Physiological Signals

Physiological signals are observable and recordable variations generated by bodily physiological processes that provide evidence about changing biological function during behavior. A physiological process is an ongoing biological mechanism or function occurring within the body. A physiological signal is a measurable time-varying manifestation associated with such a physiological process, captured through instrumentation or recording methods. A physiological measure is a quantity derived or estimated from the recorded evidence of a physiological signal. It is important to recognize that a physiological signal is not the physiological process in its entirety, is not behavior itself, and is not a direct measurement of emotion, stress, engagement, fatigue, pain, cognition, intention, diagnosis, or any other behavioral or psychological construct.

Meaning of Physiological Signals

Physiological signals exist scientifically as measurable manifestations of dynamic bodily function. Many physiological processes change continuously or episodically through electrical activity, mechanical motion, pressure, flow, conductance, volume, temperature, chemical composition, or related physical properties. Instrumentation allows selected manifestations to become observable over time, enabling physiological activity to be described quantitatively. However, capturing these signals does not imply that every relevant aspect of the underlying biological process has been recorded or represented.

It is crucial to distinguish among related concepts:

• Biological process: A broad term referring to any biological activity or mechanism at any level of organization.
• Physiological process: A specific ongoing biological function within the body, such as cardiac contraction or sweating.
• Physiological state: The condition or status of a physiological system over a relevant interval.
• Physiological response: A change in physiological state associated with an event or condition.
• Physiological signal: The recorded or recordable manifestation of a physiological process, such as an electrical potential or mechanical movement.
• Physiological measure: A summary or estimate derived from the physiological signal, such as heart rate or skin conductance level.
• Physiological interpretation: The scientific inference or claim made by connecting the measured evidence to a physiological or behavioral phenomenon.

These terms are not interchangeable and represent distinct layers of conceptualization.

Physiological signals differ from the broader term biosignal. While biosignals can include signals from many biological levels and sources—such as neural, biochemical, cellular, or other biological phenomena—physiological signals specifically refer to evidence tied to bodily physiological function relevant to human behavioral analysis. The two terms are related but not universally synonymous.

Historical Emergence of Physiological Recording

Physiological signals did not originate from a single invention but emerged gradually as physiology developed methods to convert transient bodily processes into visible, recordable, and comparable traces. This evolution arose from the need to measure processes that could not be reliably characterized through unaided observation alone, especially when timing, magnitude, waveform, or response dynamics were important.

The foundational contribution of Luigi Galvani in the late eighteenth century was pivotal. His experiments on "animal electricity" demonstrated that electrical phenomena are intrinsic to excitable biological tissue and participate in physiological processes such as nerve and muscle activity. This work established electrical activity as a measurable window onto physiology, although not all physiological signals are electrical in nature.

During the nineteenth and early twentieth centuries, the transition toward instrumental traces of bodily function became pronounced. Augustus D. Waller pioneered early human cardiac electrical recordings, while Willem Einthoven significantly advanced electrocardiographic recording by developing the string galvanometer. Einthoven’s innovations made cardiac electrical recording more practical, sensitive, standardized, and interpretable. This history highlights the scientific value of continuous physiological traces, enabling detailed temporal analysis rather than focusing solely on discrete observations.

In the late nineteenth century, Charles Féré and Ivan Tarchanoff conducted pioneering work on the electrical properties of the skin. Féré’s approach involved externally exciting the skin to measure electrical properties, whereas Tarchanoff observed endogenous skin potentials without external excitation. This example illustrates that even within the same physiological domain, different measurement principles can produce different kinds of signals.

More broadly, the historical importance of graphical and electrical recording lies in physiology’s increasing ability to represent bodily processes as time-varying traces. These traces can be compared across events, conditions, and people. In modern behavioral analysis, physiological recordings provide temporally precise evidence that is difficult or impossible to obtain through outward observation alone.

From Physiological Process to Recorded Signal

Physiological evidence follows an evidential chain: a physiological process produces a measurable manifestation, which is transduced or sensed, producing a recorded signal. From the recorded signal, derived physiological quantities are computed, which then support behavioral interpretation. Each transformation selects and modifies information. No sensor captures the entire physiological process; derived measures summarize useful structure while discarding other information.

