Abstract
Magnetoencephalography (MEG) is a critical non-invasive functional imaging technique for pre-surgical mapping. However, movement-related MEG mapping of the primary motor cortex (M1) is often challenging in patients with brain lesions and sensorimotor dysfunction. Traditional methods require a large number of trials to achieve an adequate signal-to-noise ratio, which is difficult for patients with motor impairments. This paper presents a novel Electromyography (EMG)-projected MEG source imaging technique that localizes M1 using approximately one minute of recording without trial averaging. By projecting M1 activity toward the skin EMG signal, we achieved high-resolution source images across various frequency bands. In a study of 13 healthy participants and two presurgical patients, the technique demonstrated 100% accuracy in the delta and theta bands. These results suggest that EMG-projected MEG is a highly accurate, feasible approach for M1 mapping, providing new insights into brain-muscle coupling at frequencies above physical movement.
Introduction
Before neurosurgical intervention for brain tumors or lesions, establishing a precise "road map" of the primary motor cortex (M1) is essential to avoid permanent neurological deficits. Currently, functional MRI (fMRI) and traditional MEG are the gold standards for this mapping. However, both rely heavily on "trial averaging." To extract a clear signal from background brain noise, patients must typically perform hundreds of repetitive, rhythmic movements (e.g., finger tapping).
For patients suffering from brain tumors, strokes, or tremors, this requirement is a significant barrier. Their movements are often slow, weak, or highly irregular, resulting in "noisy" data that fails to localize the motor strip accurately. Furthermore, the scientific community has lacked a full understanding of Corticomuscular Coupling (CMC)—the communication between the brain and muscles—at frequencies higher than the physical movement itself. This study introduces a faster, more robust imaging solution to bridge these clinical and scientific gaps.
Methodology: EMG-Projected MEG Source Imaging
The core innovation of this research is the EMG-projected MEG source imaging technique. Unlike traditional methods that average data in the time domain, this approach utilizes the relationship between brain activity and muscle signals to isolate the relevant neural circuits.
Data Collection and Processing
Participants performed self-paced index finger flexions (at approximately 1 Hz) for roughly one minute. During this time, MEG (brain) and EMG (muscle) signals were recorded simultaneously. The researchers then applied an enhanced version of the Fast-VESTAL algorithm. Instead of looking for a rhythmic pattern, the algorithm mathematically projected the MEG sensor data into a parameter space defined by the EMG activity.
This projection acts as a high-performance filter, amplifying brain signals specifically involved in muscle communication while suppressing unrelated background noise and artifacts. The team analyzed five distinct frequency bands to determine which provided the most reliable localization:
- Delta (1–4 Hz)
- Theta (4–7 Hz)
- Alpha (8–12 Hz)
- Beta (15–30 Hz)
- Gamma (30–90 Hz)
Experimental Results
The study evaluated the method using 26 datasets from 13 healthy controls and two clinical cases involving presurgical patients with sensorimotor dysfunction.
Localization Accuracy
In healthy subjects, the method achieved remarkable precision:
- Delta and Theta Bands: 100% accuracy in localizing M1.
- Beta Band: 76.9% accuracy.
- Alpha and Gamma Bands: Poor performance (34.6% and 0.0%, respectively). The failure of the alpha band is likely due to "mu rhythm" suppression during active movement, while the gamma band's failure may stem from the lower signal-to-noise ratio of non-invasive MEG at high frequencies.
Clinical Case Studies
The technique proved its clinical utility in two challenging scenarios:
- Patient 1: A patient with metastatic melanoma near the motor strip. The EMG-projected MEG map allowed surgeons to perform a total resection of the tumor while successfully avoiding the motor strip, resulting in no post-surgical neurological deficits.
- Patient 2: A patient with a perinatal stroke and highly irregular movement patterns. Previous attempts at 3-Tesla fMRI mapping had failed due to the patient's inability to maintain a rhythm. However, the theta-band EMG-projected MEG accurately pinpointed the motor cortex despite the chaotic EMG signal.
Discussion: The "Theta" Sweet Spot and Efficiency
A major finding of this study is that the Theta band (4-7 Hz) provides 100% localization accuracy. Because this frequency is higher than the physical movement frequency (~1 Hz), it suggests that theta-band coupling is driven by the brain's internal motor control mechanisms rather than the physical rhythm of the movement itself. This explains why the method remains effective even when a patient's movements are irregular.
Furthermore, the efficiency of this method is a significant improvement over current standards. While conventional MEG paradigms require 10–20 minutes of recording, this technique generated accurate maps using only the first 30 seconds of data (96.2% accuracy) and remained relatively successful with as little as 5 seconds of data (73.1% accuracy).
Conclusion
EMG-projected MEG source imaging represents a significant advancement in presurgical functional mapping. By reducing the task duration to approximately one minute and eliminating the need for rhythmic movement, it makes high-resolution M1 mapping accessible to patients with severe motor impairments. While further studies with larger clinical cohorts are necessary, this approach offers a promising path toward faster, more reliable neurosurgical planning and a deeper understanding of brain-muscle communication.
Summary updated