In the application of statistical process control, it is necessary to first segment data strings into homogeneous units. We will describe the 
multivariate method for determining homogeneous segments in the “ambulatory monitoring” data in which the wearer’s blood pressure 
and heart rate were measured and recorded into 4 MESOR measurements and 4 amplitude measurements. MANOVA canonical 
variables were calculated to establish whether more than one segment was present. The dynamic program algorithm was applied by 
fitting the data into the change-point model to estimate the number of segments and their boundaries. The optimal segmentation was 
provided. The conclusions of clinical issues and instrumental issues were therefore derived. The same analysis procedure was repeated 
on the subset of MESOR measurements and the subset of amplitude measurements as well.