EEE4260C: Bioelectrical Systems

Course Description

EEE 4260C Bioelectrical Systems applies the electrical engineer’s toolkit to the signals the human body generates. The heart, the muscles and the brain all produce measurable electrical activity, and this course treats those signals the way an engineer treats any other: model the source, characterise the noise, design the instrumentation, and process the result.

The Statewide Course Numbering System titles the number Bioelectrical Systems and describes it as covering “the theoretical and quantitative perspective of bioelectrical signals reflecting the activity of the brain, the muscles, and the heart. Bases of modeling, measuring, processing and analyzing bioelectrical signals and systems are discussed, as well as common clinical applications.” The statewide prerequisites are EEL 3111C, EEL 3112 and EEL 3135 — circuits and, decisively, a signals and systems course.

The University of South Florida carries the suffixed number EEE 4260C at 3 credits under the title Bioelectricity. The University of Florida teaches the same subject under the unsuffixed EEE 4260 Bioelectrical Systems, also at 3 credits, describing it in language that follows the statewide entry closely. The subject is the same; the suffix, and therefore the SCNS record, is not — see Special Information.

⚠ The word “quantitative” in the statewide description is doing real work. This is not a survey of medical technology. It is a modelling and signal-processing course that happens to take its signals from physiology, and students who enrol expecting an accessible introduction to biomedical engineering sometimes find the mathematics heavier than anticipated.

Learning Outcomes

Required Outcomes

• Explain the origin of bioelectric potentials from ionic concentration gradients, and apply the Nernst and Goldman–Hodgkin–Katz equations to compute equilibrium and resting membrane potentials.
• Model the cell membrane as an electrical circuit, relating membrane capacitance and ionic conductances to observed behaviour.
• Describe the action potential quantitatively using the Hodgkin–Huxley formulation, and explain the role of voltage-gated channels, threshold and refractory periods.
• Apply cable theory to signal propagation along axons and muscle fibres, and explain the effect of myelination on conduction velocity.
• Explain volume conduction and the relationship between the bioelectric source and the potential measured at the body surface.
• Describe the generation, characteristic waveform, amplitude and frequency content of the electrocardiogram (ECG), electromyogram (EMG) and electroencephalogram (EEG).
• Analyse the electrode–electrolyte interface, including half-cell potential, electrode impedance, polarisation and motion artefact.
• Design biopotential amplifier front ends, specifying gain, input impedance, bandwidth and common-mode rejection appropriate to the signal.
• Identify the principal noise and interference sources in biopotential measurement — power-line interference, motion artefact, baseline wander, and cross-signal contamination — and design filtering to address each.
• Apply digital signal processing to biosignals: sampling, filtering, spectral analysis, and feature detection such as QRS complex identification.
• Explain the electrical safety requirements governing devices connected to patients, including leakage current limits and patient isolation.

Optional Outcomes

• Analyse functional electrical stimulation and the design of pacemakers and defibrillators.
• Apply advanced signal processing — adaptive filtering, wavelets, independent component analysis — to artefact removal.
• Describe brain–computer interface principles and their signal-processing requirements.
• Solve the forward and inverse problems of source localisation in ECG or EEG.
• Describe bioimpedance measurement and its applications.
• Apply machine-learning methods to biosignal classification and clinical decision support.
• Describe the regulatory pathway for a medical device in the United States.

