Article

Biomedical engineering in heart-brain medicine: A review

Author and Disclosure Information

ABSTRACT

New reports have emerged exploring the use of electrical stimulation of peripheral nerves in patients for the treatment of depression, heart failure, and hypertension. Abolishing renal sympathetic nerve activity in resistant hypertension has also been described. Since nerve bundles carry a variety of signals to multiple organs, it is necessary to develop technologies to stimulate or block targeted nerve fibers selectively. Mathematical modeling is a major tool for such development. Purposeful modeling is also needed to quantitatively characterize complex heart-brain interactions, allowing an improved understanding of physiological and clinical measurements. Automated control of therapeutic devices is a possible eventual outcome.


 

References

Biomedical engineering is a rapidly growing field, as indicated by a recent threefold increase in the number of students enrolled in the more than 80 programs granting biomedical engineering degrees.1 The field is known primarily for contributing to the development of devices that aid diagnosis (such as chemical sensors and medical imagers) or help to restore lost function (such as pacemakers, cochlear implants, and artificial limbs). At the same time, biomedical engineering has also contributed to the understanding of physiology and is now a participant in the more recent molecular- and cellular-based discoveries and their potential clinical applications.

This article reviews the contributions of biomedical engineering to heart-brain medicine and looks ahead to where its future contributions may be expected. The focus is on the autonomic control of the heart, with special emphasis on stimulating or blocking the activity of peripheral nerves for therapeutic purposes.

AUTONOMIC MECHANISMS

One way biomedical engineering has contributed to heart-brain medicine is through systems physiology, attempting to characterize complex cardiac control mechanisms by mathematical models ranging from the relatively simple to the very complex.2 In particular, investigation of the effect of the baroreceptor reflex (baroreflex) on heart rate led to recording of efferent vagal activity, demonstrating that the respiratory variations in heart rate are attributable to complete stoppage of vagal efferent activity, at least in anesthetized dogs.3 The results suggested that respiratory sinus arrhythmia was a measure of parasympathetic cardiac control.4 Extensive further work investigated heart rate variability in humans, resulting in the generally accepted concept that rapid variations in heart rate are primarily due to the parasympathetic nervous system, while slower variations are primarily due to sympathetic contributions. 5 It has been amply demonstrated that in a variety disease states, a high degree of parasympathetic control is correlated with improved outcome.6,7

Heart rate variability: Much is still unknown

Because heart rate variability is derived through analysis of easily recorded single-lead electrocardiograms (ECGs), techniques are still being described for determining the degree of variability.8–10 Measurements may be based on heart rate or heart period (interbeat interval); they may be made during spontaneous, deep, or timed periodic breathing; and the recordings may last for a few minutes or for 24 hours. The results are rarely reported for different states, yet variability depends on whether the subjects are awake, in quiet sleep, or in REM sleep.11

The present incomplete understanding of the physiological basis and clinical significance of heart rate variations makes it difficult to judge what the optimum measurement is. For example, recent publications have still debated whether the respiratory variations are primarily of central origin or a result of the baroreflex.12,13 The reviewed evidence leaves little doubt that most respiratory variations are caused by vagal modulation induced by breathing, independently of the baroreflex. On the other hand, breathing affects blood pressure and thus must have some influence on heart rate through the baroreflex as well. These effects are likely to be important when considering low-frequency variations.14 It also has been suggested that the branch of the vagus that contributes to the slow variations may be different from the one responsible for the rapid changes.15

A role for mathematical modeling

Untangling the above relationships requires at least an approximate physiologically based quantitative model of the entire system. What is the “entire system”? It includes all components that significantly affect the clinical condition or physiological/biological phenomenon studied. Developing such models is challenging since they can be misleading if they are too simple. However, if they are too complex, they may obscure rather than illuminate. Even though mathematical modeling has been a major component of biomedical engineering for decades, there is still great need for physiological and clinical studies that are aided by biomedical engineers with mathematical skills and a discerning eye toward the life sciences.

For example, following numerous previous efforts to develop models of cardiorespiratory control systems, a model has been developed to describe changes in heart rate, stroke volume, and blood pressure caused by disordered breathing during sleep.16 The model was initially tested in animals but recently has been evaluated in children with obstructive sleep apnea.17 It was found to be more effective in identifying and characterizing cardiac control abnormalities than was spectrum-derived variability of heart rate or blood pressure alone.

To enhance the usefulness of heart rate variability as a clinical tool, it is necessary to go beyond observing that decreased high-frequency heart rate variations are ominous in a particular disease state because there is an “imbalance” of autonomic control. This gives the physician little guidance as to what to do. Should he or she treat the brain to get more parasympathetic outflow? Should attention be concentrated on making the heart more tolerant to the imbalance? Or should both approaches be tried?

In addition to the building of models, biomedical engineers can also contribute to heart-brain medicine by developing sensors that can measure appropriate physiological variables in animal experiments—and eventually in humans. Such variables include neural activity and chemical signals controlling the heart, as well as neural and chemical signals that arise from the heart and are sensed by the brain. These variables must be included in any model for a comprehensive characterization of heart-brain interactions.

Next Article: