Pharmaceutical & medical
when applied to pulse oximetry, two recurrence arrays are required: one for the red history and another for the infrared history. To complete the sliding transform picture, the updated contents of the recurrence buffers, which have the length of the period bin being processed, are rotated by the basis function corresponding to that period. The length of this buffer determines the overall resolution and once enough data has entered the process to fill these buffers, the transform results reach a stable limit and only change in amplitude or period as the input data changes. For the data processing reported, the recur- rence buffers held the last 10 s of data. The raw data was collected by researchers at ADI and the software used to process the data was a sliding DPT in MATLAB script. Figure 6 shows the raw data taken from one subject along with band-pass filtering of 1 Hz to 4 Hz and using a flat smoothing moving average filter with a total width of 200 ms. The spectrums that are shown in Figure 7 are those after the recurrence buffers were filled and the spectrums reached stable amplitudes. The DPT will continue to track any changes in the raw data as new data is sampled and the spectrums will be updated accordingly.
To estimate SpO₂, the commonly known equation referred to as the ratio of ratios was first used. The AC component entries used the peak values from the spectrum plot shown in Figure 7 and the DC compo- nent entries used the average values of the unfiltered signals shown in Figure 6.
The SpO₂ and heart rate collected from a Masimo oximeter using its SET algorithm were compared with the data taken at the same time with an ADI MAX30101 pulse oximeter sensor. The data from a subject was chosen at random and the results plotted in Figure 8 and Figure 9.
put to this test, with one considered the standard, the results should remain within two standard deviations or 95 per cent for the second device to be consid- ered as good as the standard for clinical use. Using the Bland-Altman method, as opposed to correlation analyses that identifies the relationship between two variables, a statistical method is applied that addresses the difference between two variables. The accuracy and precision of the DPT algorithm was evaluated by collecting data from 26 healthy adult subjects using the MAX30101 pulse oximeter sensor and comparing that with a Masimo oximeter containing its latest Signal Extraction Technology. The cohort consisted of 15 male and 11 female subjects with ages between 20 and 40 years. Our study was designed to compare the results of the two oxime- ters on each of the individuals in the cohort and not any differences men and women. Note that SpO₂ between the sexes does vary somewhat. One study has shown that, for young healthy adult individuals, the mean SpO₂ for males was 97.1±1.2 per cent, while for females the mean was 98.6±1.0 per cent. The results using the Bland-Altman criteria are depicted in Figure 10 and Figure 11, where each circle represents the Bland-Altman results for an indi- vidual subject. All of the SpO₂ comparisons met the Bland-Altman criteria.
Figure 8. Comparison of DPT processed photoplethysmographic data.
Figure 6. Raw, filtered, and smoothed photoplethysmographic pulse data from a subject using a MAX30101 PPG AFE device. The top waveforms show the raw infrared and red signals while the bottom waveforms show the filtered and smoothed data.
Figure 9. Comparison of processed heart rate data from the MAX30101 oximeter processed with the discrete period transform and a Masimo oximeter.
It is common medical practice to evaluate values produced by two different instruments that are measuring the same parameter. One of the instru- ments is assumed to produce correct results and used as the standard.
Figure 7. This graph shows the red and infrared spectrums using the sliding window DPT. The larger of the two peaks is the infrared spectrum and the smaller is the red spectrum.
84
Bland and Altman developed a method based on the agreement between two quantitative measure- ments. They did this by studying the mean difference and constructing the limits of agreement. The Bland- Altman plot analysis is a simple way to evaluate a bias between the mean differences and to estimate an agreement interval. If two medical instruments are
Figure 11. Heart rate differences in beats per minute between a Masimo oximeter and an ADI oximeter using the DPT algorithm. The Bland- Altman algorithm was met in all cases but one. An arrow shows where the analysis falls outside of two standard deviations.
January 2025 Instrumentation Monthly
Figure 10. The percentage differences of SpO2 between a Masimo oximeter and an ADI oximeter using the DPT algorithm. The Bland-Altman criteria was met.
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92 |
Page 93 |
Page 94 |
Page 95 |
Page 96
