Comments on "compression of the ECG by prediction or interpolation and entropy encoding".

A recentpaper1 discusses prediction and interpolation fordata compression of ECG. This communication points out that bothmethods canbe viewed as linearfiltering. They aretherefore equivalentand give the sameresult in terms of the amount of data compressionachieved. In a recentpaper,1 Ruttimann and Pipberger seem toclaimthat predictive coding, where the residual isobtained by inManuscript received March25, 1980; revised July31, 1980.P.-O.Bbrjesson is with the Department of Biomedical Engineering,Malmb General Hospital,Mahmi, Sweden.G. Einarsson is with the Division of Telecommunication Theory,Department ofElectrical Engineering, University of Lund, Lund,Sweden.0. Pahlm is with the Department of Clinical Physiology, Universityof Lund, Lund, Sweden.1U. E. Ruttimann and H.V. Pipberger, IEEE Trans.Biomed. Eng.,vol.BME-26, pp.613-623, Nov.1979.Author's Reply2 URS E. RUTTIMANN AND HUBERT V. PIPBERGER We agree with the comment made above [ 1 ]that both pre-diction and interpolation can be viewed aslinear filtering oper2Manuscript received July 8, 1980.U. E. Ruttimann is with the Department of Cardiology, Children'sHospital National Medical Center,Washington, DC 20010.H. V. Pipberger is with the Department of Clinical Engineering andthe Department of Medicine, The George Washington University Medi-cal Center, Washington, DC 20037. 0018-9294/80/1100-0674$00.75 X 1980 IEEE674