EGEL Plus Ingeniería Biomédica common mistakes
EGEL Plus Ingeniería Biomédica · Understand the traps and how to avoid them.
Resolving drug-dose, dilution, or unit-conversion problems by memorizing a formula (C1V1=C2V2) and applying it without checking whether the concentration units on each side match (mg/mL vs %, mg/dL vs g/L).
Chemistry and physiology content is often studied as isolated formulas, so candidates treat unit labels as decoration rather than as part of the quantity. The exam deliberately mixes % w/v, mg/dL, and mEq/L because clinical reports use all of them, and the arithmetic looks routine so the mismatch goes unnoticed.
What to do instead
Before computing, write the target unit next to the answer blank and convert every given value into that unit in a separate line. Practice ten conversions among %, mg/mL, g/L, mg/dL, mmol/L, and mEq/L, including one where the valence of the ion (Na+, Ca2+) changes the result. Make dimensional analysis the default: carry units through the multiplication and cancel them explicitly.
Applying a physics or biomechanics relation without checking static equilibrium or sign conventions - for example computing torque about the wrong point in a lever problem of the musculoskeletal system, or ignoring that a joint reaction force and a muscle force act in opposite directions.
Candidates learn formulas such as torque = force x distance but practice them on textbook levers with obvious pivots. In the body the pivot is the joint center, muscle insertion distances are short, and body-segment weights add extra terms, so the naive single-muscle calculation gives a plausible but wrong number.
What to do instead
For every biomechanics item, draw a free-body diagram first, mark the pivot explicitly, and list all forces with their lever arms and directions before substituting numbers. Do three problems that require summing moments about the joint and about a second point to confirm consistency. Keep an explicit sign convention written at the top of your scratch work.
Designing or analyzing an instrumentation chain by choosing gain from the signal amplitude alone, ignoring the electrode offset (half-cell potential), common-mode interference, and the input range of the ADC, so the amplifier saturates or the resolution is wasted.
Biopotential problems are usually practiced with clean simulated signals. Real ECG/EMG electrodes sit on a DC offset of tens to hundreds of millivolts while the signal of interest is 1-5 mV, so a gain that looks reasonable on the signal saturates the stage on the offset. Candidates also forget that CMRR matters only when the interference is actually common to both inputs.
What to do instead
For each amplifier question, tabulate signal amplitude, DC offset, interference amplitude, desired output range, and ADC resolution, then check saturation and quantization noise separately. Compute total gain as the product of stages and verify the DC-blocking or driven-right-leg stage is placed where it can actually remove the offset. Practice one full chain calculation from electrode to ADC bits.
Confusing sampling-rate adequacy with anti-alias filter design - stating that a 500 Hz sample rate is fine for a 200 Hz signal without checking what happens to higher-frequency interference, or forgetting that the analog low-pass filter must cut off below the Nyquist frequency, not at the signal bandwidth.
Nyquist is taught as a single inequality (fs > 2fmax) and the filter is treated as a separate topic. In real biopotential acquisition, EMG noise and mains pickup above Nyquist fold into the band of interest, so the inequality is satisfied while the recording is corrupted.
What to do instead
For every sampling question, name the highest frequency present at the electrode, not just the physiological band of interest, and set the analog anti-alias cutoff below fs/2 with margin for filter roll-off. Work one example where 50/60 Hz mains aliases into a low-frequency band and identify the alias frequency. State the filter order and attenuation needed, not just the cutoff.
Choosing a digital filter or spectral method by habit (always a Butterworth low-pass, always an FFT) without stating the requirement: passband ripple, transition width, phase linearity, or whether the analysis needs time localization.
Coursework drills filter design procedures but rarely drills requirement elicitation. Candidates therefore answer with a filter type when the question asks whether a filter is appropriate at all, or apply an FFT to a non-stationary signal such as a seizure onset where the spectrum smears and the event timing is lost.
What to do instead
Answer in two steps: first state the requirement in words (must preserve ST-segment shape, must not shift QRS timing, must separate two bands 2 Hz apart), then justify the filter or transform against it. Build a one-page comparison of Butterworth, Chebyshev, FIR linear-phase, and wavelet/STFT with the property each one buys. Practice identifying when zero-phase (filtfilt) filtering is required.
