*Article* **Navigated TMS in the ICU: Introducing Motor Mapping to the Critical Care Setting**

**Severin Schramm 1, \*, Alexander F. Haddad 2 , Lawrence Chyall 2 , Sandro M. Krieg 1,3 , Nico Sollmann 3,4 and Phiroz E. Tarapore 2**


Received: 23 November 2020; Accepted: 16 December 2020; Published: 18 December 2020

**Abstract:** Navigated transcranial magnetic stimulation (nTMS) is a modality for noninvasive cortical mapping. Specifically, nTMS motor mapping is an objective measure of motor function, offering quantitative diagnostic information regardless of subject cooperation or consciousness. Thus far, it has mostly been restricted to the outpatient setting. This study evaluates the feasibility of nTMS motor mapping in the intensive care unit (ICU) setting and solves the challenges encountered in this special environment. We compared neuronavigation based on computed tomography (CT) and magnetic resonance imaging (MRI). We performed motor mappings in neurocritical patients under varying conditions (e.g., sedation or hemicraniectomy). Furthermore, we identified ways of minimizing electromyography (EMG) noise in the interference-rich ICU environment. Motor mapping was performed in 21 patients (six females, median age: 69 years). In 18 patients, motor evoked potentials (MEPs) were obtained. In three patients, MEPs could not be evoked. No adverse reactions occurred. We found CT to offer a comparable neuronavigation to MRI (CT maximum e-field <sup>52</sup> <sup>±</sup> 14 V/<sup>m</sup> vs. MRI maximum e-field 52 <sup>±</sup> 11 V/m; *p* = 0.6574). We detailed EMG noise reduction methods and found that propofol sedation of up to 80 mcg/kg/h did not inhibit MEPs. Yet, nTMS equipment interfered with exposed pulse oximetry. nTMS motor mapping application and use was illustrated in three clinical cases. In conclusion, we present an approach for the safe and reliable use of nTMS motor mapping in the ICU setting and outline possible benefits. Our findings support further studies regarding the clinical value of nTMS in critical care settings.

**Keywords:** nTMS; brain stimulation; intensive care; motor mapping; ICU; neurocritical care; neuromonitoring; functional mapping; motor evoked potentials

#### **1. Introduction**

Navigated transcranial magnetic stimulation (nTMS) is an established technique for non-invasive brain stimulation [1,2]. At present, the technique is commonly applied in pre-operative mappings of patients undergoing neurosurgical procedures. In this study, we investigate whether nTMS may also play a role in the diagnosis and management of acute neurological injury in the inpatient setting. Because it is image-guided (and therefore highly accurate), it can provide detailed information about the

level of intactness in specific neuromuscular pathways (useful in spinal cord injury). Unlike transcranial electrical motor-evoked potentials, it is non-invasive and painless, and can therefore be applied at bedside without anesthesia. Furthermore, it does not require patient participation, and may thus be used in patients who are sedated, or who are awake but unable to participate in detailed neurological examination. For these reasons, nTMS, if it can be successfully implemented in the neurocritical care setting, would offer a valuable new method for quantifiable neurological examination in patients with suspected or known neurological compromise.

TMS involves a strong magnetic field pulse that penetrates the skull and causes depolarization of underlying neuronal tissue via the principle of electromagnetic induction. When using a figure-of-eight coil it is highly focal and entirely noninvasive [2,3]. By combining the capacity for focal cortical stimulation with image-based, frameless stereotactic neuronavigation, nTMS allows for highly accurate, navigated interrogation of cortical function [1,4]. The broad applicability of this basic principle has created applications for nTMS in both the basic science and clinical realms. The technique has been utilized across neuroscience disciplines such as psychiatry, in physical medicine and rehabilitation, and neurosurgery [5–7].

As mentioned earlier, a common application of nTMS is the mapping of the primary motor cortical system (MCS). Here, stimulation of the cortical motor area leads to peripheral muscle activation (motor evoked potentials [MEPs]), which is recorded via electromyography (EMG) [1]. This process allows for attribution of a specific muscle response (i.e., abductor pollicis brevis or tibialis anterior muscles) to its respective cortical location, thereby generating an individualized functional map of the MCS [8,9]. Herein, the navigational aspect adds critical capabilities to the underlying TMS technology. Compared to non-navigated TMS of the motor cortex, where the stimulation site is usually determined in relation to anatomical landmarks [10], nTMS has been shown to be superior in evoking replicable and high-amplitude MEPs [11]. Importantly, nTMS allows optimal orientation of the electric field in relation to individual gyral and skull anatomy, a critical aspect in ensuring ideal MEP generation [12]. Furthermore, for any form of longitudinal monitoring, precise revisiting of previously identified sites is vital and, therefore, necessitates accurate navigation. The accuracy of nTMS motor mapping is superior to other noninvasive techniques such as functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG), and is highly comparable in accuracy to intraoperative direct cortical stimulation (DCS), the current gold-standard method for functional mapping [1,13].