Major Topics

Required Topics

• Origin of bioelectricity — ionic composition of intracellular and extracellular fluid, selective permeability, the Nernst equation, the Goldman–Hodgkin–Katz equation, resting potential and the sodium–potassium pump.
• Membrane models — the parallel-conductance circuit model, membrane capacitance, time constant, and passive electrical properties.
• The action potential — voltage-gated channels, the Hodgkin–Huxley model and its gating variables, threshold, all-or-none behaviour, absolute and relative refractory periods.
• Propagation — cable equation, space and time constants, saltatory conduction in myelinated fibres, conduction velocity and fibre diameter.
• Volume conduction — the dipole source model, the relationship between source and surface potential, and why a surface recording is a spatially filtered view of activity.
• The electrocardiogram — cardiac conduction system, the cardiac dipole and its vector representation, Einthoven’s triangle and the 12-lead system, P-QRS-T morphology, and common arrhythmia signatures.
• The electromyogram — motor unit anatomy, motor unit action potentials, recruitment and rate coding, surface versus intramuscular recording, and the relationship between EMG amplitude and force.
• The electroencephalogram — cortical generators, the 10–20 electrode system, the classical frequency bands, evoked potentials, and the very low amplitude that makes EEG the hardest of the three to measure.
• Electrodes — the electrode–electrolyte interface, half-cell potential, Ag/AgCl and its advantages, polarisable and non-polarisable electrodes, electrode impedance and motion artefact.
• Biopotential amplifiers — the instrumentation amplifier, differential and common-mode signals, CMRR requirements, input impedance, right-leg drive, isolation amplifiers and defibrillation protection.
• Noise and interference — 60 Hz power-line interference and its coupling paths, notch filtering and its cost, baseline wander, motion artefact, and the practical limits of filtering.
• Biosignal processing — sampling rate selection, digital filter design for biosignals, spectral analysis, QRS detection algorithms, averaging for evoked potentials, and feature extraction.
• Electrical safety — macroshock and microshock, leakage current limits, patient isolation, grounding, and the standards that govern them.

Optional Topics

• Stimulation: pacemakers, defibrillators, functional electrical stimulation, deep brain stimulation.
• Adaptive filtering, wavelet analysis and independent component analysis for artefact removal.
• Brain–computer interfaces and neural prosthetics.
• Forward and inverse problems; source localisation.
• Bioimpedance, impedance plethysmography and body composition measurement.
• Machine learning for arrhythmia classification, seizure detection and gesture recognition from EMG.
• FDA regulatory pathways and design controls for medical devices.

Resources & Tools

• Medical Instrumentation: Application and Design (Webster) is the standard reference for the instrumentation, electrode and safety material and is the most commonly adopted text. Bioelectricity: A Quantitative Approach (Plonsey and Barr) is the standard treatment of the membrane, propagation and volume-conduction theory and matches USF’s course title directly. Biomedical Signal Processing (Rangayyan) covers the processing half.
• Principles of Neural Science (Kandel et al.) is the usual reference for the physiology, though only selected chapters are relevant.
• MATLAB with the Signal Processing Toolbox is the dominant environment; Python with NumPy, SciPy, neurokit2 and MNE-Python is increasingly used and is free.
• PhysioNet (physionet.org) is the essential resource for this course — a free, open archive of annotated physiological recordings including the MIT-BIH Arrhythmia Database, with the WFDB toolkit for reading them. Most course projects draw on it.
• Hardware — instrumentation amplifiers such as the AD620 and INA126, biopotential analogue front ends such as the TI ADS1299 (EEG) and ADS1292 (ECG), and teaching platforms including OpenBCI, BITalino and Backyard Brains kits.
• Standards — IEC 60601-1 (general safety of medical electrical equipment) and IEC 60601-2-25/-2-47 for electrocardiographs, AAMI EC13 for cardiac monitors, and ISO 14971 for risk management. These are the documents that actually govern a marketed device.

Career Pathways

• Biomedical engineer (SOC 17-2031, Bioengineers and Biomedical Engineers) — the direct destination.
• Medical device design engineer and biomedical instrumentation engineer, particularly in cardiac rhythm management, neuromodulation and patient monitoring.
• Biosignal processing engineer and algorithm engineer — arrhythmia detection, seizure detection and wearable health analytics.
• Clinical engineer (SOC 17-2031) — managing and maintaining medical equipment within a hospital system, a substantial and stable employment category.
• Regulatory affairs and quality engineer in medical devices — roles for which understanding both the physics and the standards is the qualification.
• Research and graduate study in neural engineering, brain–computer interfaces and neuroprosthetics.
• Florida is a strong market for this specialisation. Employers include Medtronic, Johnson & Johnson MedTech (Jacksonville and Palm Beach Gardens), Stryker, Arthrex (Naples — a major and growing medical-device manufacturer), Bausch + Lomb (Tampa), and large health systems including AdventHealth, Orlando Health, BayCare, Tampa General, Moffitt Cancer Center and Mayo Clinic (Jacksonville), all of which employ clinical and biomedical engineers directly. USF’s Tampa Bay location and its Morsani College of Medicine give students unusually good access to clinical settings.

