作者: M JBurgess , J AAbildskov , A KEvans , R FWyatt , R LLux
DOI:
关键词: Temporal database 、 Data compression 、 Redundancy (information theory) 、 Compression (functional analysis) 、 Basis (linear algebra) 、 Channel (digital image) 、 Reduction (complexity) 、 Pattern recognition 、 Basis function 、 Artificial intelligence 、 Computer science
摘要: This paper describes use of the Karhunen-Loeve expansion to identify and reduce temporal redundancy in electrocardiographic body surface potential maps (192 leads recorded simultaneously at 1 kHz/channel for approximately 600 msec). Temporal data compression about 20 was obtained with accurate representation original data. Use separate sets orthonormal basis functions QRS ST-T provided a more than derived from QRST. Combined spatial described preceding paper, overall map 320 without significant loss accuracy or appearance. With both 100,000 numbers which typically comprise single cardiac complex were accurately represented by 216 coefficients. Using training s...