CPU. We also investigate the effects of input data size on the IO / computation ratio and note that a larger dataset compensates for IO overheads associated with GPU computations. Finally, we demonstrate that an efficient cross-matching comparison can be achieved with a cost-effective GPU.

Hepatic steatosis droplet quantification with histology biopsies has high clinical significance for risk stratification and management of patients with fatty liver diseases and in the decision to use donor livers for transplantation. However, pathology reviewing processes, when conducted manually, are subject to a high inter- and intra-reader variability, due to the overwhelmingly large number and significantly varying appearance of steatosis instances. This process is challenging as there is a large number of overlapped steatosis droplets with either missing or weak boundaries. In this study, we propose a deep-learning-based region-boundary integrated network for precise steatosis quantification with whole slide liver histopathology images. The proposed model consists of two sequential steps: a region extraction and a boundary prediction module for foreground regions and steatosis boundary prediction, followed by an integrated prediction map generation. Missing steatosis boundaries are next recovered from the predicted map and assembled from adjacent image patches to generate results for the whole slide histopathology image. The resulting steatosis measures both at the pixel level and steatosis object-level present strong correlation with pathologist annotations, radiology readouts and clinical data. In addition, the segregated steatosis object count is shown as a promising alternative measure to the traditional metrics at the pixel level. These results suggest a high potential of artificial intelligence-assisted technology to enhance liver disease decision support using whole slide images. Accurate quantification of steatosis in liver biopsies is a key step in the treatment of patients with fatty liver diseases. To assist pathologists for such analysis tasks, we develop a novel deep learning-based framework to segment overlapped steatosis droplets in whole slide liver biopsy images. Quantitative measurements of steatosis at both pixel and object-level present strong correlation with clinical data, suggesting its potential for clinical decision support.

Digital imaging plays a critical role for image guided diagnosis and clinical trials, and the amount of image data is fast growing. There are two major requirements for image data management: scalability for massive scales and support of comprehensive queries. Traditional Picture Archiving and Communication Systems (PACS for short) are based on relational data management systems and suffer from limited scalability and query support. Therefore, new systems that support fast, scalable and comprehensive queries on image data are highly demanded. In this paper, we introduce two alternative approaches: DCMRL/XMLStore (RL/XML for short)—a parallel, hybrid relational and XML data management approach, and DCMDocStore (DOC for short)—a NoSQL document store approach. DCMRL/XMLStore manages DICOM images as binary large objects and metadata as relational tables and XML documents based on IBM DB2, which is parallelized through data partitioning. DCMDocStore manages DICOM metadata as JSON objects, and DICOM images as encoded attachments in MongoDB running on multiple nodes. We have delivered two open source systems DCMRL/XMLStore and DCMDocStore. Both systems support scalable data management and comprehensive queries. We also evaluated them with nearly one million DICOM images from National Biomedical Imaging Archive. The results show that, DCMDocStore demonstrates high data loading speed, high scalability and fault tolerance. DCMRL/XMLStore provides efficient queries, but comes with slower data loading. Traditional PACS systems have inherent limitations on flexible queries and scalability for massive amount of images.

LSM-tree has been widely used in data management production systems for write-intensive workloads. However, as read and write workloads co-exist under LSM-tree, data accesses can experience long latency and low throughput due to the interferences to buffer caching from the compaction, a major and frequent operation in LSM-tree. After a compaction, the existing data blocks are reorganized and written to other locations on disks. As a result, the related data blocks that have been loaded in the buffer cache are invalidated since their referencing addresses are changed, causing serious performance degradations. To re-enable high-speed buffer caching during intensive writes, we propose Log-Structured buffered-Merge tree (simplified as LSbM-tree) by adding a compaction buffer on disks to minimize the cache invalidations on buffer cache caused by compactions. The compaction buffer efficiently and adaptively maintains the frequently visited datasets. In LSbM, strong locality objects can be effectively kept in the buffer cache with minimum or no harmful invalidations. With the help of a small on-disk compaction buffer, LSbM achieves a high query performance by enabling effective buffer caching, while retaining all the merits of LSM-tree for write-intensive data processing and providing high bandwidth of disks for range queries. We have implemented LSbM based on LevelDB. We show that with a standard buffer cache and a hard disk, LSbM can achieve 2x performance improvement over LevelDB. We have also compared LSbM with other existing solutions to show its strong cache effectiveness.

