What is the Blood Pressure Monitor Sensor Looking At? How 'Amplitude' is Extracted from a Single Pulse

In the previous article, “Experiencing the Oscillometric Method through Numbers”, we simulated how the algorithm determines systolic blood pressure (SBP) and diastolic blood pressure (DBP) based on the “amplitude numerical data” obtained for each cuff pressure.

However, when you look at the simulation table again, didn’t you wonder:

“It says ‘At a cuff pressure of 120 mmHg, the amplitude was 0.95 mmHg’, but how exactly was that ‘0.95 mmHg’ measured?”

In this article, we zoom in one step further from the macro blood pressure calculation algorithm and thoroughly dissect from a micro perspective what physical phenomena are occurring inside the arm and what kind of signal processing is being performed inside the blood pressure monitor during the mere 1 second (1 pulse duration) when the cuff pressure is around 120 mmHg.

1. The True Identity of the “Raw Data” Captured by the Sensor

Inside an electronic blood pressure monitor, there is a single pressure sensor that measures the air pressure inside the cuff (armband).

The values (raw data/raw signal) captured in real-time by this pressure sensor during the process of slowly deflating air after the cuff is fully inflated are not exactly the “pulse waves” we imagine.

The raw data is mainly a combination of the following two components.

1. DC (Direct Current) Component: The pressure of the cuff itself. It is a massive, gentle slope of pressure that gradually decreases over time at a rate of 2-4 mmHg/second recommended by the AHA (American Heart Association).
2. AC (Alternating Current) Component: Minute pressure fluctuations (pulse waves) that occur with the heartbeat. The magnitude of this AC component is typically around 1-4 mmHg (Drzewiecki et al., 1994), riding like ripples on top of the massive downward slope of the DC component.

The graph above simulates the raw data captured when the cuff pressure decreases from about 122 mmHg to 110 mmHg over about 4 seconds. You can see the small jagged lines (AC component) riding on top of the massive deflation curve (DC component, dotted line) for every heartbeat.

In the oscillometric method, what we really want to see is only these “small jagged lines (AC component)”.

2. Micro Physical Phenomena Occurring in a Single Beat

Why do these small jagged lines (minute increases in pressure) occur?

Let’s look frame by frame at the physical phenomena occurring inside the arm (brachial artery) at the moment the cuff pressure is fixed at 120 mmHg. By the way, let’s assume this subject’s actual blood pressure is 120/80 mmHg.

① Diastole (When the heart is resting: Blood Pressure 80 mmHg)

During diastole, when blood is not pumped from the heart, the pressure inside the artery drops to 80 mmHg. Because the cuff is compressing from the outside with a force of 120 mmHg, the artery is completely crushed flat, and blood flow is stopped. At this time, there is no change in the volume of the arm, and the pressure sensor inside the cuff remains quiet.

② Systole (When blood rushes in: Blood Pressure 120 mmHg)

When the heart contracts and a wave of blood rushes in, the pressure inside the artery suddenly rises to 120 mmHg. At this moment, the pressure from the inside (120) momentarily counteracts the cuff pressure from the outside (120), and the flat artery is pushed open slightly.

③ Conversion from Arterial Expansion to “Increase in Cuff Internal Pressure”

This is where the magic happens. The artery expanding slightly means that the “volume of the arm” increases slightly at that moment.

The fabric of the cuff (band) wrapped around the arm does not stretch or shrink. If the volume of the arm increases while the fabric cannot stretch, the “air” inside the cuff has less space and is compressed.

According to Boyle’s Law (pressure increases when volume decreases), the compression of the air causes the pressure inside the cuff to “jump”. This is what the pressure sensor catches, which is the true identity of the “small jagged lines (AC component)” mentioned earlier.

3. What is the Shape of the Pulse Wave?

The tiny jagged shape seen as the AC component is not a simple triangular or sine wave. The actual arterial pulse wave has a distinctive shape caused by the heart’s ejection and reflections in the arterial system.

In physiological studies (Baruch et al., 2011; Rubins, 2008), a method to mathematically model the pulse wave of a single beat as a superposition of three Gaussian functions (Pulse Decomposition Analysis: PDA) has been established.

Every chart in this article generates waveforms based on this 3-Gaussian decomposition model. The parameters for the amplitude ($A_i$), center time ($\mu_i$), and width ($\sigma_i$) of each Gaussian function are set referring to the values reported in the literature above.

Why the Shape Matters: In recent studies, “Pulse Wave Analysis (PWA)”, which extracts information about arterial stiffness and cardiac function not only from the amplitude (P-P value) but also from the contour of the pulse wave itself, is attracting attention. It has become widely recognized in recent years that cuff pressure fluctuations acquired via the oscillometric method are practically the arterial pulse wave itself (Baruch et al., 2011).

