A technique for the derivation of a pulse oximetry signal using fractal dimension analysis of detected light signals. First and second light sources transmit light through the patient's finger or reflect light off the blood vessels in the patient's finger. A light detector detects light from each of the light sources and generates a measured intensity signal. The measured intensity signal includes the true intensity of light transmitted from each of the light sources as well as noise introduced during the measurement process. A data sample from each of the light sources is digitized and a set of equations developed as a function of a ratio value indicative of oxygen saturation in the patient. The fractal dimension is determined for the set of signal functions over the normal physiological range for the ratio value. Maximum and/or minimum fractal dimension values are calculated to determine the desired ratio values which are possible indicatives of the ratio of true physiological signals or noise signals. The ratio values are subsequently processed to determine the oxygen saturation within the patient.
Characteristics of time varying data, such as an electrical signal, are analyzed by converting the data from a temporal domain into a spatial domain pattern. Fractal analysis is performed on the spatial domain pattern, thereby producing a fractal dimension D.sub.F. The fractal dimension indicates the regularity of the time varying data.
Differential values, for use in blood oxygenation calculations, are determined based on multiple sample values for each channel of an oximetry system, each such value constituting a data point. In one implementation, each of these data points is defined by a sample window (220, 222 and 224), where the window includes, for example, 7-10 data points. That is, the data points within window (220, 222 or 224) are used to establish a differential value nominally associated with the data sample about which the window is centered. The differential value is calculated based on a mathematical model such as a weighted linear regression analysis. In this manner, output may be provided on a sample-by-sample basis while mitigating noise sensitivity.
A pulse oximetry method and system for improved motion correction is disclosed. The method/system provides for the use of a detector output signal to obtain a different plurality of differential absorption data sets in corresponding relation to each of a succession of measurement, wherein each of the data sets includes differential absorption values for light of a first wavelength and light of a second wavelength. The data sets are processed to obtain a relative motion estimate value for each measurement. When the relative motion estimate value for a given measurement falls within a predetermined range (i.e., corresponding with clinical motion), a corresponding blood analyte indicator value is adjusted in a predetermined manner, wherein the corresponding adjusted blood analyte indicator is employable to obtain at least one blood analyte concentration value. In one embodiment, blood analyte indicator values may be readily multiplied by a predetermined adjustment factor (i.e., when clinical motion is identified). The relative motion estimate value for a given measurement may be obtained by conducting a principal component analysis of the corresponding plurality of data sets relative to a corresponding best fit function therefor to obtain corresponding variance values V.sub.1, V.sub.2. The variance value V.sub.1, and/or V.sub.2 for a given current measurement may be employed to obtain a current motion estimate value. The current motion estimate value and the relative motion estimate value obtained for a prior low motion measurement (i.e., for which no adjustment was necessary) may be used to compute the relative motion estimate value for the current measurement. The variance values V.sub.1 and/or V.sub.2 are also employable to compute an ongoing, updated motion probability factor, wherein such factor may be used to adjust relative motion estimates values in instances of rapid tissue perfusion changes.
A system and a method monitor the useful and safe life of a pulse oximetry sensor. According to one embodiment, the system includes a sensor having a timer and a sensor life indicator. The sensor life indicator may advantageously include a visual alarm, an audio alarm, a vibrational alarm, a power down function, or the like. According to another embodiment, the system includes a sensor having a memory device storing a unique identifier. According to another embodiment, the system employs the oximeter for aspects of the timer or sensor life indicator.
A probe configured as a finger tip of a glove used to recognize tissue types in a human patient and adapted to perform both optical and electrical measurements to diagnose tissue type while scanned over the surface of the tissue. Hybrid electronic technology reduces the size of the diagnostic elements. The necessary electrical connections are led down the finger and the glove to a remotely positioned control unit. The control unit provides power to the diagnostic device and analyzes the signals received from the device.
