Signal Extraction Technology

Steven J. Barker · Anesthesia & Analgesia · 1997

To the Editor: I read with interest the editorial by V. Pollard and D. S. Prough [1], regarding the clinical evaluation of the Masimo pulse oximeter by Dumas et al. [2]. While most of their points are well taken, they make one statement that I believe to be incorrect: "The two most annoying problems of current technology are signal failure in patients with poor peripheral perfusion and signal artifact in patients who have distal extremity movement. Masimo SET does little to improve the first problem." Actually, the Masimo technology may do quite a bit to improve performance during low perfusion situations. The Masimo signal processing algorithm, which is a completely new analysis of the two-wavelength light absorbance signal, is first described in detail in our laboratory study published in October 1996 [3]. The Masimo SET effectively computes the fluctuating noise signal common to both wavelengths and then removes this noise from the total signal by a form of adaptive digital filtering. This approach should improve pulse oximeter performance in all situations of low signal-to-noise ratio. In fact, the introduction of a periodic noise signal such as the artificial motions created in our laboratory experiment should represent a greater challenge to the instrument than does purely random noise. Quantitative studies of the performance of the Masimo instrument during low perfusion conditions are currently underway. In conclusion, Masimo's new algorithm should improve pulse oximeter performance in most low signal-to-noise situations, including that of low perfusion. Regarding the last statement in their editorial: "the enormous advance [pulse oximetry] has been shown to make no apparent difference in clinical outcome," Drs. Pollard and Prough have only pointed out the serious limitations of outcome studies in today's clinical practice. Steven J. Barker, PhD, MD Department of Anesthesiology; The University of Arizona Health Sciences Center; University of Arizona; Tucson, AZ 85724-5114

Read the paper · More papers on PaperTik