Special Information

⚠⚠ The suffix problem — two institutions, two SCNS records

USF carries EEE 4260C; UF carries the unsuffixed EEE 4260. Both are 3 credits and both teach the same subject. SCNS equivalency does not cross the suffix, so a course taken under one number does not automatically satisfy a requirement written for the other.

The practical effect: a statewide inventory listing two institutions against EEE 4260C overstates its portability — only USF carries the suffixed identifier. Carry the syllabus, and request a substitution explicitly rather than assuming the numbers match.

The C suffix normally indicates integrated laboratory or design hours. Given the instrumentation content, expect USF’s version to include hands-on measurement — recording your own ECG or EMG is a standard and memorable exercise in this course. Confirm the meeting pattern, since the contact hours will exceed three per week if a laboratory is scheduled.

⚠ Course-code variation across Florida

Bioelectricity and biomedical instrumentation are taught across several prefixes in Florida, which makes transfer matching unusually unreliable here:

• EEE 4260C / EEE 4260 — USF and UF.
• EEE 4200C and EEE 4214 — Biomedical Instrumentation and Systems Design and Biomedical Imaging, carried under the EEE prefix elsewhere in the state.
• EEE 4800 Neural Signals, Systems, and Technology — University of Florida, an adjacent course on the neural side.
• Much comparable material sits under BME and EGN prefixes in biomedical engineering departments.

⚠ This is a prefix-divergence case as much as a suffix one. A student who took an equivalent course under a BME number has almost certainly covered the material, but a receiving programme naming its prerequisite as EEE 4260C will not match it. Read target-programme prerequisites literally, and raise substitutions with an advisor early.

Prerequisites

The statewide record lists EEL 3111C, EEL 3112 and EEL 3135 — circuits and signals and systems.

⚠ The binding requirement is signals and systems, not biology. The physiology in this course is developed from scratch and is genuinely accessible to a student with no biology background. What is not developed from scratch is the frequency-domain reasoning: filtering, spectral analysis, sampling and the Fourier view of a signal are assumed throughout, from the first week. A student who is shaky on signals and systems will find this course hard for reasons that have nothing to do with the body.

The instrumentation half additionally assumes fluency with operational amplifier circuits — differential gain, common-mode rejection and input impedance. Students who have not yet taken a second electronics course can manage, but should expect the instrumentation amplifier material to require extra reading.

Position in the curriculum

EEE 4260C is a senior-level elective, taken after circuits and signals and systems. It is frequently the course that determines whether an electrical engineering student pursues biomedical engineering at graduate level or in employment, and it pairs naturally with digital signal processing (EEE 4510) and with machine learning coursework.

⚠ Working with human subjects and with your own physiological data

Where the course includes laboratory recording — and most versions do — several requirements apply that students should know in advance:

• Participation is voluntary. No student is required to serve as a recording subject, and an alternative must be available.
• Measurements are not medical advice. An ECG recorded in a teaching laboratory is an engineering exercise. Students occasionally observe something in their own trace that worries them; the correct response is to consult a physician, not to interpret it from a course textbook, and not to be alarmed by an artefact.
• Classmates’ recordings are private. Physiological data about an identifiable person is sensitive, and should not be shared, published or retained beyond the exercise.
• Safety is not a formality. Any circuit connected to a person must be battery-powered or properly isolated. This is the reason the electrical-safety material is in the syllabus, and the laboratory rules implementing it are not negotiable.
• Research involving human subjects requires Institutional Review Board approval; classroom exercises are normally covered by a departmental protocol, but a student project that extends beyond the classroom may need its own.

