Blood glucose calibration system

Description

PRIORITY CLAIM AND REFERENCE TO RELATED APPLICATIONS

The present application claims priority benefit under 35 U.S.C. §119(e) to U.S. Provisional Patent Application Ser. No. 61/579,460, filed Dec. 22, 2011, titled Blood Glucose Calibration System, hereby incorporated in its entirety by reference herein.

BACKGROUND OF THE INVENTION

Medical device manufacturers are continually increasing the processing capabilities of patient monitors, specifically of patient monitors that process signals based on attenuation of light by patient tissue. In general, such patient monitoring systems include one or more optical sensors that irradiate tissue of a patient and one or more photodetectors that detect the radiation after attenuation thereof by the tissue. The sensor communicates the detected signal to a patient monitor, where the monitor often removes noise and preprocesses the signal. Advanced signal processors then perform time domain and/or frequency domain processing to determine measurements of blood constituents and other physiological parameters of the patient.

Manufacturers have advanced basic pulse oximeters that determine measurements for blood oxygen saturation (“SpO2”), pulse rate (“PR”) and pethysmographic information, to read-through-motion oximeters, to co-oximeters that determine measurements of many constituents of circulating blood. For example, Masimo Corporation of Irvine Calif. (“Masimo”) manufactures pulse oximetry systems including Masimo SET® low noise optical sensors and read through motion pulse oximetry monitors for measuring Sp02, PR, perfusion index (“PI”) and others. Masimo sensors include any of LNOP®, LNCS®, SofTouch™ and Blue™ adhesive or reusable sensors. Masimo oximetry monitors include any of Rad-8®, Rad-5®, Rad®-5v or SatShare® monitors.

Many innovations improving the measurement of blood constituents are described in at least U.S. Pat. Nos. 6,770,028; 6,658,276; 6,157,850; 6,002,952; 5,769,785 and 5,758,644, which are assigned to Masimo and are incorporated by reference herein. Corresponding low noise optical sensors are disclosed in at least U.S. Pat. Nos. 6,985,764; 6,088,607; 5,782,757 and 5,638,818, assigned to Masimo and hereby incorporated in their entirety by reference herein.

Masimo also manufactures more advanced co-oximeters including Masimo Rainbow® SET, which provides measurements in addition to Sp02, such as total hemoglobin (SpHb™), oxygen content (SpCO™), methemoglobin (SpMet®), carboxyhemoglobin (SpCO®) and PVI®. Advanced blood parameter sensors include Masimo Rainbow® adhesive, ReSposable™ and reusable sensors. Masimo's advanced blood parameter monitors include Masimo Radical-7™, Rad87™, and Rad57™ monitors as well as Pronto and Pronto-7 spot check monitors.

Innovations relating to these more advanced blood parameter measurement systems are described in at least U.S. Pat. Nos. 7,647,083; 7,729,733; U.S. Pat. Pub. Nos. 2006/0211925; and 2006/0238358, assigned to Cercacor Laboratories of Irvine, Calif. (“Cercacor”) and hereby incorporated in their entirety by reference herein.

Such advanced pulse oximeters, low noise sensors and advanced blood parameter systems have gained rapid acceptance in a wide variety of medical applications, including surgical wards, intensive care and neonatal units, general wards, home care, physical training, and virtually all types of monitoring scenarios.

Advanced pulse oximetry is described in at least U.S. Pat. Nos. 6,770,028; 6,658,276; 6,157,850; 6,002,952; 5,769,785 and 5,758,644, which are assigned to Masimo Corporation (“Masimo”) of Irvine, Calif. and are incorporated in their entirety by reference herein. Corresponding low noise optical sensors are disclosed in at least U.S. Pat. Nos. 6,985,764; 6,813,511; 6,792,300; 6,256,523; 6,088,607; 5,782,757 and 5,638,818, which are also assigned to Masimo and are also incorporated in their entirety by reference herein. Advanced pulse oximetry systems including Masimo SET® low noise optical sensors and read through motion pulse oximetry monitors for measuring SpO₂, pulse rate (PR) and perfusion index (PI) are available from Masimo. Optical sensors include any of Masimo LNOP®, LNCS®, SofTouch™ and BIue™ adhesive or reusable sensors. Pulse oximetry monitors include any of Masimo Rad-8®, Rad-5®, Rad®-5v or SatShare® monitors.

