System and method of data compression and non-linear sampling for implantable and battery-powered devices

A compression system for adapting the effective sampling rate of an analog signal as a function of variations of said analog signal, comprising:

• sampling means for sampling said analog signal at a predetermined sampling rate;
• block means for holding a selected number of said samples;
• delta means for determining a delta value for each said held sample, said delta value being representative of the difference between each said sample and the prior sample;
• max means, for determining for each said block, the maximum delta value of said block;
• threshold means for setting a plurality of thresholds corresponding to predetermined different delta values;
• comparison means for comparing said maximum value with said thresholds and, on the basis of said comparing, keeping some of said delta values and discarding the others;
• and
• writing means for writing compressed data made up of said maximum value and said kept delta values for each said block, wherein said threshold means comprises means for storing a plurality of respective discrete threshold values, and said comparison means comprises means for determining which two of said threshold values said maximum delta value is between, and for discarding delta values as a function of said determining.

An implantable pacemaker system, said system having pacing means for pacing a patient, sensing means for sensing physiologic analog signals, and data means for deriving from said analog signals compressed digital data representative of said analog signals, said data means comprising:

• variation means for determining a measure of the variation of said analog signal;
• non-linear means for obtaining sample data from said analog signal at a non-linear sample rate, said non-linear means comprising means for determining said non-linear sample rate as a function of said variation measure;
• and
• writing means for writing data representative of said non-linear rate and said obtained samples, wherein said non-linear means comprises clock means for generating clock signals, sample means for sampling said analog signals coincident with clock signals, converter means for converting samples into digital data, and sample skip means for causing said sample means to skip sampling corresponding to Portions of said analog signals which contain reduced variations with respect to said variation measure, whereby the number of sampling operations is reduced to less than the number of said sample clocks further comprising means for obtaining data representative of said skipped samples, and skip compression means for compressing said skipped sample data.

A method of obtaining a compressed data representation of an analog signal, comprising:

• establishing threshold data relating to anticipated values of said analog signal;
• sampling said analog signal at an effective non-linear rate to obtain non-linear sample data, said non-linear sampling including comparing said analog signal to said threshold data and deriving said non-linear rate from said comparing;
• obtaining a measure of said non-linear rate corresponding to each sample, and converting said measures into rate data;
• losslessly compressing said sample data establishing discrete threshold data points, first sampling said analog signal at a fixed rate to obtain first sample data, comparing said first sample data with said threshold data points, and non-linear sampling said first sample data as a function of said comparison;
• and
• writing said rate data and said sample data.

An implantable medical device system comprising:

• comprising a pacemaker, said pacemaker having marker means for obtaining marker channel data,
• signal means for obtaining an analog physiological signal;
• and
• data means for obtaining a compressed digital data stream representative of said analog signal, said data means having non-linear means for obtaining non-linearly spaced samples of said analog signal, first compression means for providing compressed data representative of said samples, and second compression means for providing compressed data representative of the spacing of said samples, thereby providing a data stream representative of said analog signal and insertion means for inserting said marker channel data into said data stream.