Repetitive single-shot diffusion weighted (DW) imaging of the lower leg was reported to reveal spontaneous signal voids in different muscle groups with a random pattern in temporal and spatial domain ¹. Most probably incoherent mechanical activities of muscle fibers are causing this phenomenon, but the underlying physiological processes are still unknown. This study is dedicated to analyze whether the signal voids are related to electrical neuromuscular activity. Surface electromyography (EMG) recorded the electrical activity of muscles during the MR examinations in order to assess potential relations of signal voids in DW images of the right calf with coincident EMG signals.MR acquisition: Three healthy volunteers (age: 35 ± 14, BMI: 25.4 kg/m² ± 3.4 kg/m²) were examined with at least 5 min rest before the measurement was starting. The DW images were acquired with a stimulated echo DW EPI with: matrix size: 64 x 64, FoV = 200 x 200 mm², TE = 31 ms, TR = 500 ms, BW = 2004 Hz/px, TM = 145 ms, 500 repetitions and SPAIR (Spectral Attenuated Inversion Recovery). All measurements were performed on a 3 T MR Scanner (Magnetom Skyra, Siemens Healtcare, Erlangen, Germany). Transverse DW images were acquired at the position with the maximum diameter of the calf. The signal voids were detected and segmented by an automatic evaluation framework ². EMG acquisition: The EMG signal was acquired with an eight channel BrainAmp ExG MR (Brain Products GmbH, Gilching, Germany) with a sampling rate of f_sampling = 5 kHz and the BrainVision Recoder software (Brain Products GmbH, Gilching, Germany). Due to the expected high induction current from MR gradient switching, electrodes with a current limiting resistor with R = 15 kΩ were utilized to ensure patient safety during the MR measurement protocol. An inter-electrode distance of 2 cm was utilized ³. The schematic electrode placement is illustrated in Fig. 1. The EMG sampling clock was synchronized with the MR scanner gradient clock with the Brain Vision SyncBox utilities. The artifact correction was performed with the BrainVision Analyzer 2.0 software (Brain Products GmbH, Gilching, Germany), while the MR gradient induced artifacts were corrected with a sliding average window (size: 21) and the cardioballistic artifact template related to blood flow pulsation were estimated from the first 20 seconds of the measurement without active MR measurement protocol ^4,5. Additionally, a notch filter with rejection-band from 45-55 Hz and a pass-band filter with a lower cut-off frequency f_low = 0.5 Hz and an upper cut-off frequency f_upper = 250 Hz were applied to reduce environmental and high frequency distortions. An exemplary result of the EMG artifact correction is illustrated in Fig. 2. The EMG signal was measured at electrode #4. The maximum allowed signal amplitude of the measured EMG signal (±16.384 mV) was not exceeded and thus the front-end of the amplifier was not in saturation.Three different kinds of signal voids can be identified: 1) spontaneous often perceptible motion of the volunteer, 2) fasciculation related motion of muscle regions, and 3) signal voids without measureable electrical activity. In Fig. 3, the m. tibialis posterior shows an extended signal void. Due to the fact that the whole muscle area is affected and a measureable EMG event is recognizable, a reflectory and nervally coordinated motion of the volunteer is assumed. In Fig. 4, a signal void in the diffusion-weighted image of the m. soleus is illustrated. The EMG event shows a fasciculation activity measured only by electrode #2 just before the MR acquisition starts. Fig. 5 illustrates one exemplary signal void without measureable EMG event.Based on the results from simultaneous MRI and EMG measurements the signal voids can be classified in several groups: It is shown that only a part of the signal voids are related to neuromuscular electrical activity. Other spontaneous muscular activities seem to be not associated with electrically measurable events.