Advanced blood parameter measurement systems are described in at least U.S. Pat. No. 7,647,083, filed Mar. 1, 2006, titled Multiple Wavelength Sensor Equalization; U.S. Pat. No. 7,729,733, filed Mar. 1, 2006, titled Configurable Physiological Measurement System; U.S. Pat. Pub. No. 2006/0211925, filed Mar. 1, 2006, titled Physiological Parameter Confidence Measure and U.S. Pat. Pub. No. 2006/0238358, filed Mar. 1, 2006, titled Noninvasive Multi-Parameter Patient Monitor, all assigned to Cercacor Laboratories, Inc., Irvine, Calif. (Cercacor) and all incorporated in their entirety by reference herein. An advanced parameter measurement system that includes acoustic monitoring is described in U.S. Pat. Pub. No. 2010/0274099, filed Dec. 21, 2009, titled Acoustic Sensor Assembly, assigned to Masimo and incorporated in its entirety by reference herein.

Advanced blood parameter measurement systems include Masimo Rainbow® SET, which provides measurements in addition to SpO₂, such as total hemoglobin (SpHb™), oxygen content (SpOC™), methemoglobin (SpMet®), carboxyhemoglobin (SpCO®) and PVI®. Advanced blood parameter sensors include Masimo Rainbow® adhesive, ReSposable™ and reusable sensors. Advanced blood parameter monitors include Masimo Radical-7™, Rad87™ and Rad57™ monitors, all available from Masimo. Advanced parameter measurement systems may also include acoustic monitoring such as acoustic respiration rate (RRa™) using a Rainbow Acoustic Sensor™ and Rad87™ monitor, available from Masimo. Such advanced pulse oximeters, low noise sensors and advanced parameter systems have gained rapid acceptance in a wide variety of medical applications, including surgical wards, intensive care and neonatal units, general wards, home care, physical training, and virtually all types of monitoring scenarios.

SUMMARY OF THE INVENTION

One aspect of a blood glucose calibration system has an optical sensor, a composite parameter generator, a glucose estimator, a strip meter and a glucose calibrator. The optical sensor illuminates a tissue site with multiple wavelength optical radiation and outputs multi-stream data responsive to the optical radiation after attenuation by blood flow within the tissue site. The composite parameter generator is responsive to the multi-stream data so as to calculate composite parameters indicative of constituents of the tissue site blood flow. The glucose estimator calculates a glucose estimate according to a weighted and scaled sum of the composite parameters. The strip meter intermittently reads a test strip exposed to blood drawn from an individual so as to generate a glucose calibration. The glucose calibrator generates an individually-calibrated glucose estimate from the glucose estimate and the glucose calibration.

Another aspect of a blood glucose calibration system utilizes a glucose calibration method to derive pre-selected composite parameters each responsive to a noninvasive multi-stream sensor in communications with an individual's blood flow. Blood glucose values are estimated over time from a combination of the composite parameters. Invasive blood glucose calibrations are measured over time from corresponding test strip readings. Individualized blood glucose values are calculated from a combination of the noninvasive blood glucose values and the invasive blood glucose calibrations. The invasive blood glucose calibrations intermittently update the individualized blood glucose values.

A further aspect of a blood glucose calibration system has a noninvasive sensor that attaches to a tissue site so as to generate multi-stream physiological data responsive to blood constituents. Composite parameters each in the form of a mathematical combination of invasive blood panel parameters are responsive to the multi-stream physiological data. A glucose estimate is derived from a weighted and scaled combination of the composite parameters. An individualized glucose estimate is derived from the glucose estimate and intermittent invasive test strip measurements of an individual.

An additional aspect of a blood glucose calibration system attaches a noninvasive sensor to a tissue site of a person so as to generate multi-stream physiological data responsive to that person's blood constituents. Composite parameters derived from a general population and each in the form of a mathematical combination of blood constituents are responsive to the multi-stream physiological data. A population-based, blood glucose estimate for that person is derived from a weighted and scaled combination of these composite parameters. An individualized blood glucose estimate is derived from the population-based blood glucose estimate and intermittent invasive test strip measurements from that person.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a perspective view of a blood glucose monitor utilizing a blood glucose calibration system;

FIG. 2 is a general block diagram of a glucose calibration system;

FIG. 3 is a general block diagram of a glucose estimator;

FIG. 4A is a glucose calibration graph;

FIGS. 4B-C are flow diagrams of glucose calibrator embodiments;

FIGS. 5A-C are flowcharts of a glucose calibration method;

