Blume's Atlas of Pediatric and Adult Electroencephalography, 1st Edition

Chapter 2

Artifacts

Definition: Unwanted Potentials Originating from Extra-Cerebral Sources

Electrode (Figs. 2-12-22-4 to 2-9)

  • Abrupt, bizarre-appearing potentials.
  • Differ markedly from cerebrally originating background activity.
  • Superimposed upon background activity.
  • Usually confined to single electrode.

Faulty Ground (Fig. 2-10)

  • 60 or 50 Hz.
  • Other potentials that do not blend with ongoing cerebral originating activity.
  • In many or all channels.

Eye Movement (Figs. 2-11 to 2-202-47)

  • Cornea 100 mV positive compared with retina.
  • Upward rotation of ocular globe on eye blinking or closure. FP1, FP2 become more positive, create downward deflections when connected to other standard scalp leads.
  • Lateral eye movements create opposite polarities at F7, F8; that is, leftward movement increases positivity at F7 and decreases positivity (increases negativity) at F8.

Muscle (Figs. 2-21 to 2-232-25 to 2-292-442-45)

  • Very brief potentials.
  • Single or multiple.
  • May obscure EEG.
  • Principally temporal, frontal, and occipital areas; may be diffuse.
  • High-frequency filtering produces spike-like or beta-like appearance.
  • Frontal spikes rarely resemble muscle artifact.

Glossokinetic (Figs. 2-242-30 to 2-32)

  • Bursts of diffuse delta.
  • Field varies according to tongue position.
  • Usually accompanied by bursts of muscle artifact.
  • Tip of tongue has negative DC potential with respect to its base.

Regularly Repetitive Muscle Potentials (Figs. 2-33 to 2-35)

  • Tremor.
  • Focal motor seizures.
  • Segmental myoclonus, i.e., palatal.

Cardiac (Figs. 2-362-37)

  • R-wave most prominent.
  • A1 usually positive.
  • A2 usually negative.
  • Appears at O1,2 if neck is short.

Pulse (Figs. 2-382-39)

  • Periodic waves.
  • Smooth or sharply contoured.
  • Time-locked to electrocardiogram (ECG).
  • 200 ms delay to wave peak.

Metals (Fig. 2-40)

  • Abrupt, spike-like, brief single or multiphasic potentials.
  • From dental fillings moving against each other

P.16

Subgaleal Fluid (Figs. 2-41 to 2-43)

  • Attenuates potentials by “salt bridge” or increased distance of EEG generators from electrodes.
  • More apparent on bipolar montages.

External to ICU Patient (Figs. 2-462-48)

  • Artifact relating to the unique intensive care environment.
  • Morphology may resemble theta, delta, spikes.
  • Often intermittent, periodic.
  • Unlikely to have a “physiological field.”
 

Fig. 2-1. Electrode artifacts. Defects in electrical and mechanical continuity of an electrode can produce bizarre, often sudden electrode potentials that differ markedly from ongoing background activity, do not blend with other simultaneously recorded activity, and appear only in derivations involving one electrode. In the top segment, such activity appears uniquely at Fz and has various bizarre shapes. Such intermittent activity can reflect varying impedance from imperfect electrode contact to the scalp. Occasionally, such activity can be repetitive and share morphology of periodic lateralized epileptiform discharges (PLEDS) or a focally originating seizure as in the center tracing. These repetitive potentials represent FP2 artifact because they did not spread at all to adjacent channels (not shown) and are superimposed upon normal background activity. They bore no relationship to ECG (not shown). The high voltage and persistent nature of the Pz artifact in the bottom segment is the type produced by a faulty connection between the electrode and its wire or by its receptacle in the head box. Calibration signal 1 s, 50 µV.

 

Fig. 2-2. Adjacent artifacts. Electrode artifacts may occur in adjacent electrode positions, creating the false impression that arrhythmic delta activity and/or spikes are present. Thus, the delta activity at F8 and T4 here is entirely artifactual, as suggested by (a) the electropositive nature of the apiculate waves–spike-like artifacts, which are commonly electropositive and (b) confinement of each artifactual waveform to a single electrode position. Note the normal background activity underlying the F8–T4 artifact. Calibration signal 1 s, 50 µV.

