Robust design of finger probe in non-invasive total haemoglobin monitor

Abstract

The development of a non-invasive monitor of total haemoglobin has been reported previously. The paper presents the design and testing of a finger probe used in that optical haemoglobin monitor. When a finger is inserted into the probe, light is radiated onto the nail, and a silicon detector measures transmitted light. This finger probe can have different values or settings for design parameters such as the internal colour, detector area, the emission area of a light source and the distance between the light source and detector. Design of experiment (DOE) was introduced to select the best combination of design parameters that were robust to external conditions such as finger alignment and ambient light. An optimally designed finger probe from DOE analysis, compared with the initial design, increased the correlation coefficient from 0.696 to 0.869 and improved the standard deviation from 1.18 to 0.81 g dl(-1) in predicting total haemoglobin. This was under different conditions of finger-probe alignment. Under different ambient light conditions, the optimum design improved the correlation coefficient from 0.735 to 0.870 and reduced the standard deviation from 1.14 to 0.83 g dl(-1).

Photonics‐based In Vivo total hemoglobin monitoring and clinical relevance

“…The ratio of pulsatile varying signals to constant signals is then used to minimize effects from pathlength variation and interference from other analytes. This group initially reported correlation r ¼ 0.80 (n ¼ 129) and standard deviation of error (standard error) of 1.14 g/dL for Hgb relative to in vitro determined levels using the aforementioned ratio as a correlation variable [20] and has since estimated potential clinical performance to a correlation of r ¼ 0.87 and standard error of 0.81 g/dL with a more optimized finger probe [24]. This degree of performance is achieved using a five diode array with multiple wavelength ratios found to most effectively reject additional errors from saturation state and wavelength dependent tissue scattering changes.…”

Analysis of Uncertainty and Variability in Finite Element Computational Models for Biomedical Engineering: Characterization and Propagation

“…The best way to do that is by using design of experiments (DOE) (section 3.1.1), which explore the simultaneous effect of multiple input variables (factors) on the output (response) (Yoon et al, 2005; Chen et al, 2011). DOE determines discrete values within a range for each factor, called levels.…”

“…DOE is a statistical technique which studies the effect of multiple variables simultaneously defined as discrete sets of values (Yoon et al, 2005; Chen et al, 2011) (see 2.2.1). It received a special attention mainly in industries, such as electronics or chemistry, but it was recently adopted into bioengineering (Montgomery, 1997; Guvenis, 2013).…”

“…Controllable factors are part of the inner array, and uncontrollable ones are part of the outer array. Usually, the controllable factors are the ones that improve product characteristics (Yoon et al, 2005) and uncontrollable factors those that cause variations in a production process (Khuri and Mukhopadhyay, 2010). Once the value or level of the controllable factors is established, each run from the controllable factor array is tested across each run from the uncontrollable one.…”

Robust Design of Pulse Oximeter Using Dynamic Control and Motion Artifact Detection Algorithms

“…Pulse oximetry is a non-invasive and nonconstrained method of monitoring arterial oxygenation and pulse rate. Arterial oxygenation is estimated by the oxygen saturation of hemoglobin in arterial blood (SpO 2 ), which indicates the proportion of oxygenated hemoglobin (HbO 2 ) with respect to total hemoglobin [1,2].…”