Peak Detection / Integration Options
This section applies only to dynamic analysis samples. |
File Open > [.RPO file] > [Integration button]
Peak detection parameters can be customized using the Integration button on the Report Options window or the Integration button on the Peak Editor window. Peak detection options can be customized while creating the sample file or after analysis.
The TCD detects and records all deviations from baseline, but only those which satisfy the criteria established in this window are reported as peaks.
Peak detection is accomplished through a combination of noise, height, and area thresholds. |
Field or Button | Description | ||
---|---|---|---|
Standard |
The Baseline Mode affects how the Find All Peaks function works. If Best Fit Baseline is selected, the bottom of the peaks is placed in the baseline that best describes the signal outside the range of the peaks. This assumes a linear baseline between the beginning and the end of the peak. If Snap to Signal is selected, the bottom of the peaks is moved to the signal recorded, not the best fit baseline between the peaks.
Sets the minimum height for peaks to be identified and included in the peak table. This value is expressed in terms of the trace’s Y-axis units. Use a value of 0 (zero) to include all peaks.
Smoothing allows the application to average the points before using them, so that noise spikes are ignored. Specify the number of points to average into a single value during the peak picking process. The smoothing parameter can be turned off by setting the value at 1 or 0 (zero). A setting of 1 disables smoothing, and the peak edges are interpolated to the best X-axis value. A setting of 0 also disables smoothing, but the peak edges (the points where the peak begins and ends) are not interpolated. Instead, the nearest data point is used as the peak edge. |
||
Advanced Settings [group box] |
Sensitivity sets the noise rejection level for identifying the peaks in a trace. Use a value from The sensitivity can also be set to negative values to define a specific noise level (in Y units) for peak rejection. For example, a sensitivity setting of -2.5% sets the noise rejection to 2.5 V. This means that maxima with an amplitude of 2.5 V or less will be considered as baseline noise instead of as peaks. (As opposed to the Minimum Peak Height rejection parameter which eliminates refined peaks by using their height above the baseline.) A setting of 0 (zero) automatically sets a default noise level for the trace.
There can also be shoulder peaks (also called combination peaks) within a group. See Maximum group separation for a description of peak groups. Shoulders are usually small peaks that are overlapped on the front or the tail of a larger peak. These peaks can also be called leaders and followers, respectively. As with baseline groups, the areas of these peaks can be calculated incorrectly. If the larger parent peak has a long tail with a much smaller peak riding on it, most of the area under the trace belongs to the parent peak. However, if the area of these peaks was determined using baseline grouping, the smaller peak would be calculated by using vertical drop lines at the edges. This would give the parent peak too little area, and the rider peak too much. The application can detect these shoulder peaks. The areas of shoulder peaks are calculated by drawing a skimmed baseline from the leading edge to the trailing edge. Either an exponential or a straight skim line can be used. The skim type is specified by a secondary method parameter (see Methods) and the default is exponential skimming. The remaining area between the shoulder peak baseline and the group baseline is considered to be part of the parent peak. The Max Shoulder Ratio parameter is used to specify whether the peaks that are overlapped in the front or the tail of much larger peaks should be identified as shoulder peaks. To use shoulder peak detection, use a non-zero value for the Max Shoulder Ratio parameter. After a baseline group has been identified, the application looks for peaks within the group that satisfy the following shoulder peak criteria: Shoulders must have a significantly higher Y value at one edge than the other. More importantly, the height of the shoulder above the common value must be much smaller than the height of the “parent” (larger) peak above the same valley. It must be smaller by the Max Shoulder Ratio setting. For example, a setting of 33 implies that shoulders must be smaller than 33% of their parents in terms of height above the common valley. The areas for shoulder peaks are calculated by drawing a skimmed baseline from the left edge to the right edge of the peak. The remaining area between the shoulder peak baseline and the group common baseline is considered to be part of the parent peak. Shoulder peaks can only be calculated within a group of peaks. See Maximum group separation for a description of peak groups. If the Max Group Separation parameter is set to 0 (no groups), a Max Shoulder Ratio parameter value is not used. Use a value of zero to specify no shoulder peak detection. A value of 33% works well with most data. Use a setting of zero to treat shoulder peaks with a perpendicular drop to the common group baseline instead of a skimmed baseline. If the application detects unwanted baseline noise peaks, try increasing the Sensitivity setting. Conversely, if some peaks are not detected, decrease the value. The application normally calculates peak areas by drawing a valley-to-valley baseline from the leading edge to the trailing edge of every identified peak. However, in many traces, the valleys between peaks do not always drop back to the original baseline. If a valley-to-valley baseline is used for this type of peak, the calculated area does not accurately reflect the true area under the peak. The application provides a parameter that allows the calculation of Baseline Groups. A group of peaks is defined by a common baseline that extends from the leading edge of the first peak in the group to the trailing edge of the last. The areas of grouped peaks are calculated by dropping vertical lines from the peak edges down to the group baseline.
The Max Group Separation parameter specifies a percentage of the smallest of these two widths in X units. If the edges of two adjacent peaks differ by less than this value, the two peaks constitute a group and are given a common baseline. For example, if two adjacent peaks in the trace have largest half widths of 1 and 1.5 respectively, and the Max Group Separation parameter is set at 20%, then a difference of less than 0.4 X units between the adjacent edges of these peaks would make them a group with a common baseline. If the same two peaks have adjacent edges that are greater than 0.4 X units apart, they do not define a group, and each peak has its own separate baseline. parameter is used to determine which peaks in a trace have a common baseline. When using peak grouping, the application compares the width (actually double the largest half width) of every identified peak in a trace to the width of the following peak. TheThe areas for grouped peaks are calculated by drawing imaginary vertical lines from the peak edges to the common baseline. Any peaks that share common edges are automatically considered a group and are given a common baseline for any Max Group Separation setting greater than 0. To specify no peak grouping (each identified peak has its own baseline), use a setting of 0. A value of 33% for this parameter works well with most data. Use a setting of zero to force all baselines to be drawn from peak valley to valley.
This parameter sets the minimum area required for refined, processed peaks to be recognized, identified, and included in the peak table. Any peak with a calculated peak area smaller than the current setting is not detected. Values for this parameter are expressed in terms of the trace X-axis units multiplied by the Y-axis units (e.g., millivolt-minutes). Setting the Sensitivity parameter to large values (greater than 20%) can cause noise spikes (or dips) on the sides of major peaks to be identified as peaks themselves. If the areas of the peaks are smaller than the Minimum Peak Area parameter, major peaks in the trace may not be identified at all. Exercise caution when using high settings with the Minimum Peak Area parameter. If peaks are not detected by the application, lower the setting with a smaller value. This parameter must be adjusted in conjunction with the Peak Sensitivity and/or Minimum Peak Height parameters above. Peak rejection is accomplished through a combination of peak height and peak area rejection parameters. If the application is not detecting the peak(s) of interest, decrease both parameters.
The maximum baseline slope indicates how the application determines the start and end of the baseline. The algorithm scans the trace and looks for the first time when the slope of the trace is larger than the maximum baseline slope. Then it scans from the end of the trace backward and looks for the first time the slope is larger than the maximum baseline slope. It records these two points as the start and end of the baseline.
Basically, the Maximum Baseline Slope value is used to decide how flat the trace would need to be to set the baseline of the current peak. |
||
|
Copyright © 2018. Micromeritics Instrument Corporation. All rights reserved.