 

Fig. 2-3. Head movement artifact during hyperventilation. Not all single-electrode-position artifacts represent technical faults. The low-frequency arrhythmic delta activity recorded at O1 during hyperventilation is not cerebrally originating because of the lack of associated “background” abnormality such as focal theta or attenuation and the lack of any spread to P3 as noted at the C3–P3 derivation (left). Stopping the head movement abolishes the artifact (right). Calibration signal 1 s, 50 µV.

 

Fig. 2-4. Artifactual delta activity from multiple high-impedance electrodes. Suspicion that this apparently diffuse delta activity may be artifactual stems from the virtually normal background activity and from 60-Hz artifact in the P3–O1 derivation (top). Diminishing the impedance to acceptable levels eliminated the artifact (bottom). Calibration signal 1 s, 100 µV.

 

Fig. 2-5. Artifacts on referential recordings. At least some of the potentials in the tracing on the left resemble temporal spikes because the apiculate phase is electronegative. However, the similar voltage and waveform in all left-sided derivations indicate confinement of these potentials to A1; this would not occur with temporal spikes, where the involvement of F7, T3, and even Fp1 is usual. The common average reference (right) is vulnerable to head movement and to electrical charges in the environment, producing the widespread stereotyped potentials seen here. Although the first potential resembles a spike–wave, its identical appearance in all derivations identifies it as artifact; “generalized” spike–waves are always accentuated either anteriorly or posteriorly. Calibration signal 1 s, 70 µV.

 

Fig. 2-6. Effect of interchange of electrodes on background activity. Erroneous substitution of Fz for O1 (left). Note the higher amplitude, out-of-phase alpha activity in the 4th and 12th channels (P3–Fz, T5–Fz). Tracing (right) after correction of erroneous substitution; alpha activity is only slightly higher in the left hemisphere, and the complete phase reversal of alpha activity is no longer present in the parasagittal leads. Normal phase reversals of alpha activity are present in the T5, 6–O1,2 derivations. Montage labeling indicates intended linkages. Calibration signal 1 s, 50 µV.

 

Fig. 2-7. Spurious amplitude asymmetry and other oddities from interchange of electrode positions. The several bizarre features of this recording can be explained by the erroneous interchange of Fz with O1. This produces an out-of-phase alpha activity of considerably higher voltage in the 4th and 12th channels (P3–Fz, T5–Fz), out-of-phase eye blinks because of the greater electropositivity at Fz than at either P3 or T5, and a mu rhythm from Fz in the 4th and 12th channels. Note the low-amplitude lambda activity in the technically correct derivations of P4–O2 and T6–O2. Montage labeling indicates intended, not actual linkages. Calibration signal 1 s, 50 µV.

 

Fig. 2-8. Reversal of FP2, F4 electrodes. The eyeblink potentials are prominently altered in the right hemisphere leads. Labelling is the intended montage. Calibration signal 1 s, 70 µV.

 

Fig. 2-9. Effect of interchange of electrodes on frontal activity. The tracing on the left illustrates erroneous interchange of Fz with O1, producing alpha activity in “F3–Fz” and “Fz–F4” derivations. Correction (right) reveals the normal low-voltage frontal activity. Suspect erroneous electrode interchange when high-amplitude activity is recorded on a coronal montage in derivations involving Fz, Cz, or Pz. Montage labeling indicates intended linkages. Calibration signal 1 s, 70 µV.

 

Fig. 2-10. Artifact from faulty ground electrode. A 4-Hz and 60-Hz artifact from the operating room transformer appears in most channels of this bipolar montage. Changing to a new ground lead (last 4 seconds) eliminated this artifact. Artifact from a faulty ground electrode does not always appear in all derivations. Calibration signal 1 s, 70 µV.

 

Fig. 2-11. Vertical eye movements and infraorbital leads. Electropositivity at the corneae produces a sudden negativity at Fp1,2 and positivity infraorbitally upon opening the eyes (asterisk in 3rd second) and the opposite polarity upon their closure (+). Similar phenomena occur in the 6th second. IOL, IOR, infraorbital left, right. Calibration signal 1 s, 50 µV for top 8 channels. Bottom 2 channels, 1 s, 150 µV.

 

Fig. 2-12. Eye flutter with photic stimulation. Three-Hz photic stimulation with the eyes open produces synchronous eye blinks followed by eye openings; each resulting positive–negative “complex” is preceded by a synchronous apiculate wave which presumably represents periocular muscle contraction. The resulting sequential complexes resemble spike–waves, but their primarily frontal polar location and the lockstep unison with the flash distinguishes them from the photoparoxysmal response. This sequence of spike–wave-like eye flutter is punctuated by a more obvious eyeblink in the middle third of photic stimulation. Note the bisynchronous “on” response at the beginning of photic stimulation as an apiculate phenomenon. Calibration signal 1 s, 50 µV.

