Data compression method for implantable medical devices
First Claim
1. A method of performing data compression of a physiologic waveform, comprising:
- obtaining a stream of data points within a data range, the data points corresponding to successive data samples of the physiologic waveform;
distributing the data range into major data bins;
determining within which major data bin each data point falls;
determining whether each data point in the stream of data points falls within the same major data bin as an immediately preceding data point of such data point;
retaining a compressed data point for each data point, the compressed data point having a revision level component and a division value, the revision level component indicating a revision condition, and the division component indicating the data bin which the associated data point falls within;
for each data point falling in the major data bin different from the major data bin of the immediately preceding data point, setting the division value of the respective compressed data point equal to the major data bin of such data point and setting the revision level component of the respective compressed data point to indicate a non-revision; and
for each data point falling within the same major data bin as the immediately preceding data point, dividing the major data bin into which such data point falls into a plurality of minor data bins unless the major data bin has already been divided into minor data bins, determining which minor data bin such data point falls within, setting the division value of the respective compressed data point equal to the minor data bin of such data point and setting the revision level component of the respective compressed data point to indicate a first-level revision.
1 Assignment
0 Petitions
Accused Products
Abstract
A method of data compression of physiological signals for use in implantable medical devices (IMDs) in which temporal information in the input data points is retained in the compressed sample points with a potential loss of amplitude precision. The data compression method includes distributing the stream of data samples or data points of the physiologic waveform into respective major data bins within the data range of the samples. Each major data bin can then be sub-divided into minor data bins. Compressed data points associated with the data samples are assigned values indicating which major data bin the original data point falls into, or which minor data bin the data point falls into if the data point immediately follows a data point in the same major data bin. Higher levels of revision beyond the major and minor bins may also be used to achieve greater levels of amplitude precision.
67 Citations
34 Claims
-
1. A method of performing data compression of a physiologic waveform, comprising:
-
obtaining a stream of data points within a data range, the data points corresponding to successive data samples of the physiologic waveform;
distributing the data range into major data bins;
determining within which major data bin each data point falls;
determining whether each data point in the stream of data points falls within the same major data bin as an immediately preceding data point of such data point;
retaining a compressed data point for each data point, the compressed data point having a revision level component and a division value, the revision level component indicating a revision condition, and the division component indicating the data bin which the associated data point falls within;
for each data point falling in the major data bin different from the major data bin of the immediately preceding data point, setting the division value of the respective compressed data point equal to the major data bin of such data point and setting the revision level component of the respective compressed data point to indicate a non-revision; and
for each data point falling within the same major data bin as the immediately preceding data point, dividing the major data bin into which such data point falls into a plurality of minor data bins unless the major data bin has already been divided into minor data bins, determining which minor data bin such data point falls within, setting the division value of the respective compressed data point equal to the minor data bin of such data point and setting the revision level component of the respective compressed data point to indicate a first-level revision. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)
-
-
16. A method of performing data compression of a physiologic waveform, comprising:
-
obtaining a stream of data points within a data range, the data points corresponding to successive data samples of the physiologic waveform;
defining the data range into data bins, each data bin at least once further dividing itself into successively narrower data bins, each successively narrower division of the data bins defining a respective revision level, the data range divided into data bins at an initial revision level, and a last narrower division of the data bins defining a narrowest revision level;
creating a compressed data point for each data point, the compressed data point having a revision level component and a division value;
proceeding one level at a time from the initial revision level towards the narrowest revision level, comparing each data point to its immediately preceding data point until reaching the first of the level where the compared data points fall in different data bins, the narrowest revision level, or one level beyond the revision level of the compressed data point associated with the immediately preceding data point, the first such level reached being defined as the selected revision level, and setting the revision level component of the compressed data point corresponding to such data point to the selected revision level and setting the division value of such compressed data point to the data bin at the selected revision level that such data point falls within. - View Dependent Claims (17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28)
-
-
29. A system for performing data compression of a physiologic waveform, comprising:
-
means for obtaining a stream of data points within a data range, the data points corresponding to successive data samples of the physiologic waveform;
means for defining the data range into data bins, each data bin at least once further dividing itself into successively narrower data bins, each successively narrower division of the data bins defining a respective revision level, the data range divided into data bins at an initial revision level, and a last narrower division of the data bins defining a narrowest revision level;
means for creating a compressed data point for each data point, the compressed data point having a revision level component and a division value;
proceeding one level at a time from the initial revision level towards the narrowest revision level, means for comparing each data point to its immediately preceding data point until reaching the first of the level where the compared data points fall in different data bins, the narrowest revision level, or one level beyond the revision level of the compressed data point associated with the immediately preceding data point, the first such level reached being defined as the selected revision level, and means for setting the revision level component of the compressed data point corresponding to such data point to the selected revision level and setting the division value of such compressed data point to the data bin at the selected revision level that such data point falls within. - View Dependent Claims (30, 31, 32, 33)
-
-
34. A method of performing data compression of a physiologic waveform, comprising:
-
obtaining a stream of data points within a data range, the data points corresponding to successive data samples of the physiologic waveform;
distributing the data range into major data bins;
subdividing each major data bin into minor data bins;
determining within which major data bin each data point falls;
retaining a compressed data point associated with each data point;
for successive data points falling in the same major data bin, assigning the compressed data point associated with the second of the successive data points a value equal to the major data bin within which the second of the successive data points falls; and
for successive data points falling in different major data bins, assigning the compressed data point associated with the second of the successive data points a value equal to the minor data bin within which the second of the successive data points falls.
-
Specification