*2.4. Motor Mapping Protocol*

For patient motor mappings, we followed the existing guidelines [8]. We employed an e-fieldnavigated TMS system (Nexstim eXimia NBS system, version 5.0; Nexstim Plc., Helsinki, Finland). The system provides an integration of all necessary subcomponents (navigation, figure-of-eight stimulation coil, EMG). For EMG recordings, surface electrodes were used (Neuroline 700; Ambu, Ballerup, Denmark). The nTMS system was positioned next to the patient's bed. The patient's skin at electrode placement spots was cleaned with alcohol swabs. EMG electrodes were applied to all muscles of interest, chosen according to the lesion location and predicted or known motor deficits.

EMG noise optimization was performed by finding optimal electrode placement spots, avoiding crossings of cables and eliminating all non-essential sources of electromagnetic noise [23]. Recorded muscles for the upper extremity were deltoid muscle, biceps brachii muscle, flexor carpi radialis muscle, and either abductor pollicis brevis muscle or adductor digiti minimi muscle, depending on accessibility. For lower extremities, recorded muscles were quadriceps femoris muscle, tibialis anterior muscle, and gastrocnemius muscle. MEPs were considered valid according to

amplitude (above 50 µV), latency (10–35 ms for upper extremity musculature, 30–60 ms for lower limb musculature) [24], and plausible morphology.

Prepared imaging was loaded into the system and a head model was created. After successful co-registration, the head model was peeled to an individual depth allowing for visual identification of the precentral gyrus. During the entire mapping, the coil was positioned tangentially to local skull surface, aided by the neuronavigation system [3]. E-field orientation was kept at 90 degrees to local gyrus orientation [1]. A rough mapping was performed to give primary information on muscle representations of the upper extremity and possible motor hotspots. After identification of a given motor hotspot, said point was used for resting motor threshold (rMT) determination, defined as the stimulation intensity that results in a valid MEP in 50% of cases [8]. rMT determination was performed via a system-integrated procedure based on the maximum likelihood algorithm [25,26].

After determination of the rMT, the full extent of the MCS was mapped with 105% rMT. Then, lower extremity musculature was mapped with an intensity starting at 130% rMT. The distance between individual stimuli was roughly 5 mm. If no MEPs could be elicited, the stimulation intensity was raised stepwise until either MEPs were present, the patient reported discomfort, or no further raise was possible. For patients with prior hemicraniectomy (consisting of the temporary removal of a large fronto-temporo-parietal skull flap to allow for decompression of brain tissue) intensity was capped at 75% of maximum stimulator output due to safety considerations. In post-hoc analysis of motor mappings, the amplitude of EMG noise after signal optimization was measured and noted down for each mapped muscle for later descriptive statistical analysis.

The primary maxim during motor mapping was to avoid disrupting the nursing routine or clinical necessities. Extraordinary effort was made to explain the testing procedure to the bedside nurses and to ensure that testing would not disrupt their patient care responsibilities. Emphasis was made on the importance of open communication between the nursing team and the nTMS team. During testing, no active monitoring was removed from the patient and care was taken to minimize patient mobilization. If any other procedures had to be performed on the patient (e.g., neurological examination, suction, repositioning, and toileting), the examination was paused or aborted. After the examination, all nTMS equipment was removed from the patient, the room was returned to its original layout, and the system was cleaned and brought back to its designated storage place.