Evaluating medical image quality by resolution alone - quoting pixel size or matrix dimensions while ignoring contrast, noise, and the trade-off that a smaller pixel reduces the signal per pixel and raises noise, or confusing spatial resolution with contrast resolution.
Resolution is the most visible and easily memorized image metric, and matrix size appears in every specification sheet. The exam tests whether the candidate understands that SNR, contrast-to-noise ratio, dose, and slice thickness are coupled, which requires reasoning rather than recall.
What to do instead
For any image-quality question, write the four parameters (spatial resolution, contrast resolution, noise, dose or acquisition time) and explain which two the proposed change trades against. Work through one CT example (thinner slice increases noise for fixed dose) and one MRI example (smaller FOV with fixed matrix reduces SNR). Be able to define MTF and CNR in one sentence each.
Selecting a biomaterial by a single headline property - choosing a metal because it is strong, or a polymer because it is biocompatible - without matching degradation behavior, elastic modulus, fatigue life, and sterilization compatibility to the specific implant site and load history.
Biomaterials content is dense with property tables, and candidates memorize values without a selection framework. Stress shielding, corrosion in chloride environments, and heat sensitivity of resorbable polymers are exactly the interactions the exam probes, and they are invisible in a single-property comparison.
What to do instead
Use a fixed four-column screen for every material question: mechanical (modulus, fatigue, strength), chemical/biological (corrosion, degradation rate and products, cytotoxicity), processing (sterilizability, machinability), and clinical (load environment, expected service life). Practice the classic contrasts: titanium vs stainless steel in chloride, cortical bone modulus vs implant modulus, PLGA vs permanent polymer for a resorbable suture.
Treating a biomechanics problem as a static strength check when the question involves viscoelasticity or fatigue - using a single Young's modulus for cartilage, tendon, or bone under cyclic loading and concluding the tissue will not fail.
Introductory mechanics teaches linear elasticity with one modulus, and candidates carry that habit into tissue problems. Biological tissues are anisotropic, viscoelastic, and fail by cumulative damage at stresses far below ultimate strength, so a static safety factor gives false reassurance.
What to do instead
Before calculating, classify the tissue (elastic vs viscoelastic, isotropic vs anisotropic) and the loading (single load, cyclic, strain rate). For viscoelastic cases, reason with creep, stress relaxation, and hysteresis explicitly; for cyclic cases, use an S-N or damage-accumulation argument rather than a yield check. Practice explaining why a tendon at 30% of ultimate strength can still rupture after many cycles.
Answering equipment-management items with a single action (calibrate it, replace it, take it out of service) instead of a decision sequence that includes verification against a standard, documentation, notification, and root-cause analysis.
Clinical engineering practice is procedural and candidates study the concepts (calibration, preventive maintenance, incident reporting) as separate vocabulary items. Exam items describe a realistic scenario - a reading that drifts after service, an adverse event - where the correct answer depends on the order of steps and on who must be informed.
What to do instead
Memorize one generic workflow and apply it to every scenario: detect/verify the fault with a calibrated reference, remove from clinical use and label, document in the work order and equipment history, notify the clinical user and, where required, the regulatory authority, then perform root-cause analysis and corrective action with re-verification before return to service. Practice mapping adverse-event reporting timelines and responsibilities (manufacturer, user facility, regulator) onto that sequence.
Treating interoperability and health-information items as vocabulary recall (HL7, DICOM, FHIR) without reasoning about the message flow, identifier matching, or privacy rule that the scenario actually tests - for example assuming that sharing a DICOM study between systems requires no patient-identifier reconciliation.
Standards are learned as acronym lists, so candidates pick the acronym that appears most often in the question rather than the one that fits the layer being described (transport, message format, terminology, image object). Telemedicine and record-exchange scenarios also hinge on consent, identity matching, and data minimization, which are not in the acronym list.
What to do instead
Build a layer map: network/transport, message and document format, terminology and coding, image and object model, and security/consent. For each practice scenario, identify which layer the question is about before choosing a standard. Work two identity-matching examples (same patient, different medical record numbers across institutions) and one consent example for telehealth, stating who may access what and under which authorization.