(by A. Schaafsma MD PhD)
After taking over a practice in clinical neurophysiology at the Martini Hospital Groningen from my predecessor Rud Brenninkmeijer in 1998, I was confronted with a peculiar type of apparatus that I had not encountered before: the oculoplethysmograph. After instilling a drop of local anaesthetic in both eyes, cups were placed over the sclera allowing the apparatus to exert negative pressure to the patient's eyes, temporarily blocking arterial inflow. When pressure was gradually released arterial pulsation returned and could be recorded for both sides independently, thereby providing information on the arterial blood pressure wave in each central retinal artery. Pulsation returned at lower pressures in the eye ipsilateral to carotid artery stenosis depending on the amount of collateral flow.
In the era of HIV, hospital epidemiologists rapidly concluded this apparatus was no longer to be used, since the eye cups could not be properly disinfected. In search for an alternative way to assess intracranial hemodynamics we (technicians and physicians of our department) turned to transcranial Doppler (TCD). At first we recorded middle cerebral artery blood velocity (MCAV) few hours before and on days 0-3 after carotid endarterectomy aiming to detect postoperative hyperperfusion syndrome as a complication potentially leading to severe morbidity or mortality.
Sitting on the floor with the recordings of approximately 40 patients gathered around me I was struck by the great variability of our findings: sometimes the MCAV increased dramatically after surgery but in some patients it could also decrease. Soon I realized we needed to be informed about arterial blood pressure as well, under the assumption that this is the driving force behind middle cerebral artery flow velocity.
I was looking for a better understanding how the MCAFV signal related to the different physiological parameters, such as heart rate, end-tidal CO2 and arterial blood pressure. I realized that the MCAFV is a conducting artery positioned between the large feeding arteries and the outflow arterioles and cerebral capillary networks of the larger part of a cerebral hemisphere.
From a colleague, Klaas Hoogenberg, we borrowed a non-invasive blood pressure (NIBP) measuring device, the Ohmeda Finopress 2300. This apparatus allowed us to make a continuous recording of arterial blood pressure by placing a cuff around a finger. A servo mechanism maintained a continuous volume in the finger (volume clamp), requiring rapid changes in inflation pressure. The pressure required mirrored the pressure exerted by the arterial pulse. This led to detailed and synchronous information on wave morphology of both MCAV and NIBP. Combining the two signals seemed promising so these measurement techniques were used on patients with carotid stenosis and on patients with autonomic dysfunction (2000).
There was remarkably little agreement between both signals. What corresponded best was the heart rate and thus the duration of a single heart cycle. We defined the pulse flow velocity mismatch (PFVM) and later the pulsatile apparent resistance (PaR) as indices expressing the (dis)agreement between both signals. In 2004, the PaR was patented as a blood pressure corrected pulsatility index, since it expressed the difference between the apparent resistance (velocity/pressure) during end diastole and the 2nd systolic peak divided by the average apparent resistance. For quite some time we focused on the second peak in both signals since this correlated best with heart rate and CO2-reactivity.
Based on an EU and US patent on PaR technology, Neuromon B.V. successfully applied for a IAG-grant allowing the company to buy three set-ups combining a Rimed TCD, a next generation Finometer and a capnograph mounted on a single equipment rack. Singals were processed by an Apple mini-mac running customized LabVIEW software. A collaboration agreement was established with the University Hospital Groningen, the Erasmus Hospital Rotterdam and the Martini Hospital Groningen. For some time data was collected in carotid artery patients and patients with traumatic brain injury (2007).
Around this time (2009) I analyzed MCAFV signals during deep breathing by calculating bin averages based upon the RR'-interval in the ECG. It became clear that there are two peaks in systole, both in the NIBP as well as MCAV signal. I was able to show that there is a considerable difference in variance between the Sys1 and Sys2 components, suggesting they have a different physiological origin. From then on, I started to focus on the Sys1 component more. The question arose why the Sys1 was less dependent on variation in RR'-interval and thus cardiac stroke volume than the Sys2.
During carotid endarterectomy many hours of simultaneously recordings of ECG, bilateral MCAFV and ABP were obtained. I was able to document changes during clamping and declamping. Also, I was able to study changes in the two signals in response to the administration of specific drugs, such as norepinephrine, dopamine and atropin. However, application of PaR-technology in patients with traumatic brain injury (TBI) was cumbersome. Prolonged recordings were difficult to obtain since these required a fixation band mounted on the patients head. This made me return to the MCAFV signal itself, settling with intermittent monitoring: repetitive recordings of a few minutes with a hand held probe.
