Portal de Periódicos da CAPES

Carta Acesso aberto Produção Nacional Revisado por pares

BIBTEX RIS

Robotic–electronic platform for autonomous and accurate transcranial magnetic stimulation targeting

2024; Elsevier BV; Volume: 17; Issue: 2 Linguagem: Inglês

10.1016/j.brs.2024.03.022

1935-861X

Renan H. Matsuda, Victor H. Souza, Thais C. Marchetti, Ana M. Soto, Olli‐Pekka Kahilakoski, Andrey Zhdanov, Victor Hugo Malheiro, Mikael Laine, Mikko Nyrhinen, Heikki Sinisalo, Dubravko Kičić, Pantelis Lioumis, Risto J. Ilmoniemi, Oswaldo Baffa,

EEG and Brain-Computer Interfaces

Dear Editor, to improve the safety and efficacy of non-invasive brain stimulation techniques, we need to embrace automation and precise targeting of cortical structures. Multi-locus transcranial magnetic stimulation (TMS) enables the stimulation of nearby cortical regions electronically, without physically moving the coil set [1Koponen L.M. Nieminen J.O. Ilmoniemi R.J. Multi-locus transcranial magnetic stimulation—theory and implementation.Brain Stimul. 2018; 11: 849-855https://doi.org/10.1016/j.brs.2018.03.014Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar, 2Nieminen J.O. Sinisalo H. Souza V.H. Malmi M. Yuryev M. Tervo A.E. et al.Multi-locus transcranial magnetic stimulation system for electronically targeted brain stimulation.Brain Stimul. 2022; 15: 116-124https://doi.org/10.1016/j.brs.2021.11.014Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar, 3Souza V.H. Nieminen J.O. Tugin S. Koponen L.M. Baffa O. Ilmoniemi R.J. TMS with fast and accurate electronic control: measuring the orientation sensitivity of corticomotor pathways.Brain Stimul. 2022; 15: 306-315https://doi.org/10.1016/j.brs.2022.01.009Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar]. This technology opens the possibility to engage with local cortical networks at millisecond and millimeter scales and to create automated closed-loop mapping protocols [[4]Tervo A.E. Metsomaa J. Nieminen J.O. Sarvas J. Ilmoniemi R.J. Automated search of stimulation targets with closed-loop transcranial magnetic stimulation.Neuroimage. 2020; 117082https://doi.org/10.1016/j.neuroimage.2020.117082Crossref Scopus (26) Google Scholar,[5]Rösch J. Emanuel Vetter D. Baldassarre A. Souza V.H. Lioumis P. Roine T. et al.Individualized treatment of motor stroke: a perspective on open-loop, closed-loop and adaptive closed-loop brain state-dependent TMS.Clin Neurophysiol. 2023; https://doi.org/10.1016/j.clinph.2023.10.004Crossref Scopus (2) Google Scholar]. However, existing mTMS coil sets have two major issues: a limited range for electronic targeting (30-mm diameter region) and heavy construction (approximately 5 kg for a 5-coil set), mainly due to cabling. Therefore, the manual placement of the coil set on the scalp is slow and physically demanding, requiring highly trained personnel to manipulate the coil sets. Collaborative robots improve the reproducibility and accuracy of TMS coil placements [[6]Goetz S.M. Kozyrkov I.C. Luber B. Lisanby S.H. Murphy D.L.K. Grill W.M. et al.Accuracy of robotic coil positioning during transcranial magnetic stimulation.J Neural Eng. 2019; 16054003https://doi.org/10.1088/1741-2552/ab2953Crossref Scopus (18) Google Scholar,[7]Harquel S. Bacle T. Beynel L. Marendaz C. Chauvin A. David O. Mapping dynamical properties of cortical microcircuits using robotized TMS and EEG: towards functional cytoarchitectonics.Neuroimage. 2016; 135: 115-124https://doi.org/10.1016/j.neuroimage.2016.05.009Crossref PubMed Scopus (32) Google Scholar]; they can compensate automatically for patients' head movements and can be flexibly programmed to be guided by suitable algorithms. Yet, traditional robotized TMS systems can shift the stimulation focus only physically and are limited by robot velocities that are safe for human applications to avoid harmful collisions (around 0.2 m/s) [[8]Kantelhardt S.R. Fadini T. Finke M. Kallenberg K. Siemerkus J. Bockermann V. et al.Robot-assisted image-guided transcranial magnetic stimulation for somatotopic mapping of the motor cortex: a clinical pilot study.Acta Neurochir. 2010; 152: 333-343https://doi.org/10.1007/s00701-009-0565-1Crossref PubMed Scopus (46) Google Scholar]. Furthermore, commercial robotic TMS solutions rely on closed-source platforms associated with a specific robotic arm, which can be costly and difficult to implement without the necessary flexibility for researchers to incorporate novel algorithms on demand. We developed an open-source platform combining rapid mTMS electronic targeting with accurate and autonomous robotic handling. The robot control module was developed in Python 3.11 and designed to operate with the open-source neuronavigation software InVesalius [[9]Souza V.H. Matsuda R.H. Peres A.S.C. Amorim P.H.J. Moraes T.F. Silva J.V.L. et al.Development and characterization of the InVesalius Navigator software for navigated transcranial magnetic stimulation.J Neurosci Methods. 2018; 309: 109-120https://doi.org/10.1016/j.jneumeth.2018.08.023Crossref PubMed Scopus (18) Google Scholar]. The transformation matrix between the robot and InVesalius is computed by a closed-form solution, described in Supplementary Material S1. For safe robot operation, we implemented five software layers and a force and torque sensor control, described in S2. The algorithm for robotized TMS coil positioning, defined as the robot control module, is freely available at https://github.com/biomaglab/tms-robot-control. We developed the control platform with an Elfin E5 collaborative robot (Han's Robot Co Ltd, China), which has 6 joints, a 5-kg payload, an 80-cm maximum operation range, and a repeatability accuracy of ±0.05 mm. The developed robot control module can be adapted to any commercial collaborative robot thanks to the software's modular architecture. The robotic TMS coil positioning and head movement compensation were implemented with a closed-loop control, as illustrated in Fig. 1a. The algorithm defines the robot's trajectory to move the coil to the desired target. If the patient moves beyond the threshold specified in InVesalius (default is 2 mm and 2°), the control system can detect the disturbance and adjust for head movements by utilizing the targeting feedback from the neuronavigation positioning guide. The robot-control equations are described in S3. We characterized the positioning stability of our robotic mTMS system, described in S4. Also, we characterized the accuracy of the produced induced electric field by the system [[10]Matsuda R.H. Souza V.H. Marchetti T. Cruz ASD La Kahilakoski O.-P. Laine M. et al.Characterizing an electronic-robotic targeting platform for precise and fast brain stimulation with multi-locus transcranial magnetic stimulation.BioRxiv. 2024; 2024 (12)584601https://doi.org/10.1101/2024.03.12.584601Crossref Google Scholar]. To demonstrate the combination of robotized transducer placement with the mTMS electronic targeting, we performed a motor mapping experiment with the experimental setup shown in Fig. 1b. Three healthy volunteers (age range: 32–35 years) with no reported neurological disorder participated in this study, which was conducted at the ConnectToBrain Laboratory at Aalto University. The study was approved by the local ethics committee in accordance with the Declaration of Helsinki; all participants gave informed consent prior to the experimental procedure. Neuronavigation was performed with InVesalius connected to eight Flex13 tracking cameras (OptiTrack, NaturalPoint, Inc., USA) installed in the laboratory room. The tracking cameras were positioned such that the head and coil navigation markers were visible for any mTMS coil array position. T1-weighted MRIs (volumetric gradient echo sequence; voxel size 1×1 × 1 mm3; 240×240×240 acquisition matrix) were acquired in a Skyra 3T scanner (Siemens Healthcare, Germany). Electromyography (EMG) data were recorded from the right abductor pollicis brevis (APB) muscle with a NeurOne amplifier (24-bit resolution, 5-kHz sampling frequency; Bittium Biosignals Ltd., Finland) and circular surface electrodes (24-mm diameter; Spes Medica, Italy) placed on a belly–tendon montage [[11]Cavalcanti Garcia M.A. Lindolfo-Almas J. Hiroshi Matsuda R. Labiapari Pinto V. Aparecida Nogueira-Campos A. Hugo Souza V. The surface electrode placement determines the magnitude of motor potential evoked by transcranial magnetic stimulation.Biomed Signal Process Control. 2023; 84104781https://doi.org/10.1016/j.bspc.2023.104781Crossref Scopus (0) Google Scholar]. The hotspot coil placement was defined as the placement on the scalp resulting in the highest MEP amplitudes. On the hotspot, we measured the resting motor threshold (RMT) as the minimum intensity needed to elicit MEPs in the APB larger than 50 μV peak-to-peak in at least five out of ten pulses [[12]Conforto A.B. Z'Graggen W.J. Kohl A.S. Rösler K.M. Kaelin-Lang A. Impact of coil position and electrophysiological monitoring on determination of motor thresholds to transcranial magnetic stimulation.Clin Neurophysiol. 2004; 115: 812-819https://doi.org/10.1016/j.clinph.2003.11.010Crossref PubMed Scopus (88) Google Scholar,[13]Kammer T. Beck S. Thielscher A. Laubis-Herrmann U. Topka H. Motor thresholds in humans: a transcranial magnetic stimulation study comparing different pulse waveforms, current directions and stimulator types.Clin Neurophysiol. 2001; 112: 250-258https://doi.org/10.1016/s1388-2457(00)00513-7Crossref PubMed Scopus (0) Google Scholar]. Based on the hotspot, we created three physical targets: 1) the hotspot, 2) on the medial and 3) lateral side along the left precentral gyrus. Then, we created 3 × 3 square grids of brain targets for the three targets, resulting in a total of 27 brain targets. Then, the robot control module autonomously positioned the mTMS transducer on each of the physical targets and applied five single mTMS pulses, with a randomized interstimulus interval of 2–4 s for each brain target. The stimulation intensity was set at 110% of the RMT. The motor maps were generated with the average MEP peak-to-peak amplitude across the five pulses and interpolated with a gaussian interpolation method with 4-mm radius and 3-mm sharpness. Fig. 1c shows the resulting motor maps obtained with the robotized mTMS for three volunteers. We leverage the high accuracy and autonomous operation of the collaborative robot to enable effortless and accurate positioning of mTMS coil sets. Our robotized-electronic system attains higher accuracy than manual positioning and exhibits stability and accuracy comparable with existing robotized TMS systems [[10]Matsuda R.H. Souza V.H. Marchetti T. Cruz ASD La Kahilakoski O.-P. Laine M. et al.Characterizing an electronic-robotic targeting platform for precise and fast brain stimulation with multi-locus transcranial magnetic stimulation.BioRxiv. 2024; 2024 (12)584601https://doi.org/10.1101/2024.03.12.584601Crossref Google Scholar,[14]Zorn L. Renaud P. Bayle B. Goffin L. Lebossé C. de Mathelin M. et al.Design and evaluation of a robotic system for transcranial magnetic stimulation.IEEE Trans Biomed Eng. 2012; 59: 805-815https://doi.org/10.1109/TBME.2011.2179938Crossref Scopus (36) Google Scholar,[15]Shin H. Jeong H. Ryu W. Lee G. Lee J. Kim D. et al.Robotic transcranial magnetic stimulation in the treatment of depression: a pilot study.Sci Rep. 2023; 13: 1-11https://doi.org/10.1038/s41598-023-41044-1Crossref Scopus (2) Google Scholar]. The robot control allows hands-free placement of mTMS coil arrays on a target location with real-time and automatic compensation for head movements. The robotic–electronic targeting enables the automation of mTMS protocols, such as hotspot hunting and motor mapping with closed-loop algorithms with minimal dependency on user experience and subjective analysis [[4]Tervo A.E. Metsomaa J. Nieminen J.O. Sarvas J. Ilmoniemi R.J. Automated search of stimulation targets with closed-loop transcranial magnetic stimulation.Neuroimage. 2020; 117082https://doi.org/10.1016/j.neuroimage.2020.117082Crossref Scopus (26) Google Scholar,[16]Nieminen A.E. Nieminen J.O. Stenroos M. Novikov P. Nazarova M. Vaalto S. et al.Accuracy and precision of navigated transcranial magnetic stimulation.J Neural Eng. 2022; 19https://doi.org/10.1088/1741-2552/aca71aCrossref Scopus (14) Google Scholar]. Our open-source platform for combined electronic–robotic applications is an important step in increasing the safety, accuracy, and reproducibility of TMS procedures. This platform offers new prospects to create closed-loop [[17]Weise K. Numssen O. Kalloch B. Zier A.L. Thielscher A. Haueisen J. et al.Precise motor mapping with transcranial magnetic stimulation.Nat Protoc. 2023; 18: 293-318https://doi.org/10.1038/s41596-022-00776-6Crossref PubMed Scopus (15) Google Scholar,[18]Harquel S. Diard J. Raffin E. Passera B. Dall'Igna G. Marendaz C. et al.Automatized set-up procedure for transcranial magnetic stimulation protocols.Neuroimage. 2017; 153: 307-318https://doi.org/10.1016/j.neuroimage.2017.04.001Crossref PubMed Scopus (14) Google Scholar], operator-independent brain stimulation protocols capable of covering large cortical brain areas, potentially resulting in improved treatments for neurological disorders. This work has received funding from the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (grant No. 141056/2018–5), the Academy of Finland (decisions No. 307963 and 349985), and from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No. 810377, ConnectToBrain). This article was produced as part of the activities of the FAPESP Research, Innovation and Dissemination Center for Neuromathematics (grant No. 2013/07699–0, and 2022/14526–3).