FIG. 6 is a detailed block diagram of a multi-stream sensor for noninvasive blood glucose monitoring;

FIG. 7 is general block diagram of a blood glucose monitor;

FIG. 8 is a general block diagram of a sparse solver for defining composite parameters;

FIG. 9 is a general block diagram of a composite parameter generator for noninvasive blood glucose monitoring; and

FIG. 10 is a general block diagram of optical sensor signal processing.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

FIG. 1 generally illustrates a blood glucose calibration system 100 that advantageously provides relatively frequent noninvasive measurements of blood glucose interspersed with relatively infrequent invasive measurements of blood glucose so as to manage individual blood glucose levels. The blood glucose calibration system 100 has a blood glucose monitor 110, an optical sensor 120, a sensor cable 130 electrically and mechanically interconnecting the monitor 110 and sensor 120 and a monitor-integrated test strip reader 710 (FIG. 7) that accepts test strips 150 via a test strip slot 140. In particular, the blood glucose calibration system 100 individually calibrates the noninvasive optical sensor 120 measurements with intermittent test strip measurements to advantageously provide the accuracy of individualized glucose test strip measurements at a much-reduced frequency of blood draws. Reduced blood draws is a substantial aid to persons who require frequent monitoring of blood glucose levels to manage diabetes and related diseases. In an embodiment, the monitor 110 has a handheld housing including an integrated touch screen 160 defining one or more input keys and providing a display of blood glucose levels among other features. An optical sensor is described in further detail with respect to FIG. 6, below. A blood glucose monitor is described in further detail with respect to FIG. 7, below. An optical sensor is also described in detail with respect to U.S. patent Ser. No. 13/646,659 titled Noninvasive Blood Analysis System, filed Oct. 5, 2012, assigned to Cercacor and incorporated in its entirety by reference herein. A blood glucose monitor is also described in detail with respect to U.S. patent Ser. No. 13/308,461 titled Handheld Processing Device Including Medical Applications for Minimally and Noninvasive Glucose Measurements, filed Nov. 30, 2011, assigned to Cercacor and incorporated in its entirety by reference herein. A blood glucose monitor and sensor are described in U.S. Ser. No. 13/473,477 titled Personal Health Device, filed May 16, 2012, assigned to Cercacor and incorporated in its entirety by reference herein.

FIG. 2 illustrates a blood glucose calibration system 200 embodiment that has a noninvasive measurement path 201 and an invasive measurement path 202. The noninvasive path 201 has an optical sensor 210, a signal generator/processor 220, a composite parameter generator 230, a glucose estimator 240 and a glucose calibrator 250. The optical sensor 210 attaches to and illuminates a tissue site 1 so as to generate sensor signals 212 responsive to blood constituents within the tissue site 1. The sensor signals 212 are input to the signal generator/processor 220, which outputs sensor data 222 to the composite parameter generator 230. The composite parameter generator 230 derives composite parameters CP's 232 indicative of one or more tissue site blood constituents. The glucose estimator 240 derives a blood glucose estimate 242 from one or more of the composite parameters CP's 232. The glucose calibrator 250 modifies the noninvasively-derived blood glucose estimate 242 in view of an invasively-derived glucose calibration Gu_cal 272 so as to output an individually-calibrated blood glucose estimate 203.

As shown in FIG. 2, in an embodiment the sensor signals 212 incorporate optical sensor outputs responsive to multiple wavelengths of light after attenuation by pulsatile blood flow, active-pulsed blood flow and non-pulsatile fluids and tissue. Sensor signals 212 are also responsive to the relative phase differences of multiple wavelengths of light after attenuation by pulsatile blood flow. Further, sensor signals 212 are responsive to tissue site temperature, sensor temperature and the relative orientation of the tissue site. The signal generator/processor 220 generates hundreds of data streams 222 from the sensor signals 212. Composite parameters CP's 232, however, may only be responsive to, say, 40-70 of these data streams 222. The relationship between the data streams 222 and specific composite parameters CP_i is determined by a correlation engine, which stores these relationships in a composite parameter look-up table. For each selected composite parameter CP_i, a relevant subset of the sensor data 222 is identified and the specific composite parameter CP_i is calculated accordingly. An optical sensor and a composite parameter generator are described in detail with respect to U.S. patent application Ser. No. 13/646,659 titled Noninvasive Blood Analysis System, cited above.