 

Fig. 2-13. Eyeblinks. Sequential eyeblink artifacts are identifiable by their location at Fp1 and Fp2, their considerably lower amplitude at F3 and F4, and their response to eye opening (left asterisk). Note the bisynchronous downward potential in the frontal leads with an eyeblink (right asterisk), the potential due to movement of the positive end of the ocular dipole towards the frontal polar electrodes suddenly creating a positive field centered near the frontal polar electrodes which extends somewhat posteriorly(Brittenham, 1990). Movement of the eyelids across the eyeball may contribute to this potential (Fisch, 1999). At times, eyeblink artifact may combine with higher-frequency background activity to resemble spike–waves (center) but the virtual confinement to the Fp1,2 electrodes would make this interpretation most unlikely. Calibration signal 1 s, 50 µV.

 

Fig. 2-14. Prosthetic left eye. The slight, positive potential at Fp1 likely reflects the field of the right eyeblink. Calibration signal 1 s, 50 µV.

 

Fig. 2-15. Eye opening abolishes eye flutter. Eye opening abolishes not only alpha but also eye flutter. Calibration signal 1 s, 70 µV.

 

Fig. 2-16. Eye flutter abolished by holding eyes. Asking the patient to hold eyes closed helps distinguish artifact from rhythmic frontal delta activity. Calibration signal 1 s, 70 µV.

 

Fig. 2-17. Predominantly lateral eye movements on bipolar montage. These predominantly lateral eye movements are out of phase in derivations involving F7 and F8 electrodes, as an increase in positivity at one is associated with a decrease in positivity (more negative) in the other as recorded by these differential amplifiers. Calibration signal 1 s, 50 µV.

 

Fig. 2-18. Rapid lateral eye movements. Rapid lateral eye movements demonstrate the strict out-of-phase potentials between F7 and F8. Calibration signal 1 s, 50 µV.

 

Fig. 2-19. Slow lateral eye movements on a coronal montage. Lateral eye movements produce opposite potentials at F7 and F8; thus, a rightward eye movement would give increasing positivity at F8 and decreasing positivity at F7. As F7 and F8 are connected to inputs 1 and 2 of their respective derivations on this coronal montage, the opposite changes of polarity produce deflections in the same direction but summate in the F7–F8 derivation. These similar waveforms and the lack of associated disturbances in the background activity, such as excess theta, distinguish this phenomenon from temporal delta activity. Calibration signal 1 s, 100 µV.

 

Fig. 2-20. Slow lateral eye movements on a common average reference montage. The slow lateral eye movements are out of phase between hemispheres, including the frontal and central derivations. Again, the lack of regional background abnormalities distinguishes such waves from frontal delta activity. Calibration signal 1 s, 50 µV.

 

Fig. 2-21. Muscle artifact. Patient's age, 29 years. Awake. Eyes closed. Muscle activity produces very brief potentials. This muscle activity was abolished in the last half of this illustration when the patient was asked to open his mouth slightly. Calibration signal 1 s, 70 µV.

 

Fig. 2-22. Unilateral muscle artifact. Patient's age, 47 years. Awake. Eyes closed. Abolished by opening mouth slightly. Calibration signal 1 s, 70 µV.

 

Fig. 2-23. Periocular muscle artifact. Patient's age, 26 years. Awake. Eyes closed. The very brief potentials in the frontal polar regions (FP1, FP2) are produced by contraction of periocular muscles. Upward and downward ocular movements create positivity and negativity at FP1 and FP2, producing the upward pen deflections. Calibration signal 1 s, 50 µV.

 

Fig. 2-24. Glossokinetic artifact. Movement of the tongue, whose tip is electrically negative with respect to its base, may produce widely distributed, low-frequency intermittent potentials that may resemble bursts of rhythmic projected delta, i.e. “projected rhythms.” A burst of muscle potentials may precede such low-frequency waves, serving to differentiate glossokinetic potentials from “projected” activity. In the left segment, bursts of muscle potentials are accompanied by an equally long burst of diffuse rhythmic waves, both of which are characteristic of glossokinetic artifact. The first burst of glossokinetic artifact in the second segment is dominated by muscle artifact, whereas the second burst of glossokinetic artifacts is dominated by rhythmic waves. Calibration signal 1 s, 50 µV.

