Data compression for multidimensional time series data
Inventors
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11276204-B1
Publication Date
2022-03-15
Expiration Date
Abstract
Described herein are computer-implemented methods for compressing sparse multidimensional ordered series data. In particular, these methods and apparatuses for performing them (including software) may be particularly well suited to efficiently compressing spectrographic data.
Core Innovation
The invention provides a computer-implemented compression framework for sparse multidimensional ordered series data. The data are divided into a plurality of local regions, and each local region includes one or more indexed data sets, each indexed data set comprising an index (n) and one or more variables indexed by the index (n).
A predictor is identified that calculates each variable as a function of the index (n), current local region data, and at least one previous local region data. Correlated predicted data is generated from the previous local region data, and the current local region data is adjusted by subtracting a scaled predicted correlated data when a level of correlation exceeding a threshold exists between the current local region data and the previous local region data.
The adjusted current local region data is encoded into a compressed stream, including an optimum scale factor. The approach supports decompression that restores correlated portions using decoded optimum scale factors and prior-region information, with optional streaming or direct access to selected regions. The document describes high compression ratios and quality metrics for sparse multidimensional ordered series data, including multiplexed fluorescence microscopy, spectrographic data, histopathological imaging data, and Z-stack/time-sequence microscopy.
Claims Coverage
The patent document includes three independent claims: compressing sparse multidimensional ordered series data using predictor-based correlated subtraction with thresholded correlation, applying the same concept to histopathological imaging data with spatial indexing and ordered processing, and a corresponding system implemented via instructions on a processor. Across these independent claims, the core coverage centers on dividing into local regions, computing indexed variables via a predictor using current and previous local regions, conditionally subtracting scaled predicted correlated data based on correlation exceeding a threshold, and encoding adjusted data including an optimum scale factor into a compressed stream.
Dividing sparse multidimensional ordered series data into local regions with indexed data sets
Dividing a multidimensional ordered series data into a plurality of local regions, wherein the data in each local region comprise one or more indexed data sets, each indexed data set comprising an index (n) within a given local region of the sparse multidimensional ordered series data and one or more variables that are indexed by the index (n).
Predictor computing variables from index, current local region, and previous local region
Identifying a predictor that calculates each of the one or more variables as a function of the index (n), a current local region data, and at least one previous local region data.
Thresholded correlation-based adjustment by subtracting scaled predicted correlated data
Adjusting the current local region data by subtracting a scaled predicted correlated data based on the at least one previous local region data when a level of correlation exceeding a threshold exists between the current local region data and the previous local region data.
Encoding adjusted data with an optimum scale factor into a compressed stream
Encoding the adjusted current local region data, including an optimum scale factor, into a compressed stream.
Processing histopathological imaging local regions in an order
Dividing a multidimensional ordered series data comprising histopathological imaging data into a plurality of local regions, wherein the data in each local region comprise one or more indexed data sets, each indexed data set comprising a spatial index (n) within a given local region of the histopathological imaging data and one or more variables that are indexed by the spatial index (n); and processing the plurality of local regions in an order.
System implementation for predictor-based thresholded compression
A system for compressing sparse multidimensional ordered series data comprising a non-transitory computer-readable medium with instructions stored thereon, that when executed by a processor, cause the processor to divide a multidimensional ordered series data into a plurality of local regions; identify a predictor that calculates each of the one or more variables as a function of the index (n), a current local region data, and at least one previous local region data; adjust the current local region data by subtracting a scaled predicted correlated data based on the at least one previous local region data when a level of correlation exceeding a threshold exists between the current local region data and the previous local region data; and encode the adjusted current local region data, including an optimum scale factor, into a compressed stream.
Across the independent claims, the inventive concept is conditional compression of sparse multidimensional ordered series data by computing indexed variables from a predictor using current and previous local regions, subtracting scaled predicted correlated data only when correlation exceeds a threshold, and encoding the adjusted region data into a compressed stream while including an optimum scale factor.
Stated Advantages
Very high compression ratios for sparse multidimensional ordered series data, including up to approximately 330-fold.
High-quality reconstruction/representation as measured by quality metrics such as PSNR, including approximately 56.25 dB for multiplexed fluorescence microscopy.
Documented Applications
Compression of sparse multidimensional ordered series data for spectrographic data including mass spectrometry (MS).
Compression of histopathological imaging data.
Compression for microscopy data including multiplexed fluorescence microscopy and Z-stack/time-sequence microscopy.
Comparisons versus JPEG artifact reduction for image-related quality evaluation [procedural detail omitted for safety].
Interested in licensing this patent?