ABSTRACT
In order to eliminate the intrinsic noise due to special structure of the wrist, a new curve fitting algorithm based on singular value decomposition (SVD) was developed to increase the measurement accuracy. This algorithm could be subdivided into SVD and curve fitting algorithm (SCFA). SVD was used to extract the dominant component of oscillation waves at wrist. Then oscillation amplitudes of dominant component and cuff pressure were used to determine arterial blood pressure (ABP) with curve fitting algorithm. To test the performance of SCFA, 45 subjects underwent the ABP measurement with different methods. The correlation coefficient between the pooled blood pressure measured by the auscultation and those by SCFA was 0.96. Comparison the results of SCFA with those of traditional curve fitting algorithm (TCFA), we found that the proposed SCFA could be used to reduce the partial intrinsic interference and efficiently improve the accuracy of the ABP at wrist.
