MRI artifacts: Head position matters
A pilot study shows that head position during brain and inner ear magnetic resonance imaging can create artifacts that impact interpretation and patient comfort.
Dark, diamond-shaped spots in inner ear magnetic resonance imaging (MRI) scans are not always signs of pathological changes; they can simply result from how a person is positioned in the scanner. A pilot study at Austria’s Karl Landsteiner University (KL Krems) shows that characteristic flow void artifacts in the inner ear become markedly more pronounced when the head is tilted back and less when the chin is tilted down. With the head tilted back, some volunteers also reported dizziness. The work supports the idea that strong magnetic fields can drive inner ear fluid motion. It further suggests that head position should be considered when interpreting brain and inner ear MRI scans and when trying to keep people comfortable in high-field scanners.
High-field MRI at 3 Tesla and above has become standard in neuroradiology. At these field strengths, the static magnetic field can interact with tiny electric currents in the inner ear fluids. The resulting so-called Lorentz forces are known to trigger nystagmus (uncontrolled, rhythmic eye movement) and vertigo in people with a normal balance system (vestibule) of the inner ear. At the same time, MRI techniques used to image the inner ear’s labyrinth are very sensitive to even slow fluid motion. Earlier observations from Krems had linked these effects to small, sharply outlined low-signal areas in the vestibule that do not match any anatomical structure. The new study set out to test in a controlled way whether these hypointensities (labelled flow voids) really behave like flow-related artifacts and whether they change systematically with head pitch.
A team jointly led by Prof. Dr. Domagoj Javor, Head of the Institute of Diagnostic and Interventional Radiology and Dr. Béla Büki from the Division of Otorhinolaryngology both at University Hospital Krems, a teaching and research center of KL Krems, examined 20 healthy adults without known vestibular disease in a 3T scanner. The number was kept deliberately low; the authors describe the project as a proof-of-principle rather than a definitive trial. Each volunteer underwent two high-resolution inner ear scans with a so-called T2-weighted SPACE sequence: once with the chin tilted towards the chest (head flexed) and once with the head tilted back (extended). Images were reconstructed in the plane of the horizontal semicircular canal. Two experienced colleagues, working independently and blinded to each other’s results, measured what proportion of the vestibule was occupied by the hypointense flow voids.
Head back, artifact up
The pattern was clear: With the head tilted back, the low-signal area in the vestibule increased on both sides by around 15 percentage points compared with the chin-down position. In the same position, three of the 20 volunteers about 15% reported mild vertigo. None did so when their head was flexed. The findings show that these small vestibular dark spots are not fixed anatomical features but change with head position in the magnetic field, says Prof. Dr. Javor. That is exactly what one expects from a benign, position-dependent artifact rather than from inner ear pathology.
From a physics perspective, the observations fit current models of magnetic vestibular stimulation. When the head is tilted back, the main direction of ionic currents in the inner ear becomes more perpendicular to the scanner’s magnetic field. This increases a physical force acting on ions, the Lorentz force, and drives stronger endolymph flow in parts of the inner ear, notably the utricle and the lateral semicircular canal. Such motion can both deflect specific gelatinous features (cupulae), contributing to vertigo, and disturb the MRI signal enough to create a more prominent flow void.
Pragmatic workaround
For clinical work, the authors suggest a pragmatic approach. If a suspicious vestibular hypointensity appears on a T2 spin-echo sequence, checking whether it changes with head position or across different sequence types can help distinguish artifact from true lesion. Gradient-echo sequences, which are less sensitive to slow fluid motion, may serve as a useful comparison. Documenting head pitch on sagittal localizers and reconstructing images in the plane of the horizontal semicircular canal can also make left right comparisons more reliable. Radiologists should be aware that this characteristic, diamond-shaped hypointensity in the vestibule tends to increase with head extension and diminish with flexion, says Dr Büki. Seen in isolation, it can mimic a focal lesion but in many cases it simply reflects fluid moving in a strong magnetic field.
At the same time, the group is clear about the limits of their data. The study was carried out at a single center, on one 3T scanner, with one specific T2 SPACE protocol and only 20 healthy volunteers. The range of head positions was restricted by the design of the head coil, and neither eye movements nor inner ear fluid dynamics were measured directly. The authors stress that their work should be read as an internally consistent pilot, not as a new standard of care. Larger series at different field strengths and, crucially, data from patients with vestibular disorders will be needed.
Category: Education