Recent advancements in systematic analysis of high resolution whole slide images have increase efficiency of diagnosis, prognosis and prediction of cancer and important diseases. Due to the enormous sizes and dimensions of whole slide images, the analysis requires extensive computing resources which are not commonly available. Images have to be tiled for processing due to computer memory limitations, which lead to inaccurate results due to the ignorance of boundary crossing objects. Thus, we propose a generic and highly scalable cloud-based image analysis framework for whole slide images. The framework enables parallelized integration of image analysis steps, such as segmentation and aggregation of micro-structures in a single pipeline, and generation of final objects manageable by databases. The core concept relies on the abstraction of objects in whole slide images as different classes of spatial geometries, which in turn can be handled as text based records in MapReduce. The framework applies an overlapping partitioning scheme on images, and provides parallelization of tiling and image segmentation based on MapReduce architecture. It further provides robust object normalization, graceful handling of boundary objects with an efficient spatial indexing based matching method to generate accurate results. Our experiments on Amazon EMR show that MaReIA is highly scalable, generic and extremely cost effective by benchmark tests.

Systematic analysis of high resolution whole slide images enables more effective diagnosis, prognosis and prediction of cancer and other important diseases. Due to the enormous sizes and dimensions of whole slide images, the analysis requires extensive computing resources which are not commonly available. Images have to be divided into smaller regions for processing due to computer memory limitations, which will lead to inaccurate results due to the ignorance of boundary crossing objects. In this paper, we propose a highly scalable and cost effective MapReduce based image analysis framework for whole slide image processing, and provide a cloud based implementation. The framework takes a grid-based overlapping partitioning scheme, and provides parallelization of image segmentation based on MapReduce. It provides graceful handling of boundary objects with a highly efficient spatial indexing based matching method, thus avoiding loss of accuracy due to partitioning. We demonstrate that the system achieves high scalability and is cost-effective – our experiments demonstrate that it costs less than fifteen cents to analyze one image on average using Amazon Elastic MapReduce.

LSM-tree has been widely used in data management production systems for write-intensive workloads. However, as read and write workloads co-exist under LSM-tree, data accesses can experience long latency and low throughput due to the interferences to buffer caching from the compaction, a major and frequent operation in LSM-tree. After a compaction, the existing data blocks are reorganized and written to other locations on disks. As a result, the related data blocks that have been loaded in the buffer cache are invalidated since their referencing addresses are changed, causing serious performance degradations. In order to re-enable high-speed buffer caching during intensive writes, we propose Log-Structured buffered-Merge tree (simplified as LSbM-tree) by adding a compaction buffer on disks, to minimize the cache invalidations on buffer cache caused by compactions. The compaction buffer efficiently and adaptively maintains the frequently visited data sets. In LSbM, strong locality objects can be effectively kept in the buffer cache with minimum or without harmful invalidations. With the help of a small on-disk compaction buffer, LSbM achieves a high query performance by enabling effective buffer caching, while retaining all the merits of LSM-tree for write-intensive data processing, and providing high bandwidth of disks for range queries. We have implemented LSbM based on LevelDB. We show that with a standard buffer cache and a hard disk, LSbM can achieve 2x performance improvement over LevelDB. We have also compared LSbM with other existing solutions to show its strong effectiveness.

LSM-tree has been widely used in cloud computing systems by Google, Facebook, and Amazon, to achieve high performance for write-intensive workloads. However, in LSM-tree, random key-value queries can experience long latency and low throughput due to the interference from the compaction, a basic operation in the algorithm, to caching. LSM-tree relies on frequent compaction operations to merge data into a sorted structure. After a compaction, the original data are reorganized and written to other locations on the disk. As a result, the cached data are invalidated since their referencing addresses are changed, causing serious performance degradations. We propose dLSM in order to re-enable high-speed caching during intensive writes. dLSM is an LSM-tree with a compaction buffer on the disk, working as a cushion to minimize the cache invalidation caused by compactions. The compaction buffer maintains a series of snapshots of the frequently compacted data, which represent a consistent view of the corresponding data in the underlying LSM-tree. Being updated in a much lower rate than that of compactions, data in the compaction buffer are almost stationary. In dLSM, an object is referenced by the disk address of the corresponding block either in the compaction buffer for frequently compacted data, or in the underlying LSM-tree for infrequently compacted data. Thus, hot objects can be effectively kept in the cache without harmful invalidations. With the help of a small on-disk compaction buffer, dLSM achieves a high query performance by enabling effective caching, while retaining all merits of LSM-tree for write-intensive data processing. We have implemented dLSM based on LevelDB. Our evaluations show that with a standard DRAM cache, dLSM can achieve 5--8x performance improvement over LSM with the same cache on HDD storage.