4. From Analog to Digital: The Amplitude Extraction Process

The physical “minute pressure increase” captured by the sensor goes through several digital signal processing steps inside the microcomputer before it is converted into a single numerical value (amplitude: 0.95 mmHg).

Step 1: Sampling (A/D Conversion)

Real-world pressure changes are continuous analog waveforms, but the microcomputer cannot process them as they are. Therefore, the pressure is measured at a frequency such as 100 times per second (sampling rate 100 Hz), and recorded as a collection of digital points.

Step 2: High-pass Filtering (Removal of DC Component)

As seen in the earlier graph, the pulse wave (AC component) is superimposed on the steeply dropping cuff pressure (DC component). We cannot accurately measure the height of the wave like this. Therefore, a “band-pass filter” (passband approx. 0.5 to 20 Hz) based on IEEE/AAMI standards is applied to remove slow changes (DC component, below 0.5 Hz) and at the same time cut off high-frequency noise (above 20 Hz).

As a result, only the pure pulse wave (AC component), with the 0 mmHg horizontal line as the reference (baseline), is extracted.

In the above graph, the DC component is completely removed, and the characteristics of the pulse wave composed of the three Gaussian functions are clearly visible. With each beat, the pattern of steep rise (P₁: Systolic wave) → shoulder bulge (P₂: Reflected wave) → small rebound (P₃: Dicrotic wave) → silence of diastole is repeated.

Step 3: Calculation of Peak-to-Peak (P-P) Amplitude

From the filtered waveform, the “single representative value” used for blood pressure calculation is determined. In the oscillometric method, the Peak-to-Peak (P-P) amplitude method is usually used.

The waveform of a single beat (about 0.83 seconds for a heart rate of 72) is extracted, and the following calculation is performed.

1. Find the “highest point (Peak)” in that beat.
2. Find the “lowest point (Trough)” in that beat.
3. Subtract the depth of the valley from the height of the mountain (Peak - Trough).

The graph above expands the waveform for one beat (the horizontal axis is milliseconds). The difference between the apex (Peak) of P₁ (systolic wave) and the lowest point in late diastole (Trough) is the Peak-to-Peak (P-P) amplitude. This exactly is the true identity of the data point in the simulation article: “The oscillation amplitude at a cuff pressure of 120 mmHg is 0.95 mmHg”.

5. Returning to the Macro Perspective: From a Single Amplitude to the Envelope

Let’s summarize the micro processes so far.

1. Physical Phenomenon: Triggers a momentary opening of the artery, increasing the arm’s volume, compressing the air inside the cuff, and causing a minute rise in pressure.
2. Signal Processing: The raw data from the sensor is sampled, and a band-pass filter (0.5-20 Hz) is used to remove the DC component and noise, extracting a pure pulse wave (AC component).
3. Quantification: The “height from peak to trough (Peak-to-Peak)” of a single waveform is calculated as “one amplitude value”.

The electronic blood pressure monitor endlessly repeats this calculation of “extracting the wave → filtering → extracting the P-P amplitude” every time a pulse occurs (over dozens of times) while decreasing the cuff pressure.

The extracted “individual amplitude points (0.95 mmHg, 1.30 mmHg, 1.65 mmHg…)” are plotted for each cuff pressure and connected with a smooth curve to form the “Oscillometric Envelope” shown in the previous article.

A collection of points extracted from the height of the waves in the raw data (micro perspective) forms an envelope (macro perspective), from which the final blood pressure values, SBP (systolic blood pressure) and DBP (diastolic blood pressure), are finally calculated—this is the whole picture of the grand baton relay of information happening inside the blood pressure monitor.

References

1. Geddes LA, Voelz M, Combs C, Reiner D, Babbs CF. “Characterization of the oscillometric method for measuring indirect blood pressure.” Annals of Biomedical Engineering 10:271-280, 1982.
2. Drzewiecki G, Hood R, Apple H. “Theory of the oscillometric maximum and the systolic and diastolic detection ratios.” Annals of Biomedical Engineering 22:88-96, 1994.
3. Baruch MC, Warbritton DER, Babb AR, Shaltis PA, Ring R. “Pulse Decomposition Analysis of the digital arterial pulse during hemorrhage simulation.” Nonlinear Biomedical Physics 5:1, 2011.
4. Rubins U. “Finger and ear photoplethysmogram waveform analysis by fitting with Gaussians.” Medical & Biological Engineering & Computing 46:1271-1276, 2008.

Related Articles

• Experiencing the Oscillometric Method through Numbers: From Pulse Wave Data to Blood Pressure Values
• Deep Dive into the Oscillometric Method: How Electronic Blood Pressure Monitors Were Born, What They Measure, and What Are Their Limitations?