University of Amsterdam
Courtesy of Bert Molenkamp
VPA-02 - VPA/VBV Probe Assembly
The VPA-02 VPA/VBV probe assembly was designed to measure vaginal pulse amplitude (VPA) and vaginal blood volume (VBV) with high accuracy.
The probe uses a perceptible light source (LED) and a photodiode to measure the blood volume in the tissue surrounding the probe. The blood in the tissue reflects the light emitted by the LED. This reflected light is measured with the photodiode (see figure 1). The output signal of the photodiode varies with changes in the amount of blood in the tissue.
The first ideas to build our own VPA probe assembly dated from 1996. Most important reason was the poor availability of commercial probes, which often caused problems when probes broke down during an experiment. Another problem was the poor between-probe accuracy, mainly because the probes we used were hand-made.
From our experience with the older probes, we could easily formulate our design criteria:
- High between-probe accuracy
- High long term stability
- Good sensitivity for both VPA and VBV
- Very little heat production in the probe
- As comfortable as possible for the subject
The second criterion posed no problem, instead of the commonly used incandescent light source, we would use solid state LED’s. The sensitivity seemed no problem at that time. We chose red to be the LED colour, simply because blood is red, giving the highest reflection rate, and thus the highest sensitivity. Electronic accuracy is achieved by using electronic components from the same silicon dye, or, in manufacturers terms, from the same batch. By using high efficiency LED’s at low power, heat production is reduced to a minimum. A probe like this, will never be really comfortable to use, but keeping the wiring to the probe as flexible as possible, possibly reduces the feeling of being hooked up to a test set-up.
The actual probe assembly is composed of a mould, holding the electronic components, and a computer drilled epoxy glass case, into which the mould tightly fits. By using this method, positioning of the electronic components is always within a few tenths of millimetres, contributing to high between-probe accuracy.
In the first version of the probe, a high-intensity narrow band dark red LED was used as light source, with a linear visible range light-to-voltage converter. The resulting probe, VPA-01, proved to be too sensitive to blood colour. In some subjects the performance was very good, while in others the VPA traces were hardly distinguishable. Furthermore, the research staff reported an inverted VBV signal.
The current probe uses a high-intensity orange-red wide spectrum LED, which covers a wider range of blood colours. With this wider range, average VPA performance is largely improved compared to the previous version of the probe. However good the VPA signal may be, the VBV signal still is inverted, or isn’t it?
The photoplethysmography method to determine vaginal blood flow, was first used about 35 years ago by Palti and Bercovici. The method became a sort of a standard when Sintchak and Geer modified the measurement device into its present form.
The basic theory behind the photoplethysmograph is very simple: use a light source to shine into the surrounding tissue, and measure the reflected light with a light sensitive cell. The more blood is present around the device, the more light will be reflected onto the light cell. The signal of the light cell is transformed into a voltage, yielding higher voltages with higher blood volumes.
Because the amount of blood constantly changes due to the functioning of the heart, a small dynamic pulse shows on top of the reflected light. This change, which only consists of about 2% of the total blood content, is often referred to as pulse amplitude, or, in case of the vaginal probe, as vaginal pulse amplitude or VPA. Because electronically this part of the signal can easily be detected by using an AC filter to remove the constant component of the signal, the signal is also often described as ‘AC signal’. The slow changing component of the reflected light is often referred to as VBV, vaginal blood volume, or DC signal.
Many studies, in the years since the Sintchak/Geer device became common good, have indeed proven the functionality of the method, in terms of increased VPA and/or increased VBV during sexual arousal. Despite the fact that sexual arousal increases blood flow and blood volume, the VBV signal not always shows the expected characteristic. Maybe it is because of this fact, that some researchers simply omit the behaviour of the VBV traces when reporting about female sexual arousal.
The simplified vision of the light reflection being greater with higher blood volumes is clearly invalid for the DC (“VBV”) part of the signal. What the plethysmograph measures is indeed the DC component of the reflected light, but the question is whether this accurately represents the blood volume surrounding the probe.