Difficulty and time commitment

The course is demanding in an unusual way: it requires holding two vocabularies at once — the physiological and the engineering — and translating between them. The Hodgkin–Huxley model in particular is a system of coupled non-linear differential equations wearing biological labels, and students who try to memorise the biology without following the mathematics get stuck. Plan on eight to ten hours a week, plus laboratory time where scheduled.

Articulation and transfer

SCNS records this course as guaranteed to transfer to an institution offering the same course. Only USF carries the suffixed number, so read that narrowly. The course is upper-division and carries no general-education or Gordon Rule designation.

FE exam relevance

The NCEES Fundamentals of Engineering (Electrical and Computer) exam does not cover bioelectricity as a named topic. The signal-processing and instrumentation-amplifier content supports the Signal Processing and Electronics areas indirectly. Students in this specialisation who intend to pursue professional licensure should note that a Biomedical Engineering FE examination does not exist — biomedical engineers who seek a PE licence normally take the Electrical and Computer FE and then a corresponding PE examination.

AI Integration

Biosignal analysis is one of the fields where machine learning has genuinely changed practice, and where the consequences of a wrong answer are unusually serious. Both facts belong in this course.

Where AI is genuinely used in the discipline. Automated arrhythmia detection from ECG, seizure detection and prediction from EEG, gesture and intent decoding from EMG for prosthetic control, and sleep staging are all tasks where learned models now match or exceed classical algorithms, and several are FDA-cleared and in clinical use. Wearable devices performing atrial fibrillation detection are the most visible consumer example. A student entering this field will work with these systems, and understanding them is part of the qualification.

Where a general-purpose assistant helps in coursework. Explaining the physiology behind a waveform feature; walking through the Hodgkin–Huxley gating variables; generating MATLAB or Python code for filtering, spectral analysis and PhysioNet data handling; and explaining an unfamiliar clinical term.

⚠ Where it fails, and why the failure is this field’s central risk. The characteristic error is that a model will interpret a physiological signal clinically when asked — naming a rhythm, suggesting a diagnosis, or reassuring the asker — on the basis of a description rather than a recording, with no access to the patient, the leads, the artefact history or the clinical context. That is exactly the boundary this course teaches students to respect. The engineering deliverable is a measurement and its uncertainty; the clinical interpretation belongs to a clinician. A confident generated diagnosis is both wrong in method and, if acted on, potentially harmful.

A second failure has a specific technical shape. Models routinely recommend a 60 Hz notch filter for ECG without noting what it costs — the QRS complex has significant energy near 60 Hz, and aggressive notch filtering distorts exactly the feature most diagnoses depend on. That trade-off is a core lesson of the course, and it is the sort of nuance a generic answer omits.

Third, and important for anyone doing a project: a classifier trained on PhysioNet data and reported at high accuracy is usually being evaluated wrongly. Splitting records rather than patients leaks information between training and test sets and inflates accuracy dramatically. Arrhythmia datasets are also heavily imbalanced, so accuracy is the wrong metric — sensitivity and positive predictive value are what matter, because the cost of a missed event and the cost of a false alarm are not symmetric.

The engineer’s responsibility. A medical device makes a claim about a person’s body, and it is regulated because that claim can cause harm. An engineer signs for the measurement chain, its validation and its failure modes. In this field the discipline of stating what a system does not do — its false-negative rate, the conditions under which it is unreliable — is not a caveat but the substance of the work.

Academic integrity. USF and UF each maintain academic integrity policies covering AI-generated work. Analysis and interpretation are normally expected to be your own even where coding assistance is permitted, and generated physiological data is data fabrication — treated more seriously than plagiarism, and in this field a direct rehearsal of professional misconduct. Ask before you rely on a tool, and disclose its use where required.