What Are Motion Artifacts?

Motion artifacts are spurious signals or image degradations that arise when a subject moves during data acquisition, corrupting the measured output with displacement or velocity information that was not intended to be captured. They appear across a wide range of sensing modalities, from magnetic resonance imaging and ultrasound to wearable biopotential electrodes and optical sensors. The fundamental cause is consistent: a mismatch between the spatial or temporal assumptions built into the measurement system and the actual movement of the object being observed.

The consequences range from blurred or ghosted images in medical scanners to noisy waveforms in ambulatory health monitors. Because motion artifacts can mimic or mask genuine physiological and structural signals, their presence is a persistent engineering challenge in diagnostic and monitoring systems.

Artifacts in Medical Imaging

In MRI, patient motion during a scan disrupts the phase-encoding process that maps spatial information from the frequency domain (k-space) into the final image. Rigid-body motion, such as head movement, typically produces blurring and ringing patterns around edges, while periodic motion from breathing or cardiac pulsation generates ghosting replicas displaced along the phase-encode direction. A comprehensive review of MRI motion artifact mechanisms published in the European Radiology journal describes how second-order effects, including motion-induced magnetic field inhomogeneities and spin-history variations, produce artifacts that simple rigid correction cannot address. Prospective gating, where acquisition is triggered to a physiological cycle, and retrospective correction, where raw data are post-processed after acquisition, are the two broad mitigation classes. Deep learning approaches using generative adversarial networks and transformer architectures have more recently been applied to reconstruct clean images from corrupted k-space data.

Artifacts in Wearable and Biopotential Sensors

In wearable monitoring devices, motion artifacts originate at the skin-electrode interface, where physical deformation of the contact changes the electrode-electrolyte half-cell potential. For ECG sensors, the artifact spectrum overlaps substantially with the cardiac signal band (0.05 to 150 Hz), making separation by simple frequency filtering unreliable. Photoplethysmography (PPG) sensors, which measure blood volume changes optically, are similarly vulnerable: body movement introduces intensity fluctuations in the detected light that can be orders of magnitude larger than the perfusion signal. Research on adaptive motion artifact reduction in wearable ECG systems has shown that using a co-located accelerometer or impedance signal as a reference input to an adaptive filter substantially improves signal-to-noise ratio compared to single-channel processing. More recent multi-axis inertial sensor fusion methods correlate six-degree-of-freedom motion data with corrupted signal segments to improve artifact rejection further.

Detection and Mitigation Strategies

Motion artifact mitigation follows three general strategies: prevention, retrospective correction, and signal separation. Prevention relies on hardware design choices such as dry electrode materials that reduce interface impedance, motion-tolerant sensor geometries, and respiratory-synchronized acquisition sequences in imaging systems. Retrospective correction algorithms use signal processing techniques ranging from adaptive filters and independent component analysis to machine learning classifiers that identify and subtract artifact components. Signal separation, exemplified by blind source separation methods, attempts to disentangle artifact contributions from true physiological signals when a reference motion channel is unavailable.

A published study on motion artifact techniques for wearable EEG and PPG sensors in Frontiers in Electronics categorizes the dominant removal approaches by sensor type and activity level, noting that adaptive methods outperform static filters under varying movement conditions.

Applications

Motion artifact analysis and reduction has applications in a range of fields, including:

• Cardiac monitoring during exercise and ambulatory ECG recording
• Fetal MRI and pediatric brain imaging where sedation is not possible
• Intraoperative imaging where breathing motion is unavoidable
• Sports performance monitoring with wrist-worn optical sensors
• Rehabilitation tracking using body-worn accelerometers and biopotential sensors