 

Fig. 2-25. Chewing gum. Patient's age, 24 years. Awake. The most common cause of periodic muscle bursts is gum chewing; these disappeared abruptly upon gum removal (4th second). Calibration signal 1 s, 100 µV.

 

Fig. 2-26. Sucking artifact. Patient's age, 10 months. The high-voltage, brief, diffuse potentials are the characteristic artifact associated with sucking, common in awake recordings at this age. Calibration signal 1 s, 150 µV.

 

Fig. 2-27. Bursts of muscle potentials. Patient's age, 9 years. Awake. Bursts of muscle potentials may be seen in association with Gilles de la Tourette syndrome (as in this case), chorea, and other movement disorders. Calibration signal 1 s, 100 µV.

 

Fig. 2-28. Muscle potentials. On this referential montage, all of the apiculate waves represent frontal or periocular muscle artifact because of the brevity of each potential (shorter than almost all abnormal spikes) and the lack of any aftercoming slow wave except that which likely represents an ocular movement (middle of illustration). Spike waves usually appear principally at F3,4, whereas these potentials are centered at FP1,2. Calibration signal 1 s, 70 µV.

 

Fig. 2-29. Tremor-related muscle artifact. The brevity of these potentials and their minimal spread only to T5,6 suggests that these regularly repetitive (4-5 Hz) potentials represent cephalic tremor. Calibration signal 1s, 70 µV.

 

Fig. 2-30. Glossokinetic artifact. In this and the following illustration, glossokinetic artifact resembles “projected” rhythms; this is particularly prominent while the patient is talking. Such artifact may be produced by asking the patient to say “lilt.” True projected activity is often associated with a diffuse theta burst, which is not present in these samples. Calibration signal 1 s, 50 µV.

 

Fig. 2-31. Lateral glossokinetic artifact on coronal montage. The delta-free recording in the first seconds of this segment is subsequently marred by the subject wagging his tongue laterally (asterisk). Again, the artifactual nature of this delta activity is indicated by the normal background activity and its stereotyped appearance in many derivations. Calibration signal 1 s, 50 µV.

 

Fig. 2-32. Glossokinetic artifact on referential montage. Diffuse delta activity unaccompanied by excess theta should raise suspicion of artifact. A clue here is the out-of-phase delta in each hemisphere produced by the subject wagging his tongue laterally. Cessation of that activity during the last 3 s eliminated the delta. Calibration signal 1 s, 50 µV.

 

Fig. 2-33. Tremor. The 4-Hz rhythmic waves with accompanying bursts of muscle potentials seen at F4 on the bipolar montage are even more evident in the left-sided derivations on the ear referential montage. The rhythmic 4-Hz activity in synchrony with the tremor could be mistaken for cerebrally originating activity. Calibration signal 1 s, 100 µV.

 

Fig. 2-34. Myokymia. Repetitive bursts of muscle artifact may represent periodic brief tonic facial contractions. Calibration signal 1 s, 70 µV.

 

Fig. 2-35. Palatal myoclonus. Muscle artifact at 100 to 200/min from palatal myoclonus produces potentials of metronomic regularity best displayed by A1, A2. Its rate (about 110/min) and morphology differ from those of the ECG. The lower section fully displays the myoclonus in the A1–A2 derivation, whereas it appears minimally in the ear reference montage (upper) at another point in recording. Calibration signal 1 s, 70 µV.

 

Fig. 2-36. Electrocardiogram (ECG) and referential montage. Patient's age, 71 years. Since electrical fields of the heart extend to the base of the skull, they may be detected by ear electrodes and therefore may be prominent on ear referential montages. The usual cardiac electrical axis produces an R wave, which is positive at A1 and negative at A2, accounting for the out-of-phase deflections in alternate hemisphere channels. An ECG monitor helps identify the large-amplitude premature contraction as cardiac and will more readily identify unusual-appearing cardiac artifacts due to arrhythmias or aberrant electrical conduction pathways. Calibration signal 1 s, 50 µV.

 

Fig. 2-37. ECG artifact and bipolar montage. Patient's age, 74 years. Electropositive apiculate lambda-like potentials from occipital leads (O1,2) may be produced by the R wave of the ECG in patients with short necks. An ECG monitor is most helpful. Calibration signal 1 s, 50 µV.