In 2012 new parameters were proposed to better describe the pulsatile signal: Sys1, Sys2 and D560 (the diastolic blood velocity 560 ms after upbeat onset). It was soon realized that these parameters were also applicable to the ABP signal. Together with Annika de Goede software was developed for online and automated detection of these parameters allowing their realtime presentation on screen. Nowadays this is an important add-on during autonomic function testing at our lab.
In 2013 a collaboration agreement was established with Compumedics DWL allowing Neuromon to market the software for automated parameter detection. A dynamic library was built and offered exclusively to end users of DWL TCD equipment with QL software.
For the MCAFV we provided normative data obtained in over 100 Dutch volunteers ranging from 18 to 80 years old. This allowed the calculation of so-called Z-scores, describing how many standard deviations a given measurement was above or below the value expected for that age. Combining the Z-scores for Sys1, Sys2, D560 and HR (heart rate) in a single radar plot enables a more rapid interpretation of the MCAFV signal. It can also be used for follow up measurements after CEA and for monitoring changes in intracranial hemodynamics during autonomic function testing.
In 2014 I published the theory of arterial acceleration in Medical Hypotheses. This paper explains the biphasic appearance of the systolic peak by proposing a shortlasting contraction in the smooth muscle layers of conducting arteries during Sys1 that adds to the ejection phase of the heart indicated by Sys2.
To explain how arterial acceleration enhances the penetration of the pressure wave into all the bodies capillary systems a computer model was designed evolving from a simple LabView based model to the CardioVascularSimulationApp that can be downloaded from the AppStore.
In 2017 we described changes in waveform characteristics during repetitive fluid challenges in patients with sepsis. We described the newly defined parameters for the follow up in patients undergoing carotid endarterectomy. A study was undertaken showing differences in waveform during standing and sitting. Characteristic changes were described for MCAFV and NIBP during hyperventilation, normoventilation and CO2-retention. S1-PaR was shown to correlate with intracranial pressure elevation (2025).
The theory of arterial acceleration obtained little attention until in 2025 evidence was found in a group of patients with aortic stenosis showing that from proximal to distal a phase of laminar flow (Sys1) increasingly replaces the turbulent phase of cardiac outflow (Sys2). The Sys1 is an addition to cardiac ejection and represents energy added by the conducting arteries to the pressure wave from the heart.
It is Neuromon's mission to increase the awareness of the theory of arterial acceleration and to implement the new parameters for the description of pulsatile signals in medical systems. In the end these goals are expected to support medical decision making. In 2026 the new parameters are commercially available in the TCD system of Compumedics DWL Germany and, with slight adaptations, in a blood pressure device for the outpatient clinic manufactured by Uscom Australia.
Galen (129-200)
Galen as experimenter
In ancient history the Greek physician Galen (c. 129-c. 200) was the first to show that veins were connected to the heart and that arteries contained blood. Galen was an experimenter performing animal dissections and thoroughly describing anatomical details. The interpretation of his findings remained within the Greek tradition going back to Hippocrates and Aristotle.
He regarded blood to be produced in the liver where it received its "natural spirit" and from which it flowed out to the periphery of the body due to a pulling or attractive force. Transported to the heart, the blood obtained "vital spirits" moved from the right to left side of the heart (Galen supposed both halfs were separated by a permeable membrane) and finally received its "animal spirits" from the brain.
Galen embraced by the Church
Galen's medical authority lasted for almost 1500 years. His writings on human physiology, anatomy, the use of medication etc. were embraced by the church and achieved dogmatic status. It was hard for his successors to propose new ideas on the physiology of blood and blood flow.
Harvey (1578-1657)
In 1628, after 20 years of experimenting, William Harvey (1578-1657) finally felt confident enough to publish 'De motu cordis'. In this book Harvey described how the blood circulates through the body: that the blood travels through arteries to the different organs and returns to the heart via veins. Although at that stage he was unable to see them himself he postulated that the blood passes through tiny capillaries allowing the transition from arterial to venous phase. Also, he described the small circulation from heart to lungs and back. He realised that the small and large circulation are separated only by the heart septum which therefore cannot be permeable to blood.
