[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-290906-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-290906-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","decoding-motor-imagery-with-various-machine-learning-techniques-abstract","Decoding Motor Imagery with various Machine Learning techniques - Abstract","","Motor imagery (MI) supports brain-computer interface experiments by modulating activity in the motor cortex and surrounding regions. The study instructs subjects to perform right- and left-hand MI in two-class trials, collecting offline signals using Gel and POLiTag electrodes. Multiple machine-learning decoders are trained on subject-specific offline data, using evidence accumulation to predict intent in real time, while addressing EEG non-stationarity, inter-subject variability, and electrode impedance differences.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/decoding-motor-imagery-with-various-machine-learning-techniques-abstract/290906/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/decoding-motor-imagery-with-various-machine-learning-techniques-abstract/290906.png","ImageObject",300,407,{"name":42,"@type":43},"Eden","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-10-21","2026-09-17",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",6,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"What task and class structure does the study use for motor imagery decoding?","Question",{"text":63,"@type":64},"Subjects perform motor imagery split into two classes: right vs left hand intent. Trials provide cues and are structured as left/right cues for classification.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"Which electrode types are compared, and why does this matter?",{"text":68,"@type":64},"The study compares Gel and POLiTag electrodes. Electrode impedance differs by scalp characteristics for POLiTag, affecting signal quality and decoder performance.",{"name":70,"@type":61,"acceptedAnswer":71},"How is the decoder trained and used for real-time prediction?",{"text":72,"@type":64},"A decoder is trained from offline data collected during four offline runs using selected EEG frequency-channel features (4–30 Hz). Evidence accumulation is then used to predict intent during online runs in real time.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},290906,1789982877,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,105,110,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":102,"show_sort_weight":103,"slug":104},"Technology",50,"technology",{"id":106,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},7,"Healthcare",40,"healthcare",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},8,30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":111,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":106,"language":139,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":140,"faqs":141,"seo_title":142,"seo_description":12,"update_tm":143,"read_time":144,"preview_page_count":4},1374391974468,"https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0","Decoding Motor Imagery with various Machine Learning techniques  \nGiovanni Jana, Corey Karnei, Shuvam Keshari  \narXiv :2306 .07519v1 [ cs .HC] 13 Jun 2023  \nAbstract—Motor imagery (MI) is a well documented technique used by subjects in BCI (Brain Computer Interface) experiments to modulate brain activity within the motor cortex and surrounding areas of the brain. In our term project we conducted an experiment in which the subjects were instructed to perform motor imagery that would be divided into two classes (Right and Left). Experiments were conducted with two different types of electrodes (Gel and POLiTag) and data for individual subjects was collected. In this paper we will apply different machine learning (ML) methods to create a decoder based on offline training data that uses evidence accumulation to predict a subjects intent from their modulated brain signals in real time.  \nWhen working with BCIs, there is a direct correlation between how your decoder performs and how well your system is able to provide feedback to the user. This is motivation for researchers to continuously explore new methods of decoding brain signals for use in BCI application. Despite an advancement in technology, there are several key features of EEG signals that make them difficult to decode accurately. One feature is that they are non-stationary which means that from one session to another they can present differently. Because of this when decoding, channels must be chosen that display stability over time. Another feature is that EEG signals can present differently for different subjects. This means that a decoder that is built and trained for a certain individual will not perform well on a different individual. Due to this, we will be building decoders for each subject as well as testing multiple ML techniques for each subject. This study has an additional issue; namely the electrodes being used. Non-invasive EEG recordings often utilize an electrolyte gel in order to minimize impedance between an electrode and a subject’s scalp. This was possible for our Gel experiments, however using a newtype of electrodes resulted in increased impedance per channel that was dependent on the oiliness of a subjects scalp.  \nI. INTRODUCTION  \nDespite all of these obstacles, there are some assumptions we can make when designing our experiment. Literature has shown that there are some common characteristics of EEGs produced by subjects performing MI tasks. Roth et al. [1] showed that there is evidence of brain activity in the primary cortex during motor imagery tasks. Another commonality is that the sensory motor rhythms generated during MI occur within the alpha (8-12Hz) and beta (12-30Hz) bands as shown by McFarland et al. [2] . Specifically for our experiment we will be conducting motor imagery tasks that involve the arms and hands. The homunculus model in Figure 1 . depicts a representation of what areas of the somatosensory and motor cortex correspond to the hands and arms. Using this  \nFig. 1: Homunculus model  \nknowledge, electrodes will be placed in positions that can capture these areas. For BCI systems that involve the motor imagery task, this has been extensively studied which allows researchers to use a smaller subset of the total channels while still capturing the data necessary for classification.  \nII. EXPERIMENTAL SETUP  \nThe experiment was conducted over the course of two weeks. During this time three subjects each participated in recording sessions using both Gel-based and new POLiTag experimental electrodes. The structure of a recording can be seen in Figure 2 .  \nFig. 2: Recording Session Structure  \nFig. 3: Single Trial Structure  \nOn a given day, for one type of electrodes there was a single offline session as well as two online sections. Before the first  \noffline session there was an impedance check conducted in order to ensure that all 13 channels were operating withintolerable impedance ranges. This varied between the types of electrod","cbCaifunYrbZcWvs","https://ap.wps.com/l/cbCaifunYrbZcWvs","pdf",1563558,"English","# Abstract\n# Introduction\n# Experimental Setup\n## Offline\n## Building a Decoder\n## Online 1","[{\"question\":\"What task and class structure does the study use for motor imagery decoding?\",\"answer\":\"Subjects perform motor imagery split into two classes: right vs left hand intent. Trials provide cues and are structured as left/right cues for classification.\"},{\"question\":\"Which electrode types are compared, and why does this matter?\",\"answer\":\"The study compares Gel and POLiTag electrodes. Electrode impedance differs by scalp characteristics for POLiTag, affecting signal quality and decoder performance.\"},{\"question\":\"How is the decoder trained and used for real-time prediction?\",\"answer\":\"A decoder is trained from offline data collected during four offline runs using selected EEG frequency-channel features (4–30 Hz). Evidence accumulation is then used to predict intent during online runs in real time.\"}]","Decoding Motor Imagery with various Machine Learning techniques - Abstract | PDF",1789642934,18]