Further shown in FIG. 2, the invasive measurement path 202 has a blood glucose test strip 260 measured by a strip meter 270. Although the blood glucose estimate 242, described above with respect to a noninvasive measurement path 201 is responsive to a particular patient, it is calibrated across a population of many individuals. Advantageously, the strip meter 270 provides an individualized measurement of blood glucose Gu_cal 272, which is used to calibrate the blood glucose estimate 242 for a particular individual. A glucose test strip 260 is coated with a reagent that chemically reacts with the glucose in the blood. The strength of the reaction depends on glucose concentration. The strip meter 270 is responsive to the strength of the glucose-reagent reaction to determine glucose concentration. For example, the reaction strength may be proportional to a strip resistance electrically measured by the strip meter 270 and converted to a blood glucose measurement Gu_cal 272. One of ordinary skill in art will appreciate that various glucose test strip and strip meter technologies may be used to derive a Gu_cal measurement. In an embodiment, the glucose calibrator 250 may have feedback 252 that is responsive to or Gu_cal or both so as to alter the composite parameters 232 chosen for the glucose estimate 242 and/or the weights associated with composite parameters in deriving the glucose estimate 242. A glucose calibration method is described in detail with respect to FIGS. 3-5, below.

FIG. 3 illustrates a glucose estimator 300 embodiment having composite parameter inputs 301 and a glucose estimate output 302. The composite parameters are factored f_i 310 into glucose estimates 320. The glucose estimates 320 are weighted 330, and the weighted estimates 332 are summed 340 to generate a weighted parameter sum 342. The weighted parameter sum 342 is scaled 350 to generate the glucose estimate 302 according to Eq. 1, where the scaling 350 is the inverse sum of the weights.

= S ⁢ ∑ i = 1 n ⁢ w i · = ∑ i = 1 n ⁢ w i · ∑ i = 1 n ⁢ w i EQ . ⁢ 1

As shown in FIG. 3, in an embodiment the composite parameters 322 may be composed of one or more of the glucose composite parameters listed in Appendix A, attached hereto. The factors 310 are applied to the composite parameters 301 to generate factored glucose estimates 320 from the composite parameters. In an exemplar embodiment, the composite parameters are Gu/BUN, Gu*A1C and Gu, where BUN is blood urea nitrogen and A1C is glycated hemoglobin, each obtained from, say, an invasively derived blood panel. Then, the factors 310 are f₁=BUN; f₂=1/A1C; and f₃=1.

In an embodiment, the weights w_i 330 are each 1 and the scaling S 350 is 1/n. However, some composite parameters may be better estimators of glucose than others. Accordingly, in other exemplar embodiments, each weight w_i 330 is inversely proportional to the glucose measurement error σ_i² of its respective factored glucose estimate 320, and:

= ∑ i = 1 n ⁢ 1 σ i 2 · ∑ i = 1 n ⁢ 1 σ i 2 EQ . ⁢ 2

In other embodiments, weights are determined by data fitting or long-term calibration methods, such as described with respect to FIGS. 4-5, below.

FIGS. 4A-C illustrate glucose calibrator embodiments for generating an individualized measurement of glucose from a noninvasively measured glucose estimate . Shown in FIG. 4A is an exemplar scatter plot 401 of test strip measured glucose values Gu_cal 492 versus corresponding sensor measured glucose estimates 491, shown as a set of n points 493. Overlaying the scatter plot 493 is a simple linear regression 495 that fits a straight line through the n points 493 such that the sum of squared residuals is minimized. In particular, the linear regression is
=+β  EQ. 3
where α and β are the slope and y-intercept.

As shown in FIG. 4B, in a glucose calibrator embodiment 402, a gain 410 and offset 420 are applied to a glucose estimate 412 to yield an individualized glucose estimate 432 according to EQ. 3. The individualized glucose estimate advantageously converts a population-based noninvasive (sensor) glucose measurement to an individual-based sensor glucose measurement accordingly to a relatively small number of invasive (test strip) glucose measurements, as described above.

As shown in FIG. 4C, in another glucose calibrator embodiment 403, an additional composite parameter CP_n+1 452 may be used to further refine the individualized glucose estimate 432 according to EQ. 4.
=+w_n+1+β  EQ. 4

One of ordinary skill in the art will recognize that multiple additional composite parameters CP_n+2, CP_n+3 . . . may be used to further refine the individualized glucose estimate 432. In other embodiments, an individualized glucose estimate 432 may be derived from a generalized data fitting of noninvasive glucose estimates to individual test strip measurements.