To the human eye, the most important factor accounting for the light reflection of the blood is the red blood cells, or erythrocytes. In the red blood cells, the haemoglobin molecules are responsible for the red colour. Haemoglobin in the red cells is the carrier of oxygen from the lungs to the tissues. A haemoglobin molecule consists of four haeme groups and the protein globin, hence the name haemoglobin. The four haeme groups contain iron, to which oxygen atoms easily bind, causing the haemoglobin molecule to assume a new shape. The oxygenated state of the haemoglobin is referred to as oxyhaemoglobin (HbO2), the state where no oxygen is bound to the haeme groups is called deoxyhaemoglobin (Hb). When light is shone through these different states of haemoglobin, different absorption spectra occur (fig.5).
Note: the haeme groups can occur in four different states: MetHb (Methaemoglobin), CoHb (Carboxyhaemoglobin), Hb (Deoxyhaemoglobin) and HbO2 (Oxyhaemoglobin). The haemoglobin molecule can contain any combination of these groups, of which only the oxygenated (all four groups binding oxygen) and deoxygenated (the four haeme groups are unbound) states are analysed here.
From the above absorption spectra for HbO2 and Hb the bright red colour of oxygenated blood can easily be deduced. The absorption of HbO2 in the 600 – 700 nm band is very low, hence the reflection of bright red light in oxygenated blood is high. From the absorption spectra, it is fair to assume that oxygenation of haemoglobin is responsible for respiration artefacts in the VPA signal. With inhalation, the oxygenation of the haemoglobin is likely to be greater than with exhalation, causing the VPA signal to vary with respiration. Using a near-infrared light source instead of white or perceptible red light can reduce this artefact. Another deduction may be that venous blood appears darker red than arterial blood, because of the difference in oxygenation of the two. (In venous blood, the Hb curve plays a more important role in the reflection of light; Hb is more reflective with increasing wavelengths)
Respiration artefacts are not present at wavelengths where both absorption spectra cross each other. Because of the relatively low absorption at 800 nm, this wavelength is the optimal wavelength to use in a photoplethysmograph.
When visible light is used in a plethysmograph, the darkening of the venous blood has other unforeseen effects, besides the clear respiration artefact in the VPA signal. The assumption that more blood volume in the tissues yields more reflection, and thus higher output from the photocell, is too simplistic when looking at absorption coefficients of venous and arterial blood.
When blood pooling occurs, resulting in a higher blood volume, it is very likely that the micro-vascular bed in the tissue surrounding the probe is filled with venous blood, hence the darkening of the tissue. Figure 6 and 7 are visual representations of the blood content of the tissue with normal and high blood volumes. (Note that these figures are in no way an exact representation of the actual proportion of arterial to venous blood in the tissue, but merely an aid in understanding the mechanism) In both figures the size of the blood pulse is highly exaggerated, to demonstrate the existence of the VPA signal. In figure 6 the ratio arterial blood versus venous blood is almost the same: hence a high percentage of the light is reflected by the blood (due to the reflection of the arterial blood).
Figure 7 represents a situation where a high blood volume is present in the tissue. Because of the high percentage of venous blood, the overall reflection of visible light (either red or white) will be lower.
In other words, because a larger amount of deoxygenated blood is co-responsible for the reflection of the light emitted in to the tissue, relatively more light with longer wavelengths (700 – 1000nm and higher) will be reflected. Because relatively less HbO2 is present, the reflection of visible light will be reduced. This means that care should be taken with the colour of the light source.
Figure 5 shows that wavelengths between 650 and 750 nm are the least absorbed by the oxygenated blood, but the curves of HbO2 and Hb are far apart, and thus a great sensitivity to respiration artefacts is very likely. The wavelengths where the two curves are very close, or even cross each other, are likely to show the least respiration artefacts. Besides the respiration artefacts, the absorption graphs of HbO2 and Hb indicate that higher blood volumes yield decreased reflection of visible light.