Transduction is the process of converting a physiological manifestation into a measurable signal or representation. Examples include electrical potentials recorded by electrodes, optical changes associated with blood volume (e.g., photoplethysmography), mechanical displacement related to respiration, conductance changes of the skin, or motion linked to muscle contraction. Transduction does not imply direct equivalence to the physiological process; it converts specific physical properties into a form suitable for recording.

Direct and indirect physiological measurements must be distinguished cautiously. A measure can be physically close to a process without representing the biological quantity of interest directly. Many commonly used physiological measures rely on intermediate physiological and physical relationships. The term “direct” should be used only relative to a clearly specified quantity, and indirect measurement is not inherently inferior.

Major Forms of Physiological Evidence

Cardiovascular physiological evidence includes signals associated with cardiac electrical activity, cardiac cycles, pulse waves, blood-volume changes, pressure, and related circulatory phenomena. Electrocardiographic recordings and peripheral pulse-related measurements are examples. These signals are not by themselves measures of stress, arousal, fitness, or health status.

Electrodermal physiological evidence pertains to changes in the electrical properties of the skin, strongly linked to sweat-gland activity under sympathetic nervous system control. Skin conductance, resistance, and potential represent different electrical quantities and measurement approaches. Electrodermal activity should not be equated with emotion, stress, arousal, or sweating amount in totality.

Respiratory physiological evidence includes signals related to breathing cycles, thoracic or abdominal movement, airflow, pressure, gas exchange, or other respiratory phenomena depending on method. A respiration waveform is a selected manifestation of breathing, and measures such as breathing rate, depth, airflow, and respiratory effort are distinct quantities.

Electromyographic physiological evidence reflects electrical activity associated with muscle activation, recorded from suitable locations. Electrical muscle activation differs from resulting force, visible movement, joint motion, or behavioral action. Electromyographic amplitude is not a universal direct measure of muscle force or behavioral effort.

Other peripheral physiological evidence includes body or skin temperature, peripheral vascular signals, gastrointestinal activity, and other measurable bodily processes, illustrating the breadth of physiological evidence.

Physiological signals are conceptually distinct from neurophysiological evidence. Neural signals arise specifically from nervous-system activity, whereas physiological signals concern broader bodily physiological processes and peripheral manifestations. Peripheral physiological changes can be neurally regulated without constituting direct neural recordings.

Signal Components and Temporal Organization

Physiological signals can be described in terms of tonic and phasic components. Tonic components describe relatively sustained or slowly changing physiological levels over a specified interval, while phasic components describe shorter changes or responses against that tonic background. The distinction depends on the physiological system, timescale, and analysis. “Tonic” does not mean permanent, and “phasic” does not mean instantaneous.

Temporal concepts include:

• Baseline: A defined reference condition or interval.
• Response: Change relative to a relevant reference.
• Recovery: The process of returning toward a prior or alternative physiological level.
• Adaptation or habituation: Alteration of repeated responses over time.

Baseline is not a universal resting truth, and recovery does not prove that the underlying behavioral state has ended.

Descriptors of event-related physiological changes include:

• Latency: Time between event onset and physiological response onset.
• Rise time: Duration from response start to peak.
• Peak or extremum: Maximum or minimum response value.
• Duration: Total length of the response.
• Decay: Time for response to diminish after peak.
• Recurrence: Repetition of responses.
• Inter-response interval: Time between successive responses.

Physiological systems differ in characteristic response timing, and the time of a recorded response need not exactly match the timing of the associated behavioral, perceptual, or psychological process.

Physiological signals can also display oscillatory and cyclic structure. Cardiac and respiratory processes contain repeating cycles, while other signals may show slower fluctuations or rhythms. Detected periodicities can arise from physiological regulation, behavior, movement, environment, measurement methods, or system interactions, and should not be assigned a single interpretation automatically.