FIGS. 5A-C illustrate a glucose calibration method 500 embodiment that inputs optical sensor data 510 and outputs individualized glucose measurements 509. As shown in FIG. 5A, sensor data 510 is used to derive a preselected set of composite parameters CP's 520. In various embodiments, the composite parameters are selected on the basis of the highest correlation with invasively-measured glucose over a general population of interest; the highest correlation with invasively-measured glucose over a specific population matching a patient of interest; or the lowest error in the measurement of glucose, to name a few CP selection criteria. A glucose estimate 530 derives an uncalibrated glucose value from the composite parameter values CP_i. The derived glucose estimate assumes a tested individual's glucose corresponds to the average glucose across a population of individuals according to the measured CP's. Next, the population-based glucose estimate is refined by deriving an individually-calibrated glucose estimate 540.

As shown in FIG. 5B-C, glucose calibration 540 determines if glucose has been fully calibrated 542. If not, an individual test strip measurement is requested 550. A glucose meter reads the test strip 552 (FIG. 5C) so as to generate a glucose calibration value Gu_cal, as described with respect to FIG. 2, above. In an embodiment, the relationship between the (noninvasive) glucose estimate and the (invasive) test strip glucose value Gu_cal, is determined by a data fitting 554, such as a linear regression having a gain α and offset β, as described above. Multiple test strip calibration values Gu_cal derived over a period of time may be required to determine the data fitting 554. An individual glucose value is calculated 548 from the data fitting 554 so as to output 509, as described with respect to FIGS. 4A-C, above.

Also shown in FIG. 5B, calibration may take multiple comparisons of noninvasive and invasive readings over a period of time. Once an individual relationship between invasive test strip-based glucose readings Gu_cal and noninvasive multi-stream sensor-based values is established, a patient may rely on 509 for a longer period of time. This period may be pre-established or determined during the calibration phase. In an embodiment, this calibration or learning period establishes not only the data fitting relationship between noninvasive/invasive measurements but also the drift over time, i.e. the estimated period of validity between invasive updates 544. When the time period since the last update is exceeded 545, then another test strip measurement is requested 550.

FIG. 6 illustrates an optical sensor 600 for noninvasive blood glucose monitoring, as generally described with respect to FIGS. 1-2, above. The sensor 600 has LEDs (emitters) 610, detectors 620, temperature sensors 630, an active pulser 640 and an accelerometer 650. The LEDs 610 are individually activated by LED drives 612 so as illuminate a tissue site 1 with optical radiation 614. The detectors 620 receive attenuated optical radiation 618 after absorption, reflection and diffusion by the tissue site 1 and by pulsatile blood flow within the tissue site 1. The active pulser (AP) 640 has a motor that generates a mechanical “active pulse” in response to an AP drive signal 613. The motor has a “motor-on” state for starting the active (or artificial) pulse and a “motor-off” state for stopping the active pulse. Accordingly, the pulsatile blood flow may be arterial blood flow, AP blood flow, or both. The detectors 620 generate multiple channels 622 of plethysmograph and AP signals to a DSP 720 (FIG. 7) within a blood analysis monitor 700 (FIG. 7) for signal processing and analysis, as described in detail below.

As shown in FIG. 6, in a particular embodiment, the LEDs 610 are organized in two groups 616, 618 of seven LEDs each. The two groups also share a common LED 617. Hence, each group 616, 618 has eight LEDs, which are individually activated so as to emit eight wavelengths in sequence. In a particular embodiment, the temperature sensors include a T_led sensor 632 responsive to the temperature of the LEDs 610, a T_digit sensor 634 responsive to the temperature of the fingertip 1 and a T_pd sensor 636 responsive to the temperature of the photodiode detectors 620. The accelerometer 650 indicates the sensor orientation and movement and is used by the signal processor in determining valid plethysmographs (pleths). An optical sensor is described in U.S. Ser. No. 13/473,477 titled Personal Health Device, cited above.

FIG. 7 illustrates a blood glucose monitor 700, such as shown pictorially in FIG. 1, above. The monitor 700 has a plurality of processors, including a DSP 720 that performs sensor signal processing, a coprocessor 740 that assists the DSP 720 in intensive calculations, and an application processor 750 that executes medical and smart phone applications, including, for example, cell phone, Internet, entertainment, and productivity applications. A front end 730 having LED driverss 732, detector receivers 733, DACs 731, an ADC 734 and an active pulse (AP) driver 735 communicates with the sensor 600 (FIG. 6) to accomplish noninvasive sensor measurements. The DSP 720 additionally communicates with the applications processor 750 for display 752 and user I/O 754 functions. The applications processor 750 also communicates with the strip reader 710. In an embodiment, the strip reader 710 comprises a commercially available OEM strip reader. In an embodiment, the strip reader 710 includes a current detector or reader 714 and a controller 716 for determining from an inserted strip 712 minimally invasive glucose measurements. The reader 710 forwards calculated measurements to the applications processor 750, where, for example, medical applications use the data to present information to the user on the display 752.