In the clinical context, mapping of the MCS is often performed in preparation for brain tumor surgery, allowing for better presurgical planning by identifying eloquent motor regions and their relationship to the lesion [5,9,14,15]. While increasingly common in the outpatient setting, nTMS motor mapping has thus far remained underutilized in the acute inpatient setting. The use of other neurophysiological monitoring modalities in the ICU, most notably electroencephalography (EEG), has opened up a range of diagnostic innovations in recent years [16–20]. Similarly, nTMS could be a valuable tool for the inpatient setting. It can provide a highly accurate, low-cost, and relatively quick assessment of the functionality of the MCS without dedicated subject participation. It would thus be the ideal modality for evaluation of the corticospinal integrity and excitability in patients with altered mental status or coma. Alternatively, because it is painless, nTMS may also be used to assess the MCS in patients who are awake but unable or unwilling to participate in neurological examination.

However, to date, only few studies have addressed the application of nTMS motor diagnostics in the ICU context. While TMS/EEG combination setups have been used in exploring disorders of consciousness [20], nTMS motor mappings in the intensive care framework are at this time not reported in the literature. The reason for the current lack of exploration is likely the plethora of challenges faced in translating established protocols and workflows to the unique setting and patient population found in the ICU (Table 1). This study seeks to explore the feasibility of nTMS motor mappings in the intensive care framework, based on established protocol and safety guidelines [8,21]. By identifying the primary challenges, working through solutions, and sharing our results, we hope to encourage and enable further studies into the value of nTMS motor mapping in the neurocritical care setting.


**Table 1.** Translation of navigated transcranial magnetic stimulation from outpatient setting to critical care setting.

This table gives an overview over challenges in translating navigated transcranial magnetic stimulation (nTMS) motor mappings from the outpatient to the critical care setting. Other abbreviations: magnetic resonance imaging (MRI), computed tomography (CT), electromyography (EMG), motor evoked potential (MEP), intracranial pressure (ICP).

#### **2. Materials and Methods**

#### *2.1. Study Setup and Patient Inclusion*

This study was approved by the UCSF Institutional Review Board (16-20684, 1/9/2020). All research was conducted according to the Declaration of Helsinki. Informed consent was acquired from conscious patients or next of kin when known.

We conducted nTMS motor mappings on patients treated in the neurological ICU of Zuckerberg San Francisco General Hospital. Patients are referred to this unit when neurological damage necessitated continuous monitoring of vital parameters, close interval neurochecks, mechanical ventilation, vasoactive infusions, and other typical ICU indications. Available and commonly applied monitoring includes blood pressure (both invasive and noninvasive), pulse oximetry, electrocardiography (ECG), intracranial pressure (ICP), EEG, and interval neurochecks [22].

Inclusion criteria were defined as treatment in the ICU, confirmed presence of brain or spinal lesion and suspicion or presence of a motor deficit due to neurological damage. Exclusion criteria were denial of consent by patient or next of kin, uncontrolled epilepsy, presence of fixed nTMS-interacting devices close to or within the skull (e.g., cochlear implant, deep brain stimulation), acute isolation due to infectious disease, acute instability of vital parameters, and end-of-life care. The procedures were explained to the patient (if conscious) and/or next of kin (if known).

The nTMS system was deployed into the patient's room on the ICU, and motor mapping was performed. After the examination, clinical factors such as ICP and sedation were documented if applicable. Afterwards, the data were analyzed by experienced nTMS users (SS, PT) and a report on the procedure and results was written for internal use.

#### *2.2. Image Acquisition and Processing*

All CT imaging in this study was acquired with a Somatom 64-slice CT scanner (Siemens AG, Erlangen, Germany), all MRI imaging with a Magnetom Skyra 3T MRI scanner (Siemens AG, Erlangen, Germany). In order to compare the navigational accuracy of CT and MRI, we exported image stacks of both CT and MRI from 4 patients that had received both types of scans. For CT, we exported 4 scans of 2 mm slice thickness (3 sagittal, 1 axial). For MRI, we exported 4 T1-weighted scans with slice thickness ranging from 0.75 mm to 5 mm (3 sagittal, 1 axial).

To make CT scans viable for use in the nTMS system, we performed a series of preprocessing steps (removal of non-patient structures such as e.g., headrest, proper windowing, intensity rescaling) using the Aliza Medical Imaging and DICOM Viewer (Aliza 1.98.12, Copyright 2014–2020 Aliza Medical Imaging, Bonn, Germany), the details of which are available as Supplementary Material (Document S1). The exported DICOM files were loaded into the nTMS system (Nexstim eXimia NBS system, version 4.0; Nexstim Plc., Helsinki, Finland) and used to create head models (Figure 1A,B).

**Figure 1.** Head model based on computed tomography (CT). By using appropriate preprocessing, CT can serve as a viable basis for head model reconstruction in the nTMS system (**A**,**B**). Important anatomical structures such as the precentral gyrus are easy to identify (**A**,**B**, white arrows). For image testing, 12 targets were placed around the central sulcus on each hemisphere with two additional targets at the midline vertex (26 targets total per brain, **C**).