Physiological variability refers to meaningful variation across beats, breaths, responses, time, contexts, and individuals. Genuine physiological variability must be distinguished from measurement noise and artifact. Neither lower nor higher variability universally implies better or worse physiological stability or health without contextually specific interpretation.

Derived Physiological Measures

Physiological recordings often support derived quantities such as rate, interval, amplitude, area, slope, variability, response count, response latency, spectral content, coupling, and other summaries. A derived measure is useful because it focuses on a physiologically interpretable property but is not identical to the raw recording, the underlying physiological process, or the behavioral construct.

Waveform morphology includes detailed features such as peaks, troughs, intervals, slopes, shapes, and temporal relationships and can contain information discarded by scalar summaries. Conversely, scalar summaries can facilitate comparison when waveform detail is unnecessary. The appropriate representation depends on the physiological and behavioral question.

It is important to distinguish within-person change from between-person difference. Physiological levels and response magnitudes differ substantially across people due to anatomy, age, fitness, health, medication, environment, habitual activity, circadian factors, and other conditions. A measure useful for detecting change within one person may not support direct comparison across individuals.

Normalization and baseline correction are interpretive operations that express change relative to a reference, potentially reducing some between-person or session differences. However, they can discard absolute physiological information and become misleading if the reference condition is unstable or behaviorally inappropriate. These operations require careful contextual justification.

Physiological Coupling and Multisystem Responses

Physiological systems interact dynamically. Cardiac, vascular, respiratory, electrodermal, muscular, thermoregulatory, endocrine, and other processes can respond concurrently to shared regulatory, behavioral, environmental, or task-related influences. Parallel changes across signals may thus reflect shared regulation without implying redundancy or direct causation among signals.

Physiological coupling describes statistical or dynamical relationships among physiological signals or systems over time. Coupling differs from synchronization, correlation, common cause, and causation. Correlation between physiological signals can arise from direct interaction, shared regulation, common external events, respiration-related effects, movement, measurement artifacts, or coincident trends.

Physiological systems show cross-system latency and response diversity: different processes may respond to the same event with varying delays, durations, amplitudes, and recovery profiles. Simultaneous peaks or matching waveforms are not required for physiological responses to relate to the same behavioral episode.

Observation Conditions and Physiological Confounds

Physiological signals depend on both the underlying physiology and measurement conditions. Factors such as sensor placement, contact quality, posture, movement, temperature, ambient conditions, clothing, device pressure, participant movement, and instrumentation can alter recorded evidence. Apparent physiological changes may therefore be behavioral, physiological, technical, or a combination.

Careful distinction is needed between physiological variation and artifact. Movement can produce genuine physiological change while simultaneously corrupting sensor signals; respiration can be both a physiological phenomenon of interest and a source of variation in other recordings. Not all movement- or respiration-correlated components should be labeled artifact without reference to the scientific question.

Person-related influences include anatomy, fitness, medication, substances, hydration, sleep, circadian timing, recent physical activity, posture, illness, and habitual behavior. These factors contribute biological variability and interpretive context rather than diagnostic categories.

Environmental and task-related influences such as temperature, noise, physical and cognitive demands, social situations, posture requirements, speaking, movement, and breathing constraints also affect physiological signals. Physiological responses associated with experimental or behavioral conditions can arise through multiple plausible pathways and should not be automatically attributed to the intended psychological construct.

Physiological Signals and Behavioral Meaning

Physiological signals can be behaviorally informative because bodily regulation changes with physical activity, orientation, effort, environmental demands, social interaction, affective processes, cognitive demands, recovery, and many other conditions. Physiological evidence contributes to behavioral interpretation when relationships are operationalized and alternative explanations considered.

Physiological interpretation is many-to-many: one behavioral or psychological condition can produce changes across several physiological systems, and the same physiological change can occur under many different behavioral conditions. This precludes universal mappings such as “elevated heart rate = stress” or “increased skin conductance = fear.”

Physiological signals may reflect autonomic regulation, somatic motor activity, endocrine or metabolic processes, thermoregulation, and their interactions. A recorded signal should not be classified as purely sympathetic, parasympathetic, voluntary, involuntary, emotional, or cognitive without supporting physiological basis.

