Following peak detection, the pattern is matched using statistical parameters. We attempted to determine pattern parameters such as amplitudes and intervals based on peaks and valleys. We first estimated parameters using a baseline, but the results were unsatisfactory, thus parameters were determined using a zero-line reference. To prepare the sample pattern, a range of pattern data is chosen. A sample pattern is created for each peak depending on the selected peak and a given range. Maxima and minima are used to calculate amplitude and intervals. The presence of specified patterns in the sample pattern is determined by checking polarity. Waveform analysis is used in this method.
A. Waveform analysis
The waveform analysis can be performed by observing waveforms separately. After analysis it has been observed that different parameters are there in waveforms. The Characteristics of waveforms can be derived by using these parameters. Peaks and valleys are observed parameter for PPA waveform analysis. As we can see in Figure5, pulse morphologies can be differentiated based on location of peaks and valleys. Criteria of parameters may differ in different waveforms. For pulse morphologies; three peaks and three valleys are identified for analysis purpose.
By clicking generate excel sheet, user can generate an excel sheet of statistical parameters for each file and percentage of each repeated pattern.
Table 1 : Screenshot of Excel sheet1 for parameters
Table 2 : Screenshot of Excel sheet2 for percentage of patterns
Different points are introduced for understanding purpose of peaks and valleys as shown in Figure5. Point A, C and E are valleys whereas point B, D and F are peaks for typical pulse morphology. Amplitudes are taken from zero line to introduced peaks and valleys. For point A, B, C, D, E and F amplitudes are A, B, C, D, E and F respectively. Similarly intervals are taken from one point to another as AB, BC, CD, DE, EF intervals .Polarity of C point plays vital role to determine pattern from P1-P4 and P5-P8. Among three valleys, first valley (V1) is considered as a baseline point for all patterns. Second valley is for polarity check.
Pattern 1: For pattern 1, three peaks and three valleys have been identified. Firstpeak (Pk1) is ending below the baseline and height of second (multiple) peak (Pk2) is around 50% of the first peak shown in Figure 7. here we can say V2 is below the baseline point and after that it follows ‘m’ pattern.
Pattern 2: As we can see in Figure 7 for pattern 2, Firstpeak (Pk1) is ending below the base line and height of second (multiple) peak (Pk2) is around 75% of the first peak (Tallest Peak).
Pattern 3: First peak (Pk1) is ending below the base line and height of second (multiple) peak (Pk2) is around 20% of the first peak and it is the shortest peak among the three peaks.
Pattern 4: All peaks are ending at the base line and amplitude of all valleys are same.
Pattern5: First peak (Pk1) did not finish (i.e. didn’t cross the base line), there is a slur in the downward slope near bottom side.
Pattern 6: First peak (Pk1) did not finish (i.e. didn’t cross the base line) and there is a slur in the downward slope near middle side.
Pattern 7: All peaks do not touch the base line and there is more variation in amplitudes of valleys.
Pattern 8: First peak (Pk1) did not finish (i.e. didn’t cross the base line) completely and there is a slur in the downward slope near top side.