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.