There can be dissociation among physiological response, outward behavior, and subjective experience. A person may show physiological change without visible behavioral expression, visible behavior without corresponding physiological change, or subjective experience not uniquely reflected by a single physiological measure. Disagreement among these forms of evidence is scientifically interpretable rather than necessarily erroneous.

Use in Behavioral Signal Processing

Physiological signals are valuable in Behavioral Signal Processing because they provide temporally structured evidence about bodily regulation and response that complements externally observable behavior. They remain available when overt expression is limited and reveal physiological dynamics not visible through ordinary observation. Their value depends on the evidential relevance to the behavioral question, not on granting physiology privileged access to psychological truth.

Representative uses include:

• Stress-related and workload-related research: physiological signals can index bodily regulation changes linked to task demands.
• Affect-related behavior: evidence about autonomic nervous system engagement.
• Physical effort and recovery: muscle and cardiovascular activity patterns.
• Sleep and fatigue-related behavior: physiological rhythms and autonomic tone.
• Health-related behavioral assessment: monitoring vital signs and physiological variability.
• Human-computer interaction: real-time physiological feedback for adaptive interfaces.
• Learning: physiological correlates of engagement and cognitive load.
• Social interaction: autonomic and motor signals reflecting interpersonal dynamics.
• Exercise and mobility: physiological indicators of activity and exertion.
• Adaptive systems: physiological input for context-aware computing.
• Long-duration wearable observation: continuous monitoring in naturalistic settings.

Physiological evidence can serve as a predictor, outcome, reference evidence, contextual evidence, or behavioral correlate depending on the scientific question. The same heart-related, electrodermal, respiratory, or muscular quantity may play different analytical roles; these roles must be explicitly defined rather than assumed from the signal name.

Physiological signals relate to movement, facial behavior, gaze, vocal behavior, language, touch, spatial organization, task events, and environmental conditions. These related knowledge domains support interpretation but agreement across signals is not automatic validation, nor is disagreement necessarily failure.

Quantification and Scientific Limits

Physiological signal analysis uses diverse mathematical tools: time-domain measurement, frequency-domain analysis, time-frequency methods, event-related analysis, variability measures, nonlinear dynamics, statistics, probability, coupling analysis, and machine learning. These tools characterize selected properties of physiological evidence; no single universal equation defines a physiological signal or its behavioral meaning.

Equations become meaningful only when tied to a specific physiological quantity, such as rate, interval, conductance, pressure, flow, electrical potential, or spectral property. Because these relations are physiologically specific, conceptual explanation is preferred over a single formula governing all physiological signals.

Inferential distance must be recognized: claims about measured voltage, conductance, pulse interval, respiration cycle, muscle electrical amplitude, temperature, or derived physiological rate are closer to recorded evidence than claims about stress, arousal, emotion, fatigue, pain, workload, engagement, cognition, intention, diagnosis, or subjective experience. Stronger behavioral claims require explicit operationalization, context, suitable reference evidence, and evaluation.

Computational physiological analysis carries risks of unintended-information and confounding. Models may exploit participant identity, physical activity, posture, respiration, medication, device placement, recording site, task structure, sensor quality, environmental temperature, missing-data patterns, or other correlated factors while appearing to predict behavioral targets. Predictive performance alone does not establish that the intended physiological pathway or behavioral mechanism has been identified.

Greater physiological response is not universally more meaningful or severe. Larger amplitude, faster rate, greater variability, lower variability, stronger coupling, or slower recovery may have different implications depending on physiological system, baseline, person, timescale, task, and context. Universal better/worse or high/low interpretations detached from physiology should be avoided.

In synthesis, physiological signals are time-varying measurable manifestations of bodily processes, captured through specific physical relationships and converted into recorded evidence and derived quantities. Scientific interpretation requires separating physiological process, signal, measurement, derived measure, physiological state or response, behavioral cue, and behavioral claim. This evidential chain preserves conceptual clarity and avoids treating bodily change as direct access to psychological meaning.