 

Fig. 2-38. Pulse artifact, smooth type. Rhythmic delta activity confined to a single electrode position (C3 here) likely represents pulse artifact. Although the position of an electrode on or near a scalp artery may produce this artifact, the latter may also represent high electrode impedance, as reflected in the sudden high-amplitude electropositive artifactual deflection at C3 in the middle of the tracing. Calibration signal 1 s, 50 µV.

 

Fig. 2-39. PLEDs-like pulse artifact. Pulse artifact may appear as periodic sharply contoured potentials that are time-locked to ECG as depicted by the M1–M2 derivation. Note the invariable 200-ms delay between the R wave and the pulse artifact peaks. Calibration signal 1 s, 50 µV.

 

Fig. 2-40. Metals artifact. High-voltage, very brief, regional or moderately widespread artifact may occur when metals such as dental fillings rub against each other during mouth movements, as illustrated here. Such activity is usually more abrupt, higher in voltage, and briefer than muscle artifact. Thus, in the example on the right, the first burst of potentials (left asterisk) is muscular with glossokinetic artifact, whereas the last burst (right asterisk) is a metal artifact. A similar artifact occurs during electrocorticograms from “snaps” rubbing together. Calibration signal 1 s, 50 µV.

 

Fig. 2-41. Attenuation from subgaleal fluid. Subgaleal edema or blood reduces amplitudes of background rhythms, particularly in bipolar recordings. Since comatose patients are usually nursed supine, such edema collects posteriorly. Possible mechanisms include a “salt bridge” an increased distance of electrodes from generators of cerebral potentials. It is quite possible that the arrhythmic delta activity, expressed principally anteriorly, is diffuse. Calibration signal 1 s, 50 µV.

 

Fig. 2-42. Subgaleal fluid and coronal montage. Same patient as in Fig. 2-41 with posteriorly situated subgaleal fluid. This fluid likely caused the attenuated potentials posteriorly. Note the persistent diffuse delta activity elsewhere and the blunted vertex waves. Calibration signal 1 s, 50 µV.

 

Fig. 2-43. Regional attenuation from subgaleal edema. Scalp edema will attenuate EEG amplitude, as occurs here in the parietal–occipital (P4, O2) regions on this bipolar montage. The referential recording shows that this attenuation also includes the left occipital (O1) region, as indicated by the considerably lower-voltage potentials at P4, O1, and O2 compared with more anterior positions. The attenuation at O1, displayed on the referential montage, is not apparent on the bipolar montage, since P3 likely contributes most of the potentials to the P3–O1 derivation. Both members of the derivation pair must be attenuated or in complete synchrony for attenuation to occur, as seen at the P4–O2 derivation. A1 and A2 record background potentials less than parasagittal leads. Calibration signal 1 s, 70 µV.

 

Fig. 2-44. Muscle artifact obscures EEG. Sustained muscle activity obscures all but M1–M2 derivation, leaving the reader guessing from which side the spikes originate. High frequency filter (HFF) 70 Hz. Calibration signal 1 s, 50 µV.

 

Fig. 2-45. Filtered EEG. Same illustration as previous. Diminishing the HFF to 15 Hz reveals the T3–M1 location of these broad spikes. Calibration signal 1 s, 50 µV.

 

Fig. 2-46. Rhythmic artifact of physiotherapy (RAP) in the ICU. The saw-toothed artifact in the 4th channel is also captured on the EKG (bottom channel). Calibration signal 1 s, 150 µV.

 

Fig. 2-47. Suppressed EEG with eye movements. Enucleated left eye. Calibration signal 1 s, 150 µV.

 

Fig. 2-48. Jiggle artifact. Patient's age, 25 years. The patient met the clinical criteria for brain death following an earlier cardiac arrest. The EEG was completely suppressed except for occasional rhythmic complexes that occurred because of mechanical bed movements (note the simultaneous involvement of the EKG channel). Also note that the amplitude of the waves lessens progressively, as would be expected with a mechanical perturbation. This initially gave the erroneous impression of a generalized burst-suppression pattern. Calibration signal 1 s, 30 µV.

Bibliography

Brittenham DM. Artifacts. Activities not arising from the brain. In: Daly DD, Pedley TA, eds. Current Practice of Clinical Electroencephalography. New York: Raven Press: 1990;85–106.

Fisch BJ. Fisch and Spehlmann's EEG Primer. Basic Principles of Digital and Analog EEG. 3rd ed. Amsterdam: Elsevier Science, 1999;107–122.