Young (1773-1829)
Thomas Young was a man of many talents. In his Croonian lecture ("Functions of the Heart and Arteries", 1808) he reported on experiments with tubes of variable length and diameter. He discovered that tubes with a smaller diameter were more resistant to flow than tubes that were wider. Since the diameter of arteries becomes smaller at every junction of the arterial tree, their resistance becomes higher.
This argument, however, only holds for individual arteries and ignores the effect of their exponential increase in numbers: a replacement resistance would be smaller than the sum of the arteriolar resistances since they stand in parallel, not in series. The ever decreasing diameter of conducting arteries in the arterial tree was also erroneously taken as explanation for the often observed increase in pulse wave velocity towards periphery, again ignoring the exponential increase in numbers.
Frank (1865-1944)
The name of Otto Frank is strongly linked to the Frank-Starling law of the heart, but also to the Windkessel model of the arterial system. Frank envisaged how the aorta by its elastic properties could temporarily store the volume ejected by the heart to release its content during diastole. He devised a 2-element windkessel model depicted below to explain the characteristics of the arterial blood pressure wave.
McDonald (1917-1973)
Donald A McDonald, the well known author of "Blood flow in arteries" (1960), was the first to explain the characteristic waveform of the arterial pressure pulse from forward and backward traveling waves. He found that while the pressure pulse is increasing in size the amplitude of the flow pulse is decreasing. This was modeled as a wave traveling in a tube with a partially closed end. The resulting high resistance to outflow would cause reflecting waves traveling back in the direction of its origin and causing complex interference at any point of measurement along its trajectory.
Schaafsma (1962-..)
After performing many simultaneous transcranial Doppler (TCD) and arterial blood pressure (ABP) investigations in patients with carotid stenosis as well as patients undergoing autonomic function tests, Schaafsma came to the conclusion that the systolic waveform consists of two distinct phases that vary independently: a Sys1 thought to arise from a short-lasting contraction in the smooth muscle cells within the conducting arteries, followed by a Sys2 that related to stroke volume and peripheral resistance. For instance, during deep breathing (5 seconds of deep inhalation followed by 5 seconds of exhalation repeated 8-10 times) the variability of Sys2 was considerably larger than of Sys1.
The theory of arterial acceleration was first published in Medical Hypotheses in 2014 but for long received only little attention. It was followed by an observation of blood flow velocity recordings in patients with aortic stenosis in 2026 showing that the turbulent flow measured directly behind the stenotic aortic valve (Sys2) became increasingly preceded and overtaken by a wave with laminar flow (Sys1) at proximal to distal measurement sites. This study provided strong evidence that the conducting arteries add energy to the pressure wave from the heart.
Almost all energy required for blood flowing through our arteries and veins is delivered by the heart. A small amount is delivered by a short-lasting contraction of the arteries at every heart beat onset; that, at least, is the proposition of the theory of arterial acceleration.
It is important that we realize the heart is a volume pump, not a pressure pump. The most important argument is that the heart delivers its energy in a pulsed fashion. During systole the heart pumps its volume into the arterial tree. Part of the volume is added to the volume already present in the arterial tree allowing cardiac energy to be transformed to potential energy generated by stretch of the arterial wall. Part of the volume pushes forward the volume already present, forcing blood out into the many capillary networks. During diastole, it is the potential energy of blood volume stored in the arterial tree that forces out blood from arterioles into capillaries.
Let us briefly imagine the situation where the wall of the arterial tree is non-compliant, as if it were made of stainless steel. Then any volume pushed in on the one side of the system would inevitably and instantaneously lead to the same volume being pushed out on the other side. Broader pipes have less resistance than smaller ones and will allow more blood to flow. Longer pipes have a higher resistance than shorter ones and will receive less blood to flow. There is no pressure gradient, since fluid cannot be compressed and no buildup of pressure can occur when it leads to instantaneous relief over all the branches.
The figure above represents a caricature of the windkessel model, with a pressurized wide body capacitance and many outflow resistances to be regulated for blood supply to the different destinations. Traditionally, organs or thought to adjust their arteriolar resistance locally by means of metabolic coupling, or their blood supply is regulated by humoral or neural control. By no means can the heart preferentially direct its flow towards one or more organs, leaving other organs out. Instead, the heart forces its stroke volume into the narrow end of a funnel with an immense increase in overall cross-sectional area from proximal to distal.
Given that the arterial resistance will greatly differ from branch to branch the system risks ill-perfusion of regions with somewhat higher resistance in favor of those with a reduced resistance or a less remote localization along the arteriolar tree. How can the arterial tree ensure that all the bodies capillary systems are at least perfused with minimal flow?