FIG. 8 illustrates a correlation engine 800 for defining composite parameters 832. In particular, the composite parameters 832 correlate to the sensor data 822 described with respect to FIG. 10, below, where the composite blood parameters (CP) are of the forms:
CP=B_i; CP=B_i/B_j; CP=B_i+B_j; CP=B_i·B_j/B_k;
CP=B_i/(B_j+B_k); CP=B_i/(B_j+B_k+B_l)  EQ. 5
Of particular interest are glucose related composite parameters, such as listed and described in Appendix A, attached hereto. The correlation engine 800 has a clinical data collection portion and an optimization portion. Clinical data collection compares invasive blood draw measurements 806 from test subjects 2 to noninvasive sensor measurements 807 of the same test subjects 2. Optimization utilizes a sparse solver 801, which trains an inverse tissue site model 830 to predict composite parameters CPs 832 derived noninvasively from sensor data 822 within an acceptable error 842 of the invasively derived composite parameter 812.

As shown in FIG. 8, the clinical data collection derives an invasive blood panel 812 that generates a myriad of blood constituents (B_i) 812 such as blood urea nitrogen (BUN), high-density lipoprotein (HDL), low-density lipoprotein (LDL), total hemoglobin (THB), creatine (CRE) to name just a few. Data collection then assembles parameter combinations from the blood constituents so as to derive composite parameters 832. The invasively-derived composite parameters 812 are then compared 840 to the predicted composite parameters 832 derived from the inverse tissue site model 830 so as to optimize the inverse model. Composite parameters that do not provide a high enough correlation are rejected. Appendix A, attached hereto, illustrates results obtained for approximately 60 sufficiently correlated composite parameters having glucose as a constituent.

FIG. 9 illustrates a composite parameter generator 900 for noninvasive blood glucose monitoring having a sensor data 902 input and that advantageously generates a noninvasive blood analysis 942 output accordingly to selected blood parameters 950. Sensor data 902 includes ratios r, r-ap, tx and temp described with respect to FIG. 10, below. In an embodiment, sensor data 902 includes many data streams. However, each selected blood parameter may only be responsive to a small fraction of those data streams. The relationship between the sensor data 902 and a specific composite parameter 942 is determined by the sparse solver described above with respect to FIG. 8. In an embodiment, the relationships between a selected composite parameter 950 and the sensor data 902 that determines that selected parameter is stored in a sensor data look-up table 910. A sensor data multiplexer 930 outputs the relevant sensor data 932 for the selected parameter accordingly. The relevant data 932 for a particular parameter is weighted 922 and summed 940 so as to generate that composite parameter 942. The particular weights 922 for a selected parameter 950 is stored in a weights look-up table 920. A range of composite parameters of interest 950 is selected 901 so as to calculate a particular blood constituent, such as blood glucose, as described above.

FIG. 10 illustrates optical sensor signal processing 1000, which has signal generator inputs 1012 including pleths, temperatures, currents and gains, along with sensor acceleration. Pulse processing 1020 normalizes and validates the pleths (np, np_ap) and generates transmittances (tx) from the currents and gains 1012. In an embodiment having two groups of 8 LEDs each and 4 detector channels, as described with respect to FIG. 6, above, the signal processor 1020 generates 32 (arterial) nps and 32 (artificial) np_ap's each, or 64 total normalized pleths for each of the two groups of LEDs. Data reduction 1030 reduces the normalized pleths to ratios (r, r_ap), reduces transmittances to averaged and trimmed transmittances (tx) and reduces temperatures to averaged temperatures (temp). These data reduction outputs 1032 provide the sensor data 822 (FIG. 8) for the spare solver 801 (FIG. 8) and sensor data 902 (FIG. 9) for the composite parameter generator 900.

A blood glucose calibration system has been disclosed in detail in connection with various embodiments. These embodiments are disclosed by way of examples only and are not to limit the scope of the claims that follow. One of ordinary skill in art will appreciate many variations and modifications.