The theory of arterial acceleration proposes that the first increase in pressure in the proximal aorta triggers a myogenic response in its smooth muscle cells: a so-called stretch induced depolarization. The smooth muscle cells are electronically coupled causing this depolarization to spread from proximal to distal along the contractile layers of the arterial tree. Inside a single smooth muscle cell, this depolarization induces a Ca-spike: a brief inflow of extracellular calcium that (still to be proven) causes a slight and short-lasting contraction of the cell. Smooth muscle cells in arteries are arranged circularly and not longitudinally. Together, this wave of depolarization and the contraction within circularly arranged smooth muscle cells is thought to move along the branches of the arterial tree and bring blood into motion even in the most remote capillary systems of the body. The propagation speed of this wave (pulse wave velocity or PWV) can be calculated to increase from about 2 in the more proximal arteries to 8-10 m/s in the more distal. Its relative contribution gains in importance with respect to the pressure wave generated by heart contraction, since arterial acceleration builds upon the myogenic response in more proximal branches where the pressure wave dilutes along its trajectory: there is a great increase in overall cross-sectional area making the arterial tree a sort of funnel with the heart pumping blood into its narrow end.
Consequently, the Windkessel figure above is an over-simplification of what is really happening in the arterial tree. It leads to the common misunderstanding that a patient has only a single arterial blood pressure. Instead, it should be realized that arterial blood pressure varies with location and with time. Treatment decisions can be made on instant measurements only when the measurement location and method is standardized and when its systolic waveform is correctly analyzed for its different components as below.
It is an important observation that the two components of the systolic pressure wave vary with age. Aging takes a toll on arterial elasticity and arteries dilate and elongate when we become older. The loss of elasticity in the proximal arteries causes the well known increase in arterial stiffness. The capacitance in which the heart pumps its gradually decreasing stroke volume (aging also goes with a reduction of cardiac ejection fraction) becomes sloppy, its mechanics relying more on collagen than on elastin. More blood is required to fill the capacitance vessels, the relative contribution of stroke volume becomes less and circulation time will become larger since blood lingers longer in the arterial phase. Because of the increase in stiffness, the pressure increase by addition of stroke volume is steeper and more blood is forced out during systole. During diastole, however, due to a rapid decrease in pressure diastolic blood flow approaches zero. This causes the pulsatility index (difference between systolic and diastolic pressure or flow divided by mean pressure or flow) to increase with old age.
Arterial acceleration decreases with aging: arterial blood pressure and intracranial blood flow recordings show that the systolic waveform is Sys1 dominant in the young but becomes increasingly Sys2 dominant in the old. The dilatation of capacitance arteries, causing stretch of the smooth muscle cells, is possibly detrimental for the underlying short-lasting contraction of the myogenic response.
For a rapid interpretation of middle cerebral artery flow velocity recordings we have proposed a radar plot of age corrected Z-scores. Z-scores are calculated by dividing the difference between an actual measurement and the mean value expected for that age through the observed standard deviation. In this way any measurement can easily be weighed to what is expected for that age and the four components of the waveform (Sys1, Sys2, D560 and HR) can be compared to each other.
Above a radar plot is shown for the four components of middle cerebral artery flow velocity during hyperventilation (red), normoventilation (green) and CO2-retention (yellow).
The penetration force of Sys1 into remote capillary systems is larger than of Sys2 and diastolic blood flow since it builds upon the wave of arterial acceleration in more proximal arteries. Metabolic coupling determines whether arterioles dilate or constrict letting in more or less blood adding to the blood flow generated by arterial acceleration. Arterial acceleration also makes sure that tissue is perfused despite unfavorable conditions, such as when experiencing external pressure. A simple experiment with a blood pressure cuff positioned around the hand suggests that Sys1 flow or systolic spikes may even occur despite inflating the cuff to pressures well over 150 mmHg. Sys1 perfusion ensures that local ischemia due to, for instance, high tissue pressures will not go unnoticed to the central control centers (chemoceptors) allowing the organism as a whole to respond with an increase in blood pressure and/or respiratory activity (CNS ischemic response). Without Sys1 perfusion local ischemia may remain unnoticed as blood cannot enter into a non-perfused tissue with accumulating signs of local ischemia.
From a physiological perspective arterial acceleration is a welcome addition to the pumping activity of the heart allowing blood flow to occur into all the bodies capillary systems, including those that are most